Political Constructivism

Political Constructivism

Political Constructivism is concerned with the justification of principles of political justice in the face of pluralism. Contemporary accounts of multiculturalism, pluralism and diversity have challenged the capacity of political theory to impartially justify principles of justice beyond the boundaries of particular communities. In this original account, Peri Roberts argues that political constructivism defends a conception of objective and universal principles that set normative limits to justifiable political practice. Political Constructivism explores this understanding in two ways. First, the author engages with constructivist thinkers such as John Rawls and Onora O’Neill in order to lay out a basic understanding of what constructivism is. Second, the author goes on to defend a particular account of political constructivism that justifies a universal primary constructivism alongside the many secondary constructions in which we live our everyday lives. In doing so, he outlines an understanding of principled pluralism that accepts diversity whilst at the same time recognizing its limits. This volume will be of particular interest to students and researchers of political theory and political philosophy. Peri Roberts is a Lecturer in the Politics Department at Cardiff University.

Routledge innovations in political theory

1 A Radical Green Political Theory Alan Carter 2 Rational Woman A feminist critique of dualism Raia Prokhovnik 3 Rethinking State Theory Mark J. Smith 4 Gramsci and Contemporary Politics Beyond pessimism of the intellect Anne Showstack Sassoon 5 Post-Ecologist Politics Social theory and the abdication of the ecologist paradigm Ingolfur Blühdorn 6 Ecological Relations Susan Board 7 The Political Theory of Global Citizenship April Carter 8 Democracy and National Pluralism Edited by Ferran Requejo

9 Civil Society and Democratic Theory Alternative voices Gideon Baker 10 Ethics and Politics in Contemporary Theory Between critical theory and post-marxism Mark Devenney 11 Citizenship and Identity Towards a new republic John Schwarzmantel 12 Multiculturalism, Identity and Rights Edited by Bruce Haddock and Peter Sutch 13 Political Theory of Global Justice A cosmopolitan case for the world state Luis Cabrera 14 Democracy, Nationalism and Multiculturalism Edited by Ramón Maiz and Ferrán Requejo 15 Political Reconciliation Andrew Schaap

16 National Cultural Autonomy and Its Contemporary Critics Edited by Ephraim Nimni 17 Power and Politics in Poststructuralist Thought New theories of the political Saul Newman 18 Capabilities Equality Basic issues and problems Edited by Alexander Kaufman 19 Morality and Nationalism Catherine Frost 20 Principles and Political Order The challenge of diversity Edited by Bruce Haddock, Peri Roberts and Peter Sutch 21 European Integration and the Nationalities Question Edited by John McGarry and Michael Keating 22 Deliberation, Social Choice and Absolutist Democracy David van Mill

23 Sexual Justice/Cultural Justice Critical perspectives in political theory and practice Edited by Barbara Arneil, Monique Deveaux, Rita Dhamoon and Avigail Eisenberg 24 The International Political Thought of Carl Schmitt Terror, liberal war and the crisis of global order Edited by Louiza Odysseos and Fabio Petito 25 In Defense of Human Rights A non-religious grounding in a pluralistic world Ari Kohen 26 Logics of Critical Explanation in Social and Political Theory Jason Glynos and David Howarth 27 Political Constructivism Peri Roberts

Political Constructivism

Peri Roberts

First published 2007 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN Simultaneously published in the USA and Canada by Routledge 270 Madison Ave, New York, NY 10016 This edition published in the Taylor & Francis e-Library, 2007. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.” Routledge is an imprint of the Taylor & Francis Group, an informa business © 2007 Peri Roberts All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data A catalog record for this book has been requested ISBN 0-203-46192-4 Master e-book ISBN ISBN10 0-415-33987-1 (hbk) ISBN10 0-203-46192-4 (ebk) ISBN13 978-0-415-33987-2 (hbk) ISBN13 978-0-203-46192-1 (ebk)

Contents

1

Introduction

1

2

Constructivism and A Theory of Justice

8

3

The constructivism of Political Liberalism

46

4

Primary constructivism and O’Neill

81

5

Reasoning practically: a constructivist understanding of practical reasoning

112

Making practice reasonable: political constructivism

137

Notes Index

159 182

6

1

Introduction

Political constructivism is a theory that has developed comparatively recently which is centred on a set of ideas primarily about the justification of principles of political justice, and so also about justifying political actions and institutions. Everyday, we find ourselves in situations that require us to justify principles and our actions, either to others or to ourselves. We are faced with questions such as ‘Why did you do that?’ or ‘Was that the right thing for me to do?’ or ‘Should we have treated those other people in the way that we did?’ These questions demand answers that attempt to make transparent the reasons for our actions and principles. In addition, these answers often make claims about what reasons other people have or should recognize. The answers that we give in this questioning process might themselves be open to question, and these further answers may in turn be worthy of examination. A key question for justification is to ask whether this regress of questions ever comes to an end or are our reasons subject to an infinite regress of questions and answers and questions again. Justification is usually taken to consist in identifying the stopping point, or foundation, of such a regress, the ‘unmoved mover’ that anchors a chain of reasons.1 This foundational reason is regarded as an ‘unjustified justifier’ that necessarily occupies a position of authority in justificatory argument and is therefore the ground of our reasons and justifications.2 Identifying a foundational reason prevents further questioning and underpins the authority and objectivity of our reasons and principles. Sceptics about justification usually agree that this is what justification is all about, that a justified reason either is, or is grounded on, a foundation. However, sceptics dispute whether any such foundations are available to us and thus whether objective and authoritative reasons are identifiable. Either, they argue, the regress is infinite (so that it is always open to significant question whether and when something counts as a reason) or the chains of reasoning stop in different places for different people (perhaps dependent on subjective tastes, beliefs or experiences or perhaps on the perspective of a certain society, community or culture). These positions frame some of the central questions for this book: is objective justification possible and, if so, what form does this justification take? The basic and traditional model of justification is foundationalist. Identifying a foundation for justification involves claiming that some certain reasons have a

2

Introduction

special status independent of further consideration. Foundations are what we check our everyday reasons against to reassure ourselves that our actions and principles are justifiable. The foundational reason functions as a sort of ‘moral fact’ that can be pointed to in order to verify a normative claim. On this account, when asked ‘Why should I limit my actions in this way?’ an answer that justifies would show that this is what is required by a foundational principle and that foundational principles provide reasons for everyone. As such, foundations constitute an ‘independent moral order’ that grounds legitimate normative reasoning and thereby underpins claims about the objectivity and authority of our reasons. A useful and formal statement of the core of foundationalist justification is provided by Timmons. If the propositions of a set A of ethical propositions can be justified . . . then: i there is a subset A* of A such that each member of A* can be justified independently of (i.e., without appealing to) any other member of A; ii all other members of A (all non-A*s) must include in their justification . . . some member of A*.3 For the foundationalist, a principle is justified either if it is a foundation in itself or if it can be grounded in a foundation by some plausible chain of reasoning. Foundations are often taken to be ‘undeniable and immune to revision’ as is fitting in their role as the moral facts constitutive of an independent moral order.4 There have been many suggested foundations for normative reasoning. Foundations for principles of political justice have been variously regarded as intuitively correct, self-evident, appealing to the authority of God, embodying fundamental accounts of human nature, constitutive of the common values inherent in a way of life; or existing in some transcendental or supernatural realm. Examples of prominent types of foundational justification, but by no means a definitive list, might be Platonic foundationalism that justifies principles of justice by reference to an eternal and unchanging set of facts or ideas that are more real than the transient world in which we find ourselves. For Plato, these foundations are the Forms and ultimately the Form of the Good.5 We are generally more familiar with this sort of foundational account when it is understood in religious terms as an account of the word of God or of natural law. We are also familiar with what can be referred to as economic foundationalism.6 Here, the foundation is a particular understanding of human nature and the structure of human reasoning and motivation. Reason is understood instrumentally as a tool for the efficient satisfaction of pre-existing and motivating desires or preferences rather than as critically assessing which desires we ought to have. Here, our desires function as a foundation for reasons and justification must be grounded on these desires in order to be successful. Some desires, such as the desire for security, might appear to be motivating for everybody. On this account, perhaps as exemplified by Hobbes, we cannot give reasons for our desires, only for our actions in pursuit of their satisfaction.7 Desires are the ‘stopping points’ for chains of justificatory reasons. We are also familiar with cultural foundational-

Introduction

3

ism in which the fundamental values of a way of life are regarded as authoritative for members of the associated culture. Actions and principles are justified if it can be shown that they are required by ‘our way of life’ or by an account of our shared understandings of social meanings.8 Each of these foundational accounts of justification identifies foundations as the basis of an explanation of why certain sorts of reasons are objective and thus authoritative for us; they are God’s will; we have the relevant desire or we have been born into and shaped by the relevant way of life. Each of these foundational strategies for justification has been subject to long-standing questioning. However, the idea that reasons can be objective at all has also been subject to fundamental examination. Scepticism about the possibility of objective reasons focuses on the partiality of proposed foundations. Sceptics might draw our attention to the many ‘cultural anthropologists ready and waiting to unveil exotic tribes and bizarre rituals’ in order to question the self-evidence of our intuitions or our confidence in the universality or objectivity of our claims about abstract moral facts.9 What we regard as intuitive or self-evident may not be as evident to others as it is to us and may fail to exhibit Platonic or theological ‘deep connection with the fabric of the universe’.10 That this is the case explains our widespread experience of disagreement about basic principles, even about those that are supposed to be self-evident. On this account, the partiality of our principles of justice and conceptions of the good is highlighted by our experience of multiculturalism and the normative and political pluralism and diversity that underpin it. Our reasons and principles may come to be regarded as historical artefacts of a local history and a particular way of life.11 In this way, reasons and justifications are understood to be always partial, always reflecting the local and particular rather than the general and universal. While, strictly speaking, some of these accounts may be structurally foundationalist (they are often forms of cultural foundationalism), they are concerned to undermine the traditional, foundational understanding of objectivity. Instead of objective reasons providing reasons for everyone, reasons can only ever provide reasons for ‘people like us’ and ‘people round here’. The environment for normative justification is always limited by necessary theoretical, practical and motivational limitations on the relevance of our reasoning. This understanding of reasoning is familiar to us in critiques of universalism from postmodernism and from the mainstream of communitarian and multiculturalist literature.12 Common to each of these accounts is an understanding that the sources of authoritative reasons are plural and probably incompatible with each other. The argument goes that our continuing experience of pluralism and disagreement implies that it is not possible to subsume these diverse reasons under a single account of reasoning or reasonableness. Reasons are therefore often regarded as incommensurable in that they cannot be either translated into each other or compared by reference to some common value. Pluralism appears irreducible and this in turn implies a broad conceptual relativism (in that reasons and values count as reasons and values only relative to one of a plurality of ways of life or to one amongst several conceptual frameworks, for example).

4

Introduction

Irreducible pluralism is considered to undermine any claim that particular principles or reasons might be objective and universal. Claims to have identified universal and objective values are regarded either as justified on the basis of a mistaken overconfidence in the foundational character of what are actually partial assumptions or as not justified at all. Much of contemporary political theory accepts that the contexts for reflection and action are plural; indeed, pluralism is often regarded as a background condition against which theorists work. However, contemporary theory also witnesses consistent attempts to demonstrate that some sort of universalist account of justified political principle is possible, in spite of a general acceptance of the ‘fact of pluralism’ and the reasonableness of our disagreements.13 These accounts have tended to be ‘thin universalist’ conceptions that try to minimize partial and controversial claims, perhaps by justifying on a thin, rather than a thick, set of universal principles. Although the term thin universalism is Walzer’s, which he contrasts with the various ‘thick’ or ‘maximal’ cultures we inhabit, his own account is not as useful as we might hope.14 As is discussed below, in the final chapter of this book, Walzer’s account does not justify objective and authoritative universal reasons or principles. Instead, his thin universalism is a matter of the contingent overlap of our extended sympathies, an expression of solidarity firmly rooted in our various maximal moralities. To a significant degree, Walzer’s is not an account of universal reasons or justification at all, but an explanation of how it is possible that situated persons might come to identify sympathetically with others who do not, initially, appear to be very much like them. Constructivism is an alternative account of what a thin universalism might be. Whilst accepting both the fact of pluralism and the reasonableness of disagreement, constructivism continues to be committed to the idea of objective and authoritative justification. Where objective justification has traditionally been foundationalist, constructivism attempts to show that it does not have to be. Constructivism need not reject the possibility of a successful foundationalism but does demonstrate that foundations are not a necessary element in any objective account of reasons and principles. Part of this task involves working out what an appropriate conception of objectivity is where normative judgements are concerned. We might ask if there is something about the circumstances in which persons find themselves, in the relations between them and in the way they reason practically, which enables us to regard objective standards for normative judgement as a human construct.15 As will become clear, in exploring these possibilities, constructivism will draw heavily on an everyday understanding of objectivity as involving unbiased and impartial universal reasons for everyone. Constructivism argues that at least some universal reasons of this sort are justifiable. In sharp contrast with the perspectivism of particularists such as the communitarians and many multiculturalists, which regards a demonstration of pluralism as the end of an argument about justification, constructivists regard pluralism as the start of an argument, as an invitation to reflection and critical reasoning.

Introduction

5

Not surprisingly, given its title, this is a book about constructivism in political theory; however, it is about constructivism in two different ways. In part, it focuses on prominent constructivist positions, most notably those of John Rawls and Onora O’Neill, in an attempt to better understand what political constructivism involves. Here, we find that there is a broadly shared account of constructivism’s key features and that these can be drawn out in a general account of what constructivism is and of what it entails for the justification of political principles. This book also takes the first steps in a more ambitious project. This involves providing an independent set of arguments, related to but markedly distinct from those of Rawls and O’Neill, which develops and justifies a constructivist approach to principles of political justice in a simple but plausible way. Both of these tasks are important. Although ideas of constructivist justification have been prominent in contemporary political theory for several decades they have been the subject of few extended studies and only a small number of important papers.16 On the one hand, taking a detailed and systematic look at several constructivist positions is a contribution to this much-needed discussion. On the other hand, developing an independent argument for an innovative constructivist position shapes the way that this discussion must develop in the future. Chapters 2–4 form a distinct group focused on the constructivism of Rawls’s A Theory of Justice, on the constructivism of Rawls’s Political Liberalism and on O’Neill’s constructivist writings, respectively. The main body of each chapter is concerned to refine our understanding of what constructivism might involve by examining a prominent constructivist position. This necessitates fairly detailed textual engagement, and there are times when this process is unavoidably pedestrian. However, just as pedestrians make real progress towards their destination with their small steps, so will we. It becomes clear that a general understanding of constructivism develops and that, importantly, something like this basic position holds steady throughout the breadth of both Rawls’s and O’Neill’s positions. In A Theory of Justice, Rawls outlines a constructivist project where objective principles of justice are justified by subjecting uncontroversial assumptions to reflective examination in an account of practical reasoning based on a conception of reflective equilibrium. Chapter 2 lays out a straightforward and largely mainstream account of Rawls’s justice as fairness and shows how its justification is consistent with this basic constructivist account. However, it also becomes clear that Rawls expects too much of his constructivism as he tries to over-determine its conclusions. His constructivism cannot bear all the weight that justice as fairness appears to require. Rawls attributes this failure to a lack of awareness concerning the extent of pluralism and the implications of this for justification, a mistake he attempts to rectify in Political Liberalism. Chapter 3 examines this reworking of constructivism in order to proof it against continued failure. He proceeds by explicitly outlining a conception of political constructivism that draws on ideas fundamental to the public political culture of democratic society as resources available to

6

Introduction

constructivist justification, notably ideas of society and the person. This move has encouraged many commentators to regard political liberalism as a local and particularist theory that accepts broadly communitarian limits on justification. This chapter innovatively demonstrates that a continued universalist account of justification can be maintained by Rawls if he recognizes his implicit distinction between two levels of constructivism, a universal primary construction and many local and secondary constructions. It also becomes clear that much of Political Liberalism, and the criticism that it draws, is focused on Rawls’s account of a specific form of secondary construction suitable for democratic societies and ignores the possibility of a universal primary construction. The chapter concludes by briefly showing that Rawls’s secondary constructivism tries, as he did in A Theory of Justice, to over-determine constructivist outcomes and so is legitimately subject to criticism. Chapter 4 deals with O’Neill’s constructivism as laid out predominantly in Towards Justice and Virtue and Constructions of Reason. Her reworking of Kantian philosophy encourages us to focus more closely on a constructivist account of practical reasoning and its objectivity and on its necessary presuppositions. In doing so, we find that not only does O’Neill recognize the distinction between primary and secondary constructivism but also her account of a universal primary constructivism is very similar to the general account we found in Rawls. However, like Rawls, O’Neill may push her constructivism too far. Whilst she outlines a broadly convincing account of the justification of objective principles of justice, her attempts to fill out the space between these principles with an account of virtue may not be so successful. Whereas Chapters 2–4 make progress towards a general account of constructivism by building from Rawls and O’Neill, Chapters 5 and 6 develop an independent set of arguments for the political constructivist position. Chapter 5 concentrates on developing a constructivist account of practical reasoning, its necessity, authority and objectivity within the constraints of practicality. Chapter 6 deploys this account of practical reasoning to underpin an account of the limits of reasonable practice, a theory of political constructivism. This is used to show that at least some universal principles of political justice can be justified and goes on to explain how these are related to the many secondary constructions that are the contexts for our everyday lives. It becomes clear that, rather than pluralism setting limits to the scope of principles of justice as many have argued, primary constructivist principles of justice set the limits of a justifiable understanding of principled pluralism. As well as the usual thanks to family in the development and writing of this book, I have incurred several further debts of gratitude that demand recognition. First, I must thank Mark Evans and Roland Axtmann of Swansea University’s Department of Politics and International Relations for allowing me the use of their facilities in 2006. I would also like to thank Cardiff’s School of European Studies, specifically David Boucher and Paul Furlong, for arranging some very necessary research leave. I must acknowledge with gratitude the continued encouragement of Mike ‘Hong Kong’ Foxy, a number one super guy. Thanks

Introduction

7

are also owed to Peter Sutch, not only for many years of ongoing argument (sorry, discussion) but specifically for a particular conversation that started a long and exhausting train of thought. Finally, my largest debt is to Bruce Haddock for support and encouragement generally and for being a patient sounding board for half-developed ideas.

2

Constructivism and A Theory of Justice

Rawls’s A Theory of Justice is an important constructivist response to a concern with pluralism. On first reading, this may not be immediately obvious as his explicit intention is to provide a systematic alternative to the dominance of utilitarianism as the moral doctrine forming the basis of a constitutional democracy. Dissatisfaction with utilitarianism is most often expressed by pointing out the many ways in which its conclusions appear to directly contradict common moral sentiments or intuitions; that it cannot provide a satisfactory account of basic rights and liberties, that it does not take seriously the claim that we all lead separate lives and that we should therefore not be sacrificed for the benefit of others, that its characterization of the good is overly simplistic, for example. Rawls takes our moral intuitions very seriously. If a principle consistently contradicts our intuitions, then this is a good reason to look closely at its justification. As we shall see, matching our considered intuitions is not what he takes justification to consist in, but his constructivism takes intuitions seriously as starting points for justification. As we shall also see, he does so in a way that avoids regarding these intuitions as foundational. If we are to understand the justification of our political principles and institutions, Rawls claims that we might think that we are forced into a choice between utilitarianism and intuitionism, the ‘doctrines which have long dominated our philosophical tradition’.1 Utilitarianism, Rawls accepts, forms a tacit, if not always explicitly acknowledged, background against which principles and institutions are justified. Whereas utilitarianism provides a systematic and unified account of moral and political justification, the intuitions opposing utilitarian conclusions are many and varied. We have a range of intuitions about the injustice of utilitarianism, but it is difficult to subsume them in a unified account providing a systematic and powerful alternative to utilitarianism. Intuitionism posits an irreducible pluralism of basic intuitions or first principles that may conflict in particular cases. As we have neither the ‘supreme’ intuition that acts as an organizing principle nor an obvious set of priority rules that would establish ordered relations between intuitions, each intuitive principle must be regarded as a separate first principle. Different priority relationships between principles could be established by different people accepting different relative weightings of principles dependent on their conflicting interests. In this way, pluralism and

Constructivism and A Theory of Justice

9

conflict at the level of intuitions are reinforced in a more concrete pluralism and conflict at the level of political and personal relations. It is this political and social pluralism that is understood to underpin the circumstances of justice and for which conceptions of justice propose principles of regulation. This pluralism of its basic principles prevents an intuitionist conception of justice from providing a systematic justification for any particular set of political principles or institutions and therefore prevents it from being a proper and effective alternative to utilitarianism. Rawls’s response to the problems of pluralism and conflict is to propose his conception of ‘justice as fairness’ as just such an alternative to both utilitarianism and intuitionism. Whilst he does not at this stage describe the justification of justice as fairness as constructivist, he is explicitly motivated by the identification of ‘constructive’ criteria or principles that would establish the appropriate priority and emphasis between conflicting intuitions.2 Whereas the plurality of first principles undermines the ability of intuitionism to provide clear guidance in difficult moral circumstances, the aim of justice as fairness is to provide determinate principles with definite priorities established that can in turn provide the necessary guidance despite intuitional and political pluralism. Rawls’s constructivism is therefore at least partly motivated by the search for determinate principles in response to moral and political pluralism. It will become apparent that aiming for determinate principles is precisely what endangers the success of that constructivist project. For intuitionism, the plurality of intuitions as first principles function as a set of independent moral facts that principles must ‘match’ or ‘fit’ in their attempt to establish priority rules between them. These moral facts are independent of the rules that establish their relative weighting and therefore hold a special status; we do not establish them but intuit them as independent of our engagement with them. Intuitionism regards these intuitions as foundations that must be accounted for in any acceptable set of principles. In this way, the priority problem can be regarded as a side effect of dealing with the ‘complexity of already given moral facts which cannot be altered’.3 Rather than accepting this foundationalist approach that posits an independent moral order that we must recognize, Rawls takes as his point of departure the central ideas of the ‘traditional theory of the social contract as represented by Locke, Rousseau, and Kant’.4 Just as social contract theorists used the idea of people in a state of nature in the identification of political principles, so Rawls outlines his idea of parties in an original position. As we will explore in some detail, rather than have principles fit a given and independent order of moral facts, ‘the moral facts are determined by the principles which would be chosen in the original position’ and it is up to the parties to that choice ‘to decide how simple or complex they want the moral facts to be’.5 Rawls appears to be rejecting the foundationalist understanding of moral and political justification altogether. Principles are not to be established by reference to a set of facts with special status; there are no moral facts outside of our justifications.6 This claim means that we will have to pay particular attention to the way in which Rawls’s constructivism outlines an appropriate conception of the objectivity of principles in the absence of foundations. If we begin to understand the

10

Constructivism and A Theory of Justice

relationship between principles, intuitions and the conception of objectivity integral to Rawls’s justification of justice as fairness, we will have made significant progress in our attempt to understand constructivism. In working through A Theory of Justice, we will inevitably concentrate on some aspects of this lengthy book rather than others.7 We will not systematically engage with the direct argument with utilitarianism, for example (or with the huge volume of scholarship that accompanies it). We will instead concentrate on laying out Rawls’s alternative account of the justification of principles of justice and political institutions. Recognizing the successes and setbacks of this account will significantly further our understanding of the possibilities for objective justification beyond foundationalism.

Understanding A Theory of Justice The intuitive idea of justice as fairness is to think of the first principles of justice as themselves the object of an original agreement in a suitably defined initial situation. These principles are those which rational persons concerned to advance their interests would accept in this position of equality to settle the basic terms of their association.8 As this quote highlights, justice as fairness is initially characterized as having two fundamental parts: first, an interpretation of the initial situation of equality that he refers to as the original position and of the problem of choice faced by its inhabitants, and second, a set of principles that would be agreed to by the inhabitants of that original position.9 This bipartite division of labour is subsequently supplemented by a procedure that becomes the third part of justice as fairness, the procedure of reflective equilibrium. As we shall see, the idea of seeking reflective equilibrium is an understanding of practical reasoning that instructs us, once the original position has issued in a set of first principles, to check the principles of justice against ‘our considered judgements duly pruned and adjusted’.10 These three parts of the structure of justice as fairness can be roughly but usefully superimposed on the three distinctive ‘points of view’ within the theory that Rawls draws to our attention in his ‘Kantian Constructivism in Moral Theory’.11 These three points of view are 1 2 3

that of the parties in the original position (this corresponds to the initial choice situation), that of the citizens of a well-ordered society (it is for such a society that the parties in the original position are to choose the principles of justice) and that of ourselves considering the basis of a conception of justice (it is us who employ the procedure of reflective equilibrium).

It is these parallel three-way distinctions that will usefully structure our brief but necessary exposition of justice as fairness and that will help us to better understand Rawls’s constructivism.

Constructivism and A Theory of Justice

11

The original position – structure and parties Rawls is very clear that the original position does not describe a situation in which human beings could ever find, or indeed have ever found, themselves, nor is his description of the parties that inhabit that position ever offered as an account of the psychological or cultural possibilities for actual people. Rather, it outlines a hypothetical or imaginary situation that is populated by hypothetical or imaginary people which, if properly understood, aids in our identification of principles of justice.12 Rawls argues that we should take the choice made in this hypothetical choice situation very seriously in any critical reflections. However, the hypothetical nature of the original position is often the source of uneasiness, with the broad thrust of Rawls’s argument. Someone might feel generally uneasy with the idea of this sort of hypothetical argument in any form, but this uneasiness is misplaced. We are often comfortable with hypothetical argument, especially in moral and political contexts. For example, we think it is a fair move in discussion to ask someone, ‘How would you like that if it was done to you?’ ‘How would you react if you were in their shoes?’ ‘What would it be like if everybody did that?’ or ‘What would society be like if our laws were slightly different?’ These are ordinary and familiar examples of the everyday use we make of hypothetical argument without discomfort, and we are not at all surprised if we are confronted with these sorts of questions ourselves. Perhaps, rather than this general concern with hypothetical argument, we should be more concerned with Rawls’s specific imaginings. As the original position is hypothetical rather than real might Rawls’s characterization of this situation and its inhabitants be arbitrary or contingent? Can we imagine just any sort of original position and have it play an important role, or are there particular limitations that make one imagined position better than others? The only way that Rawls can address this concern is to lay out his very specific understanding of the original position and the parties that inhabit it, confident that we can endorse this understanding on due reflection. The structure of the original position and the description of the parties that inhabit this hypothetical choice situation are inextricably linked.13 It is in these that what Rawls refers to as the ‘main idea of the theory of justice’ is laid out; the main idea being that the principles of justice ‘are the principles that free and rational persons concerned to further their own interests would accept in an initial position of equality’.14 The structure of the original position outlines an initial position of equality within which the parties (regarded as ‘free and rational people’) choose principles. Although it is difficult to separate the parties from the conditions of their situation, it is possible to make a few descriptive points. Rawls makes a range of ‘motivational assumptions’ about the parties in the original position. They are described as ‘rational and mutually disinterested’ and as each having their own ‘rational plan of life’, their own conception of the good.15 Their rationality is interpreted in the ‘narrow’ and ‘economic’ sense of ‘taking the most effective means to given ends’.16 As instrumental reasoners,

12

Constructivism and A Theory of Justice

‘each tries as best he can to advance his interests’.17 That they are mutually disinterested should not be taken to imply pure self-interest; instead, ‘they are conceived as not taking an interest in one another’s interests’.18 This plays out as the assumption that the parties are not motivated by envy; in the choice of principles, they are concerned only to advance their own system of ends and do not concern themselves with their position relative to others.19 Mutual disinterest also highlights the point that the parties are ‘not bound by prior moral ties to each other’.20 They do not therefore conceive of themselves as necessarily co-members of particular associations that could be taken to impose mutual claims or obligations on each other. This emphasizes the role that the original position plays in identifying the important obligations of justice. Further, the parties are considered to be ‘representing continuing lines of claims’.21 They are thought of as heads of families taking an interest in the descendants immediately following them. This is to ensure that the interests of all, whenever they come into society, are considered in the choice of principles. On top of these motivational assumptions, the parties are regarded as having their own conceptions of the good. These are the, perhaps very different, ‘rational long-term plans’ that determine their aims and interests and shape their lives, hopes and ambitions.22 The situation these parties find themselves in is a very peculiar one indeed. This original position is intended to act as the initial position of equality from which the rational parties must make their choice. However, having assumed that each party has his own personal and perhaps very different conception of the good, Rawls is faced by the problem of pluralism. Precisely because each has a different plan of life that entails different goals and desires, instead of agreement on principles, there is likely to be discord. To enable the parties to reach agreement on principles of justice, Rawls claims that the right theory ‘excludes the knowledge of those contingencies that set men at odds and allows them to be guided by their prejudices’.23 Indeed, when fully laid out, the original position is ‘the result of leaving aside those aspects of the social world that seem arbitrary from a moral point of view’.24 The intention is to avoid basing the justification of principles on contingent and arbitrary factors in an attempt to ensure that justification has a sense of objectivity. By abstracting away from the things that divide us, Rawls aims to identify principles of justice that everyone can accept. In order to flesh this out, Rawls makes what he takes to be a number of ‘widely accepted but weak’ presumptions. The aim is that ‘each of the presumptions should by itself be natural and plausible; some of them may seem innocuous or even trivial . . . [but that] taken together they impose significant bounds on acceptable principles of justice’.25 Rawls identifies a number of ‘natural presumptions’ that apply to the conditions under which principles of justice should be chosen. These presumptions rule out basing principles of justice on obviously biased and irrelevant information. Prominent amongst them are that ‘it should be impossible to tailor principles to the circumstances of one’s own case’ and that we should ensure ‘that particular inclinations and aspirations, and persons’ conceptions of their good do not

Constructivism and A Theory of Justice

13

affect the principles adopted’.26 Also suggested by Rawls is the presumption that ‘no one should be advantaged or disadvantaged by natural fortune or social circumstances in the choice of principles’.27 Rawls makes the suggestion that the best way in which to ensure that the deliberation of the parties meets these criteria is to deprive them of particular information in order to ‘nullify the effects of specific contingencies which . . . tempt them to exploit social and natural circumstances to their own advantage’.28 Rawls excludes this information on the grounds that it is arbitrary from a moral point of view and does so by situating the parties behind a ‘veil of ignorance’. This veil is an imaginary device that screens, and therefore limits or filters, the information available to the parties in their deliberation. Behind the veil, no one knows his place in society, his class position or social status; nor does he know his fortune in the distribution of natural assets and abilities, his intelligence and strength, and the like. Nor, again, does he know his conception of the good, the particulars of his rational plan of life, or even the special features of his psychology such as his aversion to risk or liability to optimism or pessimism. More than this, I assume that the parties do not know the particular circumstances of their own society . . . they do not know its economic or political situation . . . [its] level of civilisation and culture . . . [or] to which generation they belong.29 They do however need some information in order to guide their choice. Whilst the veil has screened from the parties’ knowledge of all particular information, It is taken for granted . . . that they know the general facts about human society. They understand political affairs and the principles of economic theory; they know the basis of social organisation and the laws of human psychology. Indeed the parties are presumed to know what ever general facts affect the choice of the principles of justice.30 As a result of the introduction of a veil of ignorance, the parties ‘do not know how . . . various alternatives will affect their own particular case and they are obliged to evaluate principles solely on the basis of general considerations’.31 By abstracting away from specific and subjective contingencies and relying solely on general considerations, Rawls is laying the ground work for later claims that the principles of justice are objectively justified in that they are not based on subjective considerations or assumptions and are therefore not obviously biased towards persons with particular attributes or desires. In addition to the presumptions that inform the veil of ignorance and to ensure that the parties are similarly situated, Rawls also stipulates that the parties to the original position are themselves equal. That is, they ‘all have the same rights in the procedure for choosing principles; each can make proposals, submit reasons for their acceptance, and so on’.32 Rawls stipulates this equality as he believes that people’s systems of ends and conceptions of the good are not

14

Constructivism and A Theory of Justice

ranked in value and also that each person has the ability to understand and to act upon the principles that are chosen.33 Together with the veil of ignorance, these conditions define the principles of justice as those which rational persons concerned to advance their interests would consent to as equals when none are known to be advantaged or disadvantaged by social and natural contingencies.34 Having laid out the main features of the original position, and the main characteristics of the parties, Rawls has outlined an objective standpoint from which it is appropriate to identify principles of justice. His contention is that the principles chosen from this standpoint would be objectively justified as they would not have been chosen on obviously partial or subjective grounds. However, as they are denied the particular knowledge of their circumstances, abilities, weaknesses and personal conceptions of the good, the parties can have no grounds upon which to discriminate between proposed principles of justice and on which to base their choice between them. The general facts of which the veil of ignorance permits them knowledge are insufficient on their own to determine the choice of principles. To this end, they must be supplemented by some method of identifying criteria by which the parties’ deliberations can be guided. Rawls takes a first step in this direction by introducing the ‘thin theory of the good’. The general idea here is that although the parties are denied the knowledge of their full conceptions of the good, and therefore the full range of interests they are motivated to advance, it is possible to identify the ‘bare essentials’ of any conception of the good.35 Summing up, each person has a conception of the good, and a person’s good is determined by this ‘rational long-term plan of life’. The happy person is one who is fairly successful in the carrying out of this plan, and for him, the good is the satisfaction of the rational desires that contribute to his plan.36 The parties are supposed to be aware of this general characterization of the good involved in living a life, although they are also aware that each of them will probably have a very different plan of life involving the satisfaction of widely differing desires. However, the simple but general thought that each person has desires and needs can tell us something important. It tells us that people are not complete and self-sufficient. They are needy and, as the parties are aware from their knowledge of the general character of economic and social theory, at least in part dependent (or interdependent) on the cooperation of others for the satisfaction of these needs. We cannot assume that anyone either lacks any needs or vulnerabilities at all or is capable of meeting all those needs in isolation from their fellows. This is only a first step, however. Without some idea of the concrete content of these plans of life, and the desires and ends they involve, the parties still have no guidance in their deliberations. Here, Rawls introduces a final consideration, the ‘Aristotelian principle’. Briefly, this principle holds that human beings prefer ‘activities that depend upon a larger repertoire of realized capacities and that are more complex’.37 The more complex an activity, and the greater its capacity to

Constructivism and A Theory of Justice

15

exercise our abilities, the more we will enjoy it. This principle is described as a ‘tendency and not an invariable pattern of choice’.38 It is however a ‘basic principle of motivation’, a ‘deep psychological law’; indeed, it is ‘a natural fact’.39 Rawls goes so far as to say that the Aristotelian principle is ‘a feature of human desires as they now exist’ and ‘characterizes human nature as we know it’.40 That this is so is ‘borne out by many facts of everyday life’.41 The Aristotelian principle accounts for our willingness to undergo training and education, the complexity of our games and the intricate subtleties of social interaction.42 The principle helps account for our judgements of value when choosing a plan of life.43 Indeed, when coupled with general facts about human needs, capabilities and interdependency, the Aristotelian principle allows us to draw conclusions about the plurality of plans of life, conclusions that can guide deliberation behind the veil of ignorance.44 These conclusions take the form of a list of social primary goods.45 Drawing on his formal outline of a thin theory of the good, Rawls defines primary goods as, things which it is supposed a rational man wants whatever else he wants. Regardless of what an individual’s rational plans are in detail, it is assumed that there are various things which he would prefer more of rather than less. With more of these goods men can generally be assured of greater success in carrying out their intentions and in advancing their ends, whatever these ends may be.46 ‘These are things that it is rational to want whatever else one wants . . . given human nature, wanting them is part of being rational.’47 The gist of the idea is to identify a class of goods that are wanted as part of any rational plan of life. Rawls does this by drawing on the thin theory of the good as characterizing the bare essentials of what it is to have a plan of life coupled with the motivation provided by the Aristotelian principle. The procedure commands us to examine an object for its general desirability in relation to the thin theory. Once we establish that an object has the properties that it is rational for someone with a rational life plan to want, then we have shown that it is good for him. And if certain sorts of things satisfy this condition generally, then these things are human goods.48 These general goods are primary goods as a preference for them is rational, whatever else one wants. The Aristotelean principle that consists in the claim that people prefer complex activities that require more developed capacities, greater opportunity to exercise our abilities and access to the wide range of social and material goods that makes this development and exercise possible underwrites the claim that it is rational to want more of these general goods rather than less. The aim is to ensure the general applicability of this notion of primary goods by deriving it from ‘only the most general assumptions about

16

Constructivism and A Theory of Justice

rationality and the conditions of human life’, so as ‘not to tie . . . [the theory] to a particular pattern of human interests’.49 Social primary goods can be thought of as general and all-purpose means to whatever ends we happen to have. As such, they privilege no particular ends and so are not biased or partial towards any particular person or their contingent conception of the good.50 Rawls lists the goods that he thinks fulfil these conditions as follows: rights and liberties, powers and opportunities, income and wealth and self-respect.51 This conception of primary goods enables the parties in the original position to make a rational choice despite the restrictive conditions imposed by the veil of ignorance. Although they are unaware of their particular conception of the good, knowing their preference for more rather than less of the social primary goods can serve to guide the ranking of conceptions of justice. The thin theory and the notion of primary goods outline the form or structure that all plans of life exhibit, and this can provide a rational basis for choice between principles. This information will be sufficient to allow mutually disinterested parties to understand what would advance their interests when choosing principles despite the other restrictions on available information. So, although ‘the argument supporting them uses only the thin theory of the good and its list of primary goods’, the selected principles of justice can have a definite content, rationally chosen.52 The thin theory and the list of primary goods it identifies are effectively an abstraction from our real and diverse conceptions of the good. As primary goods are needed by everyone, they can form the basis of a ‘universal’ motivation for the parties in the original position. This abstraction from contingent and plural conceptions of the good, if it is successful, would be a considerable step towards justifiable and objective principles of justice. We can now consider what principles would actually be preferable, and therefore chosen, from the point of view of the parties to the choice in the original position. Principles of justice/well-ordered society The parties to the original position take themselves to be choosing the best conception of justice for the citizens of a well-ordered society. A well-ordered society is taken to be a ‘cooperative venture for mutual advantage’ but one that is ‘effectively regulated by a public conception of justice’. A society is well ordered if it is concerned with advancing the interests of its members, its citizens all accept (and know the others accept) the same principles of justice, and the basic institutions of society generally satisfy and are known to satisfy those principles.53 To these conditions of well-orderedness we can also add that the citizens regard each other as free and equal, that they are generally compliant with the principles of justice, that those principles are stable over time in their fostering the requisite sense of justice, and finally that the circumstances of justice obtain and are known to do so.54 A further condition on the subject of choice is that it is a choice of principles of justice for only the ‘basic structure of society’. By basic structure, Rawls means to indicate ‘the political constitution and the principal economic and

Constructivism and A Theory of Justice

17

social arrangements’ that define rights and duties over the course of an entire life.55 Rawls’s constructivism does not, therefore, direct itself to the choice of a set of moral principles for use in the everyday personal decision-making of individuals. What is chosen in the original position is a set of public moral principles for setting conditions on the limited area of political decision-making. This is why, although he does not use the term until later works, it makes sense to regard this as a ‘political’ constructivism. The basic structure is made the primary subject of justice because of the profound and all-pervading influence it exerts on the success or otherwise of the plans of life of citizens. If the basic structure is just, this does not guarantee that people lead happy and successful lives, but if it is unjust, their chances of living such a life are likely to be severely diminished. Because this choice is so important it should also be a unanimous one. This ensures that the principles chosen can be freely affirmed on an equal footing as the condition of unanimity effectively gives each party a veto on that choice. Finally, to make the choice manageable, the parties are not asked to discover the best of all possible principles but are instead presented with ‘a short list of traditional conceptions of justice’ with a few possibilities added.56 The parties are then asked to rank these conceptions of justice, and their choice is made for the conception of justice that takes top ranking. Rawls argues that the parties will reason along certain definite lines in their consideration of the alternatives. First, they will view their choice as a situation analogous to one in which a maximum rule of decision-making applies.57 This rule instructs us to attempt to maximize the minimum or worst outcome in any choice where probable outcomes are uncertain. This rule is to be regarded as analogous to the reasoning of the parties because the veil of ignorance excludes anything but the vaguest knowledge of probabilities, making probabilistic calculation inapplicable.58 The parties do not know, nor can they calculate, their likely position in society nor their probable place in the distribution of social advantage and natural talent since the veil filters out the information necessary as the basis of such probability calculations. If we add to this the importance of the outcome of the choice (since the institutions of the basic structure profoundly influence life prospects), Rawls thinks the parties will be fundamentally concerned to minimize the possibility of receiving inadequate resources in order to minimize the chance of these institutions ruining their prospects for a successful life. It may be that they are wealthy and talented, but it may be that they are not. In these circumstances, Rawls believes that since the parties cannot know their position in society, they would reason so as to make the worst-off position as good as it can be. The parties would rationally consider the choice as a choice of a conception of justice ‘for the design of a society in which his enemy is to assign him his place’ and seek to avoid the dangers this presents.59 In order to make the worst-off position as good as it can be, the parties first need to identify the worst-off. They simplify this problem by thinking of society as made up of a number of groups, taking each to contain a representative person on whose expectations conclusions about the group can be based.60 It is here that the notion of a thin theory of the good and the idea of primary goods

18

Constructivism and A Theory of Justice

derived from it are put to use. They constitute an index of expectation and wellbeing against which to compare outcomes. The least well-off group is identified by finding the worst-off representative man defined in terms of the receipt of the social primary goods necessary for the success of any rational plan of life. Since they are necessary for the pursuit of any conception of the good and it is rational to want more of them rather than less, the fewer primary goods you have, the worse off you are likely to be. The chosen principles would then be those that provided the worst-off group with the largest bundle of social primary goods. If we recall the list of social primary goods Rawls identifies, we can note that they fall into distinct groups. Into one group we can place rights and liberties; into a second, opportunities and powers; a third group would consist of income and wealth; and a fourth, solely of self-respect.61 Rawls argues that so long as society is at a level of general economic security, the parties in the original position would place more emphasis on their rights and liberties than on financial considerations. Two of his reasons in support for this argument point at the diminishing marginal returns of further economic and social advantages over a certain level and the regulative role that an interest in determining our plans of life takes on.62 These considerations led Rawls to believe that the parties would assign a priority to liberty in their choice of principles. Rawls has set up the original position with all its constraints; he has peopled it with parties who are to choose from this constrained position the principles of justice suitable for regulating the lives of citizens of a well-ordered society. He has outlined much of the course that he believes the reasoning of the parties will take. He concludes from this investigation that the parties will choose two principles of justice. First principle Each person is to have an equal right to the most extensive total system of equal basic liberties compatible with a similar system of liberty for all. Second Principle Social and economic inequalities are to be arranged so that they are both: a to the greatest benefit of the least advantaged . . . and b Attached to offices and positions open to all under conditions of fair equality of opportunity. To these he attaches two priority rules that rank the principles in lexical order. They assign the first principle lexical priority over the second (the priority of liberty) and, within the second principle, (b) priority over (a).63 Enjoyment of the good of self-respect will be the outcome of the proper instantiation of these principles as they express a general commitment to equal citizenship for all. The application of these principles is also touched upon and must be envis-

Constructivism and A Theory of Justice

19

aged within justice as fairness as part of a four-stage process.64 This process involves a progressive lifting of the veil of ignorance at each stage intended to enable the necessarily abstract principles of justice to be brought down to earth. Each lifting of the veil allows the parties access to increasingly particular types of information that was previously denied to them. The first stage of this fourstage sequence is the original position itself, within which the principles of justice are chosen. The veil is lifted slightly to allow the formation of a constitutional convention at stage two that deals largely with the application of the first principle. It is lifted further for the third or legislative stage that is primarily concerned with the application of the second principle. Finally, the veil is lifted completely at the fourth stage, that of the judicial and administrative application of the results of stages one to three to particular cases.65 At the end of working through this four-stage sequence, the conception of justice chosen by the parties to the original position for the citizens of a well-ordered society is as complete as it can be. At this stage, we must note some of the claims Rawls makes for this extended argument from the original position to the two principles of justice as fairness. He claims that his ‘theory of justice is a part, perhaps the most significant part, of the theory of rational choice’.66 Indeed, he writes that he has ‘base[d] the theory upon a reasonably precise notion of rational choice’.67 The basic idea would seem to be that the original position is conceived of as an exercise in the use of instrumental rationality in the choice of just principles. Justice is, in this conception, a rational means to adopt in order to ensure against the failure of our plans.68 For justice, the question of justification is settled by working out a problem of deliberation: we have to ascertain which principles it would be rational to adopt given the contract situation. This connects the theory of justice with the theory of rational choice.69 Rawls’s method does set out to guide this choice by the imposition of a number of reasonable constraints on the choice situation (for example, unanimity and limited knowledge). Here, however, his concern is to put together enough of these constraints so that ‘it will become obvious that one among the alternatives . . . is to be preferred’.70 Rawls places constraints on the rational choice, the effect of which is to make the problem of choice manageable. This manageability provided by the constraints is important; it allows Rawls to claim that the argument for the two principles is a derivation.71 ‘The argument aims eventually to be strictly deductive’.72 Rawls here is ‘striv[ing] for a kind of moral geometry with all the rigor this name connotes’.73 On this reading, justice as fairness is the result of a geometric deduction, using the theory of rational choice, from a small number of axiomatic presuppositions. As such, it might appear that Rawls’s method resembles the foundationalist contractarian where our individual ends are taken as given and principles of justice are identified simply as the outcome of calculation aimed at most efficiently realizing these ends. That this is not a

20

Constructivism and A Theory of Justice

helpful (or accurate) understanding of Rawls, despite his own references to rational choice and moral geometry, should become clear. Reflective equilibrium/ourselves It is important that we maintain a clear view of the nature of this geometric deduction in order not to be misled in our attempt to understand Rawls’s constructivism. We should remind ourselves that the original position must be ‘understood as a purely hypothetical situation’ designed to lead to ‘hypothetical agreements’.74 We do not imagine that the parties are real people nor that their deliberations actually occur. ‘It is clear, then, that the original position is a purely hypothetical situation. Nothing resembling it need ever take place.’75 This, however, casts shadows over Rawls’s deduction of the two principles. Rawls’s argument, as we have examined it so far, claims that the two principles are the principles of justice for a well-ordered society and as such should recommend themselves to us. We should accept them as just principles because some hypothetical people, deliberating hypothetically, reached a hypothetical agreement resulting in a hypothetical contract that designated them as such. But Dworkin points out, ‘A hypothetical contract is not simply a pale form of an actual contract; it is no contract at all.’76 Indeed, Rawls reminds us that ‘the hypothetical nature of the original position invites the question: why should we take any interest in it, moral or otherwise?’77 The point is well made. If the original position and the agreement reached therein are merely hypothetical, why should we feel that they offer any reasons applicable to our actual situations where Rawls’s restrictions on reasoning are not in force and considerations of justice must be intensely practical? Why should our imaginings have any authority for us, especially as we have yet to see why the claim that we imagine things this way rather than another is based on anything other than our contingent whim. Instead of speaking to our practical concerns, is not Rawls’s theory of justice just an irrelevant ‘castle in the air’?78 Rawls answers his own question about the relevance of a hypothetical original position when he tells us ‘the conditions embodied in the description of this situation are ones that we do in fact accept. Or if we do not, then we can be persuaded to do so by philosophical considerations.’79 This answer is that the choice of parties in the original position is relevant to our deliberations about justice because we already, perhaps only implicitly, accept its basis. Rather than regard them as the product of contingent imaginings, Rawls argues that the particular constraints that inform the original position are ‘commonly shared presumptions’ that are both ‘widely shared’ and ‘weak’.80 Because the constraints fulfil these conditions, the conception of justice which results ‘best approximates [and extends] our considered judgements of justice’.81 Indeed, it is crucial to Rawls’s project that he be able to argue that justice as fairness matches our ‘common sense convictions more accurately than its traditional rivals’.82 This is because the final acknowledgement of the two principles as the principles of justice waits on a demonstration that they are in reflective equilibrium with

Constructivism and A Theory of Justice

21

those considered convictions about justice.83 The term ‘reflective equilibrium’ is, in some ways, a new description of a familiar process, and it is of central importance to Rawls’s justification of justice as fairness and to his constructivism in particular. The idea of a search for reflective equilibrium as part of a notion of justification involves modelling what are often very ordinary ways of thinking and reasoning. It is most simply conceived of as an account of the processes of everyday critical reflection used in moral deliberation and argument – an account of our practical reasoning. As we shall see, in this way it involves weighing all available and relevant evidence coupled with an openness to critical reflection and counter-argument as key elements in the construction of objective principles that can guide our decisions and actions. Rawls suggests that we can think of our judgements as in reflective equilibrium when, on due reflection, principles of justice accord with our considered convictions about justice. Rawls is confident in claiming that we already accept the constraints on the original position that he has outlined because they are integral to an account that identifies the principles of justice that are in line with our considered judgements after due reflection. We all, whether we always articulate them explicitly or not, have many convictions concerning justice. We can regard these convictions as considered when we are confident that they are free of avoidable distortions and therefore made in conditions conducive to moral judgement. For example, we can exclude from consideration judgements about which we are hesitant or in which we have little confidence, judgements made under the influence of fear, anger or excitement and also judgements that we make when our powers of judgement are impaired such as when we are tired, drunk or under significant time pressure.84 This notion of weighing more strongly judgements made when we are at our best, when we are calm and unhurried, than we do those made in the heat of the moment or under duress is a familiar and unproblematic attitude. Identifying the set of our considered convictions about justice places us in a situation quite similar to that of the intuitionist who has identified a range of plural intuitions for which he can develop no systematic priority rules. However, whilst identifying these considered convictions, perhaps including judgements such as slavery is wrong, as is torturing for amusement and the unnecessary suffering of children, Rawls is committed to identifying constructive principles of justice that can systematically account for these convictions whilst also drawing out their general implications. This is possible because, whereas the intuitionist identifies a range of self-evident intuitions each with the status of a first principle, Rawls is simply identifying the set of judgements that we are most confident in making. We can then visualize a process of reflection that identifies principles of justice that seem most consistent with this set of convictions. The original position is a stage in this process, laid out by reference to our convictions in order to help us clarify their implications for matters of principle. However, it is unlikely that all of our judgements are mutually consistent; some may even be in conflict with each other making the identification of principles that account for all our convictions impossible. In this situation, the intuitionist, armed with a plurality of first

22

Constructivism and A Theory of Justice

principles, faces the priority dilemma we identified at the start of this chapter. Rawls, dealing simply with our best current judgements, can do what the intuitionist cannot and envisage us making necessary reflective adjustments to those convictions where their conflict prevents the identification of systematic principles of justice that reflect them. If we can identify principles of justice that match our considered convictions about justice after appropriate adjustments have been made, then Rawls can claim that we have reached a ‘narrow’ reflective equilibrium.85 This would establish the set of principles of justice as a ‘philosophical but non-metaphysical reconstruction of everyday moral intuitions’.86 The principles of justice would match our convictions or extend them in consistent and reasonable ways. Rooted in our convictions about justice, Rawls’s starting points are embedded in our everyday methods of thinking rather than based on the identification of metaphysical foundations. This is the basis of a response to influential and important criticisms that Rawls’s liberalism is, like many foundational positions, abstract and otherworldly. Critics such as Walzer have argued that the original position and the veil of ignorance that screens the parties from all knowledge of their particular circumstances demonstrate the way Rawls’s justification is distanced from real people and the situations in which they find themselves. Such critics claim that Rawls proceeds as if he has identified a privileged standpoint, an Archimedean point, from which he can invent the right theory. In basing his argument in the original position, the claim is that Rawls builds from a foundation unconnected with our ordinary and, particularly, everyday lives. In other words, Rawls the philosopher, from his privileged epistemic position, designs the abstract theory of justice for the rest of us. He ‘walk[s] out of the cave, leave[s] the city, climb[s] the mountain’ and from this elevated stance looks down on us and ‘fashion[s] . . . an objective and universal standpoint’. Walzer argues that the political philosopher should instead ‘stand in the cave, in the city, on the ground’ and interpret what he finds there.87 It should be clear now that Rawls is, instead, concerned with providing an interpretation and systematization of our ordinary everyday convictions about justice. The approach embodied in the search for reflective equilibrium between these convictions and our general principles about justice makes plain Rawls’s intention to proceed with his feet firmly planted ‘in the city’ and ‘on the ground’ by constructing a conception of justice that ‘best approximates our considered convictions of justice’.88 Indeed, the reasonable worry at this point may be that Rawls is tying his theory too closely to our current intuitions or convictions. Is a reflective equilibrium inherently conservative, uncritically grounding principles of justice on a foundation of current judgement and practice? If foundational justifications assign a set of beliefs, facts or judgements a special status and use these to underpin further judgements, then some commentators have regarded Rawls as doing just this. It is argued that in advocating reflective equilibrium, Rawls ‘recognizes a sub-class of initial opinions which are to be seen as fixed points around which all else must be accommodated’ and that these fixed points

Constructivism and A Theory of Justice

23

‘perform the same function as basic statements do in scientific enquiry’.89 This interpretation regards Rawls as committed to a ‘fixed point’ or ‘one-way’ understanding of critical reflection similar to that commonly supposed to be used in the natural sciences where deliberative reflection can only lead to a one-way adjustment of principles to initial data. Singer argues that ‘Rawls’s view is that a normative theory is like a scientific theory . . . the aim of the theory is to explain all the data.’90 The claim is that reflective equilibrium is inherently conservative and that in tying justification too closely to our intuitions, the worry must be that principles of justice are ‘nothing more than a reshuffled pack of moral prejudices’.91 If this criticism is well made, then Rawls’s considered convictions apparently form the basis of a foundational account of justification where those convictions play the role of an independent moral order of intuitions to which principles must correspond. There are passages where Rawls appear to lend support to this one-way, fixed point interpretation of reflective equilibrium. He notes that ‘any ethical view is bound to rely on intuition to some degree at many points’ and that our convictions are ‘fixed points which we presume any conception of justice must fit’.92 In a later paper, he likens the moral theorist to ‘an observer’, seeming to confirm the interpretation of reflective equilibrium as closely related to method in the natural sciences.93 However, whilst acknowledging the ‘one-way’ interpretation of reflective equilibrium is possible, Rawls is explicit in denying that this is the only possible understanding. A ‘one-way’ reflective equilibrium is one that brings principles into line with our existing convictions except where adjustments are made to remove discrepancies. In this case, ‘we would be describing a person’s sense of justice more or less as it is although allowing for the smoothing out of certain irregularities’. This is a process that draws out the implications of our current commitments. In opposition to this understanding, Rawls outlines an alternative ‘two-way’ account of reflective equilibrium where the content of our convictions is also regarded as open to critical scrutiny in a reflective process that goes ‘both ways’. Rawls argues that on this understanding, ‘a person’s sense of justice may or may not undergo a radical shift’. He goes on to say that ‘clearly it is this second kind of reflective equilibrium that one is concerned with in moral philosophy’.94 That Rawls rejects a foundational interpretation of reflective equilibrium should be clear. He repeatedly states that there are no ‘already given moral facts which cannot be altered . . . the moral facts are determined by the principles . . . chosen’.95 Elsewhere, he argues that his constructivist justification does not aim at achieving a clear and undistorted view of a prior and independent moral order. Rather (for constructivism), there is no such order, and therefore no facts apart from the procedure of construction as a whole; the facts are identified by the principles that result.96 This argument is not the social constructionist claim that all ‘facts’ are

24

Constructivism and A Theory of Justice

constructed. Nor is it the claim that there are independent moral facts but that our knowledge of them is distorted and imperfect. Rather, it is the claim that we have no special insight into which facts have moral significance. Which facts about the human condition, our social situation or our political lives are morally significant for the identification of principles of justice is something we cannot know independently of our moral deliberations. So, whilst we may be confident in having identified fixed points of our moral judgement regarding religious intolerance or racial discrimination for example, ‘questions which we feel must be answered in a certain way’, Rawls argues that these are only provisionally regarded as fixed.97 To see how this claim impacts on reflective equilibrium, we must return to our discussion of the process of critical reflection. Rawls’s rejection of a fixed point or one-way reflective equilibrium exploits the discrepancies between our considered convictions and the principles that initially best account for them. It is unlikely that there are any principles that would exactly match all of our convictions as these convictions are undoubtedly inconsistent and conflict in places. One-way reflective equilibrium, which regards convictions as fixed points, could adjust these convictions only so far as was necessary to ensure that a coherent set of principles could be found to match them. However, Rawls argues that if we are faced with principles that provide an altogether more convincing account of justice, we may feel that it is better to revise our convictions to reflect this. Even if proposed principles do exactly fit our existing convictions, we may be moved to revise those convictions anyway, as in the case where a principled account of our convictions has implications we have not foreseen and are not willing to accept or where it draws on background theories that are unconvincing. Rawls is envisaging a two-way adjustment process by which convictions and principles are brought into line, where both convictions and principles, and the conception of the original position that mediates between them, are liable to revision.98 The process of critical reflection is therefore much more wide-ranging on this account. Deliberation drills down to our convictions and subjects them to reflective criticism in the light of principles of justice. In this way, we can be assured that the reflective process is able to assess all the relevant evidence and arguments wherever they appear and resultant principles would be in ‘wide reflective equilibrium’.99 Such principles would then be supported by the best reasons currently available to us; if they were not, then we would simply have reason to return to the reflective process until we have identified principles that are. As should be plain, where Rawls regards our considered convictions as fixed points, he does so only provisionally and not to remove them from the scope of reflection or to treat them as a moral touchstone. The greater confidence in some judgments that leads us to think of them as more or less fixed comes from not yet having been faced with good reasons to think again. When Rawls treats a rejection of slavery as a fixed point, it is because he has trouble imagining what sort of reason could make him or us think again about the merits of slavery. However, recognizing that the Greeks had equal confidence in their positive judgement of slavery means that it would be strange to claim that there could

Constructivism and A Theory of Justice

25

never be reason to think again, no matter how unlikely that seems today. In this way, we should recognize that all our considered convictions are provisional. Although in order to get started various judgments are regarded as firm enough to be taken provisionally as fixed points, there are no judgments on any level of generality that are in principle immune to revision . . .. [Indeed] it is possible that first principles should be formulated that seem so compelling that they lead us to revise all previous and subsequent judgments inconsistent with them.100 If judgements at all levels are open to the possibility of revision, then no subset can play a privileged role in justification. There can be no foundational subset of judgements with a special status against which other judgements are measured. Rawls’s justification is neither foundational nor conservative; if any principles or judgements are to be regarded as privileged, this can only be the outcome of a wide-ranging reflective process and cannot be decided at the outset. So, rather than regard our convictions as ‘moral observations’ or facts that principles, like scientific theories, must account for, convictions are regarded only as initial pointers in the direction of an acceptable or satisfactory moral theory. The process of open-ended critical reflection on all levels of judgement that characterizes this account of reflective equilibrium can be more easily understood by analogy with Neurath’s boat.101 This boat is a leaky sort of vessel in need of repair while at sea. Since we are at sea, we can remove some planks for repair at any one time, but we cannot remove all planks at once without sinking and drowning. We can, and must, caulk our leaks and repair our timbers in a gradual manner, holding firm to most of the structure in order to fix one part at a time. The important point is that sooner or later we may need to repair every part of the boat in order to stay afloat. Likewise, while seeking reflective equilibrium may involve revising any or all judgements (even those we were confident enough of to regard as fixed), it does not claim that we do the psychologically impossible and revise all of our convictions and principles all at once. Rather we can stand firm on some and examine and revise others and then, when we now stand on our revised judgements and look back, we see our initial standpoint differently too. The key is that in doing so we ensure that, as far as possible, principles are supported by good reasons. If they are not, if there are criticisms that have yet to be considered, or if we encounter ideas or practices that are novel to us then the reflective process of ‘repairing the ship’ continues. After a lengthy diversion, we can now appreciate the importance of reflective equilibrium in addressing our worries about the hypothetical nature of Rawls’s arguments for the two principles and their derivation from the original position. We need not, and probably cannot, view Rawls’s arguments as exemplars of the foundationalist geometric method, nor the two principles as solely the result of a hypothetical rational choice. This is because Rawls frames the rational choice of principles with conditions that he finds presupposed in ‘common sense . . . generally shared ways of reasoning and plain facts accessible to all’ – conditions that

26

Constructivism and A Theory of Justice

reflect our best understanding of the requirements of an impartial and objective choice of principles.102 In this way, we are the source of Rawls’s starting points, of the conditions that make the original position what it is and that guide the choice therein. It is for this reason that the claim that the two principles would be chosen in the original position is indeed a recommendation to us to accept them where matters of justice are concerned and why we should expect principles chosen in the original position to be in reflective equilibrium. However, with the stress now on the intuitions and convictions about justice that shape the conception of justice as fairness and form the starting point in seeking reflective equilibrium, what is the role of the original position? Instead of seeing it as a rational choice game, the result of which justifies certain principles, we are to regard it as an ‘expository device’ that ‘makes vivid’ the restrictions and conditions that we are willing to accept on the choice of just principles and ‘enables us to envisage . . . [this] objective from afar’.103 The original position plays a heuristic role, helping tease out the consequences of our considered convictions about justice. Indeed, the original position models ‘a certain form of thought and feeling that rational persons can adopt within the world’.104 As such, we can enter the original position at any time ‘simply by following a certain procedure . . . by arguing for principles of justice in accordance with . . . [the relevant] restrictions’.105 Given this heuristic role, the deductive argument from the description of the original position is not itself a primary source of justification for Rawls’s theory of justice’. ‘this application of rational choice theory is embedded within the assumptions defining the original position . . . and their justification depends upon the method of rational deliberation [reflective equilibrium]. It seems fair to conclude that the use of rational choice theory plays only a minor role in the justificatory argument.106 The justificatory weight is taken by the process of reflective equilibrium and not by the original position.107 Summary We are now in a position to make an initial assessment of the character of Rawls’s alternative account of justification. A Theory of Justice is regarded as embodying, perhaps conceiving of, constructivist justification, a judgement that Rawls later confirms.108 It is imperative that we lay out an initial understanding of what this constructivism consists in. As we have seen, justification is a response to pluralism and disagreement. Foundationalist theories respond to pluralism by grounding an objective and universal set of political principles. Such principles would transcend pluralism by providing objective reasons for their acceptance to everybody. In order to ground these principles, their justification relies on foundations, an independent moral order consisting of moral facts that our judgements must match. It is the existence of a moral order, independ-

Constructivism and A Theory of Justice

27

ent of the standpoint, attitude or beliefs of human beings, that underpins the foundationalist claim to objective justification. Sceptics reject these foundational justifications on the grounds that this independent moral order either does not exist or cannot be known. In either case, there are effectively no moral facts to act as foundations for objective and universal justification. As a consequence, these sceptics deny that objective principles are possible. In the justification of justice as fairness, Rawls recognizes that foundational moral facts are probably not available to him. ‘I have not proceeded then as if first principles, or conditions thereon, or definitions either, have special features that permit them a peculiar place in justifying a moral doctrine.’109 He is denying that his theory regards any moral fact, conviction or principle as possessed of independent justification. Justification must proceed without recourse to an independent moral order. But if this is the case, how can Rawlsian justification avoid sliding into scepticism; how can Rawls’s constructivism ground objective principles? Initially, the argument from the original position to principles of justice appears foundational. The original position is laid out by reference to our considered convictions, and the whole is an exercise in moral geometry with these convictions taking the part of the basic axioms foundational of geometric proof. This axiomatic role for our convictions is reinforced if we understand reflective equilibrium to be a process of one-way adjustment of principles to bring them into line with these ‘fixed’ and foundational convictions. Objectivity is thus understood foundationally; principles are objective if their justification can be shown to be the outcome of a successful deduction from accepted foundations. However, it has become clear that Rawls explicitly rejects this account of his theory and, equally explicitly, adopts an account of reflective equilibrium that highlights its two-way nature, where convictions as much as principles can undergo radical revisions. He rejects the idea of any ‘fixed point’, on which to anchor justification, arguing that critical reflection goes all the way down. Justified principles are those that we can endorse on due reflection as in reflective equilibrium with our (possibly revised) convictions about justice. This raises questions about the objectivity of resulting principles. The objectivity of these principles cannot be simply that they can be deduced from appropriate foundations or can be read off from a universal account of primary goods. Instead, it is tied up with the sort of reasons that can be taken to support principles once you conceive of the original position and reflective equilibrium in the right way. Rather than being an authoritative rational choice problem, the original position helps us to model the back and forth of debate where parties are required to offer reasons for principles within appropriate constraints of impartiality and objectivity. The parties are offering or proposing to each other reasons to choose certain principles. The original position, with the veil of ignorance and other constraints, is concerned to identify and reject obviously subjective reasons for proposing principles (that they favour people with certain physical or mental capacities, that they are biased in favour of a particular economic class, etc.). The reason that certain principles can be the object of a general choice is that they do not rest on such subjective or partial reasons. They

28

Constructivism and A Theory of Justice

are instead based on objective reasons, with objective taken to mean ‘unbiased’, ‘impartial’ or ‘count as a reason for all’, and are therefore universal. The original position is intended to rule out factors ‘arbitrary from a moral point of view’ whilst the two principles ‘express the result of leaving aside those aspects of the social world that seem arbitrary from a moral point of view’.110 Rawls uses the original position as a model of the appropriate social point of view that rules out the arbitrary, partial and subjective and identifies the universal and objective. This reflects our usual conception of objectivity. After all, our ordinary everyday moral thinking does not use objective to mean ‘from the point of view of the universe’ or from a ‘God’s eye view’ but rather to mean exactly impartial, unbiased and applying to everyone (think of the objectivity of judges or referees).111 Our confidence that the original position can play this role is dependent on our understanding that the conditions that shape it reflect our considered convictions. We can think of our considered convictions as those in which we are currently most confident of their objectivity. If we regarded them as obviously wrong or as merely selfish or biased, then we would not find them convincing. This does not mean that they directly express moral facts or principles but rather that they are those we currently think are best justified and could therefore underpin reasons for everyone. For example, on our current best understanding of those to whom we think justification is owed (all people), we are confident that no one could reasonably reject our rejection of slavery. In this way, our considered convictions can be regarded as the, always provisional, outcome of the sort of basic universalizability test that reasonable rejection consists of. Where in the past people have been confident that slavery is justifiable, we can regard this as based on a different and partial understanding of the appropriate audience for our reasons (the objections of natural slaves do not count as they are not fully human and therefore are not part of the scope of justification in the same way). The ‘special status’ that our convictions have is simply that they express our best current understanding of what would pass/fail the basic test of providing reasons to everyone. If it is apparent that our convictions do not pass this test, that we have good reason in the face of evidence and argument to find our convictions less convincing, then they are liable to revision in the process of critical reflection. So the considered convictions underlying the original position may be ‘privileged’ insofar as we confidently regard them as expressing objective reasons, but this privilege can never be more than provisional and open to revision. They are ‘moral’ facts only in this provisional sense, and their affirmation and continuing status as ‘moral’ facts would depend on our continuing confidence in them in the face of challenging reasons thrown up by the spotlight of reflective equilibrium. In this way, Rawls can regard the facts that are to count as moral facts as picked out by the continuing reflective process, not as an independent set of data points that his theory has to draw the right line through in order to connect them all up. As we noted above, when we understand justified principles as those that are in reflective equilibrium, then it is the concept of such an equilibrium, rather than the original position, that bears the weight of justification. This has led

Constructivism and A Theory of Justice

29

some commentators to question the place of the original position in Rawls’s argument.112 However, taking Rawls at his word that the original position is an expository device that makes vivid our intuitions and recognizing the central role he gives the original position, we can see the original position and reflective equilibrium as tied up in the same constructivist notion of objectivity. The original position is a particularly useful way of understanding the implications of our convictions for the choice of principles of justice. Unavoidably, having to start from our current best understandings, any other starting point would be arbitrary for us, we outline the original position as our best conception of a moral point of view. This is the point of view that we think best rules out partial, arbitrary and subjective reasons and rules in only impartial and unbiased reasons for everyone. From this best understanding of the moral point of view, we can regard the parties as choosing universal and objective principles of justice. That the parties in this moral point of view would unanimously choose a set of principles is a strong indication of their objectivity. If they could not agree on a set of principles, this would give us good reason to suspect the principles of a hitherto unnoticed partiality. Reflective equilibrium, the process of ongoing critical reflection, is simply the process of continually reassuring ourselves that our best current understandings are as objective and appropriate as we thought. One way of finding out that they are not is if the original position laid out on their basis issues in principles which we suspect are partial and biased, perhaps on the basis of the objections of others. In these cases, principles could reasonably be rejected as not objective and therefore in need of revision. We would then have good reason to think again about (and perhaps revise) our convictions and the original position that can be laid out by reference to them. In this way, we can understand the original position as tied into the process of reflective equilibrium whilst not denying that reflective equilibrium has the dominant role. There remains a further residual question over the objectivity of Rawls’s account that draws on the problems posed by pluralism. We must ask whether there can be multiple reflective equilibria; will different principles be in reflective equilibrium for different people or societies? If it is the case that there are conflicting sets of principles in reflective equilibria, how can we claim that any one of these sets, or perhaps all of these sets, is objective? Since reflective equilibrium draws on our convictions as starting points, is it not necessarily relativist? It is obviously the case that there can be multiple narrow reflective equilibria. Since a narrow reflective equilibrium is one between a person’s convictions and the principles that best express them, the convictions of different people may very well be best expressed in competing principles.113 However, Rawls’s justification depends on the notion of wide rather than narrow reflective equilibrium, and it is far from obvious that there can be multiple wide reflective equilibria. This criticism of relativism might seem to be at its strongest when we consider equilibria from different societies or cultures. Would not the principles in wide reflective equilibrium with the convictions of people from one society be expected to be different from those similarly supported in another society? This

30

Constructivism and A Theory of Justice

criticism is simply a reformulated version of the claim that the search for reflective equilibrium is necessarily conservative. The answer to this criticism is therefore the same one as given before. When two societies encounter each other their practices, institutions and ideas pose mutual questions of justification. Justified principles, and the convictions they draw on, are revised not simply in the light of internal inconsistency but also in response to external encounter and critique. We are given good reason to look again at the justification of our principles and convictions by encounters with novel ideas and cultures; new and conflicting ideas effectively destabilize our current equilibrium. We cannot regard ourselves as having achieved wide reflective equilibrium until we have re-examined whether our principles are justified given the new considerations raised by familiarity with the ideas of that novel culture. If we are aware of conflicting considerations but ignore them then we will have failed to take into account all available and relevant evidence and argument in our moral enquiry. Once we recognize this then we cannot accept that even our communal traditions are automatically authoritative as they too are as open to the process of revision as everything else. It is for this reason that Nielsen claims that reflective equilibrium ‘is not in any plausible sense relativist’.114 He goes on to argue that there cannot be distinct wide reflective equilibria or rather that one will always be the widest and more adequate equilibrium. Where different cultures, worlds or ideas have contact with each other then one equilibrium will give the best account of the new total set of considerations. ‘At least where the worlds are in any way in contact there can only be one WRE [wide reflective equilibrium].’115 While Rawls recognizes that ‘it is doubtful whether one can ever reach this state’, we can regard it as an ideal. By continually subjecting our principles to the objectivity tests of the original position and of seeking reflective equilibrium, we can ensure, as far as possible, that our principles are less and less subjective and therefore get ‘closer to the philosophical ideal’ of wide reflective equilibrium.116 Summing up Rawls’s account of ‘nonfoundationalist’ justification, it is clear that he explicitly aims to avoid identifying an independent moral order on which to ground principles.117 He does, however, uphold the possibility of objective justification through a process of ongoing critical reflection aimed at achieving reflective equilibrium. So, teasing out an initial general understanding of constructivism, it, at least as far as Rawls is concerned, involves the rejection of any appeal to moral facts coupled with a conception of objectivity dependent on an account of critical reasoning. As we examine other plausible constructivist positions, we will be in a position to modify or reject this basic understanding of constructivism, if necessary. However, with suitable refinements, I believe that this characterization will accurately capture central aspects of constructivist justification. Rawls’s specific construction of political principles, justice as fairness, is the culmination of a search for determinate principles that will be accepted by all citizens as the basis of their liberal democratic regime. The two principles of justice as fairness are chosen in an original position laid out by reference to our

Constructivism and A Theory of Justice

31

considered convictions. Their objectivity and universal application are ensured by the constraints on the original position and the notion of primary goods that it is rational for anyone to want no matter what their conception of the good. The veil of ignorance over the original position can then be progressively lifted so that we can determine the proper application of these principles to the production of a constitution, then to legislation and then to executive and judicial decision-making.118 At this point, Rawls claims that we have fully described and justified a just basic structure for society, and the political construction is complete.

Some problems in the construction of justice as fairness Outlining the basics of an exciting new constructivist understanding of justification is a great achievement of A Theory of Justice. As we have noted, Rawls uses this powerful understanding to develop an account of his own specific theory of justice as fairness. The following sections focus less on the general account of justification than on possible problems with his specific development of a constructivist justification of the two principles. It is important to explore the difficulties that Rawls’s specific construction faces, not only for our understanding of justice as fairness but also because this will help us to appreciate the possibilities of constructivist justification more generally. In particular, a better understanding of Rawls’s project will enable us to refine our account of constructivism and to ensure that we better appreciate its limitations. Intuitions and the original position The relationship between the original position and the intuitions that inform it is a difficult one, and a very important one for any discussion of Rawls’s constructivism. So far we have addressed this relationship in the abstract, but we also need to examine Rawls’s particular account and concerns it raises about the way convictions are utilized in the laying out of the original position. First, however, we must highlight a more formal concern. Attention is drawn to a potential problem by Seung who finds in Rawls an implicit distinction between monistic and pluralistic understandings of the original position as a scheme for the comparison of competing conceptions of justice.119 Drawing on very real ambiguities in the text, Seung claims that Rawls equivocates between these two understandings that differently characterize the relationship between the original position, our intuitions and the conception of justice as fairness. If interpreted as a monistic scheme, the original position is regarded as the standpoint from which the various competing conceptions of justice are evaluated. Here, the parties are seen as choosing between the two principles and, to use Rawls’s examples, utilitarianism, mixed conceptions, intuitionism and egoism.120 The parties in the original position deliberate between these competing conceptions of justice and choose the most just principles. Support for this interpretation is easily found. Rawls explicitly says the parties are ‘given a short list of traditional conceptions

32 Constructivism and A Theory of Justice of justice. . . . I try to show that the two principles would be chosen from the list.’121 The original position has usually been understood in this monistic way. Seung’s alternative interpretation, the pluralistic scheme, involves recognizing that there are as many different interpretations of the initial choice situation as there are conceptions of justice. Each conception of justice is in a ‘one-to-one correspondence’ with a specified choice situation. So, whilst the two principles are the result of a choice in the original position, utilitarianism is the choice of an ‘impartial observer’, perfectionism can be envisaged as the result of choice from a perfectionist original position and so on. In this vein, Rawls writes, ‘There are . . . many possible interpretations of the initial situation . . .. In this sense, there are many different contract theories. Justice as fairness is but one of these.’ He goes on to make the point even clearer. ‘We may conjecture that for each traditional conception of justice there exists an interpretation of the initial situation in which its principles are the preferred solution.’122 The original position mediates between our intuitions and the principles of justice. In the monistic scheme, intuitions are interpreted to produce only one choice situation, the original position. This scheme would seem to be implied by Rawls’s many clear statements that the original position as laid out by reference to our considered convictions, convictions that are widely recognized and uncontested. But if Rawls is advancing the monistic scheme of comparison as the legitimate standpoint from which to judge competing conceptions of justice, what exactly is he claiming for the original position? In this, monistic scheme is the claim to have identified all the intuitions that are relevant to the discussion, that is, all of our intuitions/convictions about justice (or at least all of the important ones). Surely this is implausible. Rawls proceeds by calling on our convictions and intuitions as and when it becomes necessary, and this would have to be supplemented by some sort of argument for their comprehensiveness or exclusive special relevance. Nowhere does he make this argument, and it would be difficult to see how the argument would go if he did. We will briefly examine later in this section the controversial nature of the intuitions Rawls does draw on, but we must also be aware of intuitions he sets aside. Prominent amongst these would be a utilitarian intuition that a society that satisfies more preferences of more of its people is better than one that does not have this concern. Also, a principle of desert is ruled out, but as Tom Campbell points out, we surely have a place in our intuitions for the notion that those who do praiseworthy and socially beneficial work should be rewarded as a consequence.123 The list could be extended to include intuitions behind each of the traditional conceptions Rawls mentions and further. The point here is not just that there are all sorts of intuitions that fall by the wayside in the deliberations of the parties and during the search for reflective equilibrium. Instead, we must recognize that these would seem to be legitimate intuitions about justice that do not get taken up into the original position, although there is nothing to suggest that they are not as equally widely shared and accepted as those Rawls does use. As such, the conceptions of justice that they favour are ruled out by the very formulation of the original position and not

Constructivism and A Theory of Justice

33

by the reasoning of the parties within it.124 Since these would seem to be genuine widely shared intuitions about justice that do not find their expression in the mapping of the original position, we might expect them to come into play when we check to see if the outcome of the choice in the original position is in reflective equilibrium with our considered convictions about justice. Rawls, however, in his sparse comments on this process does not seem to consider that they may impinge even here.125 The existence of these alternative intuitions about justice encourages us to consider the pluralistic scheme, in which Rawls implicitly admits their existence. Here, these intuitions would find their expression in different choice situations that would result in different conceptions of justice. Rawls does talk of this pluralistic scheme as a general analytic method for the comparative study of conceptions of justice. One tries to set out the different conditions embodied in the contractual situation in which their principles would be chosen. In this way one formulates the various underlying assumptions on which these conceptions seem to depend.126 If, however, each conception of justice is characterized as the outcome of a different form of choice, and choice situations mediate between conceptions of justice and intuitions, then each conception of justice is also characterized as related to a specific intuitional set. Instead of treating our convictions as a single set, each conception would appear to draw on a different, but perhaps overlapping, set. If this is the case and if each conception of justice is related intimately to a distinctive set of intuitions, what is the status of the original position? On face value, it looks very different to the account developed in the first part of this chapter. Here, it cannot be seen as the neutral, impartial or independent ground upon which to compare conceptions of justice or their characterizations of our intuitions about justice and so make an objective choice between them. Instead, it can only be a device that serves to remind us, in our deliberations about justice, of intuitions and considerations we might otherwise pass over. The original position ‘makes vivid’ one set of intuitions in order to facilitate their comparison with the intuitions underlying different principles. On this pluralistic account, deliberation about the merits of different theories of justice cannot be characterized as, in any way, a deductive exercise in moral geometry. Instead, deliberation and argument would have to be a largely intuitive matter whereby we evaluate, on the basis of the attractiveness of the set of intuitions they draw upon, the relative appeal of the alternative conceptions of justice. Is Rawls’s understanding of the relationship between convictions, the original position and conceptions of justice best characterized as monistic or pluralistic? The text of A Theory of Justice does not really force us one way or the other between these two very different conceptions of the relationship between intuitions and conceptions of justice. As we have seen, there is textual evidence for both interpretations, although it must be said that the general tone of the book

34

Constructivism and A Theory of Justice

leans towards the monistic scheme. This general monistic tone gains support if we note that Rawls occasionally writes as if he can combine the two schemes. He acknowledges the many interpretations of the initial situation but thinks that ‘if one interpretation is philosophically most favoured, and if its principles characterize our considered judgements’, then we can legitimately get from the initial pluralistic scheme to a monistic one.127 If any original position fulfilled these two conditions, then this would elevate the pluralistic scheme from a general tool of analysis to part of a procedure for justification. Does one of the plurality of interpretations of the initial situation satisfy these conditions so as to constitute a justification of the conception of justice that is associated with it? First, it would have to be ‘philosophically most favoured’, and second, it would have to ‘characterize our considered judgements’ (be in reflective equilibrium). These conditions, however, collapse into one. What could philosophically most favoured mean in the context of Rawls’s work other than ‘a reasonable approximation to and extension of our considered judgments’ or ‘in reflective equilibrium’.128 This is bringing us back closer to the understanding of constructivist justification previously developed, where the sole condition to be satisfied by an interpretation of the initial situation is then that of being in reflective equilibrium, and this is Rawls’s claim for the original position. This, however, is an argument Rawls never closes; he basically asserts its success. We will briefly highlight below some of the more controversial of Rawls’s specific intuitions, but a quick example would be appropriate. Consider the question of desert. Rawls does address this question in a piecemeal fashion.129 In doing so, he draws our attention to one intuition, that desert is not a major concern of justice as we do not deserve our native endowments or our initial starting place in society, and it is this that determines distribution. We saw above that Campbell drew our attention to a second of our intuitions about desert that does regard it as a legitimate concern for justice. Here, our intuitions clash and Rawls does not close the argument for the greater importance of his intuition, although he does chart some of its implications.130 Instead, as mentioned above, his intuition finds expression as a condition on the original position (through the veil of ignorance), where it is used to rule Campbell’s concerns out of order.131 Surely, if we bear in mind the pluralistic scheme of comparison, Campbell can just return the favour and use his interpretation of the initial situation to rule Rawls out of order. The gist of Seung’s point is that utilitarianism and perfectionism can do the same resulting in a justificatory impasse.132 On the distinction between monistic and pluralistic schemes, there is no short cut for Rawls, no substitute for justification through the process of reflective equilibrium. He cannot attempt to justify justice as fairness as philosophically most favoured by any other route. This means that at points where it can be demonstrated that Rawls’s intuitions clash with other widely accepted intuitions, the area of disagreement must itself be thematized within the critical deliberation that characterizes the reflective procedure. The response to controversy should be recourse to the procedure of seeking reflective equilibrium. The pluralist scheme can then be interpreted as part of the evolving reflective process of

Constructivism and A Theory of Justice

35

examining and ruling out partial and subjective reasons until we hit on the account that is in reflective equilibrium. Rawls’s particular original position is part of the monistic account that Rawls regards as the outcome of this reflective process. However, this reflective equilibrium may not settle where he wants. This would depend entirely upon the general acceptability of Rawls’s account of our convictions, the principles that result and the partiality of alternative accounts. Unless this is the case, some of Rawls’s own intuitions may lose out in the reflective process. If this is so, then the legitimate conditions on the original position would be altered and so too therefore would the resultant conception of justice. Rawls’s two principles may be the only ones in reflective equilibrium with our considered convictions about justice, but they may not, and then so much the worse for Rawls. The procedure of seeking reflective equilibrium is potentially a highly corrosive process, one from which Rawls’s principles may not issue unscathed. Bearing this in mind, this section concludes by briefly laying out several of Rawls’s key intuitions, over and above those already mentioned, which are perhaps more controversial than Rawls imagines but which still inform his conception of the original position and justice as fairness. These include the ‘common sense convictions’ that give priority to liberty over aggregate welfare, that welfare considerations should not override individual inviolability and that justice is not subject to bargaining or the calculus of social interests.133 There is no disputing that such convictions are a part of common sense, but common sense is notoriously contradictory. These convictions can be matched not only by utilitarian ones that stress aggregate welfare, by communitarian and multiculturalist convictions that stress the community over the individual, but also by intuitions central to some theories in the social contract tradition of which Rawls sees himself as a modern exponent. Gauthier, for example, would question the conviction that justice is not subject to bargaining; indeed, he might argue that it is founded on such a bargain.134 Finally, Rawls takes as provisional fixed points of our intuitions the injustice of religious intolerance and the justice of equal liberty of conscience.135 Whilst we might be loath to disagree with Rawls on these points, in considering their generality, we should note their apparent peculiar suitedness for (and perhaps rootedness in) western liberal democracies. These intuitions may be more or less intimately tied to such regimes, and we must recognize that they may be by no means universal and uncontroversial. If Rawls has underestimated the potential multiplicity of intuitions, this is puzzling as he identifies the need for a ‘constructive’ theory precisely because of the way in which intuitionist theories identified an irreducible plurality of intuitions. Rawls’s constructivism is conceived of precisely as an attempt to impose a conception of priority and order on our intuitive judgements about justice. When theorizing the relationship of intuitions to the original position, we would do well to bear in mind these potential controversies. Can Rawls convincingly claim his interpretation of the initial situation is ‘philosophically most favoured’ in its matching our considered convictions in reflective equilibrium? Instead, we may be warranted in taking his injunction that we ‘think of the

36

Constructivism and A Theory of Justice

interpretation of the original position . . . as the result of such a hypothetical course of reflection’ with a pinch of salt.136 The twin considerations, first that Rawls’s intuitions are controversial and second that he seems to pass over legitimate and widely shared intuitions about justice, must surely imperil at least some aspects of his characterization of the initial conditions of choice and the specific reflective process by which they are identified. Reasoning in the original position This section will highlight just some of the controversies regarding the methods of reasoning Rawls envisages the parties utilizing in the original position. It will raise a small, and obviously not exhaustive, range of problems centring on the notion of primary goods and the related Aristotelian principle. That primary goods are very important for Rawls is illustrated by his claim that the arguments both for the two principles and for a definite content to those principles ‘uses only the thin theory of the good and its list of primary goods’, both of which are ‘necessary to support the requisite premises from which the principles are derived’.137 Primary goods are defined as those goods that are generally necessary for the satisfaction of any rational plan of life. Because they are general goods, they are ‘human goods’.138 It is this generality that Rawls utilizes to bolster the argument for the objectivity of his principles; everybody needs these goods and so assuming them is not partial or biased. Rawls claims that his list of primary goods can be arrived at by assuming only that men have a conception of the good that translates as a plan of life. It is important that this thin theory of the good assumes only a conception and not any particular conception so that the list of primary goods that follow ‘imposes no prior constraint on these conceptions’ and can ‘cover all persons with rational plans of life, whatever their content . . . [and so] puts no prior limitations on what men may desire’.139 The important point is ‘not to tie . . . [the theory] to a particular pattern of human interests’.140 This characterization of primary goods has however drawn criticism. Sandel notes the objection that Rawls’s characterization is ‘biased’ and would ‘contest Rawls’s claim that the list of primary goods really is equally or nearly equally valuable to all ways of life’. He stresses that Rawls’s theory ‘introduces assumptions that are not universally shared, that it is implicated too deeply in the contingent preferences of, say, Western liberal bourgeois life plans’.141 Brown concurs saying, ‘Rawls thinks his primary goods are neutral amongst plans of life but plainly this is not so.’142 He goes on to mention the Christian and the ascetic plans of life, both of which are rational but both of which may be compromised by the difference principle. There may be many people whose needs as they understand them, like those of a certain sort of Christian, await satisfaction in the afterlife. Finally, some plans of life are inherently communal. For such community-based conceptions of the good life, primary goods like liberty and freedom of opportunity may be inherently corrosive of the central core of their identity.143 Rawls’s characterization of primary goods is criticized as potentially

Constructivism and A Theory of Justice

37

stifling to plans of life that do not conform to liberal individualist expectations or are not based on a liberal conception of the person. Indeed, even as sympathetic a critic as Baynes is led to the view that it is ‘clear from his specific list of primary goods that they rely heavily upon the tradition and institutions of liberal democratic societies’.144 In this way, the original position does impose constraints on legitimate conceptions of the good, ‘for it does indeed contain some notion . . . of preferred lifestyles’.145 The notion of primary goods does appear to be tied to a particular pattern of interests and preferences not attributable to all people. This is further highlighted if we return to examine one example of a basic principle of motivation that Rawls stresses feeds into the derivation of primary goods, the Aristotelian principle. This principle states, ‘human beings enjoy the exercise of their realized capacities (their innate or trained abilities) and this enjoyment increases the more the capacity is realized, or the greater its complexity.’146 Campbell draws the obvious parallel with Mill’s conception of higher and lower pleasures.147 Compare the statement of the Aristotelian principle above, which Rawls calls a ‘natural fact’, with Mill’s statement ‘Human beings have faculties more elevated than the animal appetites, and once made conscious of them, do not regard anything as happiness which does not include their gratification.’148 Now it is an unquestionable fact that those who are equally acquainted with, and equally capable of appreciating and enjoying, both [higher and lower pleasures], do give a marked preference to the manner or existence that employs their higher faculties.149 Mill’s notion of higher and lower pleasures seems fairly close to Rawls’s Aristotelian principle. This is remarkable given Rawls’s express wish to stay clear of controversial or partial ideas of the good or of a partisan interpretation of plans of life. If we take seriously Mill’s claim that it is ‘better to be Socrates dissatisfied than a fool satisfied’, then this sort of reasoning is going to be obviously contentious.150 Mill is quite clearly convinced that such a principle grades or judges plans of life and Rawls’s view seems implicitly to do likewise, to normatively privilege the complex over the simple. The charitable alternative is to interpret Rawls’s principle instead as an attempt to describe the preferences that people do exhibit, and even this is fraught with potential controversy.151 It is not immediately clear that Rawls’s description of human activity is entirely accurate. I will raise just one alternative description and hope that this is objection enough to at least cast doubt on the Aristotelian principle. I submit that it may be the experience of many people that we most exhaustively exercise our capacities at work, which we often do not for pleasure or enjoyment but in order to facilitate our leisure time. We then fill this leisure time with simple pleasures. Leisure often becomes a chance to rest, unwind and recuperate. Most of us enjoy a lazy Sunday doing nothing more complex than watching television, listening to music, spending time on the beach with our families or reading a book that is not too taxing. We do not find this dissatisfying, and it can be when our lives are at

38

Constructivism and A Theory of Justice

their most pleasurable. Indeed, we have names for people who are incapable of enjoying the simple pleasures of rest and relaxation, names such as ‘workaholic’ that mark such people as unusual or out of the ordinary. It is not that the Aristotelian principle does not describe a pattern of human interests; it is just that it is not obvious that this pattern is general and, as such, the Aristotelian principle must be regarded as a definite point of potential controversy for Rawls’s theory of goods. This potential lack of generality in a central aspect of Rawls’s determinate construction may severely weaken his justification of justice as fairness. However, despite Rawls’s claims to the contrary, this controversial principle is probably dispensable in the derivation of a list of primary goods. He implicitly acknowledges this in his discussion of a mathematician whose only pleasure is counting blades of grass on a variety of geometrically shaped lawns. Rawls affirms that this may constitute a rational lifeplan that obviously requires the primary goods but ‘does not require the truth of the Aristotelian Principle’.152 The partiality of the Aristotelian principle need not fatally compromise the theory of goods. However, since we have already expressed doubts that the list of primary goods can remain neutral between conceptions of the good, this is not much of a saving grace.153 Indeed, as we shall see below, the account of motivation built into the Aristotelian principle lends support to the claim that Rawls’s constructivism is in fact based on a specific and ‘foundational’ conception of the person. Determinacy problems A significant driving force behind justice as fairness is the problem of pluralism. Whilst not underplaying the opportunities for mutually advantageous enterprises that pluralism proffers, Rawls notes that it is just as often a source of conflict.154 Despite the complementarity of interests that makes possible the production of a cooperative surplus (material and intellectual), there is an accompanying conflict of interests over the distribution of this surplus and on the way in which cooperation is organized. To put the central problem bluntly, ‘Men disagree about which principles should define the basic terms of their association.’155 Indeed, it is the existence of this plurality of points of view and conceptions of the good, each in potential conflict with each of the others, which makes the concepts of justice and justification necessary. That pluralism, and the conflict that it often engenders, is a central motivation for justice is demonstrated by Rawls’s account of the circumstances of justice, the circumstances in which the concept of justice is applicable. The objective conditions of justice, coexistence, rough physical equality and moderate scarcity are the impersonal factors that lead people to conflict. These are matched by the subjective conditions of justice; that each has his own plan of life and the conflicting claims on available natural and social resources that these different plans entail.156 The objective and subjective conditions of justice conspire to ensure that pluralism in conceptions of justice and the good results in conflicts over resources and over the organizational principles for the basic structure that determine the success or failure of the plans of life embodied in each conception.

Constructivism and A Theory of Justice

39

In the face of the conflicting plurality of conceptions of justice, Rawls’s reaction seems to be the search for a single justifiable conception of political justice, a conception that can navigate its way through such conflict to produce a determinate account of the reasonable and acceptable institutional structures. Rawls’s method aims at determinacy in the solution of the problem posed by pluralism. We can see this in Rawls’s continual concern to identify the two principles as the principles that would be chosen by the parties.157 If we also bear in mind the four-stage sequence of decision-making procedures that, by progressively lifting parts of the veil of ignorance, gets us from the muddle of conflicting intuition all the way to the application and administration of justice by judges and bureaucrats to particular cases, then it is plain that Rawls intends his conception of justice to be a fairly determinate remedy to the conflicts of pluralism. Rawls confronts conflict at the level of first principles and believes his method limits that conflict to largely the technical problems raised by the administration of a redistributive constitutional democracy. The intended deduction of the two principles corrects the arbitrariness of the world.158 This drive for a determinate argument with which to face pluralism that manifests itself in the language of moral geometry highlighted in earlier sections is tied closely to the structure of Rawls’s theory, most particularly to those features that we have identified as problematic. In the choice between monistic and pluralistic schemes of comparison for conceptions of justice, Rawls’s search for a determinate solution to the problem of pluralism would naturally lead him to favour the monistic scheme. The possibility of developing a single standpoint from which the advantages and disadvantages of each conception of justice could be weighed is too good to pass over. However, in order to produce the coherent description of the original position that such a project requires, Rawls glosses over the limited nature of the set of convictions that informs the structure of his choice situation. We noted above that the monistic scheme involves an implicit claim to have comprehensively identified and considered the full set of our intuitions about justice. We also noted that Rawls had failed to produce an argument or demonstration to this effect. Indeed, the point at which we would legitimately expect the intuitions about justice that are not taken up into the original position to play a part, in an account of how a specific description of the original position actually develops through critical reflection, is marked only by the claim that we should think of the original position as already the product of such a hypothetical course of reflection. Competing intuitions, intuitions of which Rawls implicitly acknowledges the existence in those passages when he entertains a pluralist scheme, again do not get explicit consideration. In his concern to identify an account of the original position that produces determinate principles of justice, Rawls prematurely closes the account of reflective criticism on convictions that a commitment to reflective equilibrium would seem to imply. This pattern is mirrored closely in some of the specific features of the original position. It is the problem of conflicting claims to resources based on a plurality of plans of life that leads Rawls to regard the adoption of the veil of ignorance as

40

Constructivism and A Theory of Justice

a necessary constraint on the original position. It is necessary because of Rawls’s focus on reaching a determinate outcome. The restrictions on particular information in the original position are, then, of fundamental importance. Without them we would not be able to work out any definite theory of justice at all. . . . The veil of ignorance makes possible a unanimous choice of a particular conception of justice. Without these limitations of knowledge . . . [e]ven if theoretically a solution were to exist, we would not . . . be able to determine it.159 However, the veil alone is not enough to bring deliberation to a determinate conclusion. It screens out all particular information in order to prevent the tailoring of principles to personal circumstances. In doing so, however, it deprives the parties of any considerations upon which to base their deliberation. Without knowledge of their aims, needs and capacities, how can they sensibly reason towards principles that advance their interests? So, whilst the veil of ignorance was introduced as a solution to the problem of indeterminacy caused by pluralism, it introduces an indeterminacy of its own. The indeterminacy introduced by the veil is counteracted by the introduction of Rawls’s conception of primary goods that greatly facilitates the argument that a determinate set of principles would be chosen in the original position, but at what cost? As we have seen already, this notion of primary goods is potentially partial. It may be embedded in a vision of liberal individualism that is by no means universally shared and is explicitly rejected by many of the plurality of conceptions of justice and the good. For some of the major competing conceptions of justice, Rawls’s list of the primary goods, and for some the very idea that there could be an identifiable list, is in dispute.160 For Rawls, for whom justification ‘always proceeds from what all parties to the discussion hold in common’, embracing such obvious controversy is an incongruous move.161 Elements of Rawls’s specific construction, introduced in order to identify determinate principles of justice, lead to problems of potential partiality. Their function is to collapse the indeterminacy of choice into a manageable choice situation in which the possibility of a rational choice of a determinate conception is central. However, it is these elements, introduced to limit indeterminacy, which we have seen to be amongst the most controversial aspects of Rawls’s theory. If we remind ourselves that justification through the search for reflective equilibrium is fundamentally concerned with objectivity, the avoidance of partiality and subjectivity in its judgements, then the search for determinacy gives Rawls problems. It is those aspects of Rawls’s specific construction that are introduced in the pursuit of determinacy that seem most partial; it is therefore those aspects that endanger its success.

Constructivism and A Theory of Justice

41

Does Rawls’s constructivism assume a ‘foundational’ conception of the person? It is worth briefly considering the partiality of these elements of the construction and trying to understand why they are controversial. As we shall see, the most straightforward explanation of this partiality not only endangers the specific justification of justice as fairness but also casts doubt on the properly constructivist nature of Rawls’s more general account of constructivist justification. The account of primary goods appears to propose a list of basic human needs necessary for any sort of human life. We have already noted that this list seems to reflect a western, liberal understanding of ‘human life’. These appear to be goods peculiarly suited to the needs of the autonomous individuals of liberal democratic theory. Theories and cultures that do not accept this account of persons would very possibly produce a different list of basic human needs. They are also likely to disagree with aspects of Rawls’s account of human motivation, perhaps with the claim that people are, or ought to be, motivated by the Aristotelian principle to develop and exercise individual talents for personal benefit and to engage in complex and reflective activities. On this view, ways of life that advocate simple activity and accepting your station in life with all the duties it entails have no place at all, ‘human beings will find . . . [this] culture and form of life dull and empty’.162 Rawls seems to take for granted a ‘progressive’ or ‘developmental’ conception of individuals and society that ties in well with liberal democracies characterized by high levels of social mobility and the absence of either traditional hierarchy or rigidly defined social roles. We have noted already that the basic convictions that Rawls draws on to lay out the original position and the veil of ignorance, the separateness of persons or the priority of liberty over communal welfare, for example, seem to embody particularly ‘liberal’ values. That justice as fairness is built out of a prior commitment to a specifically liberal conception of the self is clear to critics such as Sandel. He points to the veil of ignorance in particular to highlight how Rawls assumes the priority of the subject to its ends. The subject chooses and rejects its ends rather than discovers them in an understanding of their relationship to a broader constitutive community. Coupled with the motivational assumption that the parties in the original position are mutually disinterested, this reinforces Sandel’s further claim that Rawls also prioritizes individual goods over social or communal goods. Each of these priority relations is explained by the basic and prior assumption of a controversial account of the ‘deontological self’ embodying a conception of liberal autonomy. This liberal conception of the person contradicts and rules out alternative intersubjective, social or communal accounts of the self and its relationship to its ends and the broader community. For our purposes, Sandel’s key claim is that ‘implicit in Rawls’s theory of justice is a conception of the moral subject. . . . Rawls’ main discussion tends to take the nature of the moral subject as given and argue through the original position to the principles of justice.’163 This criticism is vitally important. If Rawls does indeed take a partial conception of the

42

Constructivism and A Theory of Justice

person ‘as given’ and then build from there to principles of justice, then it appears as if the conception of the person plays a foundational role in his justification. If Rawls’s theory contains a foundational ‘given’, then the claim that there is no independent moral order that constructivism has to fit may be undermined. Showing that a principle did not match the liberal conception of the person would be reason enough to reject the principle. It is difficult to see how, on this understanding of the conception of the person as given, you could be in a position to reject the conception of the person at all. This would undermine that conception’s provisional status and its openness to rational revision and, in turn, undermine the claim of constructivism to be distinct from foundational accounts of justification. Brink concurs, claiming that ‘Rawls’s argument for constructivism requires him to assign to ideals of the person an evidential role’ that is inconsistent with his general constructivist theory of justification.164 If Rawls’s constructivism is linked to a particular conception of the person, we will need to be clear about this. Importantly, in later work, Rawls goes on to embrace the link between his constructivism and the liberal democratic conception of the person, going so far as to agree that justice as fairness had implicitly assumed this all along. He argues that justice as fairness is a form of Kantian constructivism and that this is distinguished by the way it, specifies a particular conception of the person as an element in a reasonable procedure of construction, the outcome of which determines the first principles of justice. . . . The leading idea is to establish a suitable connection between a particular conception of the person and first principles of justice, by means of a procedure of construction.165 The conception of the person does appear to take on a special status in justification that was hitherto not noted in A Theory of Justice. In Rawls’s later account of constructivism, the person is regarded as free and equal and as having two moral powers of reasonableness and rationality coupled with corresponding highest-order interests in developing and exercising these powers. These moral powers are further understood as the capacity for a sense of justice (the ability to understand and act on principles of justice) and the capacity to form revise and pursue a conception of the good. Persons also have a determinate conception of the good and a higher order interest in advancing it.166 The original position is laid out to reflect the freedom and equality of persons and to enable the conception of the person with its moral powers and regulative interests to determine principles of justice. The veil of ignorance, for example, is necessary to model the equality of persons in the fair choice of principles. This restatement of his constructivism also leads Rawls to rethink his account of the primary goods. No longer are they based on a psychological account of the goods necessary for any human life but are explicitly linked to the conception of the person. Rawls now argues that ‘the conception of moral persons as having certain specified highest-order interests selects what is to count as primary goods’.167 Primary goods are now conceived of as those goods necessary for the

Constructivism and A Theory of Justice

43

exercise of the moral powers associated with this Kantian conception of the person. Likewise, the argument for the priority of liberty is also reconceived so that it rests on ‘a conception of the person that would be recognized as liberal’.168 It appears that the controversial aspects of A Theory of Justice identified by critics such as Sandel as implicitly assuming a particular conception of the self actually do so, at least as Rawls comes to understand them. What is the status of this conception of the person? Is it the only model conception possible? Rawls rules this out explicitly when he states that ‘there are numerous conceptions of the person . . . and of its requisite intellectual, moral, and active powers.’169 As this implies, ‘there are various forms of constructivism’ corresponding to the numerous model conceptions.170 Rather than claim to find principles of ‘justice suitable for all societies’, Rawls’s constructivism is concerned to settle disagreement ‘within a democratic society’.171 Rather than identify universal principles, it seems that Rawls’s constructivism takes on a far more particular task. It is, ‘a procedure for singling out principles most fitting to the conception of the person most likely to be held, at least implicitly, in a modern democratic society’.172 Whereas in A Theory of Justice, it appeared that constructivism was concerned to avoid partial justification, in his later writings, Rawls seems to embrace it accepting that whilst he may identify determinate principles, this conception of justice has limited scope and application. Justice as fairness does not seem to be generally justified but only locally so and to the extent that people identify with the conception of the person. Most importantly, if the liberal democratic conception of the person is accepted as uncritically ‘given’, then it is unclear that Rawls’s Kantian constructivism is anything other than a repackaged form of foundationalism.

Conclusions At the mid-point of this chapter, we summarized an initial understanding of constructivism. On this account, it involves refusing to assume the existence of an independent order of moral facts on which to ground justification. Instead, constructivism accepts that we can only start from where we are and so works with our considered convictions, regarded not as intuitive perceptions of first principles but rather as pointers in the direction of an adequate and acceptable set of principles. Our convictions can therefore be regarded as provisional but revisable ‘fixed points’ in a procedure of critical reflection. This reflection is characterized as a search for reflective equilibrium between our convictions and principles. It involves utilizing the critical power of devices such as the original position to test the universality and impartiality of our principles. If we can develop principles that, upon critical examination, do not seem to be partial or subjective, then we can regard them as objective. Thus, the key ideas to take forward from Rawls’s initial constructivism both call on the notion of reflective equilibrium. First, this reflective form of practical reasoning understands that any starting conviction or principle is revisable. This ensures that our reasons

44

Constructivism and A Theory of Justice

and judgements are not necessarily tied to any subjective or partial assumptions. Second, this enables us to decouple the notion of objectivity from the idea of a ‘foundation’. Instead of objectivity being established by tracing reasoning back to an unquestionable ground, objective reasons may be conceived of as those that are the best we currently have when comprehensively tested for partiality and subjectivity. Identifying this alternative account of objectivity will be central to understanding the plausibility of a constructivist position, and this in turn relies on an appropriate account of the possibilities of practical reasoning. Critics object that Rawls’s constructivism is only able to identify principles of justice because of partial and foundational assumptions about persons and reasons built into the original position. Is this the case? Is Rawls’s Kantian constructivism simply a repackaged form of foundationalism? Rawls does accept that a conception of the person plays a key role in his constructivism. Critics argue that this Kantian conception of the person, its reasons, capacities and motivations, is both contestable and often contested. Instead of being a ‘thin’ and uncontroversial assumption, it appears that Rawls assumes a ‘thick’, or full, conception of the person. What is the status of this thick conception; does Rawls argue that people necessarily conceive of themselves in this way, thus making it a suitable assumption for an objective construction? He does not. Instead, he seems to accept that there are possibly many competing conceptions of the person. In his response to the context of pluralism, does Rawls build in exactly the type of controversial claim that underpins pluralist conflict, undermining the objectivity and universality of his construction in the pursuit of determinate principles? To avoid this charge, Rawls would need to produce strong demonstrations that the conception of the person and considered convictions he draws on, despite their apparent partiality, can be generally accepted. Alternatively, the argument may be that we instead have to recast our initial understanding of constructivism to reflect the way Rawls links it to a certain conception of the person. In this way, constructivism may have to be reconceived as a much thicker and local form of justification that makes strong assumptions about particular contexts in order to justify within those contexts to people who accept those assumptions. Perhaps constructivism might be compared to Walzer’s understanding of the role of the political philosopher standing within a community interpreting its shared meanings to their fellow citizens. In this way, the ‘local’ conception of the person embodied in, or accepted by, that community is a more or less ‘fixed point’ for justification. This interpretation appears to recast constructivism as a particular form of foundationalism. Perhaps we do not need to be quite so quick to give up on our initial understanding of a constructivist position distinct from foundationalism. Rather than accept a conception of the person as a fixed point, constructivism could open that conception up to the same critical attention as previously discussed ‘fixed points’. This enables us to regard the conception of the person as itself provisional and subject to the objectivity test of reflective equilibrium. The consequence of this however, as we have already seen reason to suspect that

Constructivism and A Theory of Justice

45

Rawls’s conception of the person is partial, may be highly corrosive to Rawls’s arguments to a determinate set of principles or to any principles at all. It is possible that reflection would modify the conception of the person as it does other considered convictions and so lead to revisions in the principles that reflect it. In principle, Rawls appears to accept the possibility of revising the conception of the person but simply finds it ‘hard to imagine’ what changes might be necessary.173 Opening the full range of assumptions up to critical reflection in this way, accepting provisionality and revisions, may lead to weaker and more general assumptions. From thinner starting points, it may be more difficult to identify a determinate set of principles; so, an alternative understanding of constructivism may sacrifice determinacy in order to maintain the possibility of objectivity and universal scope. In this way, it may be possible to maintain a distinct constructivist position, one that does not collapse into foundationalism on closer examination. These are the options: a strong but perhaps partial constructivism that issues in determinate principles of justice or a thinner, possibly universal constructivism that may not be able to determine outcomes. These options must be carried into the next chapter as we try to refine our understanding of constructivism and its limitations. Is constructivism most plausibly limited by locality and context or limited in determinacy instead? Either way, understanding the relationship between constructivism and conceptions of the person will become important.

3

The constructivism of Political Liberalism

In 1993, Rawls published Political Liberalism, his second major book.1 Political Liberalism marks the culmination of a major process of rethinking the project of justifying justice as fairness. Rawls is explicitly updating and correcting the arguments and ‘mistakes’ of A Theory of Justice. In doing so, he makes significant alterations to his account of constructivism. This chapter will give a clear account of changes as they bear on constructivism and will attempt to assess their impact on our general account of constructivist justification. The changes that Rawls makes to his theory are motivated by a concern to take seriously the impact of a proper recognition of the extent of pluralism. Since constructivism is intended as an adequate normative response to pluralism, if Rawls has underestimated its extent, we can expect revisions to be significant. Most importantly, for our purposes in Political Liberalism, Rawls explicitly discusses his constructivism, finds his constructivist starting points not in our considered convictions but in ‘public political culture’ and introduces the notion of overlapping consensus. The result is a ‘political’ constructivism that, at least at first glance and to many commentators, is very different from the constructivist position of A Theory of Justice. How different it actually is remains to be seen. On the one hand, Rawls tells us very explicitly and plainly what changes he has made and why. But on the other hand, there is much in both Political Liberalism and his other work of the time that should lead us to think carefully before we accept this straightforward picture.2 Thinking carefully about Rawls’s account of constructivism will encourage us to look again at what constructivism is more generally. For example, it became clear towards the end of the last chapter that, since the publication of A Theory of Justice, Rawls has become increasingly aware of the importance of a particular conception of the person to his constructivist justification. In his ‘Kantian Constructivism in Moral Theory’ Rawls had been absolutely explicit concerning the key role of this conception as a key element of construction, utilized in laying out the original position.3 However, we also became aware that this conception of the person, its reasons, capacities and motivations, is contested. It appears to be a thick or full conception, similar to Kant’s, and is perhaps to be regarded as one conception amongst the many possible. Whilst it may be necessary for the determination of principles, assuming

The constructivism of Political Liberalism 47 such a thick and contested conception may introduce an element of partiality into the construction. Why should we privilege this conception over others and how should we regard those who contest this account of the person? As we shall see, Political Liberalism appears to recast Rawls’s constructivism in such a way so as to accept the partial and local nature of its basic assumptions. We privilege this conception as it is the conception built into ‘our’ political culture, and we must adopt a tolerant, liberal attitude to those who disagree with us. Many commentators have interpreted Rawls, in this ‘local’ mode, as making strong assumptions based in a particular context in order to justify within that context to the people who accept those assumptions. Often Rawls seems to regard his conception of the person as a sort of fixed point for constructivist justification. We also noted at the conclusion of the previous chapter that there may be an account of Rawls’s constructivism that accepts the key role played by a conception of the person but that refines rather than redefines our understanding of constructivism. It may be possible to retain a universal and impartial understanding of constructivist justification that accepts no ‘fixed-points’. This chapter takes both these possibilities seriously and so is partly about properly understanding the relationship between political constructivism and conceptions of the person. The central conclusion will be that there is a legitimate place for both accounts of Rawls’s constructivism and that we must refine our general understanding to incorporate the notion of two levels of construction, primary and secondary. As in A Theory of Justice, Rawls outlines his conception of constructivist justification by contrasting it with intuitionism and foundationalism. Rational intuitionism is discussed as an example of a moral realist position. It regards moral principles as true statements about an independent order of moral values and moral knowledge as analogous to geometrical or mathematical knowledge. As a conception of the person, it assumes only that persons are capable of perceiving the truth and are then motivated by it.4 The key contrast is again intuitionism’s conception of an independent order of moral truths that our principles must match. In this, we can regard it as an example of foundationalism more generally that Rawls sets aside in favour of a constructivist account of principles as the outcome of a reasonable procedure of construction. He does so because, given pluralism, citizens do not obviously agree on any moral authority or order of moral facts.5 Rawls’s constructivism regards principles as the outcome of a constructivist procedure based on practical reason, utilizes a complex conception of the person and society in that construction and sets aside the concept of truth in favour of an idea of reasonable principles, persons and conceptions.6 In addition to these contrasts, Rawls also distinguishes his ‘political’ constructivism from a Kantian form of ‘moral’ constructivism. As we shall see in greater detail below, Rawls regards his constructivism as limited in scope, built out of and focused on ‘the political’. By contrast, moral constructivism is a comprehensive moral view that gives moral autonomy a central role for all of human life by drawing on a metaphysical conception of person and society.7 As will become clear, Rawls regards moral constructivism as one controversial comprehensive doctrine amongst many and therefore unlikely to gain general acceptance for its

48

The constructivism of Political Liberalism

claims to moral authority. Instead, he proposes that a more limited ‘political’ constructivism could underpin a generally accepted ‘political’ conception of justice. Unlike adopting any single comprehensive doctrine (sometimes referred to as conceptions of the good), adopting a political conception could be an impartial response to conditions of normative and political pluralism.

Outline of the new theory of justice In the two decades between the publication of A Theory of Justice and Political Liberalism, some of Rawls’s arguments have been left more or less untouched whilst others have undergone radical revisions. In these texts, there has been little significant change in the argument from the original position to the two principles of justice, but there has been substantial upheaval in what the original position models and what and how it justifies.8 As noted above, the basic materials by reference to which the original position is laid out have changed greatly; most obviously, the language of ‘considered convictions about justice’ has been set aside. The basis for social primary goods has also changed as has the account of stability. Most significantly, Rawls has apparently retreated from the substantive universalism of A Theory of Justice. Justice as fairness is now advanced as ‘a conception of political justice for a constitutional democratic regime’. This continues in the general direction of revisions initiated in the lectures on Kantian constructivism. Rawls acknowledges the partiality of his starting points in A Theory of Justice and recognizes that as a result he must take the problem of pluralism seriously or run the risk of instability.9 In A Theory of Justice, pluralism is a background assumption; in Political Liberalism, it is firmly to the fore. He summarizes the problem of pluralism as faced by democratic society in what he takes to be three general facts. 1

2 3

The diversity of reasonable comprehensive religious, philosophical and moral doctrines found in modern democratic societies is not a mere historical condition that may soon pass away; it is a permanent feature of the public culture of democracy. That a universally shared acceptance of any one comprehensive doctrine can be maintained only through the oppressive use of force. That given 1 and 2, an enduring and secure democratic regime must be willingly and freely supported by at least a substantial majority of its politically active citizens.10

These general facts characterize, along with the ever-present problem of moderate scarcity, the circumstances of justice for a modern democratic regime. We must note here that Rawls is concerned with a pluralism of reasonable doctrines, or reasonable pluralism, not with a plurality of reasonable and unreasonable doctrines, pluralism as such or simple pluralism.11 This is significant because, as we shall see below, it is very important for Rawls’s political constructivism that he

The constructivism of Political Liberalism 49 be able to account for and accept reasonable disagreement rather than regard everyone who does not accept justice as fairness as irrational, mad or stupid. It is this apparently permanent fact of even reasonable pluralism that ensures the partiality of the starting points of A Theory of Justice and encourages Rawls’s revisions. Most importantly, the problem of political justification in the face of such pluralism leads Rawls to make a fundamental distinction between comprehensive and political conceptions of justice. The difference between comprehensive and political conceptions is one of scope, of the range of subjects to which they apply. A doctrine that is comprehensive may include conceptions of what is of value in human life, personal character and conduct; it covers life as a whole in one articulated system.12 Citizens are assumed to hold more or less fully comprehensive views, and reasonable pluralism is characterized by the coexistence of at least several of these comprehensive conceptions. A political conception, on the other hand, can be identified by the three ways in which it differs from comprehensive doctrines.13 First, and in common with justice as fairness as presented in A Theory of Justice, a political conception of justice is worked out for a limited political subject, the basic structure of society. Second, a political conception is presented as a free-standing political view or module, as a conception that ‘stands free’ from comprehensive conceptions so that ‘accepting the political conception does not presuppose accepting any particular comprehensive doctrine’.14 Third, ‘its content is expressed in terms of certain fundamental ideas seen as implicit in the public political culture of a democratic society’.15 The central concern that Rawls has with A Theory of Justice is that it understands a well-ordered society to be one in which all citizens endorse justice as fairness ‘on the basis of . . . a comprehensive doctrine’.16 However, this contradicts what Rawls now regards as the general facts underpinning pluralism, that universal endorsement of a single comprehensive doctrine can be achieved only through the exercise of force. Recasting justice as fairness as a political conception of justice not reliant on the truth of any single comprehensive doctrine is central to justification in conditions of pluralism. This distinction between comprehensive and political conceptions of justice encourages Rawls to develop a two-stage justification of justice as fairness.17 At the first stage, justice as fairness is set out as a free-standing political view. It draws only on those values and ideas already implicit in the public political culture of a democratic regime and so does not assume the truth or falsehood of any of the comprehensive moral, religious or philosophical doctrines of citizens. It works these ideas up by a process of political construction into a political conception of justice for the basic structure and public reason. At the second stage, Rawls hopes to demonstrate that political liberalism is possible by showing that this political conception can and will be endorsed by citizens with reasonable comprehensive doctrines as the focus of an overlapping consensus. He believes that a wide range of comprehensive views could support the political conception as a valid political expression of the values they endorse. In this sense, their endorsement would ‘overlap’ on the political conception, a conception that nonetheless was constructed without reference to any of those comprehensive

50

The constructivism of Political Liberalism

views. This two-stage process aims to construct a free-standing political conception from ideas in public political culture that can gain the support of a wide range of reasonable and yet incompatible comprehensive conceptions of justice. If Rawls can do this, he feels that the political conception has done enough to be regarded as justified. Stage one – constructing the free-standing political conception At this initial stage, it is Rawls’s prime contention that the public political culture of a democratic regime can be ‘seen as a fund of implicitly shared ideas and principles’.18 Indeed, he posits as a fourth general fact ‘that the political culture of a democratic society . . . contains, at least implicitly, certain fundamental intuitive ideas from which it is possible to work up a political conception of justice suitable for a constitutional regime’.19 Despite the permanent nature of a reasonable plurality of answers to questions of justice, there is, in the political institutions as well as in the associations of the background culture to which comprehensive doctrines belong, a ‘tradition of democratic thought, the content of which is at least familiar and intelligible to the educated common sense of citizens’.20 These fundamental ideas form the basic material, out of which a political conception of justice can be constructed that can be acceptable to all and thus preserve stability despite pluralism. Rawls hopes these ideas will form an uncontroversial basis for justice as fairness, precisely because they are part of a shared or common fund of ideas available to all. The fundamental ideas that Rawls claims to identify in public political culture are the following: 1 2 3

the idea of society as a fair system of cooperation; the idea of the political conception of the person; and the idea of a well-ordered society.21

The first fundamental idea, that of society as a fair system of cooperation, is the fundamental organizing idea of justice as fairness.22 The form this idea takes in the political thought of democratic culture specifies an idea of society as organized for reciprocal advantage and regulated by publicly recognized rules and procedures.23 Alongside this, Rawls adopts his second fundamental idea, a conception of persons as citizens and of citizens as free and equal persons.24 Just as in his Dewey Lectures, a conception of the person is explicitly linked with constructivism. The specific conception inherent in the public political culture of democratic societies is of persons viewed as free and equal because of their possession to a requisite minimum degree of two moral powers, their capacity to be both reasonable and rational.25 Reasonableness here designates the capacity for a sense of justice, which expresses a willingness to propose and abide by fair terms of cooperation that others can publicly endorse.26 The ‘rational’ here designates the capacity to hold and revise a conception of the good, a plan by which ends are to be pursued. Persons are viewed also as actually having a determinate

The constructivism of Political Liberalism 51 conception of the good, an either fully or partially comprehensive doctrine that codifies the values, ends and ideals that they hold across all spheres of activity. Last, and much the same as in A Theory of Justice, a society is well ordered if it satisfies the following conditions. It must be a society in which it is publicly known that everyone accepts the same principles of justice, in which the basic structure is publicly known to be just according to those principles and also that citizens have a normally effective sense of justice such that they generally comply with the dictates of the basic institutions.27 When specifically applied to democratic society, the idea of well-orderedness consists of a combination of the first two ideas. Implicit in democratic public political culture is an idea of society as well ordered if it publicly embodies a fair system of cooperation between citizens regarded as free and equal.28 Analogous to his attempt in A Theory of Justice to systematize and make coherent our considered convictions about justice, Rawls here attempts to systematize and make coherent these basic materials at his disposal in Political Liberalism. In order to make a start at this project, he combines the fundamental ideas taken from public political culture with the general facts about such a culture, and he does so in two ways. First, he regards them as posing the question that political liberalism attempts to answer: ‘How is it possible that there may exist over time a stable and just society of free and equal citizens profoundly divided by reasonable though incompatible religious, philosophical, and moral doctrines?’29 Second, he views them as being modelled in the original position, the choice situation from within which the political conception of justice is constructed, which is the most appropriate and reasonable answer to that question, for us, here and now in a democratic society. Whereas in A Theory of Justice the original position was laid out with reference to our considered convictions about justice, it is now the fundamental ideas that are united in a single conception of fair choice. Each fundamental idea is modelled either by some aspect of the original position or by a part of the process that is conducted within it. Its basic features are laid out with reference to the fundamental democratic idea of persons as free and equal citizens, which takes a central role. The freedom of citizens is associated with the second moral power, the rational capacity; that is, the capacity to hold, to revise and to act on a conception of the good. This capacity is modelled in the original position by treating the parties to the choice as rationally autonomous.30 That the parties are rationally autonomous manifests itself in two ways: in their freedom within the constraints of the original position to agree on whatever principles of justice they perceive as most advantageous to the citizen they represent and also that whilst doing so they consider those citizens’ higher-order interests in the development and exercise of their two moral powers.31 The parties, in their deliberations, as in A Theory of Justice, are guided by the notion of primary goods. These were introduced in A Theory of Justice to provide substantive grounds for choice in the face of the indeterminacy behind the veil of ignorance and the formal nature of the thin theory of the good. Here, they similarly provide the necessary material grounds for choice, since the higher-order interests are purely

52

The constructivism of Political Liberalism

formal; the notion of primary goods provides at least a minimal substantive content to these formal interests. Now, however, primary goods are no longer conceived of as those goods necessary for the realization of any rational plan of life. They are instead essential for the realization of the higher-order interests of citizens conceived of as free and equal.32 As such, they are recognized as a specifically democratic conception of primary goods and no longer as universal ‘human goods’. In this, Rawls would seem to accept the criticisms of primary goods outlined in the last chapter. As we noted above, citizens are regarded as equal in virtue of their possession of the two moral powers to a minimum degree.33 In laying out the original position, this equality is associated with the first moral power, the capacity to be reasonable; here, it implies the willingness to honour fair terms of social cooperation and the possession of a sense of justice. This capacity is modelled by the structural aspects of the original position, which also model what we, here and now in democratic society, regard as fair conditions under which to make the choice of principles of justice suitable for the conditions of reasonable pluralism.34 These constraints are broadly familiar from A Theory of Justice, foremost of which is the veil of ignorance. As we recall, this shields the deliberations from differing bargaining strengths that might result from differences such as those in social position and wealth and also shields knowledge of such things as particular conceptions of the good. The parties are thus equal in that they are symmetrically situated with regard to one another, their only relevant features being their equal status as possessors of the moral powers to the required level and their holding some or other conception of the good. In addition to this, the original position has two requirements of the parties. First, they must select principles that may be stable given the fact of reasonable pluralism, principles that could possibly form the focus of an overlapping consensus. Second, their choice of principles must be unanimous and thus in the (higherorder) interests of all citizens.35 This ensures that the terms of cooperation are fair to all and impartial between them. In laying out the second fundamental idea to form the original position, the freedom of citizens is modelled by the parties in their rational autonomy and the choice that they make. Their equality is modelled by the constraints within which that choice is made. When these parties then deliberate on the first fundamental idea, that of society as a fair system of social cooperation, they construct Rawls’s two principles as the most reasonable and appropriate political principles of justice for the basic structure of a well-ordered society (the third fundamental idea). When the principles of justice adopted by the parties are freely affirmed and acted upon by equal citizens in society, the citizens then act with full (political) autonomy as they live under principles of justice they effectively give themselves. To some extent, in the transition from A Theory of Justice to Political Liberalism, the status of the original position has remained unchanged. It is still characterized as the choice situation that serves to identify the principles of justice. It is still taken to be a device of representation, a product of both hypothetical and

The constructivism of Political Liberalism 53 non-historical artifice.36 In other ways, however, its role has changed radically as has what it models and how it justifies. As we have already noted, the original position now mediates between the fundamental ideas and principles of justice instead of between considered convictions and principles. In doing so, it models peculiarly political values and ideals and, indeed, only those inherent in the traditions of democratic public political culture. As such, the openly universalist tone of A Theory of Justice has apparently been sacrificed. Now, the original position seems to justify only within the democratic political tradition. We can view Rawls as taking this path in response to the partiality in starting points identified in A Theory of Justice coupled with a fuller understanding of the fact of pluralism. It is indeed the existence of such far-reaching pluralism that points towards a purely political basis for the justification of principles and leads Rawls to construct a free-standing justification for the two principles as a political conception of justice. He believes that a theory of justice for a pluralist society must be constructivist since any appeal to the authority of an independent moral order will meet reasonable disagreement.37 Indeed, Rawls takes the ubiquitous nature of disagreement to be a fifth general fact, stressing that ‘many of our most important judgements are made under conditions where it is not to be expected that conscientious persons . . . will all arrive at the same conclusion’.38 He labels these conditions ‘the burdens of judgement’ and defines them as ‘the many hazards involved in the correct (and conscientious) exercise of our powers of reason and judgement’.39 Rawls’s six ‘burdens’ are as follows: a The evidence – empirical and scientific – . . . is conflicting and complex, and thus hard to assess and evaluate. b Even when we agree fully about the . . . considerations that are relevant, we may disagree about their weight, and so arrive at different judgements. c To some extent all our concepts . . . are vague . . . and this indeterminacy means that we must rely on judgement and interpretation . . . within some range . . . where reasonable people may differ. d . . . citizens’ total experiences are disparate enough for their judgements to diverge, at least to some degree, on many if not most cases of any significant complexity. e Often there are . . . normative considerations . . . on both sides of an issue and it is difficult to make an overall assessment. f . . . any system of social institutions is limited in the values it can admit so that some selection must be made from the full range of moral and political values that might be realized.40 The burdens of judgement provide, at the level of ideal theory, for the recognition of specifically reasonable disagreement, as opposed to unreasonable disagreement based on such things as self-interest, prejudice and bias.41 They allow Rawls to explain disagreement without recourse to labelling one or both disputants unreasonable, mad, irrational or stupid. Importantly, the burdens do not

54

The constructivism of Political Liberalism

imply either scepticism or an affirmation of the moral value of pluralism.42 Both of these are components of controversial comprehensive doctrines. Rather, the claim is simply that pluralism and disagreement are the result of the ordinary exercise of human reason in everyday conditions. Since all decisions, about truth as well as just moral or political values, are hard to make conclusively and indisputably, we rely on judgement calls, and we must expect those judgements to issue in an array of different and possibly conflicting positions on any complex issue. The ‘evident consequence’ of the burdens is that we cannot expect reasonable persons to ‘all affirm the same comprehensive doctrine’, or even compatible ones.43 Reasonable comprehensive doctrines, and the reasonable people who affirm them, recognize this and see that the burdens of judgement limit what we can expect to justify to each other.44 These limits Rawls hopes to draw at the edges of the political construction. Thus, in response to the burdens of judgement, Rawls’s political constructivism aims to construct a hopefully impartial conception of justice that can be publicly affirmed by a wide plurality of reasonable yet incompatible comprehensive doctrines. It is argued that the political conception is not controversial because of its free-standing nature, because it is not tied to any particular comprehensive and therefore controversial conception of justice or the good. Rather, it is chosen by the parties to the original position, itself laid out on the basis of the fundamental ideas found in democratic public political culture. It is political constructivism’s use of ‘political’ ideas that gives it its free-standing nature. Since the political ideas are taken from public political culture, they are part of a shared fund, not part of any particular comprehensive doctrine. Political constructivism in this way aims to avoid reliance on the truth or falsity of any reasonable doctrine. In virtue of this, the political conception should not explicitly clash with any of these reasonable comprehensive doctrines, and it is hoped that citizens can affirm the political conception alongside their other beliefs.45 It is in order to persuade us that many reasonable doctrines are consistent with such an affirmation that Rawls turns to the second stage of justification, overlapping consensus. Stage two – overlapping consensus The political conception of justice forms the basis of Rawls’s approach to a central question of politics; since political power is always at base coercive power, how can we justify its exercise? Since the political conception regards citizens as reasonable and rational, free and equal, and likely to have widely differing views, Rawls will regard political power as justified only when it is exercised in ‘accordance with a constitution the essentials of which all citizens . . . may reasonably be expected to endorse’. This is ‘the liberal principle of legitimacy’.46 Therefore, in order to justify a political order, Rawls needs to show that the political conception of justice as fairness can be reasonably affirmed by all citizens. A new stage of justification must be introduced because Rawls faces a new problem of justification. In A Theory of Justice, Rawls argued from our

The constructivism of Political Liberalism 55 considered convictions to the justification and general affirmation of justice as fairness. All citizens were connected to principles as they were justified using the convictions of those citizens as starting points of construction. However, in Political Liberalism, Rawls has disengaged the argument for the political conception from our convictions and in doing so has cut the direct connection from persons to principles. Principles are now regarded as justified from the starting point of ideas fundamental to public political culture rather than to the convictions of citizens. Overlapping consensus is a necessary second stage of justification as Rawls needs to establish a new link between citizens and the principles justified independently of their convictions. Only if Rawls can successfully show how the full range of reasonable citizens could support the political conception can he regard the political conception as justified, legitimate and stable in the face of pluralism. Rawls’s account of stability and the second stage of justification centres on the notion of an overlapping consensus. However, he is not concerned with just any sort of stability; an overlapping consensus exhibits a certain kind of stability, ‘stability for the right reasons’.47 The claim here is that having successfully completed the first stage of justification, then, given the condition of reasonable pluralism, we can aspire to an order more stable than a traditionally Hobbesian conception could envisage. Hobbesian stability is characterized by Rawls as a mere modus vivendi, where each citizen realizes that their selfinterested judgements lead to the sacrifice of a significant portion of the potential present satisfaction of their conception of the good for the sake of enduring political order and the long-term satisfaction that it can bring. Given pluralism and therefore an inability to pursue their full interest with impunity, citizens may compromise on a set of political institutions and principles for instrumental reasons. Citizens’ reasons for supporting a modus vivendi are therefore purely prudential. The main point here is that this stability is reliant on a contingent balance of power that precludes any one of the plurality of conceptions of the good subordinating the others. The fear is that if this balance is disrupted and one view gained the upper hand, then we can expect them to recalculate their interest and prudence may well encourage them now to attempt to institutionalize the conception of justice mandated by their comprehensive view. Understood as a modus vivendi, stability is therefore hostage to contingency, to changing circumstances settling on favourable conditions.48 Overlapping consensus, instead, involves the idea that each of the major comprehensive doctrines, and so most citizens, would be able to give their own reasons, from within those doctrines, for affirming the political conception as the proper conception for the political sphere. The claim is that each citizen could do so without regarding this as in any way sacrificing, or compromising, their more comprehensive views.49 Recall that the political conception is a political, not a comprehensive, conception constructed out of the fundamental ideas found in the public political culture of democratic society, and so it ‘stands free’ of particular controversial and partial doctrines. As such, it can be regarded as a module, and citizens can decide for themselves in what way the political

56

The constructivism of Political Liberalism

conception they all affirm is related to each of their particular comprehensive doctrines.50 The idea is that the free-standing module can be compatible with a wide range of comprehensive doctrines precisely because it relies on none. It is expected that each doctrine would have its own, perhaps very different, reasons for affirming the political conception. Because affirmation would then be from within citizens’ moral comprehensive doctrines, the political conception is supported on (a variety of) principled moral, and not prudential, grounds.51 Rawls can therefore hope that even if the balance of power shifted so as to favour any one doctrine, then the consensus on the political conception would hold, as any reasonable doctrine could still regard the political conception as a valid expression of its political views. Stability is therefore far less contingent on circumstance and fortune; it is not tied to a particular balance of power. As each citizen now finds reasons from within their own doctrine to affirm the political conception, citizens can now be regarded as directly connected to the principles that govern their interactions, and the overlapping consensus satisfies the liberal principle of legitimacy. Importantly, the acceptance of the political conception by comprehensive doctrines is evidence that it is properly freestanding, that it is impartial and unbiased as it is not based on any particular doctrine. In this respect, overlapping consensus is far more than a concern for stability and should be seen as a contribution to our understanding of objectivity and justification. Rawls is concerned to meet the charge that expecting this kind of consensus is ‘utopian’, that it is beyond the realm of real political interaction.52 To address this concern, Rawls outlines how the overlapping consensus could arise through two stages.53 He portrays the overlapping consensus as the second stage, built from a prior constitutional consensus. At this first stage, a consensus is envisaged on general liberal principles and the constitutional essentials they endorse, accepted simply as principles and not as grounded on any fundamental ideas of the person or society. The constitutional consensus is therefore neither deep nor wide and is narrow in scope, focused purely on the political procedures of democratic government.54 This consensus arises when liberal principles are first affirmed as part of a modus vivendi. However, over time, citizens may see the good those principles do them and society at large and come to affirm them on their own merit (if not as an outgrowth of their broader views). Citizens’ comprehensive doctrines are ‘shifted’ from affirmation of liberal principles out of pure self-interested compromise to affirmation as part of a liberal democratic constitutionalism that is valued in itself in the political sphere. A constitutional consensus on general liberal principles would, however, be pushed, by the need to work out its implications in the face of crises, the dispensation of justice, ongoing public debate and the political conception’s success thus far, towards deepening, broadening and focusing into an overlapping consensus.55 This story of possible development addresses the objection that the overlapping consensus is necessarily a utopian dream. It has, however, gone very little in the way towards furthering the justification for political liberalism. What Rawls has done thus far is to define an overlapping consensus and outline in

The constructivism of Political Liberalism 57 broad terms how it might possibly develop. He has opened our eyes to the possibility of a different kind of consensus and stability for the right reasons. But for the notion of overlapping consensus to clinch justification, far more is needed; the historical process described above actually has to be carried out by real people. Rawls recognizes that this instantiation is necessary for a complete justification and seems optimistic about its realization.56 All his efforts are recognized as ‘speculative’, but Rawls feels confident making ‘educated conjectures’ as to the success of this method.57 These conjectures consist in outlines of two examples of a ‘model consensus’ aimed at demonstrating the possibilities for affirmation. In the first, Rawls argues that the religious doctrine of free faith, a comprehensive liberalism such as Kant’s or Mill’s, and a pluralist view could partake in an overlapping consensus on the political conception of justice.58 In the second, he performs broadly the same demonstration for again free faith, a pluralist and a Kantian, but to these he adds utilitarianism as espoused by Bentham and Sidgwick.59 Rawls feels that in providing these models, he is demonstrating the breadth of affirmation an overlapping consensus on his principles could expect. The claim is that each of these doctrines has good reason, within its own set of values and judgements concerning politics, to affirm the political conception. If he is successful in showing that many of the major comprehensive doctrines present in democratic society would or could affirm the political conception, then he has greatly advanced his case that political liberalism can be justified in this way. Despite this, we must remember that nothing has yet been settled for justification through overlapping consensus. Rawls can merely demonstrate its possibility and hope that his optimism is justified in the long run. As a justificatory process, stage two remains, and is likely to remain, incomplete. In this sense, overlapping consensus is probably best seen as a regulative ideal against which the justification of justice as fairness can be assessed. To the extent to which the political conception can be accepted by many of the major comprehensive doctrines held by citizens in democratic society, we can regard it as reasonably justified. Summary Rawls is explicit that it is the necessity of responding to pluralism and the attendant problems of partiality that mandates the ‘political’ turn that is evident in his later work.60 He argues that justice as fairness, as the comprehensive doctrine to be found in A Theory of Justice, fails to take into account the modern conditions that render appeal to a substantive and universal comprehensive moral doctrine an unconvincing and impractical foundation for a political order. His response is to stress that justice as fairness is no longer to be regarded as a comprehensive doctrine. Instead, it is advanced as a doctrine applying only to the political realm and drawing only on publicly available starting points. His ‘political constructivism’ starts with several fundamental ideas taken to be implicit or inherent in democratic public political culture.61 By starting with public political culture, Rawls is stressing that these fundamental ideas are familiar to and also available

58

The constructivism of Political Liberalism

to all the citizens of a democratic regime. In this way, Rawls’s strategy is to bypass controversial comprehensive doctrines as the matter of justification gets underway. By utilizing only publicly available political ideas, Rawls hopes to abstain from questions of the truth or falsity of citizens’ comprehensive doctrines making his political conception properly freestanding.62 This move, coupled with the idea of the burdens of judgement and concomitant acceptance of reasonable disagreement, allows Rawls to more plausibly deny that the starting points for construction are partial. Also to combat the controversy his views initially aroused, Rawls makes less demands on the construction that builds from those starting points. He no longer constructs from our considered convictions, which may differ, all the way to determinate universal principles and institutions. Instead, he now finds the starting points for construction in the public political culture of a democratic regime. This is a less-demanding construction for two reasons. First, the starting points are already much closer to the outcome of the construction. It is not as far to move from what is already publicly political to an acceptable political conception of justice as it is from the personally political (considered convictions) to an acceptable conception of justice. Second, as we have seen, Rawls now appears to restrict his construction to the political sphere of democratic regimes. It would seem obviously less taxing on a construction to reach liberal conclusions from democratic starting points and only for application in democratic society than to reach those conclusions from personal starting points and expect those conclusions to apply universally. The idea of an overlapping consensus and especially the two examples of consensuses that Rawls gives are used to demonstrate the less-demanding nature of the construction. The model consensuses are Rawls’s attempt to offer us reasons for concluding both that the political conception would be acceptable to reasonable citizens of a well-ordered society and that it is plausibly in reflective equilibrium from the point of view of us deliberating, here and now, on the question of justice.63 These moves appear to have been designed to establish the uncontroversial nature of Rawls’s recent argument. Alongside this, however, there has been no similar rethink of his concern for determinacy. Rawls is still concerned to show that effectively the same substantive two principles would be the outcome of a political construction.64 Rawls also retains the ‘four stage sequence’ (original position, constitutional convention, legislation and judiciary) as the proper framework of thought to govern the application of those two principles in a well-ordered society.65 Rawls’s construction is still intended to be determinate of outcomes not only at the level of political principles but also at the institutional and judicial levels. He is also attached to a monistic interpretation of construction. That is, the fundamental ideas present in political culture can legitimately find expression in only one basic form. He has hoped to defend this determinacy by strengthening the basis of, and arguments for the suitability of, his starting points and also by bringing them closer to the desired outcome. In response to the problems set by the shortcomings of his constructivism in A Theory of Justice, we can regard Rawls as attempting to retain the determinacy of out-

The constructivism of Political Liberalism 59 comes and at the same time to reduce the controversiality of his starting points. If he has achieved this, then perhaps he has identified the form in which constructivism is a viable form of justification and an adequate response to the problems of foundationalism and the challenge of scepticism.

The constructivism of ‘political liberalism’: distinguishing between primary and secondary constructivism Is political constructivism a secondary constructivism? Rawls’s ‘turn’ to political constructivism appears to have fundamentally altered his basic project in response to perceived shortcomings of A Theory of Justice. In order to strengthen the starting points for construction and therefore to avoid partiality, he has appealed to the ideas fundamental to the public political culture of a democratic regime. However, whilst this might have the effect of reducing controversy surrounding justice as fairness within democratic societies, it may only serve to increase controversy and to introduce further partiality on a wider scale. In societies whose public political culture is not democratic, a basic appeal to democratic values either will fall on deaf ears or may even be offensive, whilst an appeal to their own public political cultures will probably garner little support for the liberal conception of justice as fairness. One of the central insights of A Theory of Justice is the introduction of a non-foundationalist conception of objective justification, at least as a significant possibility. This conception is centred on an understanding of critical reasoning that applies ‘all the way down’ to intuitive convictions, and no aspect of theory or practice is thought immune to reasoned revision. In Political Liberalism, the emphasis placed on the public political cultures of democratic regimes appears to dramatically alter Rawls’s own understanding of justification. It often appears as if the fundamental ideas found in public political culture assume a foundational role. ‘Political’ justification comes to consist in an attempt to show that a principle or institution is a reasonable expression of the fundamental ideas (suitably mediated through the original position), not that the fundamental ideas are themselves justified or reasonable. Public political culture, and the ideas fundamental to it, becomes the touchstone of justification and is not itself subject to critical reflection or revision as it must be for the constructivism of A Theory of Justice. Rawls appears to support this reading of political liberalism. He is clear that ‘in democratic society there is a tradition of democratic thought’ and that ‘justice as fairness starts from within . . . [this] political tradition’.66 It is this political tradition that is the source of the fundamental ideas that are in turn laid out in an original position that ‘models what we regard – here and now – as fair conditions’ for the identification of a ‘conception of justice that we regard – here and now – as fair’.67 Political liberalism seems firmly based on a conception of contemporary democratic culture, the essentials of which remain unexamined within the theory. Perhaps we can begin to make sense of these contradictions by refining our understanding of constructivism. We identified in A Theory of Justice a

60

The constructivism of Political Liberalism

constructivist position that was wholly non-foundationalist in aspiration. In Political Liberalism, Rawls appears to treat certain aspects of democratic public political culture as foundational to political constructivism. In the interests of clarity, I want to distinguish between these two understandings of constructivism as examples of primary constructivism and secondary constructivism.68 The basic constructivist position that we laid out from A Theory of Justice is an example of a primary constructivism. Primary constructivism endorses the understanding of justification as non-foundationalist, emphasizing the critical force of reflective reasoning in outlining the possibility of a principled pluralism. From the previous chapter, we are familiar with the basic outline of a primary constructivist position. On the other hand, secondary constructivism regards certain basic issues as settled and therefore beyond reflective examination and revision. In the process of justification, some basic materials are necessarily left unexamined and are therefore assigned a form of special status. In this sense, secondary constructivism is parasitic on some pre-existing foundationalism, and the constructivist process is only necessary as the foundation does not fully determine right action or principle. If the foundations fully determined principles, there would be no need for a secondary construction at all. However, if a secondary construction is necessary, principles must be justified by reference to the set of foundations that are themselves taken to limit or constrain possible constructions. Principles of justice are hereby justified by demonstrating how they are recommended or supported by these foundations. This distinction between primary and secondary understandings of constructivism is not, in the first instance at least, intended to convey a judgement about their relative importance or success as conceptions of the nature of justification. Whilst central conclusions of both this chapter and the main argument of the book will depend on being able to defend the possibility and priority of a properly primary constructivism, this defense will follow later. In the first instance, the distinction is simply a useful way of highlighting how some understandings of constructivism regard it as a secondary process dependent on the prior identification of some set of foundations. In this way, secondary constructivism, which regards some foundation as fixed for the purposes of justification, bears a striking resemblance to the notion of a ‘fixed-point’ reflective equilibrium examined in the previous chapter. The alternative understanding regards constructivism as the primary process of justification and is closely related to the idea of wide or two-way reflective equilibrium. Whereas primary constructivism necessarily aims at universal and cosmopolitan scope (even if it may not achieve it), secondary constructivism has a curtailed conception of scope. Even if the foundations at the base of secondary construction are understood to be universal, the underdetermination that necessitates construction would lead us to expect a plurality of secondary constructions, each drawing different determinate principles from the indeterminate foundation. More often, secondary construction works on less-than-universal foundations. Often, as appears to be the case in Political Liberalism and also more broadly in the work of communitarians and multiculturalists, the scope of

The constructivism of Political Liberalism 61 secondary constructivism is fixed by the boundaries of a ‘culture’. The social, political or moral practices of a particular cultural or political community are taken as basic for the constructivist justification of principles. As a consequence, these principles have no purchase where the foundations do not apply. In this manner, secondary constructivism can usually be understood to regard partial sets of assumptions as characterizing more or less fixed environments for justification. This is the case in Walzer’s appeal to communal ‘social meanings’ as both the source of normative force in justification and as the fixed context which appropriate principles must ‘fit’.69 Rawls’s political constructivism also appears to be an example of secondary constructivism as it ‘starts from within a certain political tradition’ embodied in western democratic public political culture.70 It treats this culture as a source of fundamental ideas to be elaborated into a political conception of justice and argues that all citizens of such democratic regimes could support this conception in an overlapping consensus. It appears that Political Liberalism presents a form of secondary constructivism that works within constraints imposed by an uncritical attitude towards the fundamental ideas found in democratic public political culture. It is therefore a partial justification, applying only to those who live in, or accept the attractions of, a liberal democratic society. Many commentators implicitly concur with the conclusion that Rawls’s project in Political Liberalism is a form of secondary constructivism, as the following brief list of comments demonstrates. Kukathas and Pettit describe Rawls’s move towards ‘Kantianism in one country’.71 Kukathas goes on to claim that Rawls ‘has abandoned his search for universal moral principles and recast his theory of justice as an attempt to articulate the principles of political justice appropriate only for modern democratic societies such as the United States’.72 Jones suggests of Rawls that ‘his reason for grounding political liberalism in democratic ideals may be simply that he is aiming to solve a problem that is internal to the modern democratic state’.73 Scheffler argues that since ‘the justification of liberal principles . . . appears to presuppose a society in which liberal values are already well entrenched . . . it is not clear that political liberalism could ever provide the original justification for a society’s liberal institutions’.74 Stuart Hampshire thinks that Rawls now seeks ‘harmony within the liberal stockade’.75 Thomas Pogge notes that Rawls has ‘dramatically shrunk . . . the domain of social systems to which he takes his conception to be applicable’, claiming that ‘he seems to be delimiting what is now the scope of his theoretical concern. His conception is to apply to constitutional democracies’.76 Brower and Gray each make a further claim. Rawls ‘now pictures his theory as applying only to societies whose members share certain model conceptions. By doing so, Rawls appears to have jettisoned the project of justifying liberalism’.77 ‘We may say of liberalism as a doctrine with aspirations for universal prescriptive authority . . . that it is dead – a result tacitly acknowledged in the most recent theorizing of the later Rawls.’78 Finally, Warnke interprets Rawls’s political constructivism as a crucial part in a ‘hermeneutic or interpretive turn’ in recent political philosophy and claims that

62

The constructivism of Political Liberalism for . . . Rawls, the examination of questions of justice is an interpretive enterprise; the theory of justice . . . is supposed . . . to work out a conception of justice that issues from the traditions and social meanings already contained in a shared political culture.79

Many commentators are therefore convinced that Rawls has turned from what we might call the project of primary constructivism to a more modest project of secondary constructivism. Here, the construction assumes a secondary role in relation to the fundamental ideas that are regarded as foundational ‘moral facts’. However, this chapter is concerned to examine whether this conviction, that on face value seems eminently reasonable, is correct. Has Rawls really given up on that primary constructivist project in favour of a secondary project that relativizes political constructivism to democratic culture? In what follows, it becomes clear that this is not the only, perhaps not even the most plausible, reading of Political Liberalism. Primary constructivism in Political Liberalism Given the judgements of the commentators above, any interpretation of Political Liberalism that tried to identify and prioritize its primary constructivist and universalist credentials would appear to be on difficult ground. We must be clear that a universalist reading of Political Liberalism goes against the mainstream of Rawlsian commentary. Indeed, in the face of the considerations and comment we have already brought to light, it might seem foolhardy to read Rawls in this way. However, I am convinced that if we pay close attention to what Rawls says, what he implies and what he has written elsewhere, it is possible to reconstruct a primary constructivism to which Rawls is committed. This continued primary project has been modified and refined in the transition from A Theory of Justice to Political Liberalism, most notably in identifying two distinct levels of construction and in the acceptance of a degree of indeterminacy. We can usefully take a closer look at exactly what Rawls says about constructivism in Political Liberalism, starting with a short passage in Lecture III.80 Political constructivism is a view about the structure and content of a political conception. It says that . . . the principles of political justice (content) may be represented as the outcome of a certain procedure of construction (structure). . . . This procedure . . . shows how the principles of justice follow from the principles of practical reason in union with conceptions of society and person, themselves ideas of practical reason.81 The structure of a reasonable political construction involves joining the principles of practical reason with appropriate conceptions of society and of the person.82 We must also take into account here Rawls’s contention that a political construction ‘is not reasonable in the first place unless . . . it can win its support by addressing each citizen’s reason’.83 In these passages Rawls implies that

The constructivism of Political Liberalism 63 political constructivism, if it is to be reasonable, should link together a conception of practical reason with appropriate conceptions of society and the person in order to construct principles capable of gaining support from each citizen’s reason. Reasonableness in justification relies on the proper use of practical reason but also in part on identifying appropriate conceptions of society and person. In order to properly understand the implications of this, we need to turn to Rawls’s other 1993 publication ‘The Law of Peoples’, later expanded and republished as a short book.84 In these pieces, Rawls addresses problems of international justice, international law and human rights. Initial impressions can confuse us as to their relationship to Political Liberalism. At times, the argument openly reinforces a particularist reading. For example, the initial stage in the identification of a just law of peoples involves working out principles for regulating the interactions between states that are already internally liberal and democratic. At a second stage, Rawls shows us how some non-liberal states could ‘buy into’ this liberal regulation of international politics. Importantly, non-democratic and non-liberal states are only considered after the just law of peoples is identified by reference to liberal democracies. The argument as a whole is advanced as a way to ‘work out the ideals and principles of the foreign policy of a reasonably just liberal people’.85 On this understanding, the extension to non-liberal states might be regarded as the way in which liberal states reassure themselves of their reasonableness. If this is all as regards to the argument, then consideration of these works would do little to undermine the particularist reading of Rawls’s later work as a form of secondary constructivism. However, whilst some aspects of The Law of Peoples confirm the particularist reading of Political Liberalism, there is much here to fuel an alternative reading. Rawls says explicitly that without an extension of political constructivism to the more universal sphere, his constructivism ‘would appear to be historicist and apply only to societies whose political institutions and culture are liberal . . . it is essential to show that this is not so’.86 Indeed, he goes so far as to call his constructivism ‘universal in its reach’.87 At least, sometimes Rawls regards his project in non-particularist terms. On these occasions, Rawls is clear that constructivism cannot be limited to liberal regimes with democratic public political cultures and it must be more universal in its scope. If this interpretation is sustainable, it would be difficult to make a secondary constructivist reading stick. If, however, the two principles, or even generic liberalism, are to be the conception of justice, we have already noted that its universalism might be suspect. Not only does Rawls identify problems defending its universalism in A Theory of Justice, but it is an awareness of these problems that leads Rawls to his ‘political’ turn. We can legitimately ask how then could political constructivism be universal, and also how can it be primary? First, despite all that we have seen about appeal to the fundamental ideas of a specific political culture, Rawls insists that there is a basic form of constructivism that runs through his work. ‘There is nothing relevantly different between how, say, justice as fairness is worked out for the domestic case in A Theory of

64

The constructivism of Political Liberalism

Justice, and how the law of peoples is worked out’.88 Rawls is plainly insistent that constructivism has a common basic form throughout his work, not only between his work on international justice and Political Liberalism but also between these later writings and A Theory of Justice. I will argue that it is only possible to make sense of Rawls’s claims regarding the continuity of his project if we properly identify the continuing commitment to a notion of primary constructivism based on a conception of practical reasoning that remains unchanged in its fundamentals. Does this mean that the principles of justice are also the same throughout his constructions? Not necessarily; he argues that in his version of constructivism, ‘there is no reason to think that the principles that apply to domestic justice are also appropriate for regulating . . . a society of peoples . . . each kind of subject . . . may be governed by its own characteristic principles’.89 This introduces the important notion that the construction of principles may vary between different subjects and different contexts. Making sense of this idea of a plurality of outcomes to a basic form of constructivism is necessary to a better understanding of Rawls’s political liberalism, of constructivist justification in general and of their relationship with pluralism. Advances made here will build on the last chapter and be carried forward to deepen our understanding of a plausible constructivist account in later chapters. Rawls’s constructivism is a procedure whereby we link practical reason with appropriate conceptions of society and person. If this constructivism is to issue in a variety of outcomes, the obvious point at which constructions would vary is in the specification of appropriate conceptions of society and person. We shall see that although the basic principles of practical reason appear to remain invariant in Rawls’s notion of justification, perhaps different conceptions of society and person will be appropriate in different contexts and for different subjects. It is important for constructivist procedures that they start from suitable assumptions, ‘a correct starting point’, but there is no reason to think that these starting points should not be ‘suitably adjusted to different subjects’.90 Indeed, the internal logic of constructivist justification that tries to avoid relying on controversial or partial assumptions encourages us to recognize that in different contexts, ‘since the problem and subject are different in each case, [the starting points and therefore] the ideals and principles adopted may also be different’.91 Perhaps we can conceive of a basic primary constructivism that, though made applicable across contexts by the identification of varying particular accounts of the basic materials of construction, is not itself context based or particular. The following sections will argue that it is possible to consistently lay out such an account building closely on the text of Rawls’s later writings. In doing so, we will need to both examine Rawls’s account of practical reasoning and look more closely at his account of the fundamental ideas of society and person. Primary constructivism in Rawlsian practical reasoning Rawls’s political constructivism consists of principles of practical reasoning unified with conceptions of society and the person. If we are to identify a

The constructivism of Political Liberalism 65 primary constructivist reading of this notion, then we should start by examining the conception of practical reasoning at work in political constructivism. Rawls claims that pluralism is a general fact about modern democratic societies, and in ‘The Law of Peoples’, he is obviously also assuming pluralism in the international sphere. Rawls asks us to recognize that this pluralism is a ‘permanent feature’ of the modern condition and says it is not going to go away in the foreseeable future; therefore, he insists that our political theories need to take it into account. Our reaction could be one that bemoans the fragmentation that pluralism perhaps brings with it.92 Or our reaction may be the fundamentalist one that attempts to eliminate competing views. Alternatively, our reaction, and Rawls’s liberal one, may be to accept that pluralism is reasonable and, indeed, that it may not be something to regret. We can recognize the fact of pluralism and accept it as a feature of the modern world. Rawls accepts that ‘a plurality of reasonable yet incompatible comprehensive doctrines is the normal result of the exercise of human reason’.93 He does not regard pluralism as a failure of reason. It is not a problem that arises because of the limits of rationality or because people reason insufficiently well to reach the same conclusion. Rawls accepts that on many issues concerning human flourishing and the good life for man reason, no matter how perfect, does not mandate a single answer. Pluralism need not be the result of a flaw in man, or even an implication of his finitude; it can be reasonable in itself. The normal and free exercise of reason will lead even competent reasoners to legitimately make different judgements on a wide variety of subjects. This is the point of the burdens of judgement; they explain disagreement between fully competent reasoners and show that such disagreement, and the pluralism that it entails, should be considered reasonable. So, instead of viewing comprehensive doctrines as conceptions of the good life that are impenetrable to reason, as in some sense irrational, they are viewed as fully competent and ‘authentic exercises of reason’.94 It is important to understand that recognizing this does not imply scepticism about the sorts of things comprehensive doctrines disagree about. It is simply a recognition of the reasonableness of disagreement. There may be one true answer to questions of value and flourishing, and the political liberal need not deny this, but the conditions of knowledge are such that people can reasonably disagree about what it is. The notion of truth is therefore reserved for comprehensive doctrines, whereas the political conception is regarded as reasonable. It is equally important to understand that this recognition is not an affirmation of the moral value of pluralism either. This is a recognition of the ‘fact of pluralism’ rather than the ‘value of pluralism’. Neither Rawls nor the political liberal generally is committed to thinking that the world would be a worse place if, by some amazing coincidence, everybody came to believe the same thing. They will, however, regard this as a fantastically unlikely state of affairs.95 Recognition of the general fact of reasonable pluralism prompts Rawls to look again at his conception of practical reasoning. However, considering the centrality of practical reason to Rawls’s constructivism, it is surprising that he does not treat us to an extended discussion of the subject. Instead, we are left to

66

The constructivism of Political Liberalism

piece his views together from the scattered remarks made throughout Political Liberalism. However, several aspects of practical reasoning have spanned his writing from A Theory of Justice to the present day. He has remained clear that in order to reason practically, we must avoid partiality in starting points, and we must reason through procedures that all can follow, just as we can all think ourselves into the original position. In A Theory of Justice where Rawls constructed from ‘our considered convictions about justice’, since he felt that most of us shared similar convictions, he felt confident arguing directly for justice as fairness against alternative conceptions such as utilitarianism. Given shared, and therefore uncontroversial, starting points, he attempted to show that utilitarianism was an unreasonable expression of our convictions about freedom, equality and rights. He was clear that justice as fairness contradicted prominent comprehensive doctrines. In accepting the burdens of judgement, and thereby accepting the reasonableness of disagreement and of pluralism, Rawls’s strategy must be revised in order to preserve the general account of reasoning. Since (most) comprehensive doctrines are considered to be ‘authentic exercises of reason’, practical reasoning that confronted and contradicted them would introduce contradiction within the bounds of reason. Rawls cannot hold both that comprehensive doctrines are reasonable and that, in contradicting practical reason, they are unreasonable. If in recognizing the burdens of judgement he is recognizing the reasonableness of a plurality of doctrines concerning a subject, he cannot simultaneously deny their reasonableness with directly confrontational practical reason. In this way, he is led to what has been called a sort of ‘epistemic abstinence’ as a principle of reasoned justification.96 A comprehensive doctrine is a legitimate expression of a citizen’s reason. Since, in order to be reasonable, political constructivism must justify by addressing each citizen’s reason, then Rawls’s procedure of practical reasoning and justification must both avoid contradicting the reasonable comprehensive doctrines of citizens and gain their affirmation.97 Practical reason, to be reasonable at all, must justify political conclusions in terms each citizen can accept.98 This is, I take it, the point of Rawls’s concern with overlapping consensus and also the ‘liberal principle of legitimacy’.99 In Political Liberalism, instead of confronting a doctrine like utilitarianism, Rawls recognizes its reasonableness, sidesteps questions of its truth and attempts to show how it could affirm the free-standing political conception of justice as part of an overlapping consensus.100 This is consistent with the account of objectivity we identified in A Theory of Justice. By justifying the political conception as a free-standing module that does not affirm or contradict comprehensive doctrines, political liberalism can avoid basing principles on partial or biased grounds. That the political conception can be the focus of overlapping consensus is not primarily an argument about stability. Rather, it is evidence of the impartiality and objectivity of the political conception. In a wider context, this theoretical commitment to avoiding partiality is a vital function of the other aspect of practical reason that Rawls concentrates on, reflective equilibrium. Demonstrating that citizens could sign up to an overlapping consensus justifies from within each citizen’s reason, but the broader target

The constructivism of Political Liberalism 67 for justification is the third point of view, that ‘of you and me who are elaborating justice as fairness and examining it as a political conception of justice’.101 This is the point of view from which, after its internal justifications are sorted out, any political conception is to be assessed. In this case, the test of practical reason is one of reflective equilibrium, that is, of how well the conception either articulates our convictions or through a process of critical reflection can show us that this conception is a more reasonable account of politics than one that embodies our present convictions. If we recall discussions of reflective equilibrium from the previous chapter, it is plain that this procedure of reasoning is one of open-ended critical reflection that refuses to recognize any particular value or view as a foundation point for justification. This characterization of a constructivist account of practical reasoning is again affirmed by Rawls in Political Liberalism where he is clear that no specification of these criteria [of practical reasoning] can be . . . treated as final, any reading of them is always open to be checked by critical reflection. This is so, even if for the present, we are fully confident of our formulations of principles and cannot see how they can be in serious error.102 It is also clear that ‘the overall criterion for the reasonable is general and wide reflective equilibrium’ and that this is ‘plainly the important philosophical concept’.103 Rawls obviously continues to affirm his central account of what it is to reason practically as laid out in A Theory of Justice. That he does not regard this as a form of reasoning peculiar to democratic political culture is made clear in The Law of Peoples. Here, he is clear that reasonable justification of the law of peoples depends on whether ‘its outcome can be endorsed on due reflection’ where he says that ‘I use the phrase to mean the same as “reflective equilibrium” as explained in A Theory of Justice.’104 There is explicit continuity in Rawls’s conception of practical reasoning. We have also noted that this sort of procedure of reasoning modelled on open-ended critical reflection is paradigmatic of primary constructivist reasoning. In its refusal to rely on any unvindicated and foundational premise or reasons, Rawls’s continued use of reflective equilibrium accepts the primary justificatory project of justifying in terms that each person could accept. So, despite accusations of particularism, when Rawls argues that constructivist justification involves principles of practical reasoning in union with conceptions of society and person, he is not advancing a local or even a plural conception of practical reasoning. His conception of practical reason remains as universal as ever. It is still concerned to avoid partiality and to justify to all ‘on due reflection’. Reasserting this is the first step towards understanding how Rawls’s political liberalism is not a form of secondary constructivism. Its basic account of reasoning is still compatible with primary constructivism. Can the same be said of the materials used in the process of political construction?

68

The constructivism of Political Liberalism

Primary constructivism in starting points Rawls is committed to recognizing the reasonableness of pluralism. This is the implication of the burdens of judgement. These burdens, and the notion of the reasonableness of disagreement that they support, underpin the experience of diversity. Rawls’s acceptance of diversity is clear in his claim that principles may legitimately vary with subject and context. But, given the universality of a form of practical reasoning and therefore, it seems, of a basic understanding of constructivism, how are we to make sense of the notion that principles may vary with subject and context. If principles of justice are identified by a construction involving the principles of practical reason in union with the fundamental ideas of society and person, and practical reasoning is invariant, then variety must enter with the conceptions of society and the person. This section aims to explore the variations in Rawls’s understanding of these fundamental ideas that form the materials of construction and therefore the idea that constructed principles may themselves vary. In the democratic domestic context that is the prime focus of Political Liberalism, Rawls’s constructivism takes the conceptions of society as a fair system of cooperation and of the person as free and equal citizen and lays them out to form the procedure of choice that is the original position. In the extension to the law of peoples, Rawls again lays out conceptions of society and person in an original position, but these are appropriate conceptions laid out to form an ‘appropriate original position’.105 In total, Rawls delineates three different formulations of the original position, each appropriate for a different context and a different subject. First, there is the appropriate original position for domestic justice.106 Second, he outlines the form of the original position appropriate for the extension of constructivist justification to a society of liberal peoples.107 In this original position, the parties now represent societies well ordered by liberal conceptions of justice; these parties’ reason is guided by the fundamental interests of democratic societies, and whilst subject to a veil of ignorance, this veil now screens out different kinds of information as it is ‘properly adjusted to the case at hand’.108 Third, he outlines the original position as appropriately extended to decent hierarchical, or non-liberal, peoples. Decent peoples are those that satisfy the twin criteria of nonaggression externally coupled with a common good idea of justice internally.109 Parties now represent well-ordered hierarchical regimes and reason from interests presupposed by the particular hierarchical conception of justice.110 In each case, the form of the construction, an original position that embodies practical reasoning joined with appropriate conceptions of society and person, remains constant. The substantive original position, however, varies according to subject and varies as a function of variation in the appropriate conceptions of society and the person. Evidence to support the contention that Rawlsian constructivism varies according to variation in conceptions of society and person is forthcoming in the text. Rawls points out that although

The constructivism of Political Liberalism 69 the original position at the first level, that of domestic justice, incorporates a political conception of the person rooted in the public culture of a liberal society, the original position at the second level, that of the law of peoples, does not.111 Rawls makes it plain that the political conception of the person as free and equal appropriate at the domestic level (and to the extension to liberal states) is inappropriate at the broad international level where he describes it as too narrow a basis for a law of peoples.112 Here, instead of as free and equal citizens, persons are regarded as ‘responsible members of society who can recognize their moral duties and obligations and play their part in social life’.113 By implication, the conception of society that is appropriate also changes. First, the international society of states is not regarded as a fair system for social cooperation but as more like a fair system for just and peaceful coexistence. Second, the conception of society as well ordered is also adjusted to the international context. In Political Liberalism, well-orderedness required that the political system satisfies publicly accepted principles of justice and that these were normally observed by citizens.114 In ‘The Law of Peoples’, it requires instead that a regime be peaceful, not expansionist, legitimate in the eyes of its people and honour basic human rights.115 The fundamental idea of a society as well ordered can be seen to be a general concept to which, when considering justice in liberal societies, Rawls gives a democratic twist in the formulation of a more specific democratic conception of well-orderedness.116 In a similar way, it is helpful to regard the other two main fundamental ideas of society and the person that Rawls works with as general concepts that can be built up into a variety of different conceptions. Rawls refers to justice as fairness as a ‘special case in which members of society are citizens regarded as free and equal in virtue of their possessing the two moral powers to the requisite degree’.117 Rawls writes that more generally, and starting in the ancient world, the person has been understood as someone capable of participating properly in social life.118 As Rex Martin notes, this more general concept of the person is as possessing the two moral powers, the capacity to be both reasonable and rational. He also notes that this concept would not have been alien to the thought of Plato, Aristotle or Aquinas.119 Rawls makes a similar point when he claims that ‘no sensible view can possibly get by without the reasonable and the rational as I use them’.120 He is also very careful to point out that it is ‘since we start within the tradition of democratic thought . . . we also think of citizens as free and equal persons’.121 Regarding the two moral powers as expressed in the freedom and equality of citizens is Rawls’s democratic twist on the more general concept of the fundamental idea of the person. Similarly, Rawls works with a democratic version of the fundamental idea of society as a fair system of social cooperation, where the specifically democratic ‘twist’ is that social cooperation is for reciprocal advantage.122 In its more general form, this idea embraces many interpretations including, as Rawls mentions, one based on mutual advantage as fair cooperation. Rex Martin again notes that this general

70

The constructivism of Political Liberalism

concept of the fundamental idea of society ‘is not as such a distinctively democratic idea . . . [it] is a mainstay of classical Greek and of medieval social thought’.123 It appears that Rawls’s political constructivism cannot be limited only to democratic societies. He specifically constructs principles of justice that extend beyond the umbrella of democratic liberal culture in ‘The Law of Peoples’. Both here and in Political Liberalism, he has also shown awareness of the more general universal applicability of his fundamental ideas. For each of the three main fundamental ideas, he gives a characterization of the general concept. Then, in constructing for different subjects in different contexts, he gives each of these ideas a twist appropriate to the construction of justice for that context. In the construction of the law of peoples, he gives each an international twist, whilst in the construction of justice as fairness for democratic liberal regimes, he gives each idea a democratic twist. A reasonable political constructivism links together practical reason and appropriate conceptions of society and the person to construct principles capable of generating support from within the reason of each citizen. Rawls’s acknowledgement of the problems posed by pluralism leads him to understand that reasonableness lies in recognizing variation over which conception of society or person is appropriate depending on the arena for choice; that is, varying the conceptions identified as fundamental to particular contexts. In this way, political constructivism addresses pluralism by attempting to identify the correct starting points for construction differently when considering different subjects. Rawls’s primary constructivism We have outlined the beginnings of a universalist constructivism, largely formal rather than substantive, centred on Rawls’s statement that constructivism expresses the principles of practical reason in union with appropriate conceptions of society and person. This basic account is a primary rather than secondary constructivism. Rather than base the construction on foundational assumptions, it reaffirms Rawls’s account of non-foundationalist reflective reasoning. As The Law of Peoples shows, Rawls does not regard this as a peculiarly democratic form of reasoning. In the democratic context, it combines with appropriately liberal and democratic conceptions of society and person and may very well justify democratic principles. However, the fundamental ideas united with practical reason vary with context. In this way, they do not seem to be an arbitrary or partial imposition but are appropriate to subject matter. Given the reasonableness of disagreement, this account can help us begin to understand the relationship between constructivism and pluralism. The conception of practical reason retains its form over the variety of particular constructions. This is not to say, however, that the substantial reasons we actually offer in various situations will not differ; they will obviously do so in response to context. The substance of our attempts at practical reasoning will vary with context, and this becomes apparent in the outcome of each substantive construction. This will be, as we

The constructivism of Political Liberalism 71 have seen, different for different subjects. In democratic regimes, we construct a liberal political conception, perhaps justice as fairness, whilst at the wider international level, we construct a just law of peoples. This interpretation suggests a reconsideration of Rawls’s two basic criteria of reasonableness, a willingness to recognize the burdens of judgement and a willingness to propose and abide by fair terms of cooperation. The reasonableness of recognizing the burdens of judgement should be regarded as substantially unchanging. It is implied in admitting the possibility of legitimate pluralism and partly underpins the always provisional nature of practical reasoning. A willingness to propose fair terms, however, will mean different things in different contexts. Although the willingness itself remains invariant in this account, what counts as fair terms is related to the conceptions of society and the person under consideration. As we have seen, whilst the two principles of justice as fairness may be fair terms in a liberal democracy, this does not necessarily make it fair to propose them as terms for the reorganization of hierarchical societies or the international system. The understanding of Rawls outlined here has brought us a long way from the dominant scholarly interpretation of Rawls as a particularist or secondary constructivist, and this extended insight should not be downplayed. However, his primary constructivism, so far, seems formal rather than substantive, explaining and legitimizing the diversity of local reasons rather than identifying universal reasons of justice. Are the latter possible or are all reasons local and particular, as they would appear to be if in each context a different set of fundamental ideas must be used to orient reasoning? Are there resources in Political Liberalism that could underpin substantive but universal reasons within this broad justificationary framework? Recall that Rawls broadly outlines, in addition to the variety of conceptions of the person and society we can expect to find in different contexts, bare concepts of society and the person. He implies that these are not tied to any particular context. Reasoning about justice oriented by reference to these would, if possible, be universal. We could usefully take a closer look at these bare concepts of society and the person. Briefly, Rawls says in Political Liberalism that it is only because ‘we start within the tradition of democratic thought’ that we think of persons as free and equal and society as a fair system of reciprocal cooperation.124 There is a more basic understanding of society that is simply as a fair system of social cooperation that may be conceived in a number of ways, one of which is as organized for reciprocal advantage and another, for example, is for mutual advantage.125 ‘Thus the basic idea of society is one whose members engage . . . in activities guided by publicly recognized rules and procedures that those cooperating accept and regard as properly regulating their conduct.’126 He is also very clear about the bare concept of the person when he says, Beginning with the ancient world, the concept of the person has been understood . . . as the concept of someone who can take part in, or who can play a role in, social life, and hence exercise and respect its various rights and duties.127

72

The constructivism of Political Liberalism

Elsewhere, he clarifies this in a less historical tone. We ask: what must persons be like to engage in practical reason? To answer we say that persons have the two moral powers as well as a determinate conception of the good. Their being reasonable and rational means they can understand, apply, and act from . . . practical principles.128 Rawls is offering an account of practical reasoners, and as practical reason is itself part of the universal framework of justification, all people have these two moral powers (at least to the extent that they are ‘normally’ functioning human beings). This is a basic conception of the person intended to be universal. ‘No sensible view can possibly get by without the reasonable and the rational as I use them.’129 Responding to Habermas’s claim that his conception of the person is comprehensive and partial, going beyond the limits of the political in its assumptions of rationality for example, Rawls claims that ‘if this argument involves Plato’s and Kant’s view of reason, so does the simplest bit of logic and mathematics’.130 He even says that these conceptions of society and of the person are necessary assumptions of any conception of the right or the good; ‘they are as basic as the ideas of judgment and inference, and the principles of practical reason’.131 Now if Rawls envisages practical reasoning oriented by these bare concepts of the society and person, concepts he argues are appropriate in any context, then he should be able to justify substantive universal conclusions about justice. We should expect these conclusions to be much thinner than those about justice in a domestic society (of Political Liberalism for example) because the conceptions of society and the person used at the universal level are much less determinate than those that are appropriate in the restricted domestic (in this case democratic) context. They would however be, as Rawls claims in The Law of Peoples, ‘universal in reach’.132 Once again, the manner in which Rawls deals with the law of peoples complements Political Liberalism. In Political Liberalism, he describes justice as fairness as ‘a special case’, where ‘other societies take a different view’. The conceptions of justice adopted by these other societies are unlikely to be explicitly constructivist, and they are likely to be comprehensive. But Rawls says, ‘I assume they all contain a conception of the right and the good that includes a conception of justice that can be understood as in some way advancing the common good.’ Such a ‘common good conception of justice’ must take into account the good of all members of society. He goes on to state that this idea of justice may seem weak. Still, some such idea is necessary if we are to have a society with a legal system imposing what are correctly believed to be genuine obligations, rather than a society that merely coerces its subjects.133 He goes on to compare this idea with Hart’s minimum content of natural law and claims that a society without such a conception of justice may not be ‘a

The constructivism of Political Liberalism 73 society but something else’.134 This expansion of the category of just societies, beyond liberal democratic societies shaped by justice as fairness, to take in other societies identifies an expanded category of ‘just and decent societies’. The broad category of decent societies takes in both liberal democracies and nonliberal societies that are characterized by a common good conception of justice.135 All decent societies, Rawls contends, could sign up to a reasonable law of peoples. The most important aspect of this running together of liberal and ‘non-liberal but decent’ societies is that the discussion of both, and their relations to a law of peoples, are included under the broad heading of Ideal Theory. Non-ideal theory is reserved for dealing with outlaw states or burdened societies. The inclusion of decent societies in an international society regulated by a law of peoples is not a concession to injustice, as would be the case if we were to include outlaw or burdened societies. Including decent peoples is not a second best but a part of the ideal. Of course, Rawls, as a liberal, might prefer a world where every society was a liberal democracy (he might regard this as fully just), but he is clear that aiming for this is not necessary for an account of universal justice. Indeed, he explicitly says that ‘the aim of the Law of Peoples would be fully achieved when all societies have been able to establish either a liberal or a decent regime’.136 Rawls is, in effect, claiming that there are a variety of contexts for justification, most of them thick, local and particular, but at least one of them thin and universal. It is this thin universal justification that may underpin a universal conception of justice involving, Rawls believes, the justification of a limited set of human rights.137 Since the bare concepts of society and the person are, for Rawls, necessary assumptions of any conception of the right or good, human rights are those rights that ‘are recognized as necessary conditions of any system of social co-operation’.138 It is because of this that Rawls can claim that ‘human rights set a necessary . . . standard for the decency of domestic political and social institutions’.139 The substantive justification based on the bare concepts of the person and society both justifies a universal account of human rights and is used to set the limits to Rawls’s expanded category of the decent. What is more, as these substantive conclusions are the outcome of what Rawls claims is a universal account of practical reasoning oriented by concepts of the person and society that any conception of the right or good must assume, the conclusions are such that their universalism is based on properly universal reasons.140 In summary, it is clear that we can, and should, recast Rawls’s political liberal project as falling into two parts. In doing so, it is important that we bear in mind that these do not correspond to the two stages of Rawls’s justification of justice as fairness within democratic societies, the free-standing justification and the overlapping consensus. Instead, as a consequence of our exploration, we can recast the constructivist project more broadly as involving two distinct aspects. Rawls’s constructivism functions on two distinct levels, and this will be important for our general understanding of constructivism in later chapters. The first level, and that which has occupied us so far, is the outlining of a reconfigured but still primary constructivist project. This conceives of practical reason

74

The constructivism of Political Liberalism

oriented by conceptions of society and person. Because of the nature of practical reasoning, these conceptions must be regarded, provisionally, as the best understanding currently available to us. These conceptions are thin and so, therefore, must be the universal account of justice constructed from them. Consequently, the universal level will be indeterminate over many complex and local issues, and so this important aspect of Rawls’s constructivism stands alongside the second aspect; Rawls’s attempts to actually undertake some of the substantive construction involved outlining fair and reasonable terms of cooperation at the secondary level of a particular context, that of a democratic regime. In other contexts and other societies at the secondary level, it is clear that alternative understandings of justice will apply, perhaps common good ideas of justice in hierarchical societies, for example. It is at this secondary level, in the context of democratic public political culture, that Rawls’s arguments in favour of justice as fairness are to apply, both the argument to its status as a free-standing political conception and the argument that it can form the focus of an overlapping consensus. It is also this secondary level upon which the bulk of commentary on Political Liberalism has focused whilst, in contrast, the primary constructivist project that forms the first part of Rawls’s refinement of constructivism has received little attention. It has been the intention of this chapter thus far to bring this continued primary constructivist aspect of Rawls’s project to the surface of his work and to the attention of his commentators. This is not to say however that his democratic secondary construction is unimportant. On the contrary, it is a central focus of his work and commands the bulk of critics’ attention; so it is worth pausing for a brief exploration of Rawls’s secondary construction. Rawls’s secondary construction Much of Political Liberalism is concerned to outline the appropriate conception of justice for a democratic society. I refer to this as a form of secondary constructivism as it takes a particular context as given, democratic public political culture, and attempts to construct principles of political justice from this basis. In this instance, the construction takes second place to the prior acceptance of a limited context. A secondary construction aims to identify the appropriate fundamental ideas implicit or inherent in public political culture that is viewed as a shared fund of political ideas. That some such ideas are available is taken to be a fourth general fact.141 As we have seen, it is these fundamental ideas that constitute the basic materials out of which Rawls constructs the free-standing political conception. We are familiar with how this construction goes. Rawls confidently identifies the appropriate fundamental ideas as those of society as a fair system of reciprocal cooperation and of the person as a free and equal citizen. It is not clear how Rawls has picked out these fundamental ideas, but he treats their identification with a high level of certainty, to the extent that he does not consider alternative ideas nor provide an argument for the centrality of these particular ideas to democratic culture. They appear to function as a fixed point for Rawls’s reflections on justice. His certainty about these ideas is reflected in the determi-

The constructivism of Political Liberalism 75 nate construction they underwrite. However, numerous critics have taken issue with both the fundamental ideas and the political conception of justice. Their criticisms can be taken as pointers towards a necessary theoretical backward step. Rawlsian secondary constructivism cannot realistically claim the degree of certainty about fundamental ideas that it appears to in Political Liberalism. A brief examination of critics supports the argument against certainty in Rawls’s recovery of the fundamental ideas from democratic public political culture. Klosko draws our attention to surveys of political attitudes in Britain and America, which he takes to illustrate that the political culture of democratic societies may not be very liberal after all. He concludes that these societies may not be accurately reflected in Rawls’s account.142 Other critics are keen to point out either that there are alternative accounts of the fundamental ideas in democratic political culture or, what amounts to much the same claim, that common comprehensive doctrines could not join an overlapping consensus on the political conception. Amongst those who would dispute Rawls’s interpretation of free and equal persons, Wenar lists the followers of Bentham, Hume and Hobbes as well as the recent thought of Gauthier.143 Doppelt also draws attention to ‘rival ideals . . . no less universal or more particular than the Kantian conception’, such as the bourgeois-individualist, the patriarchal and the Judeo-Christian ideals that ‘find no rational expression or mediation within Rawls’ system’ despite informing powerful traditions that strongly influence the public political view of social justice.144 William Galston goes further and denies that Rawls manages even a correct interpretation of part of public political culture that is instead underpinned by a conception of persons as morally equal in virtue of their humanity but admits that society can legitimately reflect natural inequality and in which freedom is curtailed by claims and duties that precede and govern our choice of social institutions. He claims that Rawls’s conception . . . diverges from this American understanding of freedom and equality and leads to principles of justice significantly different from those most Americans embrace. . . . Rawls’s reconstruction of justice as fairness does not invoke – indeed it flatly rejects – the conception of the person underlying our beliefs and practices.145 Michael Walzer proposes his own rather different interpretation of the ‘social meanings’ inherent in American culture. The appropriate arrangements in our own society are those . . . of a decentralized democratic socialism; a strong welfare state run, in part at least, by local and amateur officials; an open and demystified civil service; independent public schools; the sharing of hard work and free time; the protection of religious and familial life; a system of public honouring and dishonoring free from all considerations of rank and class; workers’ control of companies and factories; a politics of parties, movements, meetings, and public debate.146

76

The constructivism of Political Liberalism

Wenar also notes that the political conception will not be compatible within an overlapping consensus with some of the comprehensive doctrines that Rawls claims it will be.147 Barry is also keen to point out the divisive nature, at the level of overlapping consensus, of the social egalitarian commitment that Rawls reads into political culture and argues that he should renounce the difference principle altogether if this project is to be successful.148 Young similarly argues that political liberalism could not gain the support of libertarianism on precisely this issue.149 Libertarian political thought is however a doctrine with a long tradition in western democracies; if it cannot join an overlapping consensus, this should concern us. So too should Rex Martin’s conclusion that Millian utilitarianism, a doctrine that Rawls specifically refers to, could not support the political conception.150 This list of criticisms does not necessarily undermine Rawls’s identification of the fundamental ideas and construction of the political conception; his critics may be mistaken after all. However, it is not clear how Rawls would defend his interpretations. The diversity of understandings of democratic culture and the fundamental ideas we have just laid out point towards general problems with any claim to have identified the fundamental ideas with any degree of certainty. Indeed, there is much in Political Liberalism that should lead us to expect this sort of disagreement, no matter how confident Rawls seems to be in his own claims. Disagreement could be the result of competing descriptions of democratic culture. Surprisingly, it seems that the possibility that he could have misdescribed the fundamental ideas is never entertained by Rawls despite his recognition elsewhere of the permanent possibility of mistakes.151 He recognizes that ‘the judgement of any agent, individual or corporate, may be mistaken’ but seems certain of his own.152 Either Rawls is here making a claim that the fundamental ideas are set apart from the other concepts utilized in political philosophy, that they have an entirely different epistemological status and that he has the capacity to accurately perceive them, or he must admit the possibility that he has misdescribed them. Obviously, he does not claim the former and so he cannot advance his reading of democratic culture uncritically. The reasonableness of this disagreement can only be supported by considering Rawls’s account of the burdens of judgement. The burdens of judgement are formulated early in Political Liberalism precisely because it is necessary to provide some sort of account of why reasonable people, who exercise their two moral powers to a sufficient degree, come to radically differing conclusions on such things as the nature of justice. We must recall that Rawls does not regard this disagreement as a problem merely in the application of his theory. Instead, such disagreement must be accounted for at the level of ideal theory, in the best of all possible cases, since diversity is a permanent feature of democratic society.153 The importance of the burdens of judgement, given the centrality of the notions of reasonable disagreement and pluralism to Rawls’s theory, is undeniable. But why, if they are of such importance to the exposition and construction of political liberalism, does Rawls at no point seem to consider the consequences of their application to his interpretation of the fundamental ideas? The correct por-

The constructivism of Political Liberalism 77 trayal of the fundamental ideas implicit in our public political culture is precisely one of those areas of activity where the conditions that limit the authority of our interpretations would be of utmost importance. If we can expect at least some dissensus on the interpretation of the fundamental ideas, then, as the possibility of the authoritative recovery of those ideas is basic to Rawls’s political liberalism, his substantive construction of the political conception of justice as fairness is threatened. We must conclude that there are diverse accounts of democratic culture and fundamental ideas and that the disagreement between them is reasonable. Rather than as certain and authoritative claims about political culture, we can only regard Rawls’s understanding of the fundamental ideas as one part of an ongoing theoretical discussion. Although Rawls does not claim that the appropriate fundamental ideas are properly identified through reasoned debate, he has every reason to. He acknowledges the possible pluralism in political culture when he says that ‘the public political culture is bound to contain different fundamental ideas that can be developed in different ways’, although he does not make anything of this important observation. He goes on to suggest that ‘an orderly contest between them over time is a reliable way to find which one, if any, is most reasonable’.154 This is a significant suggestion and points the way to a constructivist response to pluralism at this secondary level. All that is necessary is that Rawls recognize the provisional nature of his account of the fundamental ideas. Once this is acknowledged, attitudes towards them can change. Rather than regarding fundamental ideas as certain and authoritative, we can regard them as our best current interpretation of political culture whilst expecting to reflect on them as alternatives are brought to our attention.155 Again, this may introduce an element of indeterminacy as to the outcome of the constructive process, but this is offset against a significant and necessary decrease in the potential partiality of the fundamental ideas. As a consequence, Rawls’s secondary and particular construction should be regarded as an attempt either to stake a claim as to the outcome of the reflective process or to contribute to that ongoing reflection.156 This is not to deny the importance of Rawls’s secondary construction but simply to point out that, since a single account of the fundamental ideas is not obviously the most reasonable, its conclusions must always be held provisionally.

Conclusion Whilst on initial examination the account of constructivism that Rawls advances in Political Liberalism is significantly different from that we find in A Theory of Justice, this chapter contends that the difference between the two positions is not as great as is commonly perceived. Rather, there is a significant degree of continuity in Rawls’s account of constructivism. The most significant development that Rawls directs our attention to is that constructivism is recast in a ‘political’ light. However, in drawing out the impact of this development, the continuity is obscured. Refining constructivism to take account of both this continuity and Rawls’s ‘political’ understanding would be important progress.

78

The constructivism of Political Liberalism

It is important to recognize the advances that this political turn brings in the understanding and exposition of constructivism generally. The development of the idea and vocabulary of political constructivism are of great importance in drawing out the constructivist impulse to impartiality and objectivity. In the last chapter, it became clear that impartial reasons play a key role in a constructivist account of both objectivity and reflective equilibrium. Central to political constructivism is the argument that a political conception can be constructed that does not rely on, or assume the truth or falsity of, any particular comprehensive doctrine. To do so would imply being partial to certain doctrines over others; accepting the truth of one comprehensive doctrine usually implies denying many others. Of course, the political conception may coincide with the conclusions of many comprehensive doctrines; some overlap is likely to be necessary for an overlapping consensus to develop after all. The key point, however, is that it relies on none of them in the process of justification and so is not based on a comprehensive religious, moral or philosophical doctrine. Political Liberalism also makes explicit the manner and extent to which constructivism is conceived of as a necessary response to pluralism. Rawls explicitly recognizes the ‘fact’ of pluralism and argues that this has a direct impact on his understanding of constructivist justification. He argues that he underestimated the extent and persistence of pluralism in A Theory of Justice, where he had impractically and unjustifiably expected the universal acceptance of justice as fairness as a comprehensive doctrine. Understanding the implications of the fact of pluralism directly motivates the development of political constructivism and the idea of justice as fairness as a political conception of justice. However, it is not clear that justice as fairness was ever advanced as a comprehensive doctrine. Its subject has always been political, the basic structure of society, and constructivism was always conceived of as a response to the pluralism of foundational claims. Also, although reflective equilibrium starts with our convictions about justice, it can accept neither their authority nor that of any comprehensive doctrine that might inform them. The principles of justice would always be justified on the basis of reasons that have undergone a properly critical examination. It is not the constructivism of A Theory of Justice that falls foul of pluralism; rather, it is the inadequacy of a constructivism that attempts to justify the thick and determinate principles of justice as fairness as a response to that pluralism that is brought to the fore. Also made explicit in Political Liberalism is the importance of conceptions of society and person to a constructivist account. We saw in the last chapter that a conception of the person may be implied in A Theory of Justice. Now Rawls recognizes this and claims that it is only by identifying the appropriate conceptions of society and person, those that are fundamental ideas in democratic public political culture, that we can construct a reasonable political conception of justice. In doing so, he lays the groundwork for an argument about the necessity of these conceptions that will be reinforced by Onora O’Neill in the next chapter and will be central to the general account of constructivism that we are developing. Concerns have been raised about Rawls’s use of these fundamental

The constructivism of Political Liberalism 79 ideas, recovered from democratic political culture. Commentators have tended to interpret Rawls’s conceptions of society and person as a restriction on the scope of the political conception so that it only applies within already democratic societies, understanding the political conception to be the outcome of a secondary construction that takes as given an account of the appropriate fundamental ideas as democratic. It has become absolutely clear that this is not the case. There is a strong argument, built in the terms of Political Liberalism, for recognizing two distinct levels of construction, primary constructivism and secondary constructivism. This recognition is an important refinement of our understanding of constructivism and will play a key role in later chapters. Rawls puts forward a unified account of the basic mode of construction, unifying principles of practical reasoning with appropriate conceptions of society and the person. It is when this is cashed out that the distinction between two levels becomes clear. In particular contexts, it is perfectly possible for a constructivist position to be worked out by reference to fundamental ideas appropriate to that context, identified by taking for granted the relevant political culture. In this way, the construction is a secondary process with the political culture playing a foundational role. It is also clear that different contexts, characterized by different political cultures, may contain different fundamental ideas. In this way, the conceptions of society and person used to orient practical reasoning would vary and so would the outcome of this variety of secondary constructions. Behind them is a more universal account that models these secondary procedures that simply states that reasonable principles will result from the exercise of practical reason oriented by fundamental ideas appropriate to context. Primary constructivism draws on Rawls’s further arguments that there are bare or basic accounts of the fundamental ideas of society and person that must be assumed in any context and by any comprehensive doctrine. A Rawlsian account of properly primary constructivism would orient practical reason, understood in terms of reflective equilibrium, with these bare conceptions of society and person and attempt to draw out substantive and universal conclusions. Of course, given the constructivist conception of objectivity tied up with reflective equilibrium, both the bare conceptions and the substantive conclusions could only be regarded as always provisional. Drawing on the account of constructivist reasoning in the previous chapter, this is to be expected and reinforces rather than undermines their objectivity. We can posit a less-than-fully determining primary constructivism instantiated in, and perhaps limiting, a variety of secondary constructions. The bulk of Rawls’s account in Political Liberalism is an attempt to draw out a reasonable secondary construction for a particular context. Critical comment has concentrated on this secondary project and has called many aspects of it into question. I have not concentrated to any extent on these criticisms precisely because they focus on this secondary level. Even if Rawls’s secondary construction is wholly inadequate in its working out, this does not affect the broader argument. The account of primary constructivism is unaffected by the relative strength or weakness of Rawls’s arguments for a particular context. This interpretation of Rawls’s constructivism helps to make sense of both ‘The Law of Peoples’ and some passages in Political Liberalism. Whereas

80

The constructivism of Political Liberalism

Scheffler sees ‘The Law of Peoples’ as demonstrating Rawls’s lack of liberal universalism, I would instead argue that it makes clear Rawls’s universal constructivist aspirations.157 He attempts the construction of a substantive conception of justice for an international setting. He also makes it plain that many states that are not liberal are, however, decent and well ordered and are therefore just to an appropriate extent, and must be accounted for within ideal theory.158 It is hard to see how he could make this claim without some more or less universal account of decency and justice in mind. An interpretation of Rawlsian political constructivism that makes room for two levels of construction, most significantly the possibility of primary construction, makes far more sense of Rawls’s writing of ‘The Law of Peoples’ than alternative interpretations that portray Rawls’s method as tied to the particularities of democratic societies. Such an interpretation also helps to make sense of passages in Political Liberalism. They help explain how, for example, Rawls can claim that the ideas of society and the person ‘are as basic as the ideas of judgement and inference, and the principles of practical reason’.159 Most notably, however, interpretations that deny the universalism implicit in Rawls’s constructivism cannot account for the way in which he treats justice as fairness as a special case of justice, where other societies adopt different conceptions that are still just. He talks about a ‘weak’ universal idea of justice that is ‘necessary if we are to have a legal system imposing what are correctly believed to be genuine obligations’ making reference to Hart’s idea of a minimum content of natural law and hints that even this is not as thick as what Rawls considers necessary for viable and just legal systems.160 I contend that these passages are best accounted for in terms of the universal nature of Rawls’s primary construction and are not easily accounted for in any other way. We have significantly advanced our understanding of constructivism in this chapter in readiness for further exploration and are now also in a position to recognize the basic continuity in Rawls’s constructivist project. Throughout his development of constructivist justification, he has provided arguments for a universal account of constructivism drawing on the notion of reflective equilibrium and the conception of objectivity it implies. Where A Theory of Justice recognized that traditional foundational justification was an inappropriate basis for a principled liberal politics and developed constructivism as an alternative, in Political Liberalism, this basic account has been refined, not replaced, by introducing two distinct levels of construction. The basic continuity of a primary constructivist project links the two books at a deeper level than the shared commitment to a determinate account of justice as fairness and forms the heart of the project that runs through Rawls’s work. That this universal project is at the heart of Rawls’s work is illustrated by the final page of his introduction to the paperback edition of Political Liberalism, where he writes, If a reasonably just society that subordinates power to its aims is not possible . . . one might ask with Kant whether it is worthwhile for human beings to live on the earth? We must start with the assumption that a reasonably just political society is possible.161

4

Primary constructivism and O’Neill

Onora O’Neill’s work is explicitly constructivist. Towards Justice and Virtue, where O’Neill defends her ethical theory most systematically, is subtitled ‘a constructive account of practical reasoning’, while her book on Kant’s practical philosophy has the title Constructions of Reason. Her arguments can be regarded as a reaction to perceived problems in foundational justification that ‘try to orient practical reasoning by assuming idealized conceptions of persons, reasons and action, or transcendent moral ideals or real moral properties’ and so are not properly tested ‘against the limitations of the human condition and the constraints of human knowledge’.1 Confronted with the failure of foundational argument, O’Neill’s response, like Rawls’s, is to explore constructivist possibilities for justification. However, although she applauds Rawls for the vision of his explorations, O’Neill has argued that Rawls’s constructivism is unsatisfactory. She claims that her ‘more guarded constructivism’ differs from Rawls’s in two ways.2 First, although Rawls appears to base his constructivism on abstract assumptions, in actuality he idealizes rather than abstracts. Rawls assumes idealized conceptions of the person and of instrumental rationality that introduce significant elements of partiality into his constructivism. O’Neill, as we shall see, is keen not to introduce unvindicated ideals into constructivist justification and intends to rely only on legitimate abstractions.3 Second, she aims to vindicate an account of practical reasoning and its authority without relying on metaphysical argument or on the values of a particular culture. O’Neill understands Rawls’s conception of reasoning to be based on an account of democratic public political culture. Since her more ‘Kantian conception addresses “the world at large”, rather than a restricted range of fellow citizens in closed but internally democratic societies, it differs . . . from the Rawlsian conception’.4 It should be clear from the previous chapters that Rawls need not necessarily be regarded as guilty of these failings, and so Rawls and O’Neill need not be so far apart. Whilst exploring O’Neill’s constructivism, we will come to appreciate the real similarities between the reading of Rawls that we have developed and O’Neill’s own account. We will usefully also be able to focus more clearly on the constructivist account of practical reasoning. O’Neill’s account, like Rawls’s, is a response to the problems of justification in conditions of pluralism. Not only do we encounter diverse societies, cultures

82

Primary constructivism and O’Neill

and conceptions of the good, each providing us with different justifying reasons to act in different ways, but we also encounter diverse accounts of what makes a reason count as a good reason in justification. Diversity, therefore, poses particularly deep problems for justification. In our engagement with Rawls, we developed a basic understanding of a constructivist response to these conditions of pluralism, one that involved the independence of justification from any prior order of moral facts whilst retaining a commitment to objective justification. This is precisely the task that O’Neill sets herself, exploring the possibility of constructivist space between moral realism and relativism.5 She endorses Rawls’s determination not to assume any independent moral facts. Constructivists . . . reject or (at least) bracket moral realism. They assign no place or weight to distinctively moral facts or properties, whether natural or non-natural, that can be discovered or intuited and do not seek foundations for ethics in such facts.6 However, having rejected independent moral facts, the traditional basis for objective moral judgements, O’Neill does not want to sacrifice the possibility of objective justification in an immediate acceptance of relativism or conventionalism. She is adamant that to do so is to ‘give up on the entire project of justification’ or to ‘settle for conceptions of justification too weak to support a robust form of objectivity’. O’Neill is clear that ‘constructivism, as Rawls came to see it . . . [and as O’Neill also sees it], holds that, although realist underpinnings are unobtainable, (some) objective, action-guiding ethical prescriptions can be justified’.7 O’Neill is endorsing our general picture of constructivism that couples the denial that there are applicable and generally available moral facts with a firm and ongoing commitment to objective justification. More specifically, we find that O’Neill’s account of the objectivity of reasons reinforces the picture that we are developing. She argues that practical reasoning demands that we subject our various subjective reasons for acting, whether they are personal, traditional or cultural, to critical examination through a process of reflective reasoning. Just as in Rawls’s account of reflective equilibrium, this process uncovers the partial bases of our reasons and principles. Coupled with a claim that partial reasons lack the general authority of proper reasons, O’Neill argues that any general account of practical reasoning must identify properly objective reasons and that that her critical, Kantian account does exactly that. It does so by means of the ongoing reflective process that highlights and eliminates the partial assumptions underlying our reasoning and picks out less-subjective assumptions on which we can provisionally rest our reasoning. At first glance, and this judgement is confirmed below, this account is very similar to the account of objective practical reasoning we found in Rawlsian constructivism. However, as she focuses her attention on the authority and application of practical reasoning, it will become apparent that there is much we can learn from O’Neill’s account. Similarities between our tentative account of constructivism and O’Neill’s do not end here. Like Rawls, O’Neill finds it necessary to orient

Primary constructivism and O’Neill

83

the account of practical reason with relevant conceptions of society and person in order to generate substantive principles to guide political action. She also argues explicitly that there are bare and abstract concepts of the person and society that are assumed whenever we act and that these can orient practical reasoning in the generation of universal and objective principles. Of course, O’Neill also recognizes that each society makes much thicker assumptions than these abstract ones and argues that these form the basis of particular cultures. Effectively, O’Neill draws the distinction between thick and local cultures and much thinner but universal principles that underlies the distinction that we have already drawn between secondary and primary constructivism. So, while she does not use the terminology of primary and secondary constructivism, O’Neill both recognizes and reinforces the necessity of thinking of constructivism in this way. That so many aspects of our general picture resonate with aspects of O’Neill’s constructivism should encourage our conviction that we are on the right path in our attempt to come to a general understanding of constructivist justification. O’Neill’s argument for the justification of universal (primary) constructivism goes through several distinct steps. First, she argues that for a practical reason to have the authority that reasons are taken to have, they must be formally universal in that they should apply to all relevant others. Second, she argues that, at least sometimes, all others are relevant and on those occasions our reasoning must be both formally universal and cosmopolitan in scope. The third claim is that universal and cosmopolitan practical reasoning can sometimes generate substantive but thin universal and cosmopolitan principles of justice and virtue. Finally, this thin universal account is supplemented by a range of substantive particular principles generated by diverse secondary constructions at a variety of less-than-universal scopes. This chapter will follow O’Neill through these steps. We will find that O’Neill, like Rawls, reinforces and refines our general picture of primary constructivism. We will also find that she, again like Rawls, expects more of this constructivism than is wise, running the risk of partiality in pursuit of the over-determination of its outcomes.

O’Neill’s account of practical reasoning Alternative accounts of practical reasoning O’Neill is clear that there are several competing accounts of what it is to reason practically. She initially identifies three broad alternatives to her own critical or Kantian account and refers to these as Platonist, instrumental and particularist understandings of practical reason. Platonist conceptions are those that characterize reasoning practically as reason ‘oriented by objective ends, such as moral properties or metaphysically grounded ideals’. Alternatively, instrumental conceptions characterize reasoning practically as reasoning ‘oriented by subjective ends, such as subjective conceptions of the good, preferences, desires, assisted by means ends reasoning’. O’Neill also picks out a particularist conception that

84

Primary constructivism and O’Neill

characterizes reasoning practically as reasoning that is ‘informed by actual norms and commitments’.8 O’Neill is dissatisfied with these alternatives. They each exhibit shortcomings that endanger their status as general accounts of practical reasoning. The vindication of a Platonist conception relies on the plausibility of its account of objective ends. Traditionally, these ends have been identified in an account of the objective Good or as ends ordained by God or some similar mystical or spiritual source. Alternatively, these ends are revealed by a timeless and impartial reason or simply by reflection on unchanging human nature. It will be plain that these attempted vindications of the objectivity of certain ends are the same as those foundational and universalist accounts identified in earlier chapters. It was indeed the apparent problems in sustaining such claims to objective foundations that prompted the turn to constructivist accounts of objectivity and justification in the first place. Platonist accounts have struggled to support the strong metaphysical claims intrinsic to their justification. Yet without such support this account of reasoning practically is unconvincing as a general account of our practical reason. Those that do not accept the attendant metaphysical claims will find reasons that invoke them to be incomprehensible and arbitrary and their authority illusory.9 The claim here need not be, in the first instance, a claim about what makes something a reason but about what makes a reason practical. That God tells you to do something may be a good reason for you to do it. However, it is impractical to advance this more generally as a good reason for everyone to perform the action.10 O’Neill also rejects the suitability of instrumental reasoning as a general account of practical reasoning. This is not, of course, to reject instrumental reasoning entirely but to recognize its limitations.11 For instrumental accounts of practical reasoning, ends are subjective rather than objective. An instrumental account utilizes means-ends reasoning to guide us towards the achievement of ends that we already have, whatever they are. As O’Neill colourfully puts it, instrumental reasons ‘may show that omelette makers cannot reasonably refuse to break eggs, but they cannot show that making omelettes is reasonable’.12 An instrumental conception of practical reason offers no reasons to choose certain ends rather than others, and as such the reasons it does offer are barefacedly arbitrary. Instrumental reasons are explicitly conditional on the specific ends and desires at which an action aims. Those who do not share such ends already cannot be offered reasons for doing so. They will therefore find the instrumental reasons predicated on those ends arbitrary, incomprehensible and unappealing. The instrumental justification of rational action is reliant on a prior and given account of ends. These ends may be given by a specific individual’s desires or beliefs. In this case, such reasons based on such ends will obviously be arbitrary to other individuals who may not share the relevant desire or belief. Alternatively, these ends are often thought of as given by the cultural and social norms of a particular society. O’Neill refers to accounts of reasoning practically that appeal to the existing norms of a society or tradition to guide reason and to determine what rational action might entail in a given context as a form of par-

Primary constructivism and O’Neill

85

ticularist account. Reasons offered on these sorts of particularist accounts refer at some point to the fundamental values and virtues inherent in a practice or way of life. As such, particularist conceptions of practical reason cannot offer reasons in favour of a practice or society. The reasons that such conceptions offer are therefore conditional on a prior acceptance of the importance of a set of norms with limited scope; they are always dependent on our actual norms and commitments. As a result, these reasons will also seem arbitrary and incomprehensible to those outside the relevant society or tradition. O’Neill goes so far as to call this form of reasoning ethnocentric and argues that any reasons that presuppose the norms of particular times and places cannot have general authority.13 Indeed, it will become clear that such reasoning is very similar to the account of secondary constructivism laid out in the previous chapter. These thumbnail sketches of the problems of alternative accounts of reasoning practically each highlight two shortcomings. First, the reasons that each account offers are likely to remain stubbornly incomprehensible as reasons to some. Second, and as a consequence, these reasons will appear arbitrary, as will the authority they claim, to these people. Each account offers reasons that are simply not followable by others. These shortcomings are, O’Neill explains, symptomatic of a failure to understand what it is that makes practical reason practical. If practical reason is concerned to offer reasons that will both guide action and inform our deliberation about actions, then it is surely necessary that it ‘at least aim to be followable by others for whom it is to count as reasoning’.14 Reasons offered by the accounts above will not be followable by those who cannot accept the necessary metaphysical commitments, do not share the appropriate set of subjective ends or are not part of the relevant community. As such, they will appear incomprehensible and arbitrary, lacking the authority of a general account of reason, and therefore will offer no reasons to such people.15 Reasoning is not practical if its premises disregard the world as it is and people as we find them. Reasoning that ignored this requirement would be impractical; it would be utopian reasoning. When we offer someone a reason we want it to count as a reason for them, a reason they can follow or accept. Therefore, we should not predicate our reasons on features, characteristics or values that they do not embody or wish to embody or do not hold or wish to hold. If we do so, then we have no cause to hope that our reason will count as a reason for them. Obviously, and as we noted in discussion of Rawls, every piece of reasoning will make assumptions in order to ‘get going’. These assumptions can be regarded as embodied in predicates about audience, context, knowledge and capabilities and so on. With this in mind, O’Neill makes much of a distinction between idealization and abstraction. The process of abstraction brackets (but does not deny) predicates that are true of the matter at hand. Abstraction will leave things out of starting points for reasoning but will never augment a starting point. As such, properly abstract reasoning does not rely on false premises; it merely does not rely on all available information. Idealization, on the other hand, enhances starting points on the basis of ‘ideal’ predicates. As such, idealization makes illegitimate assumptions about the basic premises of reasons

86

Primary constructivism and O’Neill

(about the beliefs or capacities of the relevant agents, for example). In a trivial example, if James and Cliff like vanilla ice cream whilst Kirk likes strawberry, then any decision about what flavour to buy is made on the assumption that everyone likes vanilla will seem arbitrary to Kirk and lack authority for him. This is because it is based on an idealized account of general preferences about ice cream; the predicate ‘likes vanilla’ is not true of Kirk and has been added in an idealized account of his preferences. On the other hand, a decision to visit the ice cream parlour rather than a pizza house based on the assumption that everyone likes ice cream will not seem arbitrary to James, Cliff or Kirk. This is because it is based only on abstract predicates that are true of each of them. Whilst leaving out information about flavour, and so not relying on full detail, it adds nothing illegitimately and so does not rely on false premises. Despite its many critics, abstraction is unavoidable.16 Communitarian critics have worried about the way universalist positions abstract from actual agents and communities. However, any account of individuals, communities, societies or cultures is an abstract account. No account can reproduce every aspect or detail of its subject matter. Any account therefore makes decisions about what to include and what to exclude. For example, if we want to understand the shared political culture of a particular group, we do not provide data about the shoe size of its members (unless we are constructing a very peculiar account indeed); if we want to concentrate on understanding membership of a particular culture, we abstract from all sorts of details about the individual members and about the other cultures or groups they may belong to. Any account necessarily leaves something (most things) out and is therefore an abstract account. Universalist and particularist political theories should therefore be regarded as arguing not over whether abstraction is appropriate but rather over what level of abstraction is appropriate, about which features it is appropriate to abstract from or whether a particular account abstracts or idealizes. O’Neill argues that any attempt to generalize one of the alternative conceptions of practical reasoning outlined above will inevitably be based on a process of idealization. It will not take the world as it is and people as we find them. Instead, it will enhance or idealize some of its starting points and so sacrifice its practicality. It will result in utopian rather than practical reasoning.17 The problem here is not with each alternative account of practical reason as such but rather with the claim that any of these can fill the role of a general account of what it is to reason practically. Each alternative conception may form a plausible account of practical reasoning within a restricted domain. For those who find the requisite metaphysical claims convincing, Platonist reasoning will generate appropriate reasons. For those who share the same subjective ends, instrumental accounts will be convincing. Finally, for those who participate in the relevant practices and therefore share the relevant norms, particularist reasoning will do the job. Within their boundaries, each conception of reasoning may be appropriate. ‘Different stretches of (practical) reasoning may be aimed at or relevant for distinct and different audiences.’18 However, outside of these domains, where the respective conditional predicates no longer hold, we have

Primary constructivism and O’Neill

87

seen that the reasons offered are inapplicable and hence incomprehensible as reasons. Precisely, because each alternative account presupposes the satisfaction of certain conditional predicates, none is capable of standing alone as a general account of practical reasoning. Each account is bounded, limited or local as each is effectively a form of secondary construction; each is ‘insider’s reasoning’ taking certain assumptions as given and having authority only over the limited scope where those assumptions apply.19 Regarding any one of these conceptions as general fails to take proper account of what Rawls calls ‘the fact of pluralism’. Importantly, O’Neill also rules out the possibility that a general account of practical reason may simply be an account that lists the various conceptions of reason appropriate to each restricted domain. A general account cannot be simply the sum of all particular accounts; ‘not all practical reasoning can have such restricted scope’.20 At least some practical reasoning will have to have a more inclusive scope. First, this must be so in order to make sense of a life from the inside. Anyone will participate in several restricted tracts of reasoning, each associated with the different practices, societies and cultures of which they are a member. These several restricted domains of reason will have to be connected internally in some way; they cannot remain wholly isolated. What counts as a reason in a part of one’s life will, at some point, have to be related to the various things that count as reasons in the other parts of one’s life. A more inclusive and general account of practical reasoning is necessary to link these various reasons and to perform an internal job of justification. Second, and as we shall take up in much more detail below, an inclusive and general account of practical reasoning is called for by considerations of justification and practical reasoning in our relations with others. Recall that O’Neill argues that for reasons to be properly practical, they must be at least potentially followable by all others for whom it is to count as reasoning. It is part of her claim that the boundaries of relevance and therefore limits to the scope of our reasons and justifications are difficult to establish. As such, we need at least a formal account of reasoning practically that aims at potential generality. As O’Neill might put it, ‘not all reasoning can be insider’s reasoning’, talking to some and excluding others.21 O’Neill’s constructivism is an attempt to understand what would be required by an ‘inclusive’ account of practical reasoning. Constructivist practical reasoning So far, O’Neill has argued that a plausible conception of practical reasoning will have to pass an elementary standard of practicality. It is a necessary condition of the practicality of reason that it be reasoning that can be followable by all the relevant others. The three accounts of practical reason offered so far struggle with this elementary standard, excluding some others from the scope of reasoning. In their place, O’Neill offers a ‘Kantian’ or ‘critical’ conception and claims that it provides an inclusive account of reasoning. This critical conception of practical reason takes the necessary elementary standard of practicality as its

88

Primary constructivism and O’Neill

only standard. That is, it takes the necessary standard also to be a sufficient standard of practical rationality. This critical conception concentrates on the requirements and implications of making sure that the reasons we offer each other are followable by all relevant others. O’Neill claims that this critical conception draws our attention to the ‘doubly modal structure of practical reasoning’.22 This conception ‘demands that practical reasoning follow principles that are thought of as adoptable and followable by all for whom it is to count as reasoning’.23 The first modal aspect of this conception is that reason sets requirements. It sets up standards that must be met if something is to count as reasoned. Not just any statement is reasoned or reasonable. The second modal aspect is that reasons must be followable or adoptable by others. O’Neill breaks down the ‘followability’ of reasons by others into two distinct aspects. First, practical reasoning must be ‘followable in thought’ by the relevant others. That is, practical reasoning must be reasoning that can be intelligible to those others. Second, practical reasoning must be ‘followable in action’. Practical reasoning aims to guide action, but to ensure its practicality, the required actions must be real possibilities for those others to whom the reasons are offered. The relevant others must be capable of adopting the recommended course of action.24 These two elements of the followability of reasoning, intelligibility and adoptability, ensure that practical reason is restricted to taking the world as it is and people as we find them. This in turn necessitates that we reason using only properly available materials. This is what the process of abstraction is designed to ensure. Criteria of followability demand that we make no idealizing assumptions about the rationality or capabilities of those whom we address or the context in which they find themselves. We must ensure that any empirical premises we rely on are genuinely available and that they ‘abstract from mundane realities but cannot legitimately include premises that idealize these realities’.25 Assuming, for example, that resources are plentiful when they are not, that everyone is ideally rational when it is clear that no one is or that everyone has certain beliefs or desires when they do not will constitute (perhaps dangerous) idealizations. Reasons based on idealizations will necessarily be partial reasons. Properly abstract assumptions, since they simply leave out aspects of persons or contexts rather than adding or enhancing aspects, will not be partial. At worst, they can be incomplete. We cannot offer reasons that rely on partial or fantastical claims about matters of fact, states of affairs or consequences of action; such premises would make those reasons unintelligible as good reasons to those for whom such claims were non-sensical. As such, this critical account of practical reasoning instructs us that to be practical is ‘only to reason with all possible solidity from available beginnings, using available and followable methods to reach attainable and sustainable conclusions for relevant audiences’.26 It is clear that, despite the different language used, O’Neill’s concern with idealization is very similar to the constructivist concern, explored in the previous chapters on Rawls, with partiality. Idealized accounts and assumptions appear to be partial accounts and assumptions and exactly the sort of thing that constructivism aims to avoid. Recall that the constructivist account

Primary constructivism and O’Neill

89

of objectivity in reasons and principles depended on continually guarding against partiality or subjectivity. Reasons and principles that we regard as objective are precisely those that we are most confident do not rely on partiality in their justification. Rejecting idealization in favour of abstraction can be regarded as simply another way of expressing the constructivist attitude towards impartiality and objectivity. In suggesting that some accounts of practical reasoning are based on idealizing premises, O’Neill is highlighting their partiality and subjectivity. That they are not followable by all others is a consequence of their lack of objectivity. Only an account of practical reasoning that is not partial, therefore based only on abstract assumptions, can offer objective and general reasons and so meet the requirement of followability. Ensuring the necessary intelligibility of practical reasons is a task that falls firmly on the reasoner.27 O’Neill tells us that if our reasoning is to be practical, it is up to us to do what we can to satisfy ourselves of its intelligibility to others. This can only be done through a process of critically reflecting on the reasons we offer and expect others to accept. Through critical reflection, it is possible to uncover the conscious and unconscious processes of idealization that may underpin our reasoning.28 Uncritical reasoning is bound to make some illegitimate assumptions whether they are intended or not. As we have seen, it is these idealizing assumptions that threaten the intelligibility, followability and therefore the practicality of our reasoning. Reflecting critically on our reasoning is therefore a sensible precaution to take against unintelligibility. O’Neill is broadly critical of what she takes Rawls’s notion of reflective equilibrium to be.29 However, the conception of reflective equilibrium as a process of critical reasoning explored in the previous chapters seems peculiarly similar to the process O’Neill describes here. It is critical reflection on the assumptions and intuitions on which our reasons are based that will highlight whether they are idealizing assumptions or legitimate abstractions. Just as reflective equilibrium is (temporarily) reached when we are (provisionally) satisfied that our reasoning is not infected with partiality so O’Neill’s critical reasoning (temporarily) stops when we are (provisionally) satisfied that we rely on no partial idealizations. Reflective equilibrium effectively describes the state of our reasoning when we are satisfied with its followability, impartiality and objectivity. On the role of critical reasoning in undermining partiality and idealization, Rawls and O’Neill stand firmly together. O’Neill’s criterion of adoptability is similar to but distinct from Rawls’s criterion of acceptability. Rawls’s notion of acceptability turns on what others would, will or already do accept in order to secure justification.30 On the other hand, O’Neill’s account does not require the relevant others to actually adopt the recommended course of action or even that they actually understand or adopt the reasons that are offered to them in order to count as instances of practical reason. Instead, all that is required is that the reasons offered could realistically be understood and adopted by those others. To require anything else would make reasoning hostage to the attentiveness and willingness or stubbornness of any of those others. Lack of attention or interest when the reasons are explained, or

90

Primary constructivism and O’Neill

failure to adopt the reasons once understood, does not affect the practicality of the reasons or indeed their reasoned nature. Practical reason does not require all to adopt the reasons offered but only stresses that reasons should be, in principle, capable of such adoption. Practical reasoning offers principles which could coherently be adopted by all.31 Thus far, O’Neill’s account of practical reason has been largely methodological rather than substantive. She claims that in order to be practical, our reasons must be followable/adoptable by all relevant others. This is a formal universal principle, not a substantive one. The ‘all’ in ‘all relevant others’ is no particular ‘all’. The relevant domain is not yet specified. As such, the ‘all’ states only a formal requirement of universality. This ‘formally universal specification of scope’ waits on the identification of the appropriate constituency of reasoners before we can begin to make this account substantive.32 How we go about identifying the appropriate domains of practical reasoning will be picked up in the next section. However, we can pause here and beneficially consider some broad similarities between Rawls’s and O’Neill’s project. We identified a two-part conception of objectivity in justification built into Rawls’s primary constructivist account. There is an almost identical basic form in O’Neill. First, there is a description of a choice situation. In Rawls, this is the original position where just principles are those that would be chosen by the parties in that position. So far, although this is remedied below, we have examined only the formal outline of O’Neill’s choice situation where justified reasons are those that could be chosen by all relevant actual others. The role of this choice situation is, in both accounts, to understand what principles we have reason to accept as general principles, which principles are not partial and subjective but objective and impartial and thereby give each of us reason to accept them. As good reasons cannot be based on partial, subjective or idealized assumptions, both Rawls’s and O’Neill’s accounts require a second element, a conception of reflective reason that is necessary to subject our reasons to critical examination (we might call it a critique of reasons). This examination is the ongoing process of reassuring ourselves of the non-idealizing, impartial nature of our reasons. This ongoing critique ensures that the nature and outcome of the choice situation, for both Rawls’s and O’Neill’s, can only ever be regarded as always provisional. Also, as on both accounts all our assumptions are subject to critique and all our conclusions are provisional, neither can appeal to any independent moral fact to ground their objectivity. Rather, objectivity, for O’Neill as well as Rawls, is a matter of guarding against partiality. O’Neill’s closer focus on what goes on in practical reasoning does sharpen our understanding. Her distinction between idealization and abstraction, and her use of this distinction to critique alternative accounts of practical reasoning, serves to draw out and illuminate the ways in which our reasoning may be partial. At least, at first glance, Rawls and O’Neill share the basics of a primary constructivist understanding of justification. Of course, Rawls goes much further than this and links his conception of practical reasoning with conceptions of

Primary constructivism and O’Neill

91

society and person when laying out his choice situation. This enables him to draw substantive conclusions from his universal constructivist argument. As we shall see, although again the language and the presentation is very different, O’Neill argues for substantive and universal principles of justice in a surprisingly similar fashion.

The scope of practical reason O’Neill’s argument has so far identified a ‘formally universal specification of scope’ as all others within the relevant domain of discourse. As noted above, this is not a substantive specification but a purely formal one. However, O’Neill, as a practical philosopher, is concerned to draw substantive conclusions; she is looking for principles to guide action, and in order to accomplish this, the formal account must be fleshed out. However, before any substantive conclusions can be drawn, we need to fix the scope of our reason. The formal account of scope alone is not enough to enable O’Neill to generate any substantive content. The formal account waits upon identification of the appropriate constituency of reasoners. This is no simple task. Given the acknowledged pluralism of cultures and societies, it is possible that there could be many such constituencies. From the particularist perspective adopted by communitarians and many multicultural theorists the realm of relevant others is coterminous with the boundaries of culture; so the scope of reasoning is regarded as local. This perspective would multiply the number of different domains of practical reason, and so different substantive accounts will be appropriate in different contexts. This understanding of the scope of reasons underpins a picture where a formal and primary account of practical reasoning is the basis for a plurality of substantive secondary constructions, each applying to a different context. This entails that, despite the formal account of practical reasoning, there would be no universal substantive moral principles. This is not O’Neill’s position as she is clear that the scope of ethical discourse is not entirely constrained by pluralism. Her argument goes through several steps, the first of which was the identification of the formally universalist account of reasoning (that reasons apply to all relevant others). The next step, undertaken in the rest of this section, will involve demonstrating that this formal account can have cosmopolitan scope (that at least sometimes all others are relevant). Only then can we go on, in the remainder of the chapter, to explore the possibilities for generating substantive reasons from this formal account, whether these are universal or particular. Fixing scope for primary constructivism Fixing the scope of reasoning is necessary, as this fixes the scope of practical ethical concern and, more specifically, answers the question of whom we must regard as having ethical standing. In the particularist position outlined above, ethical standing is attributed to those with whom we share membership of a particular domain of reasoning. In this case, the boundaries of ethical concern

92

Primary constructivism and O’Neill

follow the borders of cultural membership. O’Neill is concerned to show that, at least sometimes, the scope of ethical concern must be broader than this. Hers is the cosmopolitan claim that all persons have ethical standing. If convincing, this will significantly advance our understanding of the constructivist project as this is one of the key points at which constructivism of the sort that we are concerned with distinguishes itself from more communitarian or particularist positions. As constructivism is to deal with the world as it is and people as we find them, it must deal with the problem of scope as a practical problem. This approach must regard questions of standing and concern as ‘questions that arise for and must be addressed by particular agents who need to determine to which other beings they must accord the standing of agent’.33 Instead of seeking a general and a priori theoretical criterion to fix scope, O’Neill’s constructivism asks us to consider whom we must acknowledge as an agent whenever we act. Here, we must consider the many and varied, conscious and unconscious, assumptions upon which our activities are based. In acting, we all of us make a great many assumptions about many other people: that they can understand what we are doing, that they will not interfere, that they are capable of responding in an appropriate manner or that they may judge us worthy of praise or blame, for example. All actions commit us to assumptions of some kind about various others. Many of these assumptions are arbitrary, but some are less so and some are ‘disciplined by the demands of acting and responding effectively in the world’.34 O’Neill’s contention is that ‘three rather abstract and deeply interconnected aspects of the countless assumptions which structure all activity are particularly relevant for fixing the appropriate scope of ethical consideration’.35 These are the assumption of plurality; that there are others; connection, that these others are connected to the agent in that either can act on the other; and finitude, that those others have limited but determinate powers.36 Whether or not an agent is conscious of them, these three abstract assumptions are presupposed by (almost) all activity.37 Our activities presuppose plurality, the existence of some other agents. If we act on the assumption that there are others who will or may respond, or whose own actions may be affected, then we assume plurality. This assumption is factored into our reasons for action and the expected consequences of such action. Several things must be borne in mind here. First, the assumption of plurality may or may not be a conscious one. We may explicitly assume the existence of others, for example, when we advertize a job vacancy. Alternatively, this assumption may be unconscious as it often is when we drop litter or turn on our light switch. Here, we assume that someone will clean up after us or is working to produce electricity. Second, action can assume plurality even when it actually affects no one and impinges on nobody’s consciousness. For example, if we are stranded on a desert island and send a message in a bottle, or, more mundanely, have locked ourselves in the bathroom and cry for help, we assume plurality. Even if the bottle sinks without trace or our cries go unheard, our actions were premised on the possibility that some one could come to our aid. Even the

Primary constructivism and O’Neill

93

hermit assumes that others will leave him alone in his pursuit of solitude (and if he actively conceives of himself as a hermit, then he is implicitly contrasting his lifestyle with the more social relations of others). O’Neill claims that as our activity is based on this assumption of others who can act and react as agents, ‘the standing of those others cannot coherently be denied’.38 We cannot both assume their relevance in our practical reasoning (when we assume that they act and react) and at the same time deny their relevance to our reasoning (in claiming that they are not amongst those to whom we owe a reason, the domain of ‘relevant others’). The second presupposition of activity that O’Neill draws our attention to is connection. If we assume a plurality of others when we act, we also assume that we are connected to them in some way. This manifests itself as the assumption that our activity will affect them or that their activity will affect us. If we premise our actions on the assumption that others will pay attention or respond in some way, then we implicitly assume that we are connected to them. Again, activity and practical reasoning that assumes the relevance of that connection to others (in the ways we can affect each other) cannot deny the relevance of that connection to our reasoning when we delineate the domain of ‘relevant others’. If we premise activity on the assumption of our connection to a plurality of other agents, then we cannot deny the relevance of those agents when we deliberate about that activity and about to whom we should attribute ethical standing. O’Neill recognizes that there may be instances where assumptions of connection are not made and ethical standing is not and need not be attributed. This, however, is where there is no possibility of interaction and no possible causal connection that links the two parties. She uses the example of the Anglo-Saxons in England and their contemporaries, the T’ang Chinese. Neither group needs to attribute ethical standing to the other when reasoning practically, since, for all intents and purposes, they lived in separate worlds. There is no practical sense in which these two groups could be said to be connected, to share a world and therefore to owe each other reasons.39 This sort of case is however of limited applicability. Today, we can sensibly be said to ‘share a world’ with a great many people.40 When we act, we often assume the existence of connections to a greater plurality than we realize. When we post a letter, for example, we assume the agency not only of the addressee but also of the many people involved in the collection and distribution of mail, in the manufacture, distribution and sale of the paper, pen, envelope and stamp, and in the maintenance of the postbox and delivery vans, of the taxpayers who subsidize the postal service and of the administrators who organize this vast undertaking. Almost any one of our actions, such as buying a packet of coffee, publishing an article, making a telephone call or driving a car, assumes the existence and agency of far more people than we can even begin to list. Activity usually assumes, often unconsciously, the existence of a vast and intricate network of agents and their various interactions, and it also assumes our connection to that network. As our activity relies to some extent on the agency of these others, we are given good reason to take them into account in any deliberation about the domain of ethical relevance.

94

Primary constructivism and O’Neill

The assumption of plurality and connection also involves the assumption of the finitude of ourselves and others. Our activity is premised on an understanding not only of what others are capable but also of what they are incapable. Inherent in action are assumptions about our capacities and vulnerabilities and about the capacities and vulnerabilities of those with whom we are connected. When reasoning about ethical standing, we cannot then consistently replace these accounts of finitude with idealized accounts of ‘rational choosers’, ‘preference satisfaction maximizers’ or ‘economic man’.41 In attempting to reason practically, we try to make reasonably accurate assumptions about the powers and weaknesses of ourselves and those around us (taking people as we find them). To do otherwise would be to invite failure. Again, if our actions are premised on these assumptions about others, we have reason to include such considerations in any reasoning about the relevance of those others. This is a common theme here. For O’Neill, ‘activity speaks louder than words, and displays the assumptions actually made about others more accurately and more honestly’ than our vocal utterances often do.42 She goes on to argue that when considering the ethical standing of others, ‘What is assumed for the purposes of activity must also be assumed in fixing the scope of ethical consideration.’43 To do otherwise would be plainly and practically inconsistent. We cannot both premise an activity on an assumption and deny that assumption in justifying our actions. ‘Whenever . . . [our] activity assumes a plurality of finite and connected others, [we] . . . are also committed to including those others within the scope of [our] . . . ethical consideration.’44 It is by utilizing these considerations that O’Neill closes in on a constructivist account of scope. We have already noted our connection to vast and intricate networks of interaction. As we saw with the example of posting a letter, and as should be obvious if we think about buying a packet of third-world coffee from the supermarket or driving a petrol-fuelled car, these networks of connection do actually stretch globally in almost any of our activities. Beyond this, O’Neill is at pains to point out that it is not actual interaction but the possibility of interaction that counts here. If there is a real possibility of action by one party bearing upon another, then those parties are connected and should be viewed as being so.45 This is so, as in the contemporary world, any activity, no matter how local or trivial it seems, will be premised on fairly strong assumptions about even the most distant strangers. We assume, for example, that these strangers will not attack us or interfere with our activity (even an assumption of their inactivity is an assumption about others). We also assume that the relevant governments will prevent their settling in our part of the world and impinging on us more directly.46 There are also more abstract, but no less important, considerations of uncertainty that must be taken into account. We cannot with any certainty know the limits of the effects of our actions. All actions have unintended consequences and may affect many people in ways in which we had not planned and in ways which it is inconceivable that we could have foreseen. There is ongoing uncertainty over the boundary of effect for any action. As a consequence, it is hard, in

Primary constructivism and O’Neill

95

principle, to fix an outer limit to the scope of our connections. The boundaries of relevance are uncertain and fluid, although this is not to deny that we are obviously connected more closely to some than to others. Still, that wider connection is always and inescapably present. A constructivist account of scope is concerned to outline a method by which agents can guide their thought about scope and ethical standing when they act.47 O’Neill demonstrates that we are given reason to think that today we ‘share a world’ in the relevant way with just about everyone. As such, our activities are premised on some assumptions about all those others. As a point of consistency between the premises of our action and deliberation about that action, we are therefore given some reason to take some account of everyone in our ethical deliberations. Considerations of the basic assumptions upon which we premise our activity leads to a formally cosmopolitan account of scope. It is a formal account in that it does not provide a strict list of who counts and for how much. It only gives us reason to think that everyone counts for something, that everyone should be accorded some ethical standing.48 O’Neill has, as yet, constructed an argument only for this formal cosmopolitan account of scope – that everyone counts for something and that all others are relevant. So far, we can draw no solid conclusion about priority rules that would privilege one level of discourse over another, the cosmopolitan over the particular, for example, or vice versa.49 All we have shown is that local and particular reasoning assumes connection to a broader plurality and, as such, gives us reasons to think that everyone counts for something in our deliberations. In the last section, we outlined the first step of O’Neill’s constructivist account that practical reasoning must be universal in the realm of ‘relevant others’. We are now able to conclude that, and this is the second step, at least sometimes the scope of ‘all relevant others’ is both universal and cosmopolitan. We have gone beyond an account of scope as formally universal to an account of scope as formally universal and cosmopolitan. How to construct a substantive primary construction The next step in O’Neill’s constructivism is to move from the formally cosmopolitan account of practical reasoning to draw out substantive conclusions in the form of principles to guide our practice. As noted above, the obvious route to substantive conclusions would appear to rely on the particular values appropriate to particular domains of discourse. To the extent that we can talk of a society or culture having values, it may be useful to regard them, as Rawls does, as assuming and embodying certain conceptions of society and person, differing in substance as societies and persons differ. It is difficult to imagine moral or political rules of justice that do not make some such assumptions. As was also noted, this is likely to lead to the formal construction being instantiated in a variety of substantive secondary constructions. Such a secondary construction might be explicit as Rawls’s is in Political Liberalism. Alternatively, the construction may simply manifest in the habitual acceptance of a traditional culture

96

Primary constructivism and O’Neill

or in a culture based on the assumption of a particular religion. These constructions are secondary in that they are based upon (what we might loosely term) acceptance of certain partial ideas or values that appear unavailable or intelligible to others. Whilst within their relevant domains modes of reasoning based upon particular conceptions of the person or society may be unproblematic, such conceptions are an unsuitable basis for a primary construction. We must recall that a primary construction must hold all its premises up to open-ended critical examination by all and therefore to potential revision. A mode of reasoning that accepted certain foundations uncritically as beyond possible revision cannot be the basis for a properly primary and substantive construction. It may be that this is an argument for the impossibility of a substantive primary constructivism. However, O’Neill does not think that this ends the argument, and it is useful to follow her reasoning. If the obvious route to substantive primary conclusions is thus closed to us because of the partial, and to that extent contingent, nature of the particular assumptions upon which it relies, then the alternative that remains is to examine the possibility of working not with particular assumptions but with broader and more abstract assumptions. If we can regard the particular assumptions of each society or culture as embodying claims about the appropriate conceptions of society and person, as Rawls does, then it may make sense to cast these broader and more abstract assumptions in a similar role to the bare concepts of society and person that we uncovered in Rawls. It will become clear that O’Neill reasons in a comparable manner. Indeed, O’Neill is explicit about the ‘bare’ nature of her starting points, and as we shall see, she is not at all shy about drawing substantial conclusions from them. Again, as developed in our reading of Rawls, we find the basic constructivist account whereby practical reasoning is oriented by bare concepts of society and person in order to produce substantive and universal principles. O’Neill’s most explicit use of these bare concepts is in Constructions of Reason. In identifying these concepts, O’Neill is first and foremost concerned only to legitimately abstract from particular societies and accounts of persons and to avoid idealizing accounts altogether. Here, the bare concepts of society and person are drawn from a minimal construal of the circumstances of justice. First, justice is only a concern where there is a ‘plurality of at least potentially interacting agents’.50 This is the most minimal construal possible of the fundamental idea of ‘society’. It presupposes no ‘pre-established harmony’ of interests between the members of the plurality or any common commitment to a value or set of values. It is difficult to see that the notion of society could be drawn any thinner. Where there is no possibility of interaction between several agents, then there can be no problem of justice. Situations where there is only one agent, or where agents cannot interact, are outside the circumstances of justice. It is also difficult to see that any particular conception of society could convincingly claim not to presuppose at least this, although almost all would presuppose a much ‘thicker’ account. After all, a society implies an interacting plurality, but any particular society is based on much more than this. However,

Primary constructivism and O’Neill

97

in this sense, O’Neill is able to provisionally advance this as a plausible candidate for the bare and abstract concept of society. Likewise, O’Neill also proposes a candidate for the bare and abstract concept of the person. This consists of two parts. First, ‘a meager and indeterminate view of rationality’. On this interpretation, persons are credited only with ‘the capacity to understand and follow some [or other] form of social life and with a commitment to seek some [or other] means to any ends . . . to which they are [happen to be] committed’.51 At first glance, an agent’s capacity to follow a social life and the ability to pursue whatever ends they have closely resembles Rawls’s concept of the person as possessing the two powers of reasonableness and rationality. Second, ‘a meager and indeterminate view of the identity and of the mutual independence of agents’. This assumes only that ‘agents have capacities for varying sorts and degrees of dependence and interdependence’.52 Again, this would appear to be a minimal construal of the concept of the person. What could a person be if incapable of participating in some sort of social life, of having some ends, of having some needs and of at least some degree of independence? No assumptions are made here about what sort of social life or what ends, needs or degree of independence is either necessary or normal. Again, it is difficult to see how any conception of the person could avoid presupposing this general concept of the person, although, once again, they are very likely to make much thicker presuppositions. As such, this account stakes at least a plausible claim about the content of a bare concept of the person. As is always the case in primary constructivist discussions, these accounts of the bare concepts are advanced only as plausible, not as established. They are not regarded as unrevisable foundations but as themselves open to the possibility of revision if necessary in the face of further critical reflection on the two fundamental ideas. This could work in either direction. Perhaps this account makes hitherto unnoticed idealizing assumptions and should therefore be thinner; perhaps, it abstracts further than necessary, and a thicker account would equally avoid idealization. Whether either criticism is accurate could only be explored with critical practical reason. Either way, the provisional accounts of these concepts may be adjusted to remove idealization and partiality until reflective equilibrium is attained. Having outlined them, we find that O’Neill uses these concepts consistently throughout her work and to good effect.53 These bare concepts of society as uncoordinated others and of persons as indeterminate in their independence, neediness and rationality play their important roles also in Towards Justice and Virtue. Here, we find that the bare and abstract concepts of society and person that form the basis of a sparse account of the circumstances of justice are equivalent to the basic assumptions of any activity. The fundamental idea of society as uncoordinated but potentially interacting with others is represented by the ubiquitous assumption in action of plurality and of connection that there are others and that our actions may affect each other. The fundamental idea of the person, as indeterminate in its attributes but as being to some degree needy, rational and dependent, is represented by our similar assumption in action of the

98

Primary constructivism and O’Neill

finitude of both ourselves and others. O’Neill also describes the ‘weak view of the circumstances of justice’ as holding when a ‘plurality of finite agents share a world’ in ‘Justice and boundaries’.54 Here, she explicitly links the argument from the circumstances of justice used in Constructions of Reason with that based on the fundamental assumptions of action in Towards Justice and Virtue. Since the bare concepts of the fundamental ideas are the same as the basic assumptions of plurality, connection and finitude, they are assumed in almost all our actions (and certainly in all those that involve, rely on or affect other people). This clarifies what O’Neill means when she says that our basic assumptions in action are ‘disciplined by the demands of acting and responding effectively in the world’.55 These assumptions are simply an acknowledgement of the circumstances of justice and the constraints on practical and reasonable action that they impose. The point of drawing this conclusion is that the way forward is now clear for the attempt to construct substantive conclusions from the formal primary constructivism. Since the bare concepts of the fundamental ideas are a part of the basic assumptions underlying any significant action, if any substantive conclusion could be drawn from these bare concepts linked with the formal account of practical reason, we would have at least a prima facie reason to regard the substantive conclusion as an important one, a reasonable response to the circumstances of justice and action. This account is the basis of O’Neill’s claim to construct universal and substantive conclusions about just principles for action. Exploring her account of this construction will complete the next step of O’Neill’s argument and will be a further step on our own path towards a fuller understanding of constructivism more generally.

Constructing substantive primary principles When it comes to the procedure for constructing substantive and universal principles, it would be difficult, on the basis of the interpretations outlined above, not to notice the similarities between Rawls and O’Neill. Recall that Rawls’s project is to construct a political conception that ‘shows how the principles of justice follow from the principles of practical reason in union with conceptions of society and person’.56 He couples an outline of bare accounts of society and person that any conception of justice or the good must assume with a universal account of critical practical reasoning built on the idea of reflective equilibrium. On this basis, Rawls lays out the original position as a choice situation that identifies universal principles of justice. It is clear that O’Neill, in her minimal account of the circumstances of justice, similarly identifies bare concepts of society and the person that she argues must be underlying assumptions of any activity. The basic concept of the person is of an agent of indeterminate rationality and varying capacities for independence. The basic concept of society is of a plurality of potentially interacting agents with no necessary pre-established harmony. Any form of activity, social life or conception of justice must assume at least these abstract accounts of society and person. O’Neill also couples these conceptions with an account of practical reasoning in which authoritative prac-

Primary constructivism and O’Neill

99

tical reasons are those that are followable and adoptable by all relevant others and with the secondary claim that, at least sometimes, all others are relevant. O’Neill, like Rawls, has outlined a choice position as a thought experiment that enables us to identify justifiable principles to guide and constrain our actions. Justified principles will be those based on reasons that could be followed and adopted in a situation laid out by reference to the bare accounts of society and person. Put simply, we ask, ‘What principles can a plurality of agents of minimal rationality and indeterminate capacities for mutual independence live by?’57 As we shall see below, the choice situation is a form of universalizability test, and O’Neill argues that this very limited requirement manages to identify a thin but universally applicable set of constraints of justice and of virtue. Constructing justice Before O’Neill is willing to make progress in constructing the principles of justice, she addresses the structural question of why there are theoretical advantages in a focus on obligations rather than rights.58 A first step towards showing this involves drawing out a distinction between perfect and imperfect obligations. Perfect obligations are those for which there are correlative rights that tell us exactly to whom the obligation is owed, and in these cases, there is a prefect correspondence between obligation and rights. There are universal perfect obligations that are held by all and owed to all others. Obligations of this type tell us not only who is obliged to act/refrain from acting in a certain way (everyone) but also exactly to whom the obligation is owed (everyone else). It is this that constitutes the perfection of the obligation; all relevant parties are identified by the obligation itself.59 There are also special prefect obligations that are less than universal as they depend on a special relationship between people who imposes obligatory ways of behaving towards each other. The relationship may involve the mutual membership of a family, club or society or perhaps identify a relationship of responsibility such as one that exists between teacher and pupil or ruler and ruled. Again, these obligations are held by a specific person and owed to specified others who can be thought of as having a right to the performance of the obligatory action.60 Perfect obligations, universal or special, involve corresponding rights, whether it is a right not to be killed by anyone or a right to a pay packet from your employer at the end of each week. Imperfect obligations, on the other hand, do not identify specific recipients and so cannot entail correlative rights. A clear example of an imperfect obligation is an obligation to charity. We may feel that we are under an obligation to give charitably, but it is clear that no particular needy person has a right to our charity. If we regard rights as fundamental in the regulation and justification of action, then all obligations will be perfect. Justified rights will identify correlative perfect obligations. However, such rights-talk will be blind to imperfect obligations as they are not identified by any corresponding right. Imperfect obligations would drop out of the picture altogether. On the other hand, if we regard obligations as fundamental, then the resources associated with both perfect and imperfect

100

Primary constructivism and O’Neill

obligations are available. O’Neill’s initial claim is that the scope of obligations is broader than rights. Her second claim is that the construction of obligations will be more successful than, and superior to, the construction of rights. The superiority of obligations is demonstrated by two additional arguments, an argument from indeterminacy and an argument from human neediness. First, theories of rights suffer from a crippling indeterminacy and are therefore inadequate guides for moral or political action.61 O’Neill argues that theories of rights generally aim to identify an equal set of rights held by each person that constitute a ‘sphere’ within which they are free to act. Of the many sets of rights that meet this requirement, the preferred set is the one that maximizes the size of the ‘sphere’, effectively giving each the greatest possible liberty compatible with a like liberty for all.62 O’Neill contends that a maximal set of rights cannot be identified as this necessarily assumes a metric for liberty which is unavailable. Judging the relative size of sets of rights assumes that we can identify whether one right is bigger than another or whether a certain freedom of speech outweighs further freedom of association. O’Neill does not think that this sort of question is resolvable. It is not that we do not yet possess this metric, but rather that there is no unique method of combining or accommodating rights into an identifiably maximal set. Rights theories are indeterminate, as, lacking the relevant metric, no definitive choice can be made between competing accounts of rights. O’Neill goes further and argues that this failure prevents the identification of any single rights independent of a set of rights as individual rights are dependent for their justification on their being part of the maximal set.63 Theories of rights are crippled by indeterminacy. If we can construct a determinate account of obligations, then this will be superior to any account of rights. The second argument against rights draws on the unavoidable recognition that the finitude of human beings is far-reaching and deep, and the human condition is characterized by vulnerability and neediness.64 Recognizing the limited and vulnerable nature of human capacities involves understanding that, at least some of the time, our projects will fail without assistance and that there may be times when the help and care of others is a necessary condition of any project or action at all. After all, we all start life as helpless infants, and we are all likely to fall ill at some point. O’Neill claims that theories of rights do not take seriously the deep nature of human vulnerability. Universal neediness cannot be met by a theory of rights as no one can meet a universal demand for help in alleviating need. The same finitude that necessitates aid ensures that nobody can offer more than selective help. More importantly, despite our recognition that we are all needy at some time or other, we cannot realistically know precisely who is, or is about to become, needy at present and so can demand our selective aid as a right. The claim is that, as we cannot specify recipients, we cannot specify a right to aid. Only the notion of an imperfect obligation, where we know we are obliged to alleviate need but no one can claim our aid as a right, can adequately capture the problems of human neediness. As rights theories cannot offer an account of imperfect obligations, they are deficient in comparison with theories of obligation, and O’Neill describes them as ‘perilous’, ‘unnecessary and dam-

Primary constructivism and O’Neill

101

aging’, ‘crippling’, ‘disabling’, ‘unsettling’ and ‘dangerous’.65 Such theories exhibit a ‘poverty’ that obligation-based ones do not.66 In the light of these considerations, we should renounce the ‘rancorous rhetoric of rights’.67 The construction of principles of justice that identify universal obligations is a function of the universalizability test constituted by the choice situation outlined above and framed by the question, ‘What principles can a plurality of agents of minimal rationality and indeterminate capacities for mutual independence live by?’68 Universalizable principles are those that meet O’Neill’s modal constraints on practical reasoning of being principles that can be principles for all relevant others.69 In this situation, since it is characterized by abstract conceptions of person and society that are assumed in any activity, all others are relevant, and so principles that meet this universalizability test will be both universal and genuinely cosmopolitan. Principles that are not rejected in this situation are permitted principles of action for everyone. Principles that cannot be adopted by all fail this basic modal test of reasoning and must be rejected. The systematic and iterative application of this test to proposed principles of action will identify a body of principles whose rejection is obligatory. The language of rejection is important here, and we can characterize O’Neill’s test as a form of ‘negative discipline’.70 This process rules out certain principles of action rather than recommends them. The operative idea here is modal, not metric. Principles must be either rejected or permitted; there is no question of maximization or balancing as there is in the construction of rights. Negative discipline proscribes rather than prescribes, constrains rather than dictates. The test’s function is to guard us from error by pointing to the limits on principles that can be shared as principles for action by all members of a plurality. If a principle is not consistently universalizable, then the test proscribes action on those terms. Since there are an indefinite number of principles of action that could be submitted to this test, and each is potentially a candidate for obligatory rejection, then the list of obligations will probably never be complete; the process of iteration is necessarily open-ended.71 O’Neill hopes to iteratively pick out single obligations in isolation from questions concerning other obligations. In this way, she attempts to avoid the indeterminacy that infects the attempted identification of maximal rights sets. She also believes that such a method will pick out imperfect obligations that focus on the second problem faced by theories of rights, that of adequately characterizing and addressing the nature of human neediness. O’Neill argues that the principles that are rejected by this test, as not all can follow or adopt them as principles, are primarily principles of injury. A principle is not universalizable if, when universally adopted, it would injure capacities and capabilities for action of some relevant others, consequently assuring that the injured parties cannot adopt those principles. O’Neill goes so far as to argue that ‘the core of injustice is action on principles which if universally acted on would foreseeably injure at least some; the core of justice is rejection of principles whose attempted universal adoption would foreseeably injure at least some’.72 A plurality of rational and vulnerable agents cannot consistently follow

102

Primary constructivism and O’Neill

any principle that destroys an agent’s capacity to act on principles at all. A plurality cannot adopt principles whose aim is to destroy, undercut or erode the agency of some of its members. It is not that the victims of action on these principles do not act on the principles of their victimizers; rather it is that they cannot do so. The victims of such principles cannot share the very principles that ‘destroy or limit their very capacity to act on principles’, and as principles that cannot be acted on by all must be rejected, so the rejection of principles of injury is obligatory.73 The main examples that O’Neill provides of such principles of injury, and of their necessary rejection, are principles of violence, coercion and deception. Violence and coercion directly injure their victims. Violence is straightforwardly non-universalizable.74 The widespread adoption of a principle of violence would necessarily injure some and undermine their ability to act on the same principle. Likewise, coercion depends on (at least temporarily) preventing some persons from acting at all. The coerced, who cannot act, cannot make coercion the principle of their action, and so coercion cannot be universally followed. Deception, although a form of indirect injury, is injury nonetheless as it creates or extends vulnerabilities and needs and undermines the trust that informs the social fabric necessary if these needs are to be met. Principles of deception, exemplified in lying, defrauding and cheating, are also straightforwardly non-univerzalizable. Deception depends on a general fabric of trust which, if deception were adopted as a universal principle, would soon cease to exist. Deception can only work if at least some others make trust rather than deception their principle. If universally adopted, deception would be impossible. Each of these principles, violence, coercion and deception can be followed only on the assumption that at least some others do not follow them. These principles depend on the assumption that adoption is not universal. Successful action on these principles depends on victimization, on ‘destroying, paralyzing or undercutting . . . [the] capacities for agency’, of at least some people. Consequently, these principles are available to some only on the condition that others do not successfully adopt them. As they cannot be adopted by a plurality of agents, then their rejection is obligatory.75 Finally, O’Neill adds that we must also reject, as examples of indirect injury, principles of damage to the environmental fabric that forms the material basis of life on the basis of arguments that parallel those concerned with the damage to the social fabric perpetrated by deception.76 O’Neill is clear that these obligations of justice are universal and perfect, but this entails that, despite her earlier protestations that rights are indeterminate, her universalizability test identifies rights as well as obligations. As the perfection of obligations lies in the existence of corresponding rights, any method of identifying universal perfect obligations also effectively constitutes a method of identifying universal rights. Universal perfect obligations must be adhered to in all situations and in all our interactions with others. If our observance of these obligations must be universal, then all those with whom we interact can quite rightly expect us to treat them in a manner consistent with those obligations. The way is clear for us to regard treatment in this manner as a right for those with whom we interact. Consequently, we can regard each of us as having a justifi-

Primary constructivism and O’Neill

103

able claim against all others that we not be the victims of action based on forbidden principles. If the universalizability test rejects a principle as universally forbidden in justice, then this establishes a right for everyone, a universal right, that no one behave in such a way. O’Neill explicitly states that ‘justice is a matter of perfect obligations, matched by rights; its demands fall on all, and are owed to all’.77 These are the rights that O’Neill’s earlier analysis noted are always correlative to universal perfect obligations. If her universalizability test can iteratively establish the existence of some obligations, then it can also iteratively establish some rights. In direct contradiction to O’Neill’s indeterminacy argument, some rights will be determinate and identifiable, specifically rights against killing and injury at least. Just as universal perfect obligations are iteratively identified, so are the universal rights that go with them. This effectively removes one of O’Neill’s two main arguments against theories of rights; the second will be addressed below. This is not to argue that rights are superior to obligations but simply to suggest that it may not be necessary for primary constructivism to make a decision between the two. At least as far as the primary construction of justice is concerned, there is little to choose between them. Identifying one identifies the other, and O’Neill will have made considerable progress towards a successful primary constructivism if she can justify either. Indeed, it does seem that significant progress has been made. O’Neill starts by considering the authority of practical reasoning and argues that a general account of reasons has to demonstrate that reasons are not arbitrary. If reasons are to have the general authority that we commonly regard them as having, then they must be reasons that all can follow. They must be reasons for everyone rather than partial reasons. Partial, arbitrary or biased reasons count as reasons only for some rather than for all; they are subjective rather than objective, as their authority depends on your standpoint rather than being good reasons when viewed from any position. Objectivity in reasons is pursued by progressively rooting out partiality in the form of implicit idealizing assumptions. Objective reasons are reasons that, having gone through this ongoing reflective process, we are confident are impartial in that they rely only on abstract assumptions rather than idealizations. O’Neill argues that it is possible to identify properly abstract concepts that are assumed by all reasons and so identify some cosmopolitan reasons that apply to everyone. Abstract concepts of the person and society are implicit assumptions of any form of activity or conception of justice or the good. While these concepts can also be expressed as assumptions of plurality, connection and finitude, it is as bare concepts of society (as a potentially interacting plurality without pre-established harmony) and the person (as vulnerable and needy with limited capacities for rationality and independence) that they most easily fit into our constructivist understanding. If justice is an issue at all, these assumptions have to stand. If there is no plurality, if a plurality exists but does not interact or if the members of that plurality are not vulnerable to mutual injury or a range of needs, then we are not addressing circumstances of justice.78 Of the four broad steps in the argument, the first step in O’Neill’s argument is to show that reasons should be universal.79 The second is to show that they can

104

Primary constructivism and O’Neill

be universal and cosmopolitan. The third is to demonstrate that, once these abstract assumptions are expressed in the form of a choice situation that poses the question ‘What principles can this plurality of limited and vulnerable people live by?’ it is possible to identify thin but substantive universal and objective principles of justice. The final step in the argument is to understand the relationship between the universal principles of justice and the global diversity of societies, cultures and communities. O’Neill is clear that there is no deductive argument from the thin universal principles to a ‘right’ set of institutions, activities or practices at the local level or in any given situation.80 Principles of justice provide only a ‘scaffold’ or a ‘framework’ that constrains the variety of diverse societies with the broad limitations of justice such that ‘institutions and practices may vary in many ways’.81 Rather than knock down already existing institutions and societies in order to construct de novo, the principles of justice take on a ‘critical role’. Every society has built up political and social institutions over time, and many of these may already be just built. Indeed, it would be strange if the majority of political institutions were premised on principles of injury. However, some institutions will be unjust, and justice will require that they be remodelled or repaired, not by reference to an abstract blueprint but through local reconstruction sensitive, to some extent, to local judgements about what constitutes unjust injury. Through a process of ongoing repair, institutions can become progressively less unjust.82 The diversity of situation and pluralism of belief found in different times and places do not undermine principles of justice. Principles of justice can be justified quite generally, and can be used to judge the specific constructions of justice which, quite rightly, take different forms at different times and places.83 In this passage, O’Neill has effectively drawn the same distinction between primary and secondary constructivism that we found made good sense when considering Rawls’s constructivism and which we incorporated into our provisional account of constructivism more generally. The universal principles of justice clearly function as a primary constructivism that places limits of justice on the range of diverse particular and local secondary constructions. That the primary construction plays this limiting role for O’Neill is to be expected as it is an apparent consequence of characterizing the constructive process as a negative discipline. The question ‘Can this principle be adopted by a vulnerable plurality?’ can have only one answer, no or yes. If the answer is no, then rejection of that principle is obligatory. If the answer is yes, then the principle is a permitted principle for action. Principles are either permitted or rejected.84 There is no obvious way in which anything but the rejection of certain principles is an obligation. As a wide variety of principles may be permitted, in that they are not rejected, we should expect the adoption of a diverse range of principles, and so a pluralism of institutions and societies should be the expected norm. Thus, the thin primary constructivism limits, but does not dictate, the form of secondary constructions.

Primary constructivism and O’Neill

105

Here, we have reached a position very similar to the position that we drew out of Rawls. We have a distinction between primary and secondary constructivism. We also have an account of a justifiable primary construction. In each case, this account aims to avoid basing its claims to objectivity on independent moral facts. Instead, both Rawls and O’Neill focus on the capacity of critical reason, oriented by abstract concepts of society and person, to identify the impartial and universal requirements of justice. That these two powerful accounts of constructivist justification come together at so many points is important, as this can only reinforce our provisional general account of constructivist justification. Constructing virtue As we have just noted, O’Neill’s negative discipline makes it difficult to understand the justification of positive obligations to perform certain sorts of action rather than the obligation to reject injurious principles. This also makes it difficult to see how the secondary level is ‘filled in’ except by means of diverse local constructions. However, O’Neill seems to be committed to both justifying positive obligations to perform and arguing that their scope is universal. She identifies a range of obligations of virtue that are justified in addition to the obligations of justice already outlined. These obligations are the imperfect obligations that were introduced to show that rights cannot account for the full range of obligations. Recall that obligations of virtue are imperfect obligations as they do not identify appropriate and correlative rights-holders in the way that perfect obligations of justice do.85 Despite this obvious difference, O’Neill is insistent that obligations of virtue can be identified in the same way as obligations of justice by showing that, as they cannot be universalized, certain principles must be rejected.86 Even when no direct or indirect injury is inflicted, a person’s capacity for agency may be undermined. Human beings are persistently vulnerable to an extent that justice does not address. Principles of indifference to or neglect of this pervasive neediness, while not in themselves unjust, are not universalizable.87 Vulnerable human agents are subject to illness and death and need regular nourishment and at times in their lives the care and attention of others. They need these things not merely for some arbitrary purpose but as conditions of becoming and remaining agents. Among such rational and needy beings, a principle of self-centered indifference to others’ needs could not be universally adopted.88 If indifference to the needs of others were adopted as a general principle, O’Neill argues that ‘no one could rely on others’ help; joint projects would tend to fail; vulnerable characters would be undermined; capacities and capabilities that need assistance and nurturing would not emerge; personal relationships would wither; education and cultural life would decline’.89 People with any sort

106

Primary constructivism and O’Neill

of projects at all cannot adopt a general principle of indifference. They can do so only ‘if they were willing to accept that all relationships and action, indeed all life, including their own, were bound to fail’.90 Vulnerable beings must be willing to offer and accept help and so cannot accept universal neglect and indifference. However, rejecting indifference is different to rejecting injury as you cannot but fail to be indifferent to the needs of some people. There are just too many needy persons, and we inevitably neglect a great many of them. Unlike obligations of justice, where it is perfectly possible to refrain from injuring everyone, obligations of virtue are not capable of universal observance. Instead, we must be prepared to support at least some others in some ways whilst accepting that we cannot support all others in all ways. This is why obligations of virtue are imperfect; no person can claim our aid as an enforceable right, and such obligations are therefore better embodied in the characters of agents than in the institutions of the state.91 Just as all obligations of justice are based on the rejection of injury, so all obligations of virtue are rooted in the rejection of indifference to the needs of others. The main social virtues involve a rejection of direct indifference expressed through sympathy, charity and acts of rescue, for example. A second set of social virtues are associated with a rejection of indifference to the social fabric, perhaps expressed through toleration and participation. Finally, there are virtues implied by rejecting indifference to the environmental basis of life expressed in attitudes of cultivation and conservation.92 These are the imperfect obligations of virtue that theories of rights cannot account for. I do not necessarily want to dispute O’Neill’s broad approach to obligations of virtue. After all, the project of identifying a plausible general account of constructivism may not need to address this issue at all. Whether constructivism is better conceived of as justifying obligations or rights or whether all obligations are perfect is, to some extent at least, not so important here. However these questions are settled, it will not detract from the main point; we would have generated substantive universal principles of justice through a recognizably constructivist account of justification. However, it is worth noting that O’Neill appears to move too quickly from the recognition of universal neediness to universal obligations to aid or, as she sometimes puts it, obligations to charity.93 To understand why this is the case, it is necessary to look more closely at exactly how the universalizability test operates on several proposed principles for action. First, we test the principle P – I will kill those I disagree with. When this principle for action is universalized, it becomes uP – everyone will kill those they disagree with. uP is obviously not a viable principle; if some act on it, others cannot (they will be dead). It is not possible for all to act on uP. It undercuts the agency of some, and it is therefore rejected as a just ground for action and action based on it is not permitted.94 Second, we proceed and test the principle Q – I will help those who request help. This is universalized to become uQ – everyone will help those who request help. There can be little doubt that uQ would not undercut the agency of anyone. It

Primary constructivism and O’Neill

107

is therefore a viable and just basis for action and as such would be permitted by this test. We are therefore entitled to act on our inclinations and help everyone that we wish to. Finally, we will test a third principle R – I will never help those who need help. When universalized, this becomes uR – everyone will never help those who need help. The circumstances here are somewhat different. Remembering that we are treating neediness as a necessary presupposition of all human action, O’Neill argues that it would not be possible to universalize this indifference to the needs of others. If uR is permitted, we accept the possibility that our most treasured and valuable projects will probably fail and that agencythreatening needs will not be met. It is not possible for all to adopt a principle by which the agency of some is undercut, and so R is rejected. The rejection of R rules out never helping others. In doing so, it appears to place us under an obligation to provide some help to others. It is the non-universalizability of R that O’Neill seizes upon to ground her imperfect obligation to meet need, and in this context, it is possible to see why an obligation of this sort seems mandated. Is there, after all, no avoiding an obligation to provide aid or charity? This process appears to generate obligatory actions when we test negatively formulated principles. P and Q are both phrased positively so as to announce the intention to perform actions; they involve doing something (killing and helping respectively). When tested, action on these principles is either rejected or permitted. On the other hand, R is formulated negatively; it announces the intention to refrain from acting in a certain way. This poses no problem if the inaction is permitted (as it may be if we tested the principle, ‘I will never break a promise’). However, if, as with R, an inaction is rejected, this appears to place us under an obligation to perform an action. This introduces the possibility of a third verdict issuing from the universalizability test, adding a verdict of obligatory to those of rejected and permitted. The following discussion is not an attempt to provide a general dismissal of obligations justified by the rejection of negative formulations. Instead, I will focus simply upon R and its rejection, upon the possible grounding of an obligation to aid. In doing so, it becomes apparent that the move from rejecting R to an obligation to aid is less straightforward than it appears. The rejection of P is directly expressed as the imperative ‘do not kill others’. In the derivation of a duty to help, O’Neill treats the rejection of R in the same way. When R is rejected, she claims that this can be directly expressed by the imperative ‘you must help others’. R’s rejection does not directly entail the obligation to help. Instead, the rejection of R – ‘I will never help those who need help’ – finds its initial expression in the imperative iR – ‘do not never help others’. O’Neill’s obligation to help must issue from iR. She appears to regard the obligation ‘help others’ as synonymous with iR and so treats that obligation as a direct expression of R’s rejection. It is here that O’Neill’s argument can be pressed as iR can be interpreted in a manner inconsistent with an obligation to aid meaning that iR is not synonymous with O’Neill’s obligation. If this is the case, and iR is the natural expression of the rejection of R, an obligation is produced only indirectly from R via this intermediate stage (iR). As iR lends itself to alternative readings, then there are other indirectly produced interpretations of

108

Primary constructivism and O’Neill

what iR entails that exist alongside, and with the same potential status as, an obligation to charity. These alternative interpretations identify alternative, nonobligatory courses of permitted action as responses to iR. In order for a certain action to be obligatory, we must actually be obliged to act in that way (an obligation is not simply one permitted way of acting amongst others). So, strictly speaking, no obligation to meet need can sensibly be derived from a rejection of R. The essential vulnerability of the human condition mandates that at least some needs be met in one way or another, and brief consideration highlights at least two important, apparently permitted but not obligatory, means by which this is usually done.95 A first thought is that to satisfy the presupposition behind iR – that all people have needs that must be met (do not never help) – it is enough that some help be available towards meeting those needs. O’Neill points us in a promising direction herself with an example drawn from Locke.96 Here, Locke makes it plain that he considers it an injustice if a person refuses to trade grain with starving people. However, he also makes it plain that this person is free to sell his corn at ‘the utmost rate he can get for it’ and that this is ‘no injustice’.97 Locke not only recognizes that people have needs that must be met but also recognizes that this need not disadvantage us. While O’Neill puts this forward as an example of a liberal obligation to charity, the spirit of the example is consistent with a very different interpretation. We could reformulate R to read aR – I will never help those who need help, unless it is to my advantage.98 When universalized, this would not seem to fall foul of the universalizability test. aR involves a recognition that people can and will be needy and does not rule out the possibility that I may be one of those who helps meet those needs. Indeed, aR exhibits a definite willingness to meet the needs of others in the right circumstances.99 If aR is universalized, it would provide an avenue of access to help on terms usually described as ‘mutual advantage’. Many important needs could perhaps be met not through charitable obligation but through the self-interested exercise of freedom within the realm of permitted action delimited by the universalizability test. Tellingly, O’Neill expects many needs that do not threaten agency to be met by mutually advantageous, ‘market’ solutions. There seems to be no fundamental reason why such a method should not successfully satisfy many needs including ones that threaten agency. The potential for mutually advantageous solutions in such contexts should not be underplayed. After all, a very large proportion of needs are already met exactly in this way. A second possibility for dealing with human needs that does not rely on imperfect obligation stresses the many ways in which our connectedness to others is important. While O’Neill emphasizes our ‘connection’ to others as a basic assumption of action, she takes less account than we might expect of the real and substantial ways we are actually connected. The public and personal relationships that we inhabit and through which we live so much of our lives must surely have a central place in any discussion of the satisfaction of needs. Many needs will be met for reasons based in particular contexts, local and institutional solutions that rely on substantial connection. Family relationships

Primary constructivism and O’Neill

109

between parent and child, between siblings and within the extended family of grandparents, aunts, uncles and cousins are often especially important in attending to desperate needs. So too will be our wider connections, either personal, as expressed in friendships, or associational, through the communities, clubs and organizations of which we are a part. Perhaps even more important are the wider connections that exist at state and cultural levels. In many contexts, it is seen as part of the expected and legitimate role of the state to fortify, support and aid us in moments of severe need. Such a role will be expressed in a variety of legal or cultural institutions. The form such institutional aid takes may differ widely from society to society and from case to case. An obvious example is the welfare state; less immediate is the role any state assumes when faced with a natural disaster or catastrophe.100 These various paths of connection, through family, friendship, association, culture or citizenship, cannot be emphasized enough. It is through such relationships, and the institutions that embody them, that we live most of our lives. It is to such institutions and specific relationships that we turn when faced by severe need. It is likely that our connection with others, and their connection with us, will make us aware of many instances of human vulnerability and also move us to ameliorate it. Such connections constitute a huge fund of need-meeting potential that is called on day after day and year after year. Acknowledging the existence of such avenues of help is entirely consistent with the rejection of R. Many of these mutual advantage and contextual (or connection) solutions can be conceptualized under the rubric of special perfect rights/obligations. Indeed, it is precisely these things that constitute our special rights and obligations. These would reflect our specific undertakings towards specific others. These special rights and obligations may be incurred or adopted through private contracts between individuals and organizations or through membership of, and acceptance of, communities or states and the norms and policies that they enact. The construction and adoption of these special rights and obligations, coupled with institutional, political, social, cultural and familial ties, is exactly what makes up the secondary construction, constrained by the framework of justice provided by the primary construction that identified our universal perfect rights and obligations. The particular context, with its particular and appropriate ‘thick’ conceptions of society and person, can provide us with ‘thicker’ conceptions of human neediness and vulnerability that are missing in a primary construction, heightening our awareness of, and sensitivity to, the real and experienced needs of those around us. Out of these local particulars are constructed secondary constructions, fuller responses to the problem of need than primary constructivism could ever hope, or need, to provide. Examining the way O’Neill’s test works more closely, it is clear that the rejection of indifference to needs does not directly issue in an obligation to aid. Instead, it issues in an imperative to ‘not never help’. This imperative is satisfied by a range of ordinary and everyday activities, each based on principles of action that are permitted rather than obligatory. As most needs are already met through our connections to others or through mutually advantageous market

110

Primary constructivism and O’Neill

solutions, it is hard to avoid regarding O’Neill’s claims about life without an obligation to aid as overstated. She claims that our joint projects would fail, that children’s capacities and capabilities that need assistance and nurturing would not emerge, that personal relationships would wither, that education and cultural life would decline and that we would have to accept that all relationships and action, indeed all life, are bound to fail.101 There is no good reason to expect such general devastation. Needs are met in all sorts of ways that are not dependent on a universal obligation to aid.

Conclusion These arguments do not necessarily undermine the construction of virtue and imperfect obligations. Indeed, there may very well be a continuing universal obligation to ‘rescue’ in extreme examples of agency-threatening needs, such as in famine conditions in poverty-stricken countries, which cannot be met in other ways. We cannot and should not dismiss the construction of virtue out of hand. However, we must recognize the prudence of a general wariness about ‘thickening up’ primary constructivism or about attempts to ‘fill in’ secondary constructivisms at the level of abstract theory. The further we go down the road characterized by thickness and determinacy, the greater the risk that we will make partial and biased assumptions, building in more than we intended and becoming too prescriptive. We should be wary of trying to construct overdeterminate conceptions, of doing more than the primary construction is suited to. The thin universalism of a properly primary constructivism may be very thin indeed. O’Neill’s thin universalism, although thin, retains the important understanding of a constructivism that generates substantive principles of justice with cosmopolitan scope. Underpinning this account is a constructivist conception of objectivity in reasoning. A general and objective practical reason is not vindicated on the basis of partial and idealizing assumptions. Rather, objective reasons will make only abstract assumptions, ensuring that they could, at least, be followable by all others. In this way, the authority of practical reason, which in O’Neill’s modal account consists in the demonstration that its reasons could be generally accepted, is dependent on the exercise of its critically reflective capacity that tests the partiality and abstract nature of its basic assumptions. The objectivity and authority of O’Neill’s practical reasoning depends on its proper exercise, in very much the same way that Rawls uses a notion of wide reflective equilibrium as the test by which our critical reason assesses the objectivity of our current assumptions, conceptions and principles. Both Rawls and O’Neill claim to have constructed properly objective accounts of justice, accounts that draw their objectivity from their generation by properly exercised practical reason oriented by conceptions of society and the person. Rawls, in Political Liberalism, is clear that there are specific conceptions of society and person built into particular political traditions, and a significant part of that book is concerned to detail those inherent in democratic public polit-

Primary constructivism and O’Neill

111

ical culture. It has become clear, however, that there is also an argument in Rawls and O’Neill that some basic concepts of society and person are necessary to any exercise of practical reasoning that makes constructivism far more than a limited and local phenomenon. Rawls outlines bare concepts necessarily assumed by any conception of the right or good and as basic to practical reasoning as notions of judgement and inference. O’Neill outlines her understanding of these concepts in Constructions of Reason, where they are said to constitute a minimal account of the circumstances of justice that frame all questions of justice and political interaction. In Towards Justice and Virtue, we find them expressed in the assumptions of plurality, connection and finitude that underlie any activity. Both Rawls and O’Neill regard these basic concepts of society and person as properly abstract and impartial assumptions of any exercise of practical reasoning. Principles of justice constructed on their basis can therefore be properly objective and universal, although their ‘thinness’ is an explicit recognition of legitimate diversity at the secondary level. While we have reached a position from which the central tenets of Rawls’s and O’Neill’s accounts of primary constructivism look very similar, what we have gained from O’Neill is a clearer understanding of the workings of practical reason, of why practical reasoning assumes a constructivist account of objectivity and of why concepts of society and person are appropriate orienting conceptions for practical reasoning.

5

Reasoning practically A constructivist understanding of practical reasoning

The last three chapters have helped us in providing some basic insight into the constructivist position. This chapter and the next will try to sketch in the outline of a plausible and general understanding of constructivism. It has become clear that a conception of practical reasoning is central to constructivism. The notion of practical reasoning can usefully be discussed under the two broad headings ‘reasoning practically’ and ‘making our practice reasonable’. I find this approach useful because I have often been confused by broad-brush discussions of practical reasoning, and although what is involved in reasoning practically and making practice reasonable are not really separate questions, these issues are distinct enough that separating out their discussion simplifies matters. This chapter will concentrate on a constructivist understanding of reasoning practically. Previous discussion of Rawls has shown us how central an account of reflective reason, in his case reflective equilibrium, is to constructivism. Similarly, O’Neill has explored what it takes for a reason to be a possible reason for others. This chapter will help us understand why practical reason is necessarily reflective and why it must offer reasons to others. It will help us to see how a constructivist account, similar to that we have found in Rawls and O’Neill, can be plausibly and independently arrived at through a series of quite natural steps. In doing so, we will draw into one place and systematize our thoughts about practical reason and objectivity. This will involve working through, and developing a clear understanding of, the necessity of practical reason, the constraints of practicality, the boundaries to our reasons and the objectivity and authority of practical reason.

The necessity of practical reasoning People are always doing something or other, and in doing so, they are always doing that something with or around, perhaps as a response to, other people. People also have all sorts of reasons and convictions that, when they act and interact, inform their actions. It may usually be possible to give an explanation of people’s action in terms of their convictions. This sounds straightforward, but we must bear in mind that there is always the possibility of asking a further

Reasoning practically

113

question of justification. The basic picture of people acting for reasons must be modified in response to a recognition that we may always change our minds about the convictions that we have and, therefore, the reasons that we have for doing one thing rather than another. We can also always encounter new situations and new ideas that our current convictions may appear unsuited, or unable, to account for. In this way, we are often faced with questions such as ‘What do I do next?’ or ‘What should I do next?’1 People may also face the collective question ‘What should we do next?’ Finally, we are often faced with challenges: first, from ourselves as in ‘Was I right to do that?’, and second, from others as in ‘Why did you do that?’ or ‘What are you going to do, and why?’ That we might change our minds or face new situations and challenges to our reasons means that we face questions not simply of explaining our reasons and actions but also of justifying them. These justificatory questions are an unavoidable but everyday part of all our lives. We find ourselves asking, and answering, these sorts of questions to ourselves and to others all the time. Our answers, and the answers we expect from others, take the form of reasons, and these reasons concern actions and principles for action. In short, asking and answering these questions are exercises in practical reasoning. Just as such questions are an unavoidable aspect of everyday interaction, so is practical reasoning. An account of practical reasoning is necessary, as it both constitutes and explains how we think about questions like this, how we should think about these questions and perhaps what sort of reasons are proper and legitimate for action. Recognizing the necessity of practical reasoning encourages the exploration of its nature, possibilities and limitations. This is the focus of the remainder of this chapter.

The constraints of practicality2 We are quite used to drawing a simple distinction between the practical and the idealistic. We understand idealists to be those people with their heads in the clouds and who are somehow less than at home in the ‘real world’. On the other hand, we commonly understand the practical man to be someone who is most definitely at home in the world, someone whose plans and actions are sufficiently connected to ‘the world’ that they have some chance of success. Likewise, an idea or a plan is labelled idealistic if it is considered to be impractical, to pay insufficient attention to the way the world is, or works, to be of any use in guiding our thought and actions. If idealistic reasoning is in some way disconnected from the ‘real world’, practical reasoning must at least be concerned with, adapted to, make reference to or relate to ‘the world’. It is important not to place too much emphasis on the ordinary usage of these terms. However, and as we shall see, there are important reasons for starting our enquiry in the place that we find ourselves, not least because when discussing practical reason it is all too easy to be either confused or confusing. In starting on familiar ground and seeing how far we can work from here, I hope we can make sense of practical reasoning without being either.

114

Reasoning practically

Our first thought has been that reasoning practically involves giving reasons that are disciplined by, or connected to, the world. This is simply the recognition that practical reasons, in order to be practical rather than idealistic, should identify reasons that are adapted to the situation in which they are to count as reasons. This is not to advance the claim that there is a privileged description of the world out there and that this description presents itself to all people with adequate powers of perception. This may or may not be the case, but I do not think that constructivism needs to commit itself either way on this controversial issue. This suggestion that reasons should be connected to their context is equally consistent with the claim that the world we perceive is intersubjectively constructed or with the claim that our perceptions of the world are constrained by our standpoints. All I am concerned to suggest is that our reasoning, if it is to be practical, must avoid inconsistency with available information about the world, however that is presented to us. If it did not, then it would ordinarily be regarded as either idealistic or simply mistaken. However, this suggestion that we secure the practicality of our reasoning by ensuring that the world disciplines or constrains our reasons must consist of more than the injunction to avoid wild flights of fancy or against indulging in the construction of fantasies. The main body of this chapter will explore the implications of this first thought about the discipline involved in reasoning practically and will help us gain an understanding of the full range of constraints that practicality imposes. The first step towards this is to recognize that we are required in this way to offer reasons that are at least possible. Our reasoning does not function in a vacuum but always within particular environments, and so reasoning could not be practical if it proceeded from assumptions or expectations of these environments that were untrue of them or could not be true of them. Nor would our reasoning be practical if it failed to recognize, and so ignored, the appropriateness of certain environments. Practical reasoning must ‘fit’, be guided by, or at least be appropriate to, the environment(s) within which it is to function as reasoning, as it is concerned to offer reasons for action or for the adoption of principles for action. Since actions at least possibly impinge on others, our reasoning must be responsive to the relevant normative and political environments. We must therefore be prepared to offer reasons that are at least possible reasons given the range of expectations we can legitimately have of the environments in which these reasons are to count. This constraint on practical reasoning has been widely recognized in political theory in the assumption of an account of the circumstances of justice. An account of these circumstances involves making a claim about the assumptions that it is appropriate to make about the normative and political environments in which the associated political theory can be regarded as justified. In effect, it is a recognition that the reasons for action a theory offers hold only insofar as the account of the circumstances of justice accurately reflects ‘the world’.3 The impetus for this study has, in large part, been the recognition that both the normative and the political environments in which practical justification takes place are plural, where once they may not have been. As we have seen, the

Reasoning practically

115

normative environment for practical reasoning consists of a plurality of competing and probably incompatible conceptions of the good life. If the constitution of this environment were different, and if there were an overwhelmingly and obviously correct conception of justice or morality, then our practical reasoning may be different in reflection of this. We do not currently appear to be in a position to dismiss all but one of these understandings of the good life, nor, given an understanding of Rawls’s burdens of judgement, should we ever expect to be in such a position. As such, if our reasoning is to be practical, to be disciplined by ‘the world’, it must take account of this plural normative environment. It is also clear that we must take seriously contemporary political pluralism. It may once have been unproblematic for political theorists to assume that political environments were homogenous; today, it would be problematic not to acknowledge their heterogeneity. The state is not now, and it may be doubtful that it has ever been, easily identified with a single nation, tribe, ethnic, cultural or political grouping. Instead, political environments reflect plurality where at least several different such groupings coexist in any given territory. This is reflected in the increasingly widespread political and theoretical recognition that we inhabit multicultural societies and also in the widespread concern for minority rights.4 Reasoning that is practical must acknowledge this political plurality and take some account of it; a failure to do so would be impractical. Recognizing that in order for our reasoning to be practical it must be responsive to the normative and political environments in which the reasons that it offers are to count as reasons is an important first step in ensuring that our reasons are disciplined by the world. This recognition involves the acknowledgement that we must start from where we are and that responding to our plural normative and political environments is one part of starting from where we are in our practical deliberations. As we are without access to the alternative, a self-evident range of normative considerations whose special epistemic status is widely acknowledged, we have little choice as to where to start at all. If we cannot initially identify a self-evidently correct standpoint then we can only stand where we already find ourselves as choosing to stand elsewhere would be arbitrary. In reasoning practically, we have to make some initial assumptions about normative and political environments and, even when faced with plural environments, we can only start with the resources available to us, our convictions or intuitions or basic judgements.5 To do otherwise would be to arbitrarily attempt to use resources that we do not have, and there could be nothing practical about that. If our reasoning is to guide actions, or justify the principles that are to guide action, then it must bear some sort of ‘relation to who we are or what we believe’.6 It does this by keeping its feet firmly on the ground through its use of only resources that are ‘to hand’ or available. As the practical person works with what is at hand instead of trying to use what is not and without forcing a tool or material to a use for which it is not suited, so does reasoning that is practical start from available starting places. Our convictions, basic assumptions about our moral reasoning, form a significant part of the normative environment for our

116

Reasoning practically

reasons, and so taking them seriously is a significant part of being guided by the world in which those reasons are to function. Reasoning in such a way that our reasons are disciplined by the world means that we must avoid making further assumptions that are not accurate of that world. Assumptions that are inconsistent with available information about the world are no basis for practical reasoning. It is in this light that we should regard, for example, Onora O’Neill’s injunction to avoid making assumptions about other people or agents that are untrue of them. If we do so, she claims that we run the risk of producing reasons that are ‘inapplicable to the human case’.7 That disciplining reason with the world, in the form of avoiding assumptions that are inconsistent with available information, is necessary for reasoning to be practical seems a straightforward point.8 Such discipline is obviously necessary if we are to ensure that whatever reasons for action, or principles for action, our attempts at practical reasoning produce are at least possible actions. This is the point of making our reasoning responsive to plural normative and political environments, of using only resources that are at hand and of avoiding assumptions that are untrue of the world. We must be attentive to at least these considerations in order for our reasoning to have any chance of being properly practical rather than arbitrary, fanciful or utopian. Bearing this in mind, the recognition that the environments for our reasoning are plural is of significant consequence for justification. If reasons are to ‘fit’ environments when environments are plural, then reasons may have to be adapted to suit the variety of environments we necessarily encounter. In order for our reasoning to be practical, we may have to offer different reasons in different environments. What might be a reasonable expectation of one normative or political environment may not be so of another. As we have seen, disciplining reason with the world involves making basic assumptions in order that our reasoning can get started. Reasoning does not take place in a void but responds to questions or problems as and how they arise for us. Therefore, reasoning that ‘fits’ environments may have to make different basic assumptions in different environments in order to ensure the appropriateness of its starting points and thus of the resultant and justified reasons. Some assumptions about the ‘facts’ or relationships that constitute each environment must be regarded as ‘fixed’ loosely in any justification appropriate to that environment. Of course, in the process of reasoning or of offering reasons to others, it may become apparent that our assumptions are mistaken and need revision. The injunction to be disciplined by the world would mandate any necessary revision. Still, with this proviso, we can provisionally regard a set of basic assumptions as constituting a ‘fixed’ environment for a particular justification.9 This is roughly equivalent to recognizing that different justifications may be appropriate for different audiences, that we address different audiences differently. Whenever we attempt to justify an action or a principle, the reasons we give call on a wide range of assumptions and presuppositions about the audience. The assumptions that we make when offering reasons to one group of people with whom we share a certain relationship may differ from those we make when offering reasons to

Reasoning practically

117

another group with whom we share a different relationship. For example, in most families, a basis in a profound and shared love for each other can be assumed, and therefore general altruism in crisis situations can be presupposed by any reasons we offer in justification to members of our family. Assuming such altruism of total strangers on the other hand might be dangerous as well as misguided, and so alternative basic assumptions would be more appropriate. Likewise, it is fairly safe to assume that there are good grounds for presupposing a greater homogeneity of interests at the national than at the global level. Basic assumptions, and therefore reasons and justifications, vary with intended audience.10 Justification, in marked contrast to proof, is always targeted at an audience, even if the audience is sometimes only ourselves. Ensuring that our assumptions are appropriate to the relevant environment and the relationships that compose it is a necessary part of ensuring that the reasons we offer are potentially followable by the people in those relationships and to whom we offer reasons and justifications. This is not to say that these people will follow our reasoning but that it is at least possible that they will.11 Reasons should not exclude their intended audience. If they do, then they rely on inappropriate assumptions, utilize mistaken chains of reasoning or their real audience is other than that which is claimed and so their justification is deceptive. We have so far briefly explored the claim that reasoning practically involves being responsive to the relevant environment and audience for justification. We have, however, not yet touched upon an absolutely central aspect of this responsiveness and therefore of reasoning practically in general. This is the recognition that environments are not static. It is clear that the relationships and groupings that constitute normative and political environments are not themselves fixed but shifting; they change and evolve over time, and our account of the conditions on practical reasoning must acknowledge this basic point. We must acknowledge that disciplining reason with the world is not a once-and-for-all process but a continual one. Since relationships change, so too must the assumptions that we make as appropriate to justificatory environments. Even those assumptions that we provisionally regard as ‘fixed’ may become, over time, inappropriate or less appropriate than an alternative set of assumptions. It may be that the fundamental nature of these relationships alters or that we come to include others in our relationships that were previously excluded. Alternatively, we may become dissatisfied with the appropriateness of our basic assumptions, perhaps if they are challenged by those who feel excluded by them or by encounters with novel ideas or practices previously unconsidered. On the other hand, we may just think new thoughts or develop new practices. However, this actually comes about; if the basic assumptions underlying our reasoning become inappropriate, then the practicality of the reasons and justifications based on those assumptions becomes suspect. Suspicion that such an evolution in environment is occurring gives us reason to re-examine our basic assumptions in order to ensure their appropriateness. We are also given reason to change our basic assumptions if we confirm that a change in environment has already occurred and we are faced by novel circumstances. This would seem to be necessary if our reasoning is to

118

Reasoning practically

maintain its practicality. Reasoning practically will demand the recognition of the provisional nature of basic assumptions. That we regard them as ‘fixed’ for the purpose of a particular justification does not mean that they really are fixed or refer to fixed features of an environment. To dogmatically maintain the practicality of past reasons in the face of change, when it becomes apparent that the assumptions that underpin them are no longer appropriate, is a mistake that undermines the practicality of our reason more generally. Recognition of the necessity of revising basic assumptions is the recognition that our reasons and justifications may also change. This highlights the provisional or perhaps temporary nature of any properly practical reasons. Their dependency on environment ensures their provisional standing, as all environments and relationships change over time. No environment can be assumed to be fixed for all time, and therefore even our most strongly held reasons must be regarded as only provisionally held (even if it is difficult to imagine what criticism or novel encounter could cast doubt on them). What we have naturally and plausibly drawn out here is the necessity to practical reasoning of an ongoing reflection that continually revisits the appropriateness of basic assumptions and starting points, the processes of reasoning built on them and the reasons that are justified as a result. In other words, we have drawn out the necessity of regarding our reasoning as, or at least regarding it as subject to, an ongoing process of critical reflection. This is an appropriate response to both novelty and criticism. Both criticism and encounters with novel ideas or cultures are effectively challenges to the appropriateness of our reasons. They both pressure us to demonstrate, to ourselves and our critics, that our justifications and reasons are not based on partial and inappropriate assumptions but rather that they can encompass both novelty and criticism within their scope. If they are not able to do so, then this implies that they are in need of revision. That reasoning about the justification of normative principles may require this process of revisiting our reasons is central to Rawls’s notion of seeking reflective equilibrium. He claims that justification is a matter of seeking a reflective equilibrium between considered convictions as ‘fixed’ starting points and principles of justice. As we have seen in previous chapters, attaining such equilibrium involves making basic assumptions in order to get reasoning off the ground and also a commitment to revisiting those assumptions in the process of justification. Just as we have found necessary in our account of reasoning practically, Rawls recognizes that ‘even the judgements we take provisionally as fixed points [of moral reasoning] are liable to revision’.12 He goes so far as to recognize that ‘there are no judgements on any level of generality that are in principle immune to revision’.13 We are now in a position to understand why this critically reflective attitude, and therefore the insistence on the provisionality of reasons, is so important. Where Rawls’s critics have regarded reflective equilibrium as a controversial and problematic notion, we have arrived at a similar account in a simple and straightforward way. Such an account appears absolutely necessary to practical reasoning, a central part of what reasoning practically must involve. Summarizing our first steps, we have found that reasoning practically

Reasoning practically

119

involves recognizing two simple injunctions. First, it involves ensuring that we offer reasons that are disciplined by the world to the extent that these reasons are at least possible reasons. Our reasons, if practical, should respond appropriately to the normative and political environments in which they are to function. Responding appropriately to the environments in which we find ourselves is a matter of starting from where we are and, therefore, of making assumptions that are appropriate to our environments. We must also recognize that such environments are plural and that different assumptions might be appropriate in different environments, and so different justifications may have to be offered to different audiences. Second, reasoning practically involves recognizing that the environments in which our reasons function are continually changing. Reasoning practically in a changing world is a matter of a willingness to revisit and revise basic assumptions, reasons and principles to ensure that they remain appropriate as that world changes. Responding to these changes, coupled with a sensitivity to the possibility that our basic assumptions may be less appropriate than we thought, leads us to an understanding of reasoning practically as involving a central and basic commitment to a process of ongoing critical reflection.

Boundaries to reasons It is important that we have gained some significant understanding of the centrality of critical reflection to practical reasoning. However, in doing so, we have had to stress the way in which reasons are linked to contexts and how processes of reasoning have to take this seriously. One natural conclusion that might be drawn is that this implies the partiality of all reasons. If reasons are linked to contexts in that they rely on assumptions appropriate to environment, then they may be regarded as holding as reasons only relative to that context. Have we come to an understanding of practical reasoning that is necessarily relativist, partial or subjective in some significant way? It is important that, although this conclusion seems straightforward, we do not move too quickly in drawing out the relativist implications of our account of reasoning practically. Before this conclusion is warranted, we need to examine its grounds. While it may be that reasoning must make assumptions appropriate to normative and political environments, we have yet to explore whether these reasons are thereby limited to those environments or, to put the matter in another way, whether all environments are less than universal. The common grounds for drawing a relativist or subjective conclusion from an account of practical reasoning will rely on an account of the limits of practical reasoning. The argument goes that our reasoning is constrained within lessthan-universal limits, and so our reasons must hold only relative to those limits. This section of the chapter will outline and examine three key ways of understanding the limits that, it might be claimed, constrain our reasoning. These limits may be theoretical limits, they may be practical limits or they may be motivational limits, and I will address each in turn. It becomes clear that, on closer examination, the usual theoretical, practical and motivational boundaries

120

Reasoning practically

taken to apply to our reasons need not apply; our reasons are not necessarily partial. It is important to rule out the necessity of partiality so that, in the next section, we can understand how our reasons may be objective. Theoretical limits on reasons14 The argument that there are necessary theoretical limits on our reasons is most commonly found, in contemporary political theory, in the work of communitarians and proponents of multiculturalism. This claim, as we saw, was central to the critique of foundational political theory and drew much of its intuitive strength from our experience of cultural and political pluralism and the further claim that this pluralism is irreducible. The basic argument is that there are certain assumptions that necessarily underpin our reasoning and that these assumptions are different for different people at different times and places. These assumptions function as ‘givens’ for all our reasons and justifications, and they are ‘culturally bestowed or socially imposed’.15 They ‘constitute the given of my life, my moral starting point’, and I inherit them from my cultural context.16 My specific cultural context provides the unchosen starting points of my reasoning and the ends for which I aim; they ‘define the person that I am’.17 The dispositions and tendencies that individuals acquire when part of a specific culture are ‘inseparable and ineradicable from their being’ and form ‘an inseperable part of their personality’.18 The attitudes, desires, attachments, commitments, dispositions, sentiments and ends that constitute our particular identity are necessarily fixed by culture and placed beyond re-evaluation.19 If our reasoning necessarily has certain basic assumptions given by context and our self-understanding is closely tied to a particular culture, then other cultures become unintelligible to us. We have no resources that enable us to make sense of what those in strange and unfamiliar contexts are doing or saying. The actions and thick reasons for action of members of other cultures are literally unintelligible to us, as we are not embedded in their practices in the same way that they are or that we are in our own.20 We cannot, either temporarily or permanently, escape from or even step outside of our particular context. Nor, as they are given, not chosen, are the features of that context and the way they shape and constrain us available for rational reassessment. There are simply boundaries to the intelligibility of reasons and values that track the boundaries of community, and the recognition of this involves a rejection of the possibility of universal reasons that are not dependent on particular contexts. Recognizing the theoretical constraints on intelligibility, reason and justification, ‘the most philosophy can hope to do is summarize our culturally influenced intuitions about the right thing to do’.21 Understanding the theoretical limits on reasoning this starkly may be putting things too strongly. Rather than claim that context determines identity and reasons, the argument for theoretical limits is often recast as the claim that it ‘structures and shapes the individual’s personality . . . and gives it content’ so that while we are capable of participating in the reflective formation of personal

Reasoning practically

121

values, ends and reasons, this reflective process always remains constrained within cultural limits.22 Individuals are always ‘defined to some extent by the community of which they are a part’ but can never be more than ‘co-authors of [their] . . . own narratives’.23 Some ties are always beyond rational criticism and therefore beyond revision.24 If this claim can be sustained, then this is all that is necessary to show that our reasoning is subject to theoretical limits that prevent it from ever escaping partiality and condemn it to remain always relative. If any basic assumptions of our reasoning remain, in principle, beyond critical examination, and therefore unrevisable, then all constructivism will be secondary constructivism. All constructivism will take, as a fundamental basis, some set of cultural givens. However, if the claim regarding necessary theoretical limits is not taken to rest on the strong claim that context determines reasons but instead on the weaker claim that our reflective engagement with reasons is always culturally constrained, the argument gives too much ground to the constructivist in ways that are not often noted. If the individual is regarded as part author of his values and identity, then it is very difficult to make the argument that this process is necessarily constrained. On this modified and weaker claim, individuals are regarded as drawing their conceptions of themselves and the good, and the reasons and values these imply, from the shared understanding of a particular community or culture. They are regarded as making reflective choices from a set of available cultural resources, but only within the limits set by these shared understandings. In making these choices, individuals obviously reflect on which ends from the sets available they have reason to affirm or endorse and which they have reason to modify or reject. Note that, in the course of their individual reflection, they are at the same time reflecting on the social understandings that present those ends and reasons as available options. If someone has reason to reject certain cultural values as ones that they feel no reason to endorse, then this also gives them reason to reject or modify the social understandings tied up with those values. Obviously, an individual reflection on, and rejection of, a set of social or cultural understandings will not radically alter those understandings (although the reflection of many individuals might), but it does give a foothold to the idea that these understandings are under constant examination and may be subject to reasoned pressures to adapt or change. The relationship between cultural and individual reasons does not obviously involve a process of one-way determination of the individual by the cultural but may instead be a reciprocal relationship of mutual influence. Indeed, when we stop and think about it, it seems that cultural understandings must bear some reflective relation to the reasons and choices of individuals. Cultural understandings could not be regarded as such in the first place if they did not form part of the individual views of a significant section of the community. The reverse is also the case; if a value is rejected by a significant majority of the individual members of the community it would be very odd to claim that it was fundamental to their shared cultural understandings. Even if we allow that these understandings do not present themselves to individuals at the conscious level, it is unlikely that they

122

Reasoning practically

could persist in a central role if the values that they entail are systematically rejected by the members of the community. Individual exercises of reflective reason are not theoretically prevented from leading to the reflective revision of cultural ‘givens’. Is establishing this enough to show that reasons are not necessarily limited by context? It makes sense of at least some ways reasons can transcend specific contexts as it enables us to understand how reasons can alter the context in which they function. This in turn makes some sense of how we know that cultures change as they encounter novel ideas or challenges thrown up by individuals or groups that are a part of that culture. However, this does not get to the core of the claim regarding theoretical limits and is not yet enough to secure the conclusion that there are no such limits. The core of the claim is that different communities provide radically different contexts for reasons to the extent that reasons are unintelligible across cultural or communal boundaries, that the limits of context necessarily delimit the boundaries of reason and justification. The effective claim is that different contexts constitute different conceptual schemes.25 That individual exercises of reflective reason are not determined by but rather can themselves influence communal understandings is only one step in a reply to these wider claims. The next step is to think about our reflective engagement with novel ideas and challenges to our current ways of thinking that originate from outside our cultural horizons. Is there any reason why this is any different from our engagement with similarly novel ideas and challenges that, although we were previously unaware of them, originate from within our own culture, broadly speaking. We have already seen that, even though they have not formed part of our context in the past, novel ideas of this second sort do not sit on the other side of a necessary limit. Why, then, should we think that other novel ideas of which we were equally unaware be treated as if they were on the other side of such a limit?26 I want to argue that there is no reason to regard these two reflective exercises differently and will do so on the basis of two arguments. The first draws on the, by now familiar, idea of reflective equilibrium. We have examined this notion extensively in previous chapters so there is no reason to rehearse the arguments here. It is enough to remind ourselves that discussion of reflective equilibrium demonstrated that a ‘given’ can best be regarded as meaning ‘a starting point for reflection’ rather than ‘immune from revision’. Cultural givens are not necessarily the theoretical limits to reasoned reflection that some claim. Kai Nielsen explains this using the metaphor of Neurath’s boat. Although there can be no simple stepping out of our societies and traditions . . . our more or less local identities . . . [but] they need not ethnocentrically hobble us . . . we are not imprisoned by our traditions . . .. No belief is in principle immune to criticism and rejection and whole traditions, plank by plank, can be transformed as we repair and even rebuild the ship at sea.27 Rawls’s account of wide reflective equilibrium describes this process.

Reasoning practically

123

The argument from the possibility of wide reflective equilibrium can be supplemented by an argument from Donald Davidson; indeed, it may presuppose it. Davidson claims that the very idea of mutually unintelligible conceptual schemes makes no sense.28 The claim that there are different conceptual schemes implied by different cultural contexts must depend on the claim that these conceptual schemes do not overlap, that they do not share a fund of common concepts or ideas that enable translation between the two. If translation can take place, then the conceptual schemes are not mutually unintelligible. They would instead be better thought of as different languages for expressing a single conceptual scheme. In principle, mutual understanding would be possible and reasons would not be subject to necessary theoretical limits. Therefore, the claim must be the stronger one that there is no significant overlap between conceptual schemes. This much is necessary to substantiate the claim that reasons, truth and claims to objectivity are relative to conceptual scheme. We might be led to this conceptual relativism if we are over-impressed with the differences we encounter when we move between cultures or with the apparent breakdowns in translation that are an everyday part of these encounters. Davidson argues that a breakdown in translation can only be local, as a background of generally successful translation is what makes local failure intelligible.29 The differences we encounter are not as radical as experience might suggest, and we encounter them against a background of shared belief. It is just that ‘what is shared does not in general call for comment; it is too dull, trite, or familiar to stand notice.’30 Whatever plurality we take experience to consist in – events like losing a button or stubbing a toe, having a sensation of warmth or hearing an oboe – we will have to individuate according to familiar principles. A language that organises such activities must be a language very like our own.31 So, whenever we encounter others with whom we apparently disagree radically, ‘whether we like it or not . . . we must count them right in most matters’.32 Davidson is here referring to the everyday mundane things that we all do and encounter all the time, whether we are comfortably ensconced at home or abroad in alien and frightening surroundings. We have no trouble in understanding what is going on in our own and other cultures 99 per cent of the time. We recognize birth, sex, eating and drinking, use of the toilet, sleeping, walking, waiting, impatience, anticipation, fear and so on. Davidson is asking us to consider the ‘limits of conceptual contrast’.33 What are claimed by some people to be radically different conceptual schemes are far better thought of as necessarily overlapping to an enormous degree. That 1 per cent of the time understanding or translating is difficult does not establish the claimed necessary theoretical limits on our reasons and justifications.

124

Reasoning practically

Practical limits on reasons It may be that there are no necessary theoretical limits on our reasoning, but this does not rule out a claim that our reasoning either does or should only rely on local and limited assumptions. This may involve the claim that the extent of our interaction, and by extension our reasoning, is only ever local. Alternatively, the claim may be about the real extent of our experienced moral concern, that our only very real and tangible obligations are local and particular. This is a claim about our motivation, in which case it will be addressed in the following section, or it is a claim based on the theoretical limits of reasoning, in which case it was dealt with in the last section, or it is the first and practical claim that our interactions are local and so, therefore, are our reasons. To the extent that this is not a claim about theoretical or motivational limits, it should not much occupy our attention. There may indeed be very good reason for focusing the bulk of our reasoning locally. This is where our family and friends are, and we may feel that we owe them a special sort of concern and attention and that particular and very important obligations arise for us in this context. Also, we may feel that we have a range of particular and special obligations of limited extent that we incur as members of clubs and associations, as contractors with other individuals and groups, as citizens of particular states or as compatriots in the same community. Our attention really is overwhelmingly focused locally. However, there is no reason to think that the everyday limits that we quite naturally impose upon our reasoning are ever necessary limits, and it is this question of necessity that concerns us here. Recall the arguments that O’Neill makes regarding the extent of our connection with others. We highlighted the vast range of assumptions about distant, as well as local, others that we necessarily make every time we do something as simple and straightforward as posting a letter, buying a packet of coffee or driving a car. In almost every activity, we currently necessarily make a range of global assumptions. Even if this is sometimes an exaggeration, the key point is that it is very difficult for us to, in principle, set limits to the possible implications of our actions and the scope of their effects. It is not practical or reasonable to pretend that we make only local assumptions. We necessarily make assumptions that cross our associational or political boundaries all the time. O’Neill does raise the possibility of practical limits when discussing the way there was no interaction between the Anglo-Saxons and the medieval Chinese.34 As these groups were actually not aware of each other’s existence, and their actions did not in any way affect each other, we can attribute practical limits to the extent of their respective exercises in reasoning. But assuming such practical separation, and therefore limits, is just not plausible in current circumstances. There is no one on earth to whom we are not practically connected and whom our actions might not practically impinge.35 Also, to highlight the practical separation of the Anglo-Saxons and the Chinese is not to say that, if they ever met, the Saxons would have no obligation to the Chinese or that the Chinese world view would not be a challenge to that of the Saxons. It is just to say that, as they

Reasoning practically

125

had not met, nor their actions impinged on each other, the question of whether their principles for action are partial given the challenge such a meeting would pose and the question of whether the scope of those principles extended to the Chinese simply did not arise as a practical question. Today, all others stand in some form of relationship with us, and we therefore make some sort of practical assumptions about them whenever we act. We might ask ourselves, ‘What momentous events would have to occur for this to no longer be the case?’ Under current and foreseeable circumstances, there are no necessary practical limits to our reasoning.36 Motivational limits on reasons Perhaps there is another form of limit to our reasons that is based on an understanding of the place of motivation in arguments for principles of justice. Perhaps it is necessary, in order to understand justice, to consider a person’s motives for action, in particular, their possible motivation for acting justly. Is a commitment to justice premised on a coincidence between the dictates of justice and our pre-existing interests? Contractarians, who are sometimes referred to as a type of constructivist, are among those who take motivating interests very seriously in their discussion of justice, and it will be useful to understand how the contractarian position is radically different from the constructivism that we are concerned with.37 Broadly speaking, on this view, principles of justice are understood to be justified when a plurality of self-interested people are motivated to agree to them as principles that further the interests of each.38 Principles of justice are just those principles that are mutually advantageous. If a principle or reason does not further our interests or desires, then we may regard it as illegitimate, and its imposition in our case would be unjust. Principles are only applicable if each person has their own interest in supporting them. The basic claim concerning the limits on reasons is a claim about what counts as a good reason. A good reason must be motivating and can only be so if appeal to it presupposes a pre-existing interest or desire that I have, such that acceptance of that reason as a reason for me would further that interest or desire. If I do not have the desire that is presupposed, then that reason cannot be mine. Reasons are limited by the extent to which people have the presupposed desire. Practical reason is effectively ‘a slave to the passions’, and its role is the instrumental one of helping us to most efficiently achieve the ends those passions set us.39 Reasons are therefore always limited in scope. An account of moral principles, so this argument goes, implies some sort of motivation to be moral, and an account of universal moral principles would rely on everyone being so motivated; if the motivation is not general, then the principle is not universal. Importantly, this point should not be mistaken for an argument about the practical problems of generating support for already justified principles. All theories of justice will face this sort of practical problem, but the claim about motivational limits is much more than this. It is a claim that principles of justice can only be justified in that first place if they are

126

Reasoning practically

motivationally sufficient, and they can only exhibit this characteristic if they coincide with a pre-existing desire. I shall address this broad criticism in three ways. First, I will suggest that this is a sort of non-question, as issues of justification and motivation come apart. Second, even if constructivism concedes (although I do not think it has to) that the motivational claim is important for justification, it can still go on to show that it is a non-question in a second way (as far as constructivist justification is concerned); constructivist justification finds its starting points in the moral and political practices that we are already tied into. Third, I am impressed with Scanlon’s argument that the specific claims about the necessary role of interest or desire in motivation and justification that are assumed in the argument about motivational limits to reason are problematic. Together, these arguments will show that there do not appear to be any obvious motivational limits to reasons. The claim that the justification of principles depends on the available motivational resources just seems plain wrong. We do not ordinarily regard principles in this way. Instead, we regard principles as providing us with reasons that we ought to be motivated by even if we are not motivated by them at present rather than the other way round. Of course, we might be mistaken about this, but this would be a difficult conclusion that struggled to make sense of our ordinary experience of practical reasoning. We are used to making categorical judgements of principle; for example, that genocide is wrong and so are sexual relations with very young children. If someone, for whatever reason, experiences contrary motivations, we do not regard the principles as unjustified in their case and so stand back and let them get on with the behaviour that they are motivated to pursue. Nor do we just recognize that the principles do not apply to them but restrain them anyway because we disagree with their behaviour. Instead, we speak of such people as failing to recognize justified principles that ought to constrain their actions whatever their motivation. I might be motivated to do all kinds of strange things to people. This does not constitute a justification for my actions nor would it if a majority of (or even all) people were similarly motivated. There is nothing unusual in recognizing that we do not always do what we ought to do, even when we accept, on grounds that we regard as justified, that we ought to do it. The motivational criticism makes little sense of this everyday experience of a gap between justification and motivation. My claim does not rely on particular substantive judgements about what sort of actions or principles are justified but on the more general thought that we could all come to believe anything at all and why should justification be thought to depend on this. If a totalitarian plan of global indoctrination instilled in us all a desire to see the world peopled solely by a race of Aryan supermen, this would not justify actions to bring this about, even if everyone wholeheartedly supports the action. This suggests that motivation and justification come apart, that they are separate issues. Justification is, as we have seen, about good reasons rather than what we presently take to be good reasons. What each of us takes to be a good reason, what we are motivated by, is a question that is broadly sociological or psychological, and it is difficult to see that anything normative is implied by

Reasoning practically

127

how that question is answered. The justification of principle, on the other hand, is a normative and philosophical issue and a separate concern altogether. In this way, I regard the problem of motivating principles of justice as a matter of fostering a general acceptance of the good reasons that underpin their justification, but I do not think that their justification depends on the success of this discursive and persuasive activity. However, the constructivist can recognize that a sharp distinction between motivation and justification may be controversial, set the distinction aside, and still think that the motivational challenge misses the point. In order to understand this claim, it is only necessary to refresh our memories about the way in which constructivist justification works. Important here is the account of critically reflective reasoning, the notion of reflective equilibrium this implies and the attitude towards constructivist starting points implied in turn. Recall that constructivism addresses questions of justification as they arise for us in everyday situations. The starting points of constructivist justification are the moral and political practices in which we already find ourselves (and we are always already involved in some). When these practices are unchallenged, our practical reasoning works within the limits of these practices that we are already tied in to. Even when aspects of these practices and the principles that underpin them are challenged, constructivism adopts a critically reflective attitude that holds up to scrutiny the elements of our reasoning and practice that are in doubt whilst standing firm on the remainder. Of course, the reflective process may be radical and far-reaching, and an eventual wide reflective equilibrium may be significantly different from our initial position. However, this should not distract us from the key issue here: that we start within practices and holding reasons and principles that we are broadly committed to and go on to adapt them under the impetus of good reasons. The appeal of the broadly Humean and contractarian understanding of the relationship between justification and motivation, one that shows how principles are relevant to us by intimately linking them to us by means of motivationally sufficient desires, is attractive because it is a response to otherworldly foundational justifications to which we may not feel connected and which we experience as an imposition from the outside. By making the justification of principles dependent on our motivational experience, we are shown how we have a stake in them, a reason to endorse principles rather than to regard them as imposed on us. Constructivist justification, however, is very different from foundational justification for reasons that we have just alluded to. We have a stake in the principles that are justified by constructivist reflective reason because we are already intimately linked with moral and political practices that presuppose them. Our involvement in these practices is already evidence of our moral concern or commitment. Of course, and as we have explored above, justification does not depend on this link between principles and our experienced moral commitments, but, if you take motivational issues seriously in justification, this prior commitment should ease concerns about constructivist justification being otherworldly or disconnected from actual human concerns.

128

Reasoning practically

There is one remaining question. The Humean or contractarian, claiming that principles of morality and justice are only justified if they can be shown to appeal to a pre-existing desire or interest, when faced with the preceding argument, has one last important recourse available when seeking to sustain their account of the authority of principles. Accepting that there is a prior involvement in, and commitment to, moral practices, perhaps this can only be characterized as a desire to be moral or as underpinned by a motivationally sufficient desire the fulfilment of which makes being moral the instrumentally justified course of action. Must our commitment to just principles be based on preexisting desire? If so, then this may establish necessary limits to our reasons. Of course, it is possible that, in this way, a principle of justice could be justified universally if everyone had the same desire, but such universality would be a contingent matter dependent on the coincidence of desires. More likely is a situation in which the set of desires appealed to in the justification of principles of justice are partial, and therefore only a subset of people would find that they had reason to accept them. The constructivist, if she is as I have cast her, must respond to this argument (but only if we set aside her claim that justification and motivation come apart). If justice is relative to pre-existing desire, and reason is no more than an instrument for the satisfaction or pursuit of the ends these desires give us, then desire plays a foundational role in this position. Our pre-existing desires are beyond revision in a process of critical reflection, which, simplifying somewhat, envisages reflection extending only to means and not to ends. In this case, principles of justice and morality, in order to be justified, would have to match, or ‘fit’, a set of foundational moral facts (the range of our desires). The contractarian position is a form of foundationalism, and the constructivist, if constructivism is to be possible, needs to differentiate her position from this foundational theory just as it does any other. It can only do this by questioning a basic premise of the contractarian argument, that only desires can be the basis for both motivation and justification. If reasons, as Scanlon argues, can perform this role, then we would have undermined the explanatory necessity of desire. If reason (instead of or as well as desire) can be motivational, then the practical normative force of accepting and acting on good reasons that our prior commitment to moral practices may entail is not parasitic on pre-existing conviction or desire, and we will have disentangled constructivism from contractarianism. On this point, there is little to add to Scanlon’s account, which makes the argument far more convincing than I ever could.40 He is clear that desires, if they are something different from reasons, ‘have a much less fundamental role in practical thinking than is commonly supposed’.41 Indeed, ‘insofar as “having a desire” . . . is distinct from “seeing something as a reason” it plays almost no role in the justification and explanation of action’.42 Scanlon argues that a reason is a ‘consideration that counts in favour of’ something and draws a distinction between good reasons for believing something and my good reasons for believing it. The primary sense of a reason is that of good reasons in general, as this refers to a reason that really counts in favour of something rather than to what I

Reasoning practically

129

take to count in favour even though I could be mistaken about this.43 Importantly he regards reasons as ‘judgement-sensitive attitudes’ that are in some way ‘up to us’ and expressed in intentional action.44 Reasons are not something that just happens to us and we react to, as desires can seem to be in the Humean account, but are things that we can reflect on and recognize or reject as we judge them to be good or bad reasons. Scanlon’s next step is to claim that a person who judges that there is reason to believe P normally has this belief, and this judgement is normally sufficient explanation for so believing. There is no need to appeal to a further sense of motivation such as “wanting to believe”. Similarly, a rational person who judges there to be compelling reason to do A normally forms the intention to do A, and this judgement is sufficient explanation of that intention and of the agent’s acting on it. . . . There is no need to invoke an additional form of motivation [such as a desire] beyond the judgement and the reasons it recognizes.45 Recognizing something as a reason and then being moved to act on that consideration is not due to the presence of a further motivating factor (a desire) over and above that recognition. Armed with this alternative account of the relation between reasons and motivation, Scanlon suggests that it is clear that desire does not play the role in either motivation or justification that has often been thought. What is wrong with the argument that it is desires, and only desires, that motivate us to act and that justify those actions? Scanlon argues that having a desire involves, amongst other things, having a tendency to see something as a reason. For example, I experience the sensation of thirst, I believe that drinking to alleviate my thirst would bring about future pleasure, and so I conclude that I have a reason to drink. Although I am anticipating future pleasure, it seems to be the taking of this future pleasure to count in favour of action that does the motivational work; without this further move, I have no reason to act. This is more obvious when we are aware that we are often motivated to do something that we have no desire to do: our duty, take our medicine, live up to our responsibilities or return a wallet full of money that we have found in the street, for example. We act in these ways not out of desire but because we recognize that we have good reason to do so.46 So ‘we should not take “desires” to be a special source of motivation, independent of our seeing things as reasons’.47 What supplies the motivation is the agent’s recognition of something as a reason, the thought that it counts in favour of acting in a certain way. Motivation is not based on the ‘additional element of desire’ as a ‘motivating factor that is independent of such thoughts’.48 Scanlon’s argument is very similar when we turn to questions of justification. He starts by considering what goes on when he is beset by the desire to buy a new computer. Perhaps this involves always thinking about new models, reading and collecting adverts and reviews, visiting computer shops and having a tendency to judge that he has reason to buy a new computer. Scanlon asks whether

130

Reasoning practically

this entails that he actually has a reason to buy a new computer. He is clear that it does not, as he often experiences such desires when his considered judgement is that he has no reason to buy a computer as his present machine is new enough, and he will get little benefit from the additional features of the newer models. Having the desire does not provide the reason to act, and this conclusion is not understood to mean that the desire is outweighed by something (another desire) but rather that the desire is not taken to be a relevant consideration in the justification of a decision about whether or not to buy.49 What this homely example highlights is that reason and desire come apart in justifying action. Instead, the key is ‘taking something to be a relevant consideration’ and this, whether or not it is wholly self-conscious, is a ‘move within practical thinking rather than, as desires are commonly supposed to be, a state which simply occurs and is then a “given” for subsequent deliberation’.50 Indeed, Scanlon convincingly argues that there is no sense in which ‘desires’ can play the justificatory role implied by the claim that they are mental states independent of our practical reasoning that just occur to us and that, in themselves, provide reason for acting to fulfil them. The only sense in which they can play this role is when we think of them as counting in favour of the action so that we take them to be a relevant consideration. But this is now much more obviously an account of a reason than a desire. This makes sense of our experience of practical reasoning, not as the ‘weighing’ of various desires and then acting on whichever we happen to experience most intensely at present, but rather as ‘concerned with figuring out which considerations are relevant to a given decision, that is to say, with interpreting, adjusting, and modifying . . . [our] more general framework of principles of reasoning’.51 So, rather than being necessary to explain motivation and justification and therefore constitute a limit on their scope, a desire plays a role only insofar as it means the same sort of thing as a reason. Reason need not be regarded as a mere slave to the passions. Not only does this conclusion seem to give constructivism the ground that it requires to respond, if we think it necessary, to critics of its supposed motivational deficit, but it also shows clearly why constructivism is very different from contractarianism. Contractarianism is a theory of mutual advantage that assumes that desire plays its traditional role, whereas constructivism need not assume that desire necessarily plays this foundational role at all. Undermining the explanatory necessity of desire disentangles constructivism from contractarianism.

The objectivity and authority of reasons The preceding arguments go some significant way towards responding to the claim that reasons and reasoning are necessarily relative. The claim was that since our reasons are constrained within less-than-universal limits, our reasons must hold only relative to those limits. However, if there are no necessary limits, then relativism need not be necessary either. It became clear when considering the practicality of our reasons that their practical nature depended on their being appropriately related to the relevant context. However, if there are no necessary

Reasoning practically

131

theoretical, practical or motivational constraints on our reasons, then we can draw either of two conclusions. This implies that either reasons need not necessarily be constrained within the contexts that they initially arise or not all contexts are necessarily limited, local and particular. If either conclusion is warranted, and if they are not equivalent it may be that both are, then reasons need not be partial and subjective. This undermines the claim that since all reasoning is constrained, then all constructivism must be a form of secondary constructivism. By undermining the claims about necessary limits, it is clear that there is at least a possible space for a primary constructivism in which reasons are not limited to the particular. As we have undermined the basis of claims about necessary subjectivity, we have, at the same time, paved the way for an account of the objectivity of a properly critical and reflective practical reason. In previous chapters, we have drawn out some aspects of a constructivist account of the objectivity of reasons. Building in the space that we have just identified, it would be helpful to bring these aspects together and to try to piece them together into a fuller account. One way of thinking about objectivity that has been, and still is, very prominent is the idea of objective judgements somehow corresponding to, or giving a proper account of, some set of facts. Many people have taken the objectivity of scientific judgement to consist in the way that it seems to match the facts of the world. This broad approach to objectivity is at the heart of a positivist account that limits objectivity to those statements that are capable of empirical verification (or of withstanding attempts at falsification). It is also, importantly for us, the model for a conception of objectivity that is assumed by foundationalist justification. Foundationalism regards objective judgements as those that fit or correspond to a set of independent and morally significant facts, whether they be Platonic forms, an account of human nature, social understandings or the given desires of individuals. On this understanding, objectivity is a matter of faithfulness to an independent order of moral facts that enjoy a special epistemic status, and all justification must be traceable to a foundation in these facts. An important first step towards an alternative account of objectivity in the justification of principles of justice is the recognition that not even all of our non-moral reasoning regards objectivity as centrally concerned with fidelity to a set of facts. For example, mathematical and geometrical statements such as ‘2 + 2 = 4’ and ‘the sum of the internal angles of a three sided polygon is 180 degrees’ are not regarded as objective because they obviously correspond to anything that we may regard as a fact. We do not, however, ordinarily doubt their objectivity.52 This demonstrates our everyday and ordinary acceptance of a conception of objectivity that seems to be decoupled from claims about fidelity to a set of facts. Constructivism claims that just as the ‘traditional’ account of objectivity can be inappropriate in one case, it might also be so in another.53 So, what conception of objectivity might be appropriate to practical reasoning? In an earlier chapter, we began to identify another account of objectivity available in, and familiar to, our everyday understanding. We are quite used to the idea of objective judges or referees, and their objectivity is a matter of taking

132

Reasoning practically

all relevant factors into consideration, of weighing available evidence and argument and of coming to impartial or unbiased decisions. On this everyday understanding, objectivity is taken to be undermined by partiality, bias, subjectivity, arbitrariness and idealization and reinforced by their opposites. What we expect of objective judges and referees is that their decisions are not based on particular facts about themselves or peculiar attitudes that they have. Rather, we think of them as adopting an impartial standpoint. Traditionally transcending particular standpoints has been thought of as a matter of approximating an ‘Archimedean point’ or a ‘God’s eye view’, both of which involve the idea of standing nowhere on earth.54 However, any standpoint involves standing somewhere rather than somewhere else; this much is implied in the metaphor of ‘standpoints’. The key to constructivist objectivity is to recognize that this is the case and then to do everything that you can to ensure that where you stand is not partial or subjective. If you identify a standpoint that is not partial, then it will be available to all. Partial standpoints are partial because some are excluded from standing there; eliminating partiality removes this exclusion. Rawls recognizes that an objective choice of principles must be made ‘unencumbered by the singularities of the circumstances in which we find ourselves’.55 Nagel expresses this idea well when he encourages us to think of objectivity as a matter of degree. A view or form of thought is more objective than another if it relies less on the specifics of the individual’s makeup and position in the world, or on the character of the particular type of creature he is . . . the less . . . [the identification of right principle] depends on subjective capacities – the more objective it is.56 Minimizing partiality is the task of an ongoing process of critical and practical reflection that kicks in whenever we have reason to suspect that our reasons may be partial in ways that we had not anticipated. Ensuring that our standpoints are not partial is a central feature of the constructivism of both Rawls and O’Neill and is what underwrites their claims concerning objective justification. Both develop a sort of two-stage understanding of objective reasoning. Rawls lays out the original position as his first stage. This is intended to be an objective point of view from which to judge proposed principles of justice. In laying it out, Rawls is clear that he is attempting to identify a choice situation that, as far as possible, screens out partial and subjective influences on choice and thus prevents the justification of principles from being made for partial reasons. This is the role of the veil of ignorance that screens knowledge of conceptions of the good, position in society, natural endowments, etc. As a second stage, Rawls nests this choice situation within a broader notion of reflective equilibrium as an account of critical and reflective reason. The role of this process of reflection is to demonstrate to ourselves, and others, that the original position, and therefore the principles of justice, actually does not depend on subjective reasons. This is especially important if the impar-

Reasoning practically

133

tiality of our principles is threatened by novel encounter or fierce criticism. By holding all aspects of our basic assumptions, trains of reasoning and principles up to critical reflection, as reflective equilibrium entails, we can reassure ourselves that we are not fooled by the seeming impartiality of actually biased reasons. O’Neill, likewise, outlines a meagre choice situation and a fuller account of constructivist practical reasoning that underpins it. She is clear that her account of choice is laid out only by reference to abstract, rather than idealizing, accounts of the person and society. The point of abstraction is that it is not partial, and the reason for avoiding idealization is that it is a partial and exclusionary basis for justification. Ongoing reflection is necessary to ensure that we avoid the subjectivity of idealization. Both Rawls and O’Neill stress what we might regard as a ‘critique of reasons’. Objective justification is a matter of attempting to show that our reasons are not partial and then of accepting the necessity of an ongoing critique of those reasons, although this comes to the forefront only when our objectivity is challenged. This critique has no room for arrogance about our justifications and entails that we can only accept reasons provisionally and humbly as our current best understandings. It is always the case that we might be wrong about the impartiality of our reasons, as they may be partial in ways that we have not yet been able to anticipate. If it becomes clear that this is the case, then we must re-examine our reasons and try to more successfully remove their subjectivity in an effort to make our current understandings as good as they can be. The constructivist accepts the central role of provisionality in a conception of objectivity and, therefore, the centrality of reflective practical reason as a continual reassurance of impartiality. This is obviously in marked contrast to traditional accounts of objectivity that stress the way objective reasons and judgements must be ‘anchored’ to foundational facts that are regarded not as provisional and revisable but as solid and unmovable. The constructivist, in opposition to both the foundationalist and the sceptic, must regard the notion of matching an independent order of moral facts as the very opposite of objective practical justification. A foundational moral fact, in that it is treated as an independent ‘given’ that practical reasoning must accept as its base, is exempt from the reflective process. As such, it is not possible for us to assure ourselves of its objectivity when challenged. By rejecting the notion of moral facts, constructivism accepts that nothing is beyond the scrutiny of our practical reasoning. Particularity, therefore, can always be challenged, and impartiality is guarded by reflection. For the constructivist, it is the attitude whereby all seemingly impartial reasons are held only provisionally rather than foundationally that is central to objectivity. There is a peculiar irony in the conclusion that this account of objectivity, with its suspicion of certainty and of assumptions beyond reflective examination, reverses the traditional relationship between objectivity and what are considered to be ‘facts’. It is clear here that there are questions to be answered concerning the status of reason in constructivism. Most important is the question of whether reason itself is constructed. If it is, what could this mean? If not, does it not then

134

Reasoning practically

constitute a foundation of sorts?57 In the first instance, I am not certain what it would mean to claim that reason is constructed or even how we could tell if it was; after all, what would we use to test this claim if not reason itself? Obviously, the constructivist insists that reasons are constructed, but about reason itself things may not be so clear. On the one hand, we can say that, to the best of our knowledge, reason is a capacity that developed in humans over time; we evolved the ability to reason just as we evolved the ability to walk upright and to grip with opposable thumbs. However, understanding the origin of a capacity to reason, about which I am unqualified to comment, is a separate and very different sort of question from considering the authority of reason, about which I have something to say briefly. It is important for primary constructivism that reason is not understood to gain its authority in the wrong way. Constructivist reason should not be regarded as authoritative because, for example, either it is a reliable aid in discerning an independent order of moral facts or it has been planted in the minds of each and every human being by God. Even if such an external foundation for reason could be identified, then testing whether reason matched it would be something that could be achieved only by the application of reason. Rather than rely on such an external authority, both Kant and O’Neill are clear that reason, if it is to be objective, must somehow underpin its own authority.58 In this vein, and given the constructivist account of objectivity as involving the ongoing critical examination of our always provisional, specific reasons, we can understand the authority of reason in quite a straightforward way. The authority of reason depends not on its relation to something outside of itself but on its proper exercise. I want to suggest that practical reasoning can be thought of as ‘a way of proceeding’. This is the way of proceeding that has been explored in this chapter and this involves marshalling the reasons about which we are currently most confident and exposing them to critical reflection concerning their partiality in the light of challenges and alternative reasons. Note that the ‘aim’ of proceeding this way is not to identify a ‘pattern laid up in heaven’ but simply to reassure ourselves of the ongoing objectivity of our reasons, where objectivity is taken to mean something like ‘does not rely on any partial assumptions or ways of reasoning of which we are currently aware’. Since, as we noted above, reason cannot be tested by anything other than reason there is always a worry that arguments about the authority of reason must be circular. Constructivism should be understood to embrace this worry and to transform it into a positive argument. Accepting that a foundation outside reason is difficult to secure (and to demonstrate without using reason), the constructivist can regard the authority of reason as ‘internal’ by thinking of reason as self-authorizing.59 Reason is properly exercised when it proceeds in the manner we have described. When it proceeds in this way, we are reassured that our reasons are as objective as we can make them and therefore authoritative. If their objectivity, and therefore authority, is challenged, then reason instructs us to reassess our reasons for partiality and to revise as necessary. The revised reasons then are as objective as we can make

Reasoning practically

135

them, and they have authority. In this way, the authority of our reasons is maintained by the proper exercise of reason, and reason can be regarded as selfauthorizing. Once again, it is not that reason itself is constructed, only that its authority is constructed by its proper exercise.

Conclusions We have come to an understanding of the constructivist account of what it is to reason practically. Starting with the thought that practical reasoning is unavoidable, it becomes clear that our reasons must be related in the right way to the appropriate context and that, over time and with audience, the appropriate context may change. This leads to the conclusion that practical reasoning is necessarily reflective and that our reasons can only be held provisionally. Given the provisional nature of our reasons, and their link to contexts, it is reasonable to query whether our reasons are necessarily limited. Upon examination, it appears that the apparent limits on our reasons are not necessary limits, no matter how inescapable they may seem. If reasons are not necessarily subjective, then it remains possible that they are objective. A properly constructivist account of the objectivity of reasons also underpins an account of their authority, whereby practical reasoning, if carried out in the right way, appears to be self-authorizing. This amounts to significant progress in our attempt to characterize constructivism and constructivist reasoning. It is clear that communitarian, multiculturalist or pluralist positions misunderstand the possibilities of practical reasoning. In regarding reasons as necessarily limited, perhaps incommensurable and often dependent on particular cultures or languages, they have underestimated the way reason approaches apparent boundaries and takes the challenge that they represent as one more step in a process of properly reflective reasoning. Whereas the communitarian and friends are very impressed by the variety and difference exhibited in pluralism, and point at it as an argument against objective reasons, the constructivist is not so intimidated by boundaries as to think that they stop us from thinking about them. Rather than point at pluralism and think that this ends argument, the constructivist takes pluralism seriously as the necessary start of an argument and not as the conclusion of one. On the other hand, the foundationalist appears to work with an inappropriate and unavailable conception of objectivity as ‘fitting’ some external ‘facts’. Pluralism is characterized, more often than not, as several such foundational positions stubbornly facing off against each other, each insisting on the truth of their objective foundation to the exclusion of all others. Things need not be this stark for us to recognize that a position that necessarily appeals to controversial religious, moral or philosophical foundations (to paraphrase Rawls) is going to have trouble fulfilling the role of a general and practical response to pluralism. Basing practical justifications on appeals to foundations beyond reflective examination, and that explicitly deny the parallel commitments of others, will most likely entrench rather than respond to disagreement.

136

Reasoning practically

However, constructivism does not rely on a sceptical argument to the conclusion that foundationalism is wrong, nor does it entail the relativism that foundationalists’ worry accompanies the lack of foundation. Rather, constructivism simply sidesteps foundational questions and examines the possibilities of nonfoundational practical reasoning. In fact, the constructivist can go further and claim that, when they are challenged, foundationalists must resort to something like the constructivist account of critical reasoning if they are to respond to the challenge rather than just repeat the foundational claim (perhaps more loudly and forcefully). They must do so, first, in a direct and critically reflective engagement with the substantive reasons proposed by the alternative foundation, idea or culture that constitutes the challenge. Second, in engaging with the challenge they must, if they are not simply to descend into partial dogmatism, accept the possibility of arguments that would lead them to revise their conclusions; they must accept that their conclusions hold either provisionally or dogmatically. If the foundationalist retreats to dogmatic assertion, then they give up on the possibility of justification. The alternative is to accept that the objectivity of our reasons can only be demonstrated by a willingness to subject our reasons to challenge. This does not assume that there are no moral facts, that foundationalism is wrong, but it does suggest that the defence of their objectivity, and therefore the authority of reasons based on them, may have to look constructivist. This should not really be surprising, as the relationship between our reasons and any external foundation would, as noted above, have to be stated, explored and tested using reason.

6

Making practice reasonable Political constructivism

At the start of the previous chapter, an account of practical reasoning was divided into explorations of what is involved in making reason practical and of how we might make practice reasonable. That chapter dealt with the practicality of reasons, and so this chapter will concentrate on the notion of reasonable practice, the political implications of our account of reasoning practically. It will become clear that there are significant implications for the justification of principles of political justice. Not only can we justify some principles of justice, but the principles that we can justify are objective and universal. Of course, this primary construction amounts to a thin set of principles that are largely indeterminate of the content of the bulk of our everyday political practices. However, these principles of justice set limits to reasonable practice and therefore constrain legitimate diversity. What emerges is an understanding of principled pluralism, a diverse range of secondary constructions that respect a thin but primary account of just principles. This chapter will start by outlining a general account of the nature of political constructivism drawing on the constructivist understanding of practical reasoning. Building on this general understanding, this chapter will then lay out a primary constructivist justification of principles of justice. It will then go on to explore the important relationship between primary constructivism and the diverse range of secondary constructions that shape our everyday lives. Finally, it will address some potential criticisms of this enterprise, concluding that political constructivism, understood so as to underpin a properly primary construction, is both possible and important.

The nature of political constructivism As we are aware, the term ‘political constructivism’ is Rawls’s. He uses it in Political Liberalism to describe a form of justification that contrasts with moral realism and with moral constructivism, and that underpins the political conception of justice. Rawls’s political constructivism has four main features. First, principles of political justice may be represented as the outcome of a procedure of construction. Second, the procedure of construction is based on practical and not theoretical reason. Third, it uses complex conceptions of society and the

138

Making practice reasonable

person to give form and structure to its construction. Finally, it specifies an idea of the reasonable rather than a conception of truth (which it leaves instead to comprehensive doctrines).1 Rawls couples this account of political constructivism with his understanding of a political conception of justice that is freestanding in that it is not derived from, or dependent on, any particular comprehensive doctrine and that is constructed using fundamental ideas implicit in the public political culture of a democratic society.2 The complex conceptions of society and person necessary for the political construction find their source, for Rawls, in the fundamental ideas that are part of democratic political culture. Thus, as Rawls says, and many commentators (as we have seen) agree, the justification of his political conception of justice ‘starts from within a certain political tradition’.3 This has encouraged both supporters and critics of Rawls to regard political constructivism as a limited method of justification that applies only within already democratic societies or to people already committed to democratic politics. As it appears to rest on, and be constrained by, prior assumptions of democratic culture, Rawls’s political constructivism seems to be a form of secondary constructivism. It has been one of the main arguments of this book that this narrow understanding of the scope of political constructivism is too limited. It became clear in the previous chapter that there are no necessary boundaries or constraints on reason and justification. If this is the case, then there is no need for political constructivism to maintain a narrow focus on democratic societies. Our arguments have made space for a broader and more ambitious account of political constructivism that may be a proper form of primary constructivism. We need not, with the communitarians, accept a form of social constructivism and confine our justification to particular cultures, nor need we, with the moral constructivists, assume a comprehensive but controversial moral ideal as the basis for construction.4 Rather, a broad and ambitious account of political constructivism can grow, as we shall see, out of considerations at the heart of Political Liberalism but that extends beyond the usual interpretations. Indeed, we can capture the essentials of Rawls’s understanding of political constructivism in this broader account if we build on both the understanding of Political Liberalism that we outlined in Chapter 3 and our account of objectivity and practical reasoning laid out in the previous chapter. The broader account of political constructivism has three main aspects that capture the essentials of Rawls’s account but do not accept their apparent limitations. These are: 1

In the construction of principles of political justice political constructivism draws only on general and abstract resources, which will include conceptions of society and person, and so does not rely on, nor is it derived from, any particular conception of the good or of moral facts.

This is obviously a characteristic of Rawls’s understanding of political constructivism, and it can be seen in the conception of the original position that screens

Making practice reasonable

139

out the particular in order to ensure that decisions of justice are based on what is general and also in the very notion of a ‘freestanding’ political conception of justice. More generally, both O’Neill and Scanlon, for example, also accept the constraint on appeals to conceptions of the good or moral facts.5 2

Political constructivism also draws on an account of objectivity as impartiality or non-subjectivity.

This is the account that was laid out in the previous chapter, and this underpins the generality and authority of reasons of justice. This account of objectivity is particularly suited to a constructivist account of justification. 3

Political constructivism involves an account of reflective and critical practical reasoning that is concerned to continuously hold 1 to the standard of 2, with the effect that no conception or assumption at 1 can ever be more than provisional.

In this way, political constructivism involves the always ongoing application of practical reason to continually arising questions of the justification of principles of justice. This is entailed by the account of objectivity and the authority of reasons that draw on it. As a result, no political construction can be regarded as tied to any particular starting points (no matter what starting assumptions we actually have). At the same time, its conclusions will be provisional but objective as they will not be based on any subjective or partial assumption of which we are currently aware. As such, a political construction will exhibit the features of a properly primary constructivism. I continue to understand this constructivist position to be a form of political constructivism; first, because of its obviously close ties to Rawls’s position, and second, because what is constructed are principles of political justice. Of course, these are also moral principles, but there is no suggestion that their justification either involves or depends on the much fuller construction of morality in general. Nor does this position require us to take any particular position on the social construction of beliefs, attitudes, practices, norms or cultures. As we have noted before, such considerations are more relevant to projects of explanation than of justification, and it is the justification of principles of political justice that is at issue here. This broader understanding of political constructivism is central to two features of this book. First, this book argues that this form of political constructivism can be found in, and indeed helps us to understand, the arguments of both Rawls and O’Neill. Whether they conceive of their arguments in this way or not, it is clear that this position appears to characterize much of their work, and I think that we would be missing something significant if we try to understand their arguments without drawing on this general conception of political constructivism. Second, this book builds an independent argument to this conception of political constructivism. The previous chapter laid out the necessary

140

Making practice reasonable

conceptions of practical reason and objectivity. The next sections of this chapter will highlight the nature and place of general and abstract conceptions of society and the person in the argument.

Constructing principles of political justice Whilst in the previous chapter we concentrated on what it takes to ensure that our reasoning is practical, constructing principles of political justice is concerned with making our practice reasonable, and the broad notion of political constructivism helps us to understand how to do this. We have already mapped out an appropriate conception of objectivity. Political constructivism then invites us to construct principles of political justice using only general and abstract resources. Their generality and abstractness, and therefore their lack of partiality, is ensured (as far as possible) by our understanding of properly practical reasoning that holds our constructed principles (and the resources utilized in their construction) to its standard of objectivity. We need to try to understand what is involved in this process of construction. If we are turning our practical reasoning to the subject of principles of political justice, then we must examine what assumptions we might legitimately make about our political environment. In order to identify such principles, we must supplement our broad account of political constructivism and practical reasoning with an account of the basic assumptions underlying moral and political practices. When Rawls claims that, in constructing principles of justice, we should orient our practical reasoning with conceptions of society and person, he is pointing out that the basic assumptions underlying our practices will include at least some conception of society and some conception of the person. Some form of these conceptions must already structure our everyday thinking about morality and politics. That this is the case is obvious upon brief reflection. When we think about politics, justice and ethics, we necessarily refer to our general beliefs about what people are like and what societies are for. The sort of normative considerations that principles of justice respond to are at issue only when there is a group of people whose actions affect each other. Reasoning concerning these considerations then assumes an account of that plurality, a conception of the society. Conceptions of society help us to understand what society is and, perhaps, what it is for. Such conceptions may be very different; we might conceive of society as a way of working together for mutual gain or as a group of people united by a common religious or moral purpose or vision, for example. Whatever the appropriate conception of society, we will also assume an account of the people who make up that plurality, a conception of the person. Conceptions of the person underpin our understandings of what people, in general, are like. For example, we might conceive of people as utility maximizers, as Kantian autonomous moral subjects, as sinners or as seekers after righteousness, amongst many alternatives. That we, if we are to construct principles of political justice, need to come to an appropriate understanding of conceptions of society and person is a plausible claim since the notions of society and the person have a

Making practice reasonable

141

conceptual relationship with the circumstances of justice, circumstances that are taken to structure moral and political interaction. This must be the case as in scenarios where there is no plurality or where there is no possibility of interaction between persons, there can be no normative questions of politics or morality and we would therefore be outside the circumstances of justice. So, if we are reasoning about our moral and political practices, we will have to make a range of basic assumptions that at least include assumptions about society and the person. As we learnt in the previous chapter, our practical reasoning necessarily makes some assumptions. We have also learned that the assumptions that we make have to be responsive to environment, appropriate to context. This might be taken to imply the apparently relativist conclusion that different conceptions of society and the person may be appropriate in different circumstances. For example, in Political Liberalism, Rawls is clear that he thinks conceptions of society as a fair system of social cooperation for reciprocal advantage and of the person as a free and equal citizen are appropriate to reasoning about the practice of justice in a democratic political environment.6 We have seen that many commentators feel that such conceptions might be highly inappropriate assumptions to make of all manner of historical and contemporary societies or practices. The conceptions of society and the person that it is appropriate to assume in a particular environment will be those that, in O’Neill’s terms, it is at least possible will be acceptable for those to whom we offer reasons.7 To do otherwise would be to impose conceptions from the outside. Because, as we have seen, environments are plural and diverse, appropriate reasons and justifications may also be plural and diverse. Justification and reasoning may be complicated by the necessity of speaking differently to different people. We find ourselves apparently having arrived at a position in which political constructivism outlines a formal universal account of practical reasoning that may be ‘filled in’ in very different ways in particular societies. On this understanding, the different conceptions of society and person appropriate to different environments (communities or societies) function like a range of prisms through which our practical reasoning is refracted, and, as each prism is different, the substantive principles constructed in each context are different. In this way, political constructivism might be taken to explain our experience of political pluralism or to justify the wide variety of apparently competing principles of justice; each is justified in its context. On the face of things, political constructivism simply reproduces the pluralism with which we started. If this formal account of political constructivism, which underpins substantive pluralism, was to conclude our account of constructivist justice, then this could still be counted as some sort of achievement. Recall that the challenge to political justification posed by pluralism has often been interpreted to imply the claim that practical reasoning, like principles of justice (on this understanding), is itself plural and bounded by context. In establishing that there are no necessary boundaries to our reasons, then we would, at the least, have dispelled this claim. However, establishing the possibility of objective and perhaps universal

142

Making practice reasonable

reasons may enable us to do more to meet the challenge of pluralism. If our reasons are not necessarily limited by context, then perhaps, on some questions, the prospective audience (and therefore the relevant environment for justification) may be everybody. In casting doubt on the existence of absolute local limits to reasons, this broader possibility is opened up, the possibility of constructing principles of political justice that are the outcome of a political constructivism that is substantively, rather than just formally, universal. Since the construction of principles of political justice involves making assumptions about society and the person, if we are to construct principles that are justified to an audience of everybody, we will need to identify assumptions of society and person that are appropriate to this context. It is necessary to identify conceptions of society and the person that are generally appropriate; we must identify what amount to bare concepts of society and the person that are assumptions of any political practice. Whilst other more partial assumptions appropriate to local political cultures might underpin secondary and particular constructions, the assumptions that underpin a properly primary construction must not be partial in this way. Identifying such ‘bare concepts’ is a tall order. However, it would seem to be a necessary step if the constraints of practical reasoning, outlined in the previous chapter, are going to be more than formal at the universal level.8 As we have seen, both Rawls and O’Neill implicitly recognize the need for some such account of these concepts. Rawls understands the basic concept of society to be a system of social cooperation. This basic understanding can be ‘thickened’ in different ways. So, whilst implicit in democratic culture is a particular conception of society as a fair system of social cooperation over time and for reciprocal advantage, Rawls also notes alternative conceptions, most notably of society, as social cooperation for mutual advantage.9 He understands the concept of the person to be of someone who can take part in social life and that this involves them being reasonable and rational in that they can understand and act on practical principles and they have their own conception of the good.10 Again, this basic understanding can be ‘thickened’. In democratic societies, we embellish this by also thinking of citizens as free and equal, but there is no suggestion that this need be the case outside the tradition of democratic thought. Rather, a person is, more broadly, someone who can be a citizen, a normal and cooperating member of society.11 That these are bare concepts of society and person is implied when Rawls says that ‘no sensible view can possibly get by without . . . them’, that they are necessary for any conception of justice or the good and that ‘they are as basic as the ideas of judgement and inference, and the principles of practical reason’.12 Rawls gives us an account of bare concepts of society and person, of society as a scheme of social cooperation and of the person as capable of involvement in that social cooperation and as having their own conception of the good. Likewise, we have seen that O’Neill offers her own account of these bare concepts. No matter what our particular beliefs about persons and societies, there are some basic assumptions that we necessarily make whenever we act.

Making practice reasonable

143

We assume plurality (that there are others), connection (that our actions affect each other) and finitude (that we are all vulnerable in one way or another).13 More formally, O’Neill argues for a minimal concept of society as ‘a plurality of at least potentially interacting agents’ without ‘pre-established harmony’ or necessary common commitment. Her concept of the person is of someone with ‘the capacity to understand and follow some form of social life and with a commitment to seek some means to any ends . . . to which they are [happen to be] committed’.14 Both Rawls and O’Neill offer us accounts of the bare concepts of society and person. What is more, these accounts are very similar. O’Neill’s account of the person as capable of involvement in a social life but also as pursing some good of their own is remarkably similar to Rawls’s account of the person as reasonable and rational. The person is reasonable in that they have the capacity for a sense of justice, which is simply to understand, apply and act on the principles that govern social cooperation. The person is rational in that they have the capacity for a conception of the good, which is just to say that they can form, revise and pursue a conception of their own rational advantage or good.15 In actually having a conception of the good, Rawls’s ‘person’ also shares the vulnerability and neediness implied by O’Neill’s account of a person’s finitude. In one sense, it is remarkable that these accounts are as close as they are, and in another, it is not so remarkable. O’Neill explicitly draws her bare concepts from a minimal understanding of the circumstances of justice.16 The circumstances of justice are understood to be the circumstances that must obtain if questions of justice are to be at stake. Political theory has long understood these circumstances as those where there are a plurality of people, each with their own needs, desires or good, and where this is coupled with a scarcity of resources such that not all desires can be fulfilled and so there is the potential for conflict. As we have already seen, these circumstances have been widely assumed in political theory, notably by Hume, Hobbes and Rawls, for example. Much of political theory has been concerned with understanding the possibilities for social cooperation in these circumstances. Justice is only a concern where, as O’Neill puts it, there is a plurality of people whose actions potentially affect each other and where those people are needy and potentially competing for goods. The circumstances of justice themselves assume bare concepts of society and person, where, if justice is a possibility, there are a number of needy people capable of some sort of social cooperation or other in the pursuit of whatever they understand their (diverse) goods to be. Since this understanding of the circumstances of justice is a long-standing assumption of much political theory, it should not come as any real surprise when the bare concepts of society and person advanced by Rawls and O’Neill are similar. They are similar to each other just as they are similar to the account inherent in the circumstances of justice, as each is responding to a basic question about the possible justification of principles of political justice. We have recognized that some account of the bare concepts of society and the person is necessary if any universal principles of justice are to be justified by

144

Making practice reasonable

political constructivism. That Rawls and O’Neill offer similar accounts, and that these are both similar to the account implied by the long-standing understanding of the circumstances of justice, may give us good reason to recognize the power of understanding the bare concepts in this way. This convergence of accounts should encourage us to take Rawls’s and O’Neill’s accounts very seriously and to regard them as describing the best understanding of the bare concepts currently available. Of course, this understanding of society and person is provisional and always subject to ongoing critical examination, with the aim of uncovering ways in which it is partial or unsatisfactory. So far, however, as an account of general and abstract resources for the construction of principles of political justice, they appear considerably less partial than the specific conceptions of society and person assumed in particular societies. They are certainly thinner than the democratic conceptions Rawls finds in democratic public political culture. Indeed, we may ask what it would mean for a conception of justice to reject the assumption of the bare concepts as Rawls and O’Neill understand them. Neither Rawls nor O’Neill believe that such rejection is consistently possible as it involves rejecting either the assumption that there is a plurality of vulnerable but connected people or the assumption that social cooperation (and therefore principles of justice) is possible.17 As such, I will regard these accounts of the bare concepts of society and person, always tentatively and with the understanding that they may be partial in ways that are not yet clear, as general and abstract resources available for the construction of principles of justice. Employing these bare concepts in the construction of principles of political justice is the next step. In doing so, our first concern must be to learn the lessons gleaned from our examination of Rawls and O’Neill. In A Theory of Justice, and to some extent in Political Liberalism, Rawls oversteps the limits of constructivist possibility in the pursuit of determinate principles of justice, and O’Neill may do likewise in the pursuit of principles of virtue. In trying to draw out substantive principles of justice, we must actively avoid either building in too much to our starting points (importing unavailable resources) or attempting to construct more than is justifiable from available resources. Concerns on both these counts surfaced most clearly in our examination of A Theory of Justice. Rawls develops justice as fairness as a response to the indeterminacy of intuitionism, and his explicit intention is to identify principles of justice that will correct the arbitrariness of the world. The veil of ignorance is a significant step away from the arbitrary and particular towards the abstract and general, but, in abstracting from our specific motivations, it introduces a certain arbitrariness of its own. Deprived of important knowledge of themselves and their situation, how can the parties to the original position make reasoned decisions about justifiable principles of justice? It is precisely in order to identify a determinate set of principles of justice that Rawls introduces a range of sharply contested additional features, such as an account of primary goods and of motivation and the Aristotelean principle. The original position shaped by these features enables Rawls to conduct (what he sometimes refers to as) his ‘moral geometry’ and identify the determinate principles of justice that the parties to the original position

Making practice reasonable

145

would choose, not only at the level of principles of justice but also, through a four-stage lifting of the veil, at the constitutional, the legislative and the judicial and administrative levels too. As we have seen, it is the contention of his critics that this level of determinacy in the construction of principles is only possible because Rawls implicitly assumes a very thick and markedly partial conception of the person that possibly plays a foundational role wholly out of step with constructivist justification. On the face of things, in Political Liberalism, Rawls accepts this assessment of his work. Here, he appears to maintain the pursuit of a highly determinate conception of justice whilst accepting the partiality of his resources. His political constructivism appears to be a form of secondary constructivism that finds its foundation in democratic political culture. Even once we acknowledge the thin universal primary constructivism that lies behind Rawls’s position, in filling this in for democratic societies, he maintains the drive to identify a determinate set of democratic principles. It is this thick and determinate account that has attracted much of the critical attention. Finally, we have seen that O’Neill’s construction of virtue as a way of ‘filling in’ a thin primary construction of justice may overstep the mark in a similar way. In assuming that our neediness implies a general range of imperfect obligations, O’Neill appears to be asking her constructivism to do more than it may be capable of doing. This pattern of concerns encourages the adoption of an attitude of watchful wariness and caution about placing too much pressure on political constructivism. We should be very careful about expecting constructivism to justify too full and determinate a set of principles. Recognizing that political constructivism functions at two levels, a thin and universal primary level and a thicker and more determinate secondary level, is a major step towards acknowledging the dangers of attempts to over-determine constructivist outcomes. So, rather than strain constructivist possibilities at the primary level, it is at the secondary level, drawing on the much thicker resources appropriate in particular cultures or communities, that constructivism can satisfy a concern for thick and determinate constructions. With this division of labour in mind, we can turn to the construction of principles at the primary level. Rawls constructs principles of justice from an original position that is nested within a broader account of reflective equilibrium as complementary aspects of a two-part account of objectivity in practical reasoning. In a similar way, we can use the concepts of society and the person to lay out a minimal choice situation, which is itself informed and governed by our account of constructivist practical reasoning. We must ask, ‘What sort of situation and principles could such bare concepts underpin?’ Drawing on Rawls and O’Neill, persons are simply conceived of as capable of social cooperation and of having some conception of the good or other that they have adopted and are concerned to advance (and therefore must be thought of as having needs or vulnerabilities). The concept of society is little more than the acceptance that, beyond recognizing that human beings necessarily interact with each other, we should not assume any particular principles of social organization or cooperation from the outset, but merely

146

Making practice reasonable

assume that some sort of structured interaction is inevitable. Given these minimal understandings, we ask, ‘What principles could this plurality of persons consistently adopt as ongoing principles of justice to govern their interactions?’ Because the concepts of society and person that frame answers to this question are so thin, it is clear that a great range of principles could be consistently adopted. Any number of principles of justice could be generally acceptable to persons so conceived. For this reason, it is more interesting to focus on those principles that could not be so adopted, on what is ruled out rather than on what is ruled in. It seems clear that no principle of justice that justifies inflicting serious injury or death could be consistently adopted by all persons as a general principle for governing their interaction. Adopting such a principle would entail impairing the ability of some persons to interact at all. As such, a principle that legitimates these actions is necessarily partial, arbitrarily favouring some to the detriment of others. Proposing such a principle entails the expectation that some persons, aware that they are committed to some conception of the good or other, accept principles that ensure that they are unable to pursue it or any alternative conception they may adopt in the future. We can accept that there may be some instances where a person, in the light of circumstances and on the basis of values and beliefs that they are committed to, may find themselves in a position where they feel strongly that sacrificing their life is a justifiable course of action. However, it is important to note that this decision is made by a particular person drawing on a thick understanding of context and personal commitment. It is another thing altogether to suggest that persons in general, knowing that context and commitments vary but not what they are personally committed to, are able to collectively adopt principles of justice that make such sacrifice a matter of course. So, while a whole range of principles are ruled in, at least some are ruled out. It is important to understand that these decisions are not made on instrumental grounds. It is not that the abstract persons think, ‘given that I have these ends it is prudent for me to reject certain principles’.18 If their reasoning took the form of questions like this, then when we ask what persons could agree to, it is clear that they could agree to just about anything. However, principles licensing serious injury and death are instead ruled out, because they literally cannot be adopted by all persons. If some of a plurality of persons adopt them, then others (their victims) cannot. It may be that the principles that are ruled out in this way form a very minimal set. This should be no surprise, given the thinness and level of abstraction of the general resources informing this political constructivism. However, the claim is that at least these violent principles are not generally justifiable. This does not mean that further principles will not be susceptible to the same process of elimination. For example, O’Neill’s construction of the principles of virtue stems from an argument that persons in this broad form of choice situation would also reject a principle of general indifference. Likewise, it seems possible that principles of justice that involve relegating some persons to the status of second-class citizens would fall foul of this test. It may be that wariness about overstepping the limits of political constructivism can make us too reticent

Making practice reasonable

147

about advancing arguments that a whole range of principles will be ruled out. On the other hand, caution is warranted here, as the more claims we make about the adoption and rejection of proposed principles of justice, the more probable it is that we will be building in partiality where we should not. The same understanding of the objectivity and authority of practical reasoning that forces us to accept that all principles are only provisionally justified, and therefore always subject to ongoing critique in case of unnoticed partiality, underwrites this caution. So, in the first instance, we say only that at least some principles of individual inviolability are constructed. Constructing further principles of justice is a possibility, but this goes beyond the present aim of establishing just that some universal principles are justified by primary constructivism. It may appear that we have had to do a lot of work for very limited gains. In a certain respect, this is correct; we have expended a significant amount of argumentative energy and it may be thought that we have little to show for it. However, the little that we have justified is of the utmost importance. Recall the challenge that universal justification faces. It has been argued, as outlined in the introduction, that claims about universal principles are either made on the basis of overconfidence in the foundational character of partial assumptions or are not justifiable at all. In either case, the outcome is a radical pluralism that has fuelled widespread claims about incommensurability and the inherent limits on reason and justification. In the face of this relativist challenge, it is vitally important for constructivism to demonstrate that at least some universal principles are justifiable. Political constructivism does appear to have secured the possibility of universal justification. Further than that, it has identified and justified substantive universal principles of justice concerning a level of individual inviolability. Given the strength of the challenge to universal principles in recent years, that this constructivist account of universalism is thin is not a criticism that we need to worry about. Instead, we can accept this thin universalism and welcome the limitations on reasonable practice that it imposes. Political constructivism makes real progress in the justification of universal principles. To expect much more than this in the first instance is to run the same risks as Rawls and O’Neill, overstretching and sacrificing objectivity and impartiality for thicker and more determinate principles. It is therefore good sense and not undue reticence that presently limits constructivist claims to the justification of only a thin universalism.

The relationship between primary and secondary constructivism Political constructivism permits or, more precisely, does not rule out a great range of possible principles for governing our social interaction. It is these principles that are often already embedded in the cultures and ways of life of particular communities or societies. It is important to understand that primary constructivism has nothing more to say about either the principles or the ways of life in which we find them. Primary constructivism is focused on ruling out injustice rather than dictating the shape of just societies. Rather than attempt to

148

Making practice reasonable

identify the optimally just political regime, the project is to set the limits of reasonable practice by identifying injustice. D’Agostino, in Free Public Reason, a comprehensive critique of contemporary theories of public justification, does not distinguish the two levels of justification. As a consequence of this he expects broadly constructivist theories such as Rawls’s to pick out the optimal regime, and he holds them to account for indeterminacy when they fail to meet this standard.19 Requiring constructivism to identify the optimal regime expects an inappropriate sort of outcome. Primary constructivism effectively identifies a threshold rather than an optimum and may have nothing more to say about principles that cross that threshold. This is not to claim, however, that there is nothing that we can say about secondary constructivism, and it is important to try to be clearer about its relationship with the primary level. A secondary constructivism is any constructivism that can be thought of as building from starting assumptions which it regards itself as unable to subject to criticism, or which it exempts from critical examination. Secondary constructivism describes a form of justification that works from foundational grounding assumptions towards the identification of principles or practices. We can understand particular ways of life, and the principles they embody, in this way as constructed on the basis of a thick raft of shared communal beliefs or attitudes. It is this ‘thickness’ that describes their affirmation of one way of life amongst others. Where a community or society shares a particular set of grounding assumptions and way of life, other communities share very different sets of assumptions and ways of living that identify different, and often incompatible, principles and practices. The thickness of secondary constructions reflects the level of commitment necessary in order to ‘buy into’ the way of life; to regard the principles and practices as justified you need a prior commitment to a range of partial assumptions. The assumptions are partial, as buying into one way of life, on the basis of a particular set of beliefs and attitudes, involves having to reject many other ways of living and the beliefs, attitudes and assumptions that inform them. Recognizing pluralism is a matter of recognizing that there are many different secondary constructions. On the other hand, primary constructivism involves a refusal to exempt any assumptions or attitudes from examination by critical practical reasoning. If successful, the assumptions of primary constructivism do not require a high level of commitment prior to acceptance. This is because its assumptions are thin and impartial, held provisionally and awaiting revision if evidence of their partiality comes to light. As such, the conclusions of primary constructivism, being impartial, should be objective and be justifiable to everyone. Where there is a vast plurality of thick secondary constructions, there is only one thin primary construction. The relationship between these two levels of construction can be understood in two broad ways. The first is described by Walzer when he outlines the relationship between his notion of thin universal morality and the range of thick and particular maximal moralities that constitute different societies. Walzer argues that as maximal moralities respond to challenges and interact with each other, we may get hint that there is some sort of common experience reiterated in dif-

Making practice reasonable

149

ferent contexts. He is clear, though, that any argument for thin universalism remains tied to the secondary. ‘Morality is thick from the beginning, culturally integrated, fully resonant, and it reveals itself thinly only on special occasions’, and even on these ‘special occasions’, the logic of the argument is thoroughly particularist.20 Walzer’s thin universalism is based on the extension of feelings of solidarity, and solidarity is a construction not of reason but of the imagination. Imagination underpins a process of sympathetic identification with the situation of others.21 We recognize the suffering of others by thinking about our own. What we recognize here is a series of ‘family resemblances’, of ‘commonalities’, and Walzer tells us that ‘the sum of such recognitions is what I mean by minimal morality’.22 He argues that it is possible to keep adding to the range of people or communities with which we can identify or feel solidarity to the point where the scope will approach the universal. It is clear that Walzer’s universalism is based not on persuasive universal reasons but on imagination and on feelings of sympathy, identification and solidarity.23 Rorty understands this relationship in roughly the same way as Walzer. He argues that the ‘human rights culture’ is a function of the scope of our sympathies and the range of people with whom we can feel solidarity. ‘Universal’ morality extends only as far as we feel the boundaries of ‘our’ moral community extend. We are capable of expressing solidarity with those we can think of as ‘one of us’ through an ‘imaginative identification’ with their lives.24 The scope and application of moral principles is for Rorty, as it is for Walzer, explicitly not a function of our reason but is instead a matter of which people we are able to recognize as enough like us for our sympathy to kick in and for strong feelings of solidarity to develop. Walzer’s universalism, like Rorty’s, is particularist through and through, even on special occasions, and is only contingently universalist. Walzer’s universalism can offer reasons only to those with whom we currently feel solidarity built on sympathy and common experience. As such, the scope of these reasons is contingent upon the range of our current sympathies, offering reasons to an arbitrary subset of humanity instead of to all. Walzer and Rorty misunderstand the relationship between the two levels of construction. It is clear that they regard the very features of our common life that enable secondary constructivism, the thick identification with a set of uniting values and commitments and strong feelings of solidarity with a particular set of people, as the things that keep us apart at the primary level. It is only to the extent that we can extend the basis of our thick secondary commitments to include others that we can accept principles that apply at a broader level. Any thin account of justice always refers to, and finds its justification in, the particular and thick ways of life we inhabit. The primary constructivism is constructed on the overlap between secondary constructions. Need we understand the relationship between the primary and secondary levels in this way? This reiterative account of a thin universalism may very well provide an account of how we come to recognize and affirm universal principles. In this way, Walzer may provide a broadly descriptive and largely sociological account of the genesis of a thin universalism, but this is not necessarily an account of its justification. As

150

Making practice reasonable

such, a theory of constructivist justification should, if possible, set this approach aside. Instead, we can understand this relationship between primary and secondary constructivism in a much more helpful way. Secondary constructions, the principles, values and practices of specific societies, are the seat of particular justifications. These particular justifications are of two broad sorts. First, the various secondary constructions will embody different particular justifications of the primary construction. This is not to claim that secondary constructions are derived from the primary or even identified by reference to it. Instead, the secondary constructions that are not ruled out at the primary level will include, perhaps only implicitly, their own way of thinking about the principles of justice identified at the primary level. There are many possible ways in which a right not to be killed or injured can be justified, institutionalized and probably extended and supplemented with the thicker resources available at the secondary level. However, the universal and cosmopolitan justification for these primary principles will not always be apparent, recognized or conscious. The secondary constructions are also the seat of the particular justifications of the special rights and obligations that ordinarily exist only between the various persons and groups that inhabit particular contexts. Indeed, it is the particular relationships that exist between the group members in distinction from non-members, and the thick set of shared assumptions embodied in those relationships, that are the specific ground of such rights and obligations. These extensive commitments constitute the bulk of the thick moral contexts within which our lives are played out and, as such, can only be justified at the local level and between the members of those local contexts.25 It is the thick and local nature of secondary constructivism that has so impressed the communitarians, multiculturalists and other critics of universal principles of justice that they have claimed that this sort of justification is the only sort possible. Walzer’s account of reiterative universalism barely gets beyond this. Secondary constructivism does accurately reflect the ways in which we reason most of the time. Usually, we cruise comfortably through our day to day lives, and we are able to take for granted the bulk of our cultural, political and normative environment, content with whom we are and the judgements that we make. As a community we, by and large, regard many questions as settled and so share a settled cultural and normative environment. Our identities and values are reinforced by our neighbours and our ideals embodied in public culture. In this way, we can understand the settled set of institutions, social understandings and conceptions of the good that constitute this culture as the construction of the various members of our community, jointly constructed, possibly over a long period of time, and characterized by the evolution and development of traditional ways of doing things. However, secondary constructions are not enough. We cannot avoid being pushed from the secondary to consideration of the primary level of construction. This ‘push’ is experienced by individuals and societies when we become aware of the apparently contingent nature of our fundamental commitments or perhaps

Making practice reasonable

151

when we encounter the alternative cultural practices and principles of social organization embodied in other secondary constructions. Encountering novel institutions, principles and ideas can unsettle us, or perhaps inspire us, and is experienced as the push to re-examine our own assumptions in the light of these alternatives. If we become aware of them, alternatives to our best understanding of principles of justice demand our attention. Rather than ignore them, in order for our principles of justice to remain critically justified, we must either satisfy ourselves that we have good reason to reject the alternatives and maintain our current understandings or accept that we have good reason to modify those understandings. Doing neither involves withdrawing from the practice of justification in favour of a practice of faith. So, the secondary construction is just the set of social, political and moral principles and practices that constitute the cultural community, and we ordinarily refer to it when we need to justify ourselves and our practice. However, we cannot always only refer to this secondary level of justification. There will always be times when, responding to questions and challenges posed both by developments within our own culture and by encounters with the cultures of others, we are required to think beyond the boundaries of our current practice. Faced with the problem of justification when we are either unsure of our cultural inheritance or with people who do not share our basic cultural assumptions, we are pushed beyond the confines of our particular culture and towards the primary level. If the alternatives we encounter are diverse enough, and significantly different from our own understanding, then we may find ourselves pushed all the way to the bare assumptions about society and person and the primary construction that they underpin. It is important to understand that these bare concepts are, therefore, not just understood as the resources for primary constructivism. As the argument is that they will be basic assumptions of any principles of justice at the secondary level, they also function as the gateway from any particular secondary construction into the primary. The bare concepts do not conflict with but underlie thicker secondary conceptions and, as such, no one is excluded from primary justification. Finally, it is also important to note that this push to the primary constitutes a pursuit of continued justification; that is, the search is for better reasons rather than mutual sympathy or mutual antagonism. Identifying good reasons for universal principles of justice in no way diminishes or negates the importance of the rest of our moral and cultural obligations. The secondary constructions that are the communities and cultures within which we live our day-to-day lives are not necessarily threatened by the thin primary principles. They may indeed be supported and nurtured by the widespread recognition of our universal obligations. Universal principles of justice, generally accepted, would provide valuable levels of protection for the various secondary constructions and for their members. Recall that primary constructivism has nothing to say about the secondary constructions that meet the threshold of respect for basic principles against arbitrary violence. It is only when they adopt necessarily partial principles presupposing killing and injury that the primary framework intrudes at the secondary level at all. In this way,

152

Making practice reasonable

primary constructivism must be regarded, not as the foundation or basis of secondary constructions but, as setting minimal limits to the acceptability of whatever we construct at the secondary level. Within these limits, we would expect secondary constructions to flourish or not depending on the wishes and behaviour of their members. One reason to expect them to flourish is that, if the primary principles are generally acknowledged, the violent imposition of a culture on members of another would not be acceptable. While the primary level outlines a basic framework of security, it is certainly not thick enough in itself to constitute the basis of a way of life. Thin universal principles cannot even encompass the whole of our moral commitments. They will always be supplemented by more local principles that are themselves responses to the particular normative environments developed by groups of people, sometimes consciously but more often unconsciously, over centuries. As we have seen, it is these secondary constructions, built on our complex interrelationships with their thick understanding of our local claims on, and obligations to, each other, which are the basis for our full ways of life. There should be no suggestion that principles identified at the primary level could be expected to displace our secondary commitments. Obviously, this is not to imply that every way of life is unaffected by the justification of universal principles, as clearly some ways of living and acting fail to pass the appropriate threshold of justice. This is not simply the recognition of pluralism as such but rather the identification of a principled pluralism. Whereas critics of universal principles identify, and often celebrate, the fact of pluralism, they appear to regard it as a bar to discussion and as an unprincipled mishmash of contradictory or perhaps incommensurable positions. This approach runs the risk of emphasizing the differences between ways of life and of heightening the sense in which groups simply confront each other. In contrast, constructivism argues that recognition of the fact of pluralism is an important motor for critical deliberation on individual and group interaction. The successful justification of a primary constructivism would effectively relate the plurality of principles, positions and practices to one another in a principled manner. It identifies an acceptable and justified framework from within which the variety of ways of life need not be regarded as contradictory or antagonistic. Both Rawls and Larmore recognize the need to legitimate the reasonableness of disagreement and also the broad acceptance of diversity that follows.26 It is exactly the reasonableness of disagreement that is recognized in the notion of a principled pluralism. The alternative is to regard those who disagree with us as unreasonable or irrational, and unreasonable people do not deal with each other in a principled way. By recognizing that disagreement is both reasonable and expected, the principled pluralism that is the consequence of the justification of primary principles of justice effectively sets the limits of toleration. The only principles or ways of life that should not be tolerated are those that fail to pass the justice threshold, while we have principled reasons for tolerating those that do. So, the fact of pluralism appears, on first glance, to pose serious problems for the universal justification of principles of political justice. However, we have seen that the experience of pluralism pushes us to critically reflect on both our

Making practice reasonable

153

relationship to other societies and to the principles and values implicit in our secondary constructions, and therefore towards the idea of a primary constructivism. In distinguishing between these two levels of justification, the thin and universal and the thick and particular, we are able to understand pluralism in the context of a principled framework for interaction.

Conclusion At the beginning of this book, we identified the profound challenge to universalist theory posed by diversity. Faced by the apparent relativist implications of pluralism, many political theorists have tried to identify the foundations on which to anchor universal justification. Foundationalism regards pluralism as a mistake; it implies that a lot of people must be committed to principles that are just plain wrong. While it is quite plausible that many people hold wrong principles, it is not obvious that pluralism is necessarily a mistake. As such, denying pluralism on the basis of foundational claims may be unsuited to a project of general justification. The political constructivist position laid out here does not refute foundationalism as a strategy of justification, nor does it necessarily invalidate particular foundationalist accounts. Instead, it has set aside the pursuit of foundational givens, not because this pursuit is wrong, but because such justifications run the risk of partiality. In attempting to avoid partial assumptions, constructivist justification takes pluralism very seriously indeed. The constructivist project is an exploration of the possibility of generally justified principles in conditions of diversity and is a response to the ‘fact of pluralism’. Of course, constructivism is not claiming that pluralism is necessarily true either; to do so would be to undermine all foundational positions. Nor is constructivism to be identified with the liberal pluralist positions that value pluralism as a good in itself.27 Rather, constructivism takes the appearance of pluralism seriously and goes on to show how, despite this apparent diversity, general and universal principles can be justified whilst at the same time showing us how pluralism can be understood. It does this in the primary constructivist account of principled pluralism, a thin universalism that sets out the limits of reasonable practice, coupled with an understanding of diverse but legitimate secondary constructions. As already noted, constructivism does not either rely on or deny foundationalism. The political constructivist claim is that ‘without foundational assumptions we can justify at least this’; to foundationalism we can still ask ‘show me what else can be justified!’ Constructivism is concerned with universal and objective justification. It recognizes the apparent practical and reflective power of reason and goes on to outline an account of the authority and objectivity of reasons that goes some way to explaining its practical and reflective possibilities. It does so in terms of a general ‘critique of reasons’ that involves examining each of our reasons and assumptions for partiality and accepting the provisional authority of reasons that, upon examination, have not been shown to be partial.28 This critical reasoning is oriented by our provisionally best accounts of the concepts of

154

Making practice reasonable

society and person, since these concepts constitute our best available understanding of the minimal circumstances of justice. Laying these concepts out as a form of ‘choice situation’, constructing principles of political justice involves asking what persons, minimally conceived, would rule out as principles to govern their basic social interaction. Universal principles of justice are those that rule out only the set of prohibited principles, and these will include at least some principles of personal inviolability that protect persons from killing and injury. Constructivism yields at least some substantive conclusions about universal principles and, in distinguishing between primary and secondary levels of justification, tallies with our ordinary recognition that at least some of our principles have a categorical nature. Several analyses explore the idea that constructivism identifies justified principles with those that are the outcome of some form of choice situation or procedure of construction.29 We can accept this characterization, but only so long as it is understood that this procedure of construction is necessarily nested within a broader constructivist understanding of practical reasoning, its authority and objectivity. This is not to make the further claim, made by Milo, that constructivism identifies moral truths about rightness and wrongness.30 Our understanding of constructivism stops at showing that some principles are capable of general justification, that they can be justified to all. It makes no further claim about the relationship between the principles of political justice that govern practical interactions and moral rightness and wrongness. Obviously, once we understand constructivism in this way, it will be apparent that some theories, such as a relativist communitarian position or a form of contractarianism often associated with Hobbes, cannot be a complete account of justice. Even so, rather than being an implicit claim about moral wrongness, this is a by-product of gaining an understanding of the possibilities of practical reasoning. Despite constructivist claims to identify generally justifiable principles, it is clear that constructivism will have its critics. We have already addressed some of these. The entire book is an argument with relativists who claim that there can be no universally justified principles and with related critics who would constrain principles in line with the limitations on sympathetic identification with others.31 Likewise, the constructivist rejection of necessary limits on practical reasoning constitutes an answer to contractarian critics who argue that principles can only be justified by reference to the current desires, interests or motivations of individuals.32 A further criticism of constructivism is forcefully advanced by Cohen in an impressive and formidable paper.33 He argues that constructivism cannot provide the justification of principles of justice because it necessarily reflects facts in its claims about circumstances of justice or conceptions of the person. He goes on to claim that ‘all principles that reflect facts reflect facts only because they also reflect principles that do not reflect facts, and that the latter principles form the ultimate foundation of all principles’.34 The implication is that ‘ultimate principles cannot be justified by facts’, and so constructivism, which is factsensitive, is necessarily grounded on fact-insensitive normative principles.35 The

Making practice reasonable

155

conclusion of the argument is that constructivist justification requires foundationalism if it is to be successful. I am not at all sure how to respond to this argument other than to reiterate a basic constructivist thought about justice. Justice is only an issue under certain conditions. Only where there are potentially interacting and vulnerable persons are there concerns that injustice might occur. Only when faced with challenges that impact on the vulnerability of persons does practical reason have a problem to address and, so, only then must principles of justice be constructed. However, the claim that there are potentially interacting persons and the further claim that they are vulnerable to having their various needs to security, food, well-being, etc. thwarted by others appear to be, in an important sense, broadly factual claims of the sort that Cohen is concerned about. It is at least possible that these claims might be untrue of persons (or some persons). Perhaps Captain Kirk discovers a race of disembodied persons for whom needs and vulnerabilities are very different or perhaps humans might have evolved with hard outer shells that effectively make them invulnerable to physical attack from other humans.36 In either of these cases, the vulnerabilities of persons would be other than those that are assumed by our current account of the principles of justice; the facts would be different. It is therefore very likely that the principles that could be generally justified in these modified circumstances would be very different from those justified in circumstances as we presently understand them. In this way, justification can rightly be thought to involve or use some circumstantial or factual resources. This need not be characterized as relying on facts in justification, nor as grounding justification on fact. Instead, certain facts about people can be understood to form the context for practical thinking and for practical reason. When practical reason addresses questions such as ‘What do I do next?’ ‘How should I respond?’ or ‘What is the right thing for us to do?’ it can only make sense of these as questions in a context of human action, interaction and activity, in a context of vulnerable but connected people. To argue that an answer to these questions that invokes principles of justice implies a further normative principle ‘when there are people in circumstances C then principle P applies’, effectively that ‘justice applies when justice is a question’, seems to add an unnecessary epicycle to justification that I am not sure is helpful in understanding justice.37 As Rawls puts the point, ‘first principles are not, in a constructivist view, independent . . . nor . . . true in all possible worlds. In particular, they depend on the rather specific features and limitations of human life that give rise to the circumstances of justice’.38 This understanding of constructivism is also an important aspect of the constructivist response to further and related criticism from Bruce Haddock.39 Haddock defends an account of what he calls ‘weak foundationalism’, a recognition that there are necessary presuppositions to thinking practically and that these presuppositions can be regarded as universal and foundational.40 He has also argued that primary constructivism exhibits this necessity and would be better understood as a form of weak foundationalism. It is easy to see why Haddock reaches this conclusion. The principles of justice are justified with reference to concepts of society and the person that are universal. In this sense,

156

Making practice reasonable

the principles are neither optional nor the subject of an actual choice but are necessary conclusions of arguments drawing on universal assumptions. The idea of primary constructivism, complete with these bare concepts and an account of the nature and authority of practical reasoning appears to be foundational in exactly Haddock’s sense of a necessary foundation to thinking, and this necessity may be better captured by the terminology of weak foundationalism than by constructivism. A first, and quick response, is that this dispute is simply terminological, and, since primary constructivist principles are not the ‘foundation’ of anything, constructivist terms are more appropriate. Constructivist principles are not foundations, as nothing ‘builds’ on them. Instead, the appropriate metaphor is one of a threshold, boundary or limit to the construction of various buildings at the secondary level. A less-flippant response will draw a distinction between two understandings of necessity. First, necessity might imply that things could be no other way. Necessary foundations are foundations to any possible discussion and justification of principles of justice. It is this sort of necessity that the foundational metaphor is usually understood to refer to. There is, however, a second understanding of necessity that is drawn on by primary constructivism. When discussing Cohen’s objections to constructivism, we noted that, in the circumstances of justice as presently understood, certain principles are justified. We also noted just how dependent these principles are on the circumstances of justice and that if these circumstances were different, then the principles might be different.41 On the constructivist understanding, principles of justice are regarded as necessary to the extent that the circumstances of justice apply. However, this is a distinct understanding of necessity. Importantly, it is a contingent matter that persons are as we currently experience them. Persons could have evolved very different sets of vulnerabilities and capacities or they might inhabit a world without scarcity and would therefore have no experience of competition for resources. Far from being necessary in the foundational sense, principles exhibit a degree of contingency in that they are ‘necessary’ for persons like us; in Milo’s sense, principles of justice exhibit a form of stancedependence.42 As Rawls pointed out, constructivist principles are not true in all possible worlds, but they are justifiable from the standpoint of persons constituted in the way we are. Given this limited understanding of the necessity of principles, it may be confusing to subsume constructivist justification into foundationalism, which involves the much more strident account of necessity. Using the language of constructivism to describe the position laid out here seems entirely appropriate. Our understanding of the primary constructivist account laid out here has developed in two ways. First, it became clear as we read Rawls’s A Theory of Justice and Political Liberalism and also O’Neill’s several constructivist accounts that these prominent advocates of constructivism work with a very thin basic account of constructivism. This is not always apparent but can be drawn out of their work in a straightforward manner. Second, in the last two chapters, we have arrived at the same understanding of constructivism independently of

Making practice reasonable

157

our exposition and analysis of Rawls and O’Neill, and we have deepened that understanding by focussing exclusively on this thin primary constructivism. Both independent argument and analysis of key constructivist thinkers support the general outline of political constructivism and also the specific account of a primary constructivism that we have explored. However, it might be claimed that the thin universal construction of primary constructivism is too thin. Simply justifying principles that rule out killing and injury (in the first instance), without specifying further principles or even the manner of their institutionalization, accepts a great degree of indeterminacy in a theory of justice. The thin primary principles simply do not determine the content of the secondary constructions that both limit and empower our everyday lives. Is this a problem? It is not clear whether indeterminacy is necessarily a problem. A thicker set of universal principles that determined the structure and content of our everyday political institutions would have to regard pluralism as something to be overcome or even vanquished. It may be that primary constructivism’s indeterminacy is a virtue. That universal principles leave significant space for the local and particular is what accounts for the relationship between primary and secondary constructivism. It is also what allows the constructivist to accept diversity and to explore the possibilities for justification given the fact of pluralism. In accepting a significant level of indeterminacy the constructivist is able to consider the important idea of a principled pluralism. However, that may not be all that we can say as the very factors that push us towards thinness at the primary level might also do so at the secondary level. The experience of pluralism and disagreement within our secondary constructions may encourage a thinning out of our commitments even within our own society. It is this sort of recognition that underlies Rawls’s account of political liberalism for democratic societies. Democratic societies experience the clash of conceptions of the good as they become aware of their multiculturalism, and Rawls argues that this mandates a thin liberal political response.43 But, if virtually all societies are multicultural, does this drive to thinness in domestic politics exist more generally? Whether we ought to accept this sort of greater indeterminacy at the secondary level is an interesting prospect, but here we must focus on the key question of justifying universal principles. It is clear that justifying the thin and somewhat indeterminate primary constructivism is a limited project, but its limited ambitions are nevertheless of the utmost importance. First, as just outlined, the limited principles at the primary level constitute a threshold for justice that identifies the boundaries of reasonable practice but that also recognizes and underpins the legitimacy of a great range of secondary diversity. Second, where the very idea of universally justified principles has been continually cast in doubt, the possibility of a properly primary constructivism, however limited, is theoretically significant. Third, justifying universal principles of any sort is practically significant in a world where the experience of pluralism may undermine our commitment to reasoned and reasonable justification. Finally, although the range of principles that have been justified at the primary level is very thin in the first instance, it may not remain so. It may very well be that primary

158

Making practice reasonable

constructivist arguments can be made for further principles. O’Neill’s argument for principles of virtue is one attempt to do just this, but perhaps successful arguments for a level of basic subsistence or for the status of equal citizenship for all might also be made. The possibilities for extending primary principles are not yet clear, but what is clear is the importance of the first steps, made here, towards securing universal and objective principles of political justice and towards opening up these possibilities for consideration.

Notes

1 Introduction 1 J. Heath, ‘Foundationalism and Practical Reason’, Mind, 1997, vol. 106, pp. 451–73; pp. 452 and 461. 2 A. Ripstein, ‘Foundationalism in Political Theory’, Philosophy and Public Affairs, 1987, vol. 16, pp. 115–37; p. 136. 3 M. Timmons, ‘Foundationalism and the Structure of Ethical Justification’, Ethics, 1987, vol. 97, pp. 595–609; p. 600. 4 D. Herzog, Without Foundations: Justification in Political Theory, London: Cornell University Press, 1985, p. 20. 5 Plato, The Republic, Harmondsworth: Penguin, 1987, pp. 505d–e. 6 I refer to this as economic foundationalism as the account of human reason, motivation and action is the one generally assumed by accounts of ‘the market’. 7 See, for example, Hobbes, The Leviathan, Harmondsworth: Penguin, 1985, Chapters 6, 13 and 14, and De Cive, Man and Citizen, New York: Anchor, 1972, Chapter 1. Utilitarianism can also be regarded as another example of economic foundationalism; see J. Bentham, An Introduction to the Principles of Morals and Legislation, London: Methuen, 1982. 8 Examples of this form of foundational argument can be found in M. Walzer, Spheres of Justice: A Defence of Pluralism and Equality, Oxford: Blackwell, 1983, p. 314; J.G. Herder, J.G. Herder on Social and Political Culture, translated by F.M. Barnard, Cambridge: Cambridge University Press, 1969, pp. 27–32 and 299–314. It is assumed in Tully’s commentary on the ‘black canoe’ sculpture; J. Tully, Strange Multiplicities: Constitutionalism in an Age of Diversity, Cambridge: Cambridge University Press, 1995, p. 203. 9 Herzog, Without Foundations, p. 219. 10 Herzog, Without Foundations, p. 18. 11 For example, see B. Parekh, Rethinking Multiculturalism: Cultural Diversity and Political Theory, London: Macmillan, 2000, p. 14, and R. Rorty, Contingency, Irony, and Solidarity, Cambridge: Cambridge University Press, 1989, p. 48. 12 R. Rorty, ‘Solidarity or Objectivity?’ in Objectivity, Relativism, and Truth: Philosophical Papers, Vol. 1, Cambridge: Cambridge University Press, 1991, pp. 21–34; A. MacIntyre, ‘Is Patriotism a Virtue?’ in M. Rosen and J. Wolff, Political Thought, Oxford: Oxford University Press, 1999, pp. 269–84; pp. 274, 278 and 282, and After Virtue: A Study in Moral Theory, Second Edition, London: Duckworth, 1985 (originally 1981), p. 221; M. Sandel, Liberalism and the Limits of Justice, Cambridge, Cambridge University Press, 1982, p. 59; Walzer, Spheres of Justice, p. xiv; and positions criticized in S. Lukes, Liberals and Cannibals, London: Verso, 2003, and B. Barry, Culture and Equality: An Egalitarian Critique of Multiculturalism, Oxford: Polity, 2001.

160

Notes

13 The term ‘fact of pluralism’ is used by Rawls. See, for example, J. Rawls, Political Liberalism (Paperback Edition), New York: Columbia University Press, 1996 (originally 1993), p. xix. 14 Walzer discusses the concept of thin universalism in M. Walzer, ‘Nation and Universe’, The Tanner Lectures on Human Values, 1990, vol. XI, pp. 507–56; p. 527. This is reprinted, alongside chapters constituting an extended discussion of the idea of a thin universalism, in B. Haddock, P. Roberts and P. Sutch (eds), Principles and Political Order: The Challenge of Diversity, London: Routledge, 2006. 15 Indeed, Seung regards constructivism as the theory that regards moral facts as, in one way or another, ‘constructed by human beings’. See T.K. Seung, Intuition and Construction: The Foundations of Normative Theory, New Haven, Yale University Press, 1993, p. ix. 16 Apart from the books and papers by Rawls and O’Neill that are the central focus from this study, see Seung, Intuition and Construction, and C. McKinnon, Liberalism and the Defence of Political Constructivism, Basingstoke: Palgrave, 2002. There have also been a number of important papers. These are a few examples: R. Milo, ‘Contractarian Constructivism’, The Journal of Philosophy, 1995, vol. 92, pp. 181–204; D. Brink, ‘Rawlsian Constructivism in Moral Theory’, Canadian Journal of Philosophy, 1987, vol. 17, pp. 71–90; G.A. Cohen, ‘Facts and Principles’, Philosophy and Public Affairs, 2003, vol. 31, pp. 207–45; L. Krassnoff, ‘How Kantian is Constructivism?’ Kant-Studien, 1999, vol. 90, pp. 385–409; T.M. Scanlon, ‘Contractualism and Utilitarianism’, in A. Sen and B. Williams (eds), Utilitarianism and Beyond, Cambridge: Cambridge University Press, 1982, pp. 103–28; R. Dworkin, ‘Justice and Rights’, in Taking Rights Seriously, London: Duckworth, 1977, pp. 150–83. 2 Constructivism and A Theory of Justice 1 J. Rawls, A Theory of Justice (Revised Edition), Oxford: Oxford University Press, 1999 (originally 1971), p. 3. 2 Rawls’s use of ‘constructive’ in this way is highlighted by O. O’Neill, ‘Constructivism in Rawls and Kant’, in S. Freeman (ed.), The Cambridge Companion to Rawls, Cambridge: Cambridge University Press, 2003, pp. 347–67; p. 350. Rawls refers to the identification of constructive criteria in A Theory of Justice at 30, 35–6 and 46. It is clear that Rawls does not use the term here as he does in later writings or as we shall use it to characterize a form of justification that Rawls sometimes exemplifies. 3 Rawls, A Theory of Justice, p. 40. 4 Rawls, A Theory of Justice, pp. xviii and 10. 5 Rawls, A Theory of Justice, p. 40. 6 Rawls does not claim that there are no facts independent of our engagement with them. Rather, he is making the claim that none of these independent facts have inbuilt and necessary moral significance. Different accounts of moral principles will pick out different facts as of moral significance. We will return to this below. 7 Indeed, Rawls refers to it as ‘a long book, not only in pages’, A Theory of Justice, p. xviii. 8 Rawls, A Theory of Justice, p. 102. 9 Rawls, A Theory of Justice, p. 14. 10 Rawls, A Theory of Justice, p. 18. 11 J. Rawls, ‘Kantian Constructivism in Moral Theory’, Journal of Philosophy, 1980, vol. 77, pp. 515–72; p. 533. Also in S. Freeman (ed.), John Rawls: Collected Papers, Cambridge, Mass.: Harvard University Press, 1999, pp. 303–58. This was originally three lectures and are often referred to as Rawls’s Dewey Lectures. 12 Rawls, A Theory of Justice, p. 11. 13 That the original position is a hypothetical choice situation is demonstrated by Rawls, A Theory of Justice, pp. 11 and 104–5.

Notes 14 15 16 17 18 19 20 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 47 48 49 50 51 52 53 54

161

Rawls, A Theory of Justice, p. 10. Rawls, A Theory of Justice, pp. 12 and 123. Rawls, A Theory of Justice, p. 12. Rawls, A Theory of Justice, p. 123. Rawls, A Theory of Justice, p. 12. Rawls, A Theory of Justice, p. 124. Rawls, A Theory of Justice, p. 111. Rawls, A Theory of Justice, p. 111. Rawls, A Theory of Justice, p. 111. Someone’s ‘conception of the good’ may consist in adherence to a major religion or doctrine (for example, Catholicism or utilitarianism), but it is just as likely, perhaps more likely, to consist of the sum of a person’s understandings of what it would be for their life to go well (that they be a good parent, that they have job security and satisfaction, that they become a certain sort of person and that they have the resources they require to enable this to happen). Rawls, A Theory of Justice, p. 17. Rawls, A Theory of Justice, p. 14. Rawls, A Theory of Justice, p. 16. These presumptions are highlighted together at Rawls, A Theory of Justice, pp. 16–17. Rawls, A Theory of Justice, pp. 16 and 89. Rawls, A Theory of Justice, p. 118. Rawls, A Theory of Justice, p. 118. Rawls, A Theory of Justice, p. 119. Rawls, A Theory of Justice, p. 118. Rawls, A Theory of Justice, p. 17. Rawls, A Theory of Justice, p. 17. Rawls, A Theory of Justice, p. 17. Rawls, A Theory of Justice, p. 348. Rawls, A Theory of Justice, pp. 79–80. Rawls, A Theory of Justice, p. 377. See p. 374 for a fuller statement. Rawls, A Theory of Justice, p. 376. Rawls, A Theory of Justice, pp. 373, 375–6 and 379. Rawls, A Theory of Justice, p. 379. Rawls, A Theory of Justice, p. 378. Rawls, A Theory of Justice, pp. 376–7. Rawls, A Theory of Justice, pp. 364 and 379. Rawls, A Theory of Justice, pp. 372–3. Rawls distinguishes social from natural primary goods such as health, intelligence and imagination. However, he is primarily concerned with social and not natural goods, since these are more directly under the control of political institutions. See Rawls, A Theory of Justice, p. 54. Rawls, A Theory of Justice, p. 79. Rawls, A Theory of Justice, p. 223. Rawls, A Theory of Justice, p. 351. Rawls, A Theory of Justice, pp. 223 and 230. I refer to the conception of the good as contingent not because it is not important but because it could easily have been other than it is. Rawls, A Theory of Justice, pp. 54 and 79. Rawls, A Theory of Justice, p. 494. Rawls, A Theory of Justice, pp. 4–5. Rawls splits the circumstances of justice into two groups. Following Hume, he regards coexistence, rough equality and moderate scarcity as objective circumstances. Pluralism in life plans and conflicting claims on available natural and social resources are subjective circumstances. See the classic statement of the

162

55 56 57 58 59 60 61 62 63

64 65 66 67 68

69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85

Notes circumstances of justice in David Hume, An Enquiry Concerning the Principles of Morals, Oxford: Oxford University Press, 1975, Sec. III, Part I, pp. 183–92, and A Treatise of Human Nature (2nd edition), Oxford: Oxford University Press, 1978, Bk III, Part II, Sec. II, pp. 484–501. See also Rawls, A Theory of Justice, pp. 109–12. Rawls, A Theory of Justice, p. 6. Rawls, A Theory of Justice, p. 106. The options on the list include utilitarianism, intuitionism, egoism, Rawls’s two principles of justice as fairness and a number of ‘mixed’ conceptions; see Rawls, A Theory of Justice, p. 107. Rawls, A Theory of Justice, pp. 132–3 and 135. Rawls, A Theory of Justice, p. 134. Rawls, A Theory of Justice, p. 133. Rawls, A Theory of Justice, p. 56. Rawls, A Theory of Justice, p. 79. J. Rawls, A Theory of Justice (First Edition), Oxford: Oxford University Press, 1971, pp. 542–3. Rawls’s arguments for the priority of liberty are somewhat changed in the revised edition to reflect his response to critics. Rawls, A Theory of Justice, pp. 266–7. Strictly, this is the special conception of justice, a special case of the general conception of justice that goes: All social primary goods . . . are to be distributed equally unless an unequal distribution of any or all of these goods is to the advantage of the least favoured; see Rawls, A Theory of Justice, p. 54. Rawls, A Theory of Justice, pp. 171–6. Rawls, A Theory of Justice, pp. 172–5. Rawls, A Theory of Justice, p. 15. Rawls, A Theory of Justice, p. 510. A number of thinkers have regarded justice and principles of political order as a bargain or compromise resulting from the efficient exercise of instrumental reason and calculation in the pursuit of our ends. See, for example, Glaucon in Plato’s Republic, Harmondsworth: Penguin, 1987, pp. 358e–9b; Hobbes, The Leviathan, Harmondsworth: Penguin, 1985; and more recently, David Gauthier, Morals by Agreement, Oxford: Oxford University Press, 1986, and Moral Dealing: Contract, Ethics, and Reason, New York: Cornell University Press, 1990. This position is often referred to as contractarian. Rawls, A Theory of Justice, p. 16. Rawls, A Theory of Justice, p. 16. Rawls, A Theory of Justice, p. 456. Rawls, A Theory of Justice, p. 104. Rawls, A Theory of Justice, p. 105. Rawls, A Theory of Justice, pp. 11–12. Rawls, A Theory of Justice, p. 104. R. Dworkin, Taking Rights Seriously, London: Duckworth, 1977, p. 151. Rawls, A Theory of Justice, p. 514. A. Brown, Modern Political Philosophy: Theories of the Just Society, Harmondsworth: Penguin, 1986, p. 86. Rawls, A Theory of Justice, p. 514. Rawls, A Theory of Justice, pp. 12, 16, 111 and 214. Rawls, A Theory of Justice, pp. xviii, 18 and 171. Rawls, A Theory of Justice, p. 292. Rawls, A Theory of Justice, p. 507. Rawls, A Theory of Justice, p. 42. A narrow reflective equilibrium is that between our convictions and the principles we initially construct to account for them. This is contrasted with the wide reflective equilibrium that is the product of a more general critical process that also involves

Notes

86 87 88 89 90 91 92 93 94

95 96 97 98 99 100 101 102 103 104 105 106 107

108 109

110 111 112 113

163

testing our convictions against alternative theories of justice and more general background beliefs. As we shall see, it is this wider account of reflective equilibrium that Rawls settles on as central to his conception of justification. See N. Daniels, ‘Wide Reflective Equilibrium and Theory Acceptance in Ethics’, Journal of Philosophy, 1979, vol. 76, pp. 256–82; pp. 258–9. S. Chambers, Reasonable Democracy: Jurgen Habermas and the Politics of Discourse, London: Cornell University Press, 1996, p. 11. M. Walzer, Spheres of Justice: A Defence of Pluralism and Equality, Oxford: Blackwell, 1983, p. xiv. Rawls, A Theory of Justice, p. xviii. Brown, Modern Political Philosophy, pp. 75–6. P. Singer, ‘Sidgwick and Reflective Equilibrium’, The Monist, 1974, vol. 58, pp. 490–517; p. 493. Brown, Modern Political Philosophy, p. 78. Rawls, A Theory of Justice, pp. 18 and 35. See also p. 36 where he repeats, ‘No doubt any conception of justice will have to rely on intuitions to some degree.’ J. Rawls, ‘The Independence of Moral Theory’, Proceedings of the American Philosophical Association, 1975, vol. 48, pp. 5–22; p. 7. Rawls, A Theory of Justice, p. 43. Brown, Modern Political Philosophy, pp. 75–6, is aware of both interpretations of reflective equilibrium. However, he argues that Rawls cannot countenance the possible relativist implications of a two-way process of adjustment between principles and convictions. He therefore regards Rawls as necessarily promoting a fixed-point account. As we shall see, a two-way account of reflection need not have the relativist implications that worry Brown. Rawls, A Theory of Justice, p. 40. Rawls, ‘Kantian Constructivism in Moral Theory’, p. 568. Rawls, A Theory of Justice, p. 17. Rawls, A Theory of Justice, pp. 18 and 42–3. Daniels, ‘Wide Reflective Equilibrium and Theory Acceptance in Ethics’, pp. 258–9. Rawls, ‘The Independence of Moral Theory’, p. 8. See, for example, the entry in T. Honderich (ed.), The Oxford Companion to Philosophy, Oxford: Oxford University Press, 1995, p. 97. Rawls, A Theory of Justice, p. 188. Rawls, A Theory of Justice, pp. 16 and 19. Rawls, A Theory of Justice, p. 514. Rawls, A Theory of Justice, p. 17. D. Little, ‘Reflective Equilibrium and Justification’, The Southern Journal of Philosophy, 1984, vol. 22, pp. 373–87; pp. 374 and 385n.4. For similar conclusions, see K. Baynes, The Normative Grounds of Social Criticism: Kant, Rawls and Habermas, Albany: State University of New York Press, 1992, pp. 61 and 68–9, and W. Kymlicka, Contemporary Political Philosophy: An Introduction, Oxford: Oxford University Press, 1990, p. 69. For example, in Rawls, ‘Kantian Constructivism in Moral Theory’. Rawls, A Theory of Justice, p. 507. He is clear that ‘while some moral principles may seem natural and even obvious, there are great obstacles to maintaining that they are necessarily true, or even to explain what is meant by this’. Likewise, ‘there is no set of conditions or first principles that can plausibly be claimed to be necessary or definitive of morality and thereby especially suited to carry the burden of justification’, Rawls, A Theory of Justice, p. 506. Rawls, A Theory of Justice, pp. 14 and 63. We shall return to this everyday understanding of objectivity repeatedly. For example, see Kymlicka, Contemporary Political Philosophy, p. 69. However, strictly speaking, the principles best expressive of one person’s

164

114 115 116 117 118 119 120 121 122 123 124 125

126 127 128 129 130 131

132 133 134

135 136 137 138 139 140 141 142 143

144

Notes convictions, while different from those that best express another’s, do not contradict them. This is because, as each set claims to be the best fit with a different person’s convictions, they make claims about different subject matter. K. Nielsen, ‘Relativism and Wide Reflective Equilibrium’, The Monist, 1993, vol. 76, pp. 316–32; p. 326. Nielsen, ‘Relativism and Wide Reflective Equilibrium’, p. 327. Rawls, A Theory of Justice, p. 43. Rawls explicitly refers to his conception of reflective justification as nonfoundationalist in J. Rawls, Justice as Fairness: A Restatement, Cambridge, Mass.: Harvard University Press, 2001, p. 31. Rawls, A Theory of Justice, pp. 171–6. T.K. Seung, Intuition and Construction: The Foundations of Normative Theory, New Haven: Yale University Press, 1993, pp. 25–30. Rawls’s list of these traditional conceptions of justice is at A Theory of Justice, p. 107. Rawls, A Theory of Justice, p. 106. Rawls, A Theory of Justice, p. 105. T. Campbell, Justice (2nd Edition), London: Macmillan, 2001 (originally 1988), p. 13. For the way in which the formulation of the original position rules out considerations of desert, see Rawls, A Theory of Justice, pp. 273–4. Rawls admits that he does not work through the process of reflective equilibrium, although he says, ‘Still, we may think of the interpretation of the original position that I shall present as the result of such a hypothetical course of reflection’, A Theory of Justice, p. 18. Rawls, A Theory of Justice, p. 105. Rawls, A Theory of Justice, pp. 104–5. Rawls, A Theory of Justice, p. 171. Rawls, A Theory of Justice, pp. 57–73. See sections starting Rawls, A Theory of Justice, p. 57. Barry, in his characterization of this approach, goes so far as to say, ‘it seems clear that Rawls has hit upon a method whereby you can get anything out of the theory at the end by simply putting it in at the beginning’. See B. Barry, Theories of Justice, Berkeley: University of California Press, 1989, p. 337. Seung, Intuition and Construction, pp. 26–9. Rawls, A Theory of Justice, p. 25. See Gauthier’s Moral Dealing (especially Chapter 9, ‘Bargaining and Justice’). As noted above, this is an example of a tradition of thought in political philosophy stretching back at least to ancient Greece (as Glaucon’s contribution to Plato’s Republic demonstrates, see pp. 358e–62c). Rawls, A Theory of Justice, pp. 17–18 and 181. Rawls, A Theory of Justice, p. 18. Rawls, A Theory of Justice, pp. 349 and 494. Rawls, A Theory of Justice, p. 351. Rawls, A Theory of Justice, pp. 223–4. Rawls, A Theory of Justice, p. 230. M. Sandel, Liberalism and the Limits of Justice, Cambridge: Cambridge University Press, 1982, p. 27. Brown, Modern Political Philosophy, p. 70. This can be seen in the concern in both some traditional and some religious communities to secure group rights that protect their claims to retain members by measures including imposing sanctions on apostates or those who wish to leave a community and also to exercise significant control over the education of their children. Baynes, The Normative Grounds of Social Criticism, p. 148.

Notes 145 146 147 148 149 150 151 152 153

154 155 156 157

158 159 160 161 162 163 164 165

166 167 168 169 170 171

165

Brown, Modern Political Philosophy, p. 70. Rawls, A Theory of Justice, p. 374. Campbell, Justice, p. 108. J.S. Mill, ‘Utilitarianism’, in M. Warnock (ed.), Utilitarianism: On Liberty, Essay on Bentham, Glasgow: Fontana Press, 1990, p. 258. Mill, ‘Utilitarianism’, p. 259. Mill, ‘Utilitarianism’, p. 260. Here, we must be clear that this is exactly what Mill thought he was doing also. See ‘Utilitarianism’, pp. 259 and 260. Rawls, A Theory of Justice, p. 380. All this poses the question ‘If it is not necessary for the derivation of primary goods, what is the function of the Aristotelian Principle?’ Besides playing a part in the explanation of self-respect (Rawls, A Theory of Justice, p. 386), the principle resurfaces as a principle of institutional organization (p. 463). Here, it serves to link individual life plans to the framework of organizations and institutions in a well-ordered society. ‘Everyone’s more private plan is so to speak a plan within a plan, the superordinate plan being realized in the public institutions of society’ (p. 463). This interpretation of the principle, and the idea of a ‘social union of social unions’, leads citizens to consider a well-ordered society as a ‘greater good’ in itself (pp. 571 and 500). Together with a claim for the complementary development of talents (pp. 387 and 458), the Aristotelian principle now functions to unite the individual and social perspectives. It is to demonstrate to each of us that the ‘psychological law’ (p. 375), that is the Aristotelian Principle, gives us all a stake in the greater social whole. It accounts for (bearing in mind that it is controversial, perhaps we should say advocates) a fraternal motivation in the citizen body. The principle is here more closely tied to Rawls’s understanding of stability than to his central account of justification. Rawls, A Theory of Justice, p. 4. Rawls, A Theory of Justice, p. 5. Rawls, A Theory of Justice, pp. 109–10. See, for example, Rawls, A Theory of Justice, pp. 10–15, 40 and 102. Also, see p. 37 where he says, ‘in justice as fairness the principles of justice are not thought of as self-evident, but have their justification in the fact that they would be chosen’ (emphasis added). Rawls, A Theory of Justice, pp. 104–5 and 122. Rawls, A Theory of Justice, p. 121. See Walzer, Spheres of Justice, p. 4, and B. Parekh, Rethinking Multiculturalism: Cultural Diversity and Political Theory, London: Macmillan, 2000, p. 86, for examples. Rawls, A Theory of Justice, p. 508. Rawls, A Theory of Justice, p. 377. Sandel, Liberalism and the Limits of Justice, p. 49. D. Brink, ‘Rawlsian Constructivism in Moral Theory’, Canadian Journal of Philosophy, 1987, vol. 17, pp. 71–90; p. 73. Rawls, ‘Kantian Constructivism in Moral Theory’, p. 516. It is here that Rawls explicitly develops the constructivist nature of his account of justification. He is clear that he does so in response to internal pressures in justice as fairness and not in response to critics such as Sandel. Rawls, ‘Kantian Constructivism in Moral Theory’, p. 525. Rawls, ‘Kantian Constructivism in Moral Theory’, p. 526. J. Rawls, Political Liberalism (Paperback Edition), New York: Columbia University Press, 1996 (originally 1993), p. 290. Rawls, ‘Kantian Constructivism in Moral Theory’, p. 571. Rawls, ‘Kantian Constructivism in Moral Theory’, p. 535. Rawls, ‘Kantian Constructivism in Moral Theory’, p. 518.

166

Notes

172 Rawls, ‘Kantian Constructivism in Moral Theory’, p. 572. 173 Rawls, ‘Kantian Constructivism in Moral Theory’, p. 566. 3 The constructivism of Political Liberalism 1 Political Liberalism has been subsequently published in a paperback edition with an additional introduction and a reprint of his ‘Reply to Habermas’ (Journal of Philosophy, 1995, vol. 92, pp. 132–80). All references are to this 1996 paperback edition, J. Rawls, Political Liberalism (Paperback Edition), New York: Columbia University Press, 1996 (originally 1993). 2 Most notably, if we are to understand Political Liberalism, it should be read alongside ‘The Law of Peoples’. This was first published in 1993, the same year as Political Liberalism, in S. Shute and S. Hurley (eds), On Human Rights: The Oxford Amnesty Lectures, New York: Basic Books, 1993, pp. 42–82. A much extended version of the essay was later published as The Law of Peoples, Cambridge, Mass.: Harvard University Press, 1999. 3 See, for example, J. Rawls, ‘Kantian Constructivism in Moral Theory’, Journal of Philosophy, 1980, vol. 77, pp. 515–72; p. 533. Also in S. Freeman (ed.), John Rawls: Collected Papers, Cambridge, Mass.: Harvard University Press, 1999, pp. 303–58. 4 Rawls, Political Liberalism, pp. 91–2. 5 Rawls, Political Liberalism, p. 97. 6 Rawls, Political Liberalism, pp. 93–4. 7 Rawls, Political Liberalism, pp. 99–100. 8 There have been some changes in the argument from the original position, most notably in the way the parties compare justice as fairness and utilitarianism, in the subsequent J. Rawls, Justice as Fairness: A Restatement, Cambridge, Mass.: Harvard University Press, 2001, and some of these are reflected in the revised text of J. Rawls, A Theory of Justice (Revised Edition), Oxford: Oxford University Press, 1999 (originally 1971). However, these changes are tangential to our argument; so, we will not focus on them. 9 Rawls, Political Liberalism, p. xvi. 10 These general facts are detailed in Rawls, Political Liberalism, pp. 36–8. 11 This distinction is detailed in Rawls, Political Liberalism, pp. 63–6. 12 Rawls, Political Liberalism, p. 13. 13 Rawls, Political Liberalism, pp. 11–14. 14 Rawls, Political Liberalism, p. 175. 15 Rawls, Political Liberalism, p. 13. 16 Rawls, Political Liberalism, p. xviii. 17 Rawls discusses these two stages in Rawls, Political Liberalism, pp. 64, 133–4 and 140. For a useful discussion, see also R. Martin, ‘Rawls’s New Theory of Justice’, Chicago Kent Law Review, 1994, vol. 69, pp. 737–61. 18 Rawls, Political Liberalism, p. 14. 19 Rawls, Political Liberalism, p. 38n.41. 20 Rawls, Political Liberalism, p. 14. 21 These are laid out in Rawls, Political Liberalism, pp. 3–46. Three other fundamental ideas surface in Rawls’s exposition: the idea of the basic structure, the idea of the original position and the idea of an overlapping consensus (p. 43). Of these, the first two are relevant to the construction of a free-standing conception, but they are not taken to be implicit in public political culture; instead, they are introduced to facilitate the unified exposition of justice as fairness (p. 14n.16). 22 It is discussed in detail in Rawls, Political Liberalism, pp. 15–22. 23 See Rawls, Political Liberalism, p. 17, for the distinction between reciprocal and the more familiar mutual advantage. As Rawls characterizes it, mutual advantage

Notes

24 25 26

27 28 29 30

31

32 33 34 35

36 37 38 39 40 41

42 43 44

45

167

involves a conception of gain measured against the baseline of the situation as it is, whereas reciprocal advantage uses a conception that is measured against the baseline of an appropriate conception of equality. For discussion of the idea of the person, see Rawls, Political Liberalism, pp. 18–20 and 29–35. Rawls, Political Liberalism, p. 19. We must note here that the full conception of reasonableness involves a second criterion. A reasonable citizen must not only be willing to propose and abide by fair terms, he must also exhibit ‘a willingness to recognize the burdens of judgement and to accept their consequences’ (Rawls, Political Liberalism, p. 54). The burdens of judgement help explain the fact of reasonable disagreement, and we shall return to them below. See Rawls, Political Liberalism, p. 35. This description of well-orderedness is in Rawls, Political Liberalism, p. 103. See Rawls, Political Liberalism, pp. xviii and 47. Rawls makes a clear distinction between rational and full autonomy (as well as between both of these and an ethical conception of autonomy that might form part of a comprehensive doctrine). For details of this distinction, see Rawls, Political Liberalism, pp. 72–81. For a concise discussion of both, and of their relationship with the original position, see pp. 304–6. Rawls attributes to citizens these higher-order interests and regards them as basic and regulative of their other interests. Rawls conjoins to these a third interest, that of realizing whatever determinate conception of the good they might happen to hold. They are of higher order, because they are fundamental to the ideal of the person as free and equal (see Rawls, Political Liberalism, p. 74). Rawls, Political Liberalism, pp. 75–6. The possession of the two powers is often referred to as a minimum condition of social entry. See Rawls, Political Liberalism, pp. 19, 79, 109 and 302. Rawls, Political Liberalism, p. 78. That the choice must be unanimous at this first stage of justification is taken to be a point of the structural constraints embodied in the original position. It must be stressed however that this unanimity is characteristic only of the free-standing justification. At the second stage, overlapping consensus, we can hope for no more than the affirmation of a substantial number of citizens. Rawls, Political Liberalism, p. 24. Rawls, Political Liberalism, p. 97. Rawls, Political Liberalism, p. 58. Rawls, Political Liberalism, p. 56. See Rawls, Political Liberalism, pp. 56–7, for the full list. Rawls, Political Liberalism, p. 58. It is important for Rawls to account for disagreement at the level of ideal theory (the best of all possible cases) and not to regard it as a problem to be overcome in application, since Rawls has assumed that diversity of judgement is a permanent feature of democratic society. A similar point is made in C. Larmore, The Morals of Modernity, Cambridge: Cambridge University Press, 1996, pp. 152–5. Rawls, Political Liberalism, p. 60. Indeed, recognition of the burdens of judgement is central to Rawls’s conception of reasonable persons. As mentioned above, coupled with a willingness to propose and abide by fair terms, a willingness to accept the burdens is one of the two basic aspects of the reasonableness of persons (see Rawls, Political Liberalism, pp. 54 and 375). Elsewhere, acceptance of the burdens is referred to as one of four special features of reasonable citizens (p. 81) and as part of the basis of citizens’ ‘reasonable moral psychology’ (p. 86). Rawls, Political Liberalism, p. 97.

168

Notes

46 Rawls, Political Liberalism, pp. 136–7. 47 Rawls, Political Liberalism, pp. 142–3 and 389. 48 For Rawls’s description of modus vivendi as an account of stability, see Rawls, Political Liberalism, p. 147. 49 Rawls, Political Liberalism, pp. 169–71. 50 Rawls, Political Liberalism, pp. 12–13, 38, 140, 144–5 and 168–72. 51 Rawls describes the political conception within an overlapping consensus as a moral conception affirmed on moral grounds. Rawls, Political Liberalism, pp. 147–8 and 208. 52 Rawls, Political Liberalism, pp. 158–68. 53 Rawls, Political Liberalism, p. 158. This is to be distinguished from Rawls’s twostage method of justification for political liberalism. Here, the two stages are historical episodes in the development of overlapping consensus as stage two of the process of justification. 54 Rawls, Political Liberalism, pp. 158–9. 55 Rawls, Political Liberalism, pp. 158–68. 56 Interestingly, Brian Barry, Justice as Impartiality, Oxford: Oxford University Press, 1995, outlines the possibility of a similar historical process whereby justice as impartiality develops out of justice as mutual advantage. Barry, however, seems considerably less optimistic about its success, longevity and the speed of transformation (p. 102). 57 Rawls, Political Liberalism, p. 15. J. Habermas also draws attention to this point; see ‘Reconciliation Through the Public Use of Reason: Remarks on John Rawls’s Political Liberalism’, The Journal of Philosophy, 1995, vol. 92, pp. 109–31; p. 121. 58 This is in Rawls, Political Liberalism, pp. 145–6. Rawls’s pluralist doctrine is one that is not fully unified and is therefore only partially comprehensive. A pluralist holds different views in different spheres and has not undertaken to make these values part of a consistent system (p. 145). The existence of partially comprehensive doctrines is crucial for Rawls’s story of the development from modus vivendi through constitutional consensus to overlapping consensus. 59 Rawls, Political Liberalism, pp. 169–70. 60 Rawls, Political Liberalism, pp. xvi–xvii. 61 The fundamental ideas are described as implicit or inherent in Rawls, Political Liberalism, pp. 8, 13–14, 20, 43, 46, 97, 145 and 175. At p. 38n.41, he takes as a fourth general fact that these ideas are present in political culture and that they can be used as the foundation for the construction of a political conception of justice. 62 See Lecture III of Political Liberalism. 63 Rawls, Political Liberalism, pp. 28 and 45. 64 The two principles of justice are restated in Rawls, Political Liberalism, with some important changes but largely intact at pp. 5–6 and 271. They are described as the outcome of the construction procedure on p. 64. Strictly speaking, justice as fairness is now regarded as one of several acceptable generic liberal conceptions that might be justified in this way. However, the features of this generic conception make it so similar to justice as fairness that those who regard justice as fairness as problematic are likely to be similarly worried about generic liberalism. See, for example, Rawls, Political Liberalism, p. 6. 65 See Rawls, Political Liberalism, pp. 397–9. 66 Rawls, Political Liberalism, p. 14. 67 Rawls, Political Liberalism, pp. 25–6. 68 I have introduced this distinction elsewhere. See my ‘Identity, Reflection and Justification’, in B. Haddock and P. Sutch (eds), Multiculturalism, Identity and Rights, London: Routledge, 2003, pp. 142–57, and also in ‘Why Thin Universalism Needs Conceptions of Society and Person’, in B. Haddock, P. Roberts and P. Sutch (eds), Principles and Political Order: The Challenge of Diversity, London: Routledge, 2006, pp. 111–27.

Notes

169

69 M. Walzer, Spheres of Justice: A Defence of Pluralism and Equality, Oxford: Blackwell, 1983, pp. 5–10. 70 Rawls, Political Liberalism, p. 14. 71 C. Kukathas and P. Pettit, Rawls: A Theory of Justice and Its Critics, Cambridge: Polity Press, 1990, p. 133. 72 C. Kukathas, ‘Explaining Moral Variety’, Social Philosophy and Policy, 1994, vol. 11, pp. 1–21; p. 2. 73 P. Jones, ‘Two Conceptions of Liberalism, Two Conceptions of Justice’, British Journal of Political Science, 1995, vol. 25, pp. 515–50; p. 530. 74 S. Scheffler, ‘The Appeal of Political Liberalism’, Ethics, 1994, vol. 105, pp. 4–22; p. 20. 75 S. Hampshire, Justice is Conflict, London: Duckworth, 1999, p. 32. 76 T. Pogge, Realizing Rawls, London: Cornell University Press, 1989, pp. 213n and 266n. 77 B. Brower, ‘The Limits of Public Reason’, The Journal of Philosophy, 1994, vol. 91, pp. 5–26; p. 8. 78 J. Gray, Post Liberalism: Studies in Liberal Thought, London: Routledge, 1993, p. 314. 79 G. Warnke, Justice and Interpretation, Cambridge: Polity Press, 1992, p. 39. 80 Here, I have greatly benefited from several discussions with Peter Sutch. 81 Rawls, Political Liberalism, pp. 89–90. 82 See also Rawls, Political Liberalism, p. 114. 83 Rawls, Political Liberalism, p. 143. 84 As the texts are different, even if the argument is substantially the same, I will refer to either version or both where relevant as ‘The Law of Peoples’ and The Law of Peoples, respectively. 85 Rawls, The Law of Peoples, p. 10. 86 Rawls, ‘The Law of Peoples’, p. 44. 87 Rawls, ‘The Law of Peoples’, p. 46, and Rawls, The Law of Peoples, p. 85. 88 Rawls, ‘The Law of Peoples’, p. 67. 89 Rawls, ‘The Law of Peoples’, pp. 75–6. 90 Rawls, ‘The Law of Peoples’, pp. 46 and 76. 91 Rawls, ‘The Law of Peoples’, p. 76. 92 Such a reaction might typify that of the communitarians. See, for example, M. Sandel, Liberalism and the Limits of Justice, Cambridge, Cambridge University Press, 1982, and A. MacIntyre, After Virtue: A Study in Moral Theory, Second Edition, London: Duckworth, 1985 (originally 1981). 93 Rawls, Political Liberalism, p. xvi. 94 L. Krasnoff, ‘Consensus, Stability, and Normativity in Rawls’s Political Liberalism’, The Journal of Philosophy, 1998, vol. 95, pp. 289–92; p. 278. Cf. Rawls, Political Liberalism, p. 37. Joshua Cohen makes a similar point when he notes that ‘If we accept the idea of reasonable pluralism, then moral diversity is not simply a bare fact . . . but, rather, indicates something about the operation and powers of practical reason’ (‘Moral Pluralism and Political Consensus’, in D. Copp, J. Hampton and J. Roemer (eds), The Idea of Democracy, Cambridge: Cambridge University Press, 1993, pp. 270–91; p. 288). 95 The importance to political liberalism of a conception of reasonable disagreement rather than to an affirmation of the value of pluralism is also recognized in Larmore, The Morals of Modernity, pp. 152–5. 96 See J. Raz, ‘Facing Diversity: The Case of Epistemic Abstinence’, in Ethics in the Public Domain, Oxford: Oxford University Press, 1994, pp. 60–96. 97 See Rawls, Political Liberalism, p. 143. 98 Krasnoff, ‘Consensus, Stability, and Normativity in Rawls’s Political Liberalism’, p. 280.

170

Notes

99 This principle, as noted above, reads, ‘our exercise of political power is fully proper only when it is exercised in accordance with a constitution the essentials of which all citizens as free and equal may reasonably be expected to endorse’ (Rawls, Political Liberalism, p. 137). 100 This mode of practical reasoning is not meant to be peculiar to democratic regimes but can apply also to a society of states and by extension to each subject for construction; see Rawls, ‘The Law of Peoples’, pp. 46 and 50. 101 Rawls, Political Liberalism, p. 28. 102 Rawls, Political Liberalism, p. 102n.13. 103 Rawls, Political Liberalism, p. 384. 104 Rawls, The Law of Peoples, p. 86. 105 Rawls, ‘The Law of Peoples’, p. 64 (my italics). 106 Rawls, ‘The Law of Peoples’, p. 53. 107 Rawls, ‘The Law of Peoples’, p. 54. 108 This includes, for example, information about population, natural resources, development and military power. Rawls, ‘The Law of Peoples’, p. 54, and Rawls, The Law of Peoples, p. 32. 109 Rawls, The Law of Peoples, pp. 64–7. 110 Rawls, ‘The Law of Peoples’, pp. 64 and 67, and Rawls, The Law of Peoples, pp. 68–9. 111 Rawls, ‘The Law of Peoples’, p. 65. 112 See Rawls, ‘The Law of Peoples’, pp. 66 and 69. 113 Rawls, ‘The Law of Peoples’, p. 62. Also Rawls, The Law of Peoples, p. 66. 114 Rawls, Political Liberalism, p. 35. 115 Rawls, ‘The Law of Peoples’, p. 79. 116 In Political Liberalism, Rawls introduces a distinction between general concepts and more specific conceptions precisely when he discusses the fundamental ideas, see p. 14n.15. For the thought that Rawls’s fundamental ideas are versions of more general ideas given a democratic twist, I am indebted to conversations with Rex Martin. He interprets Rawls in a similar way in his ‘Rawls’s New Theory of Justice’, pp. 751–2. 117 Rawls, Political Liberalism, p. 109. 118 Rawls, Political Liberalism, p. 18. 119 Martin, ‘Rawls’s New Theory of Justice’, p. 751. Here, he refers only to Plato and Aristotle but added Aquinas in conversation. 120 Rawls, Political Liberalism, p. 380. 121 Rawls, Political Liberalism, pp. 18–19. 122 Rawls, Political Liberalism, p. 17. Recall that Rawls characterizes reciprocal advantage as involving a conception of gain that is measured against a baseline of an appropriate conception of equality. The more familiar mutual advantage measures gain against a baseline of the situation as it is. 123 Martin, ‘Rawls’s New Theory of Justice’, pp. 751–2. 124 Rawls, Political Liberalism, pp. 16 and 18–19. 125 Rawls, Political Liberalism, pp. 15–17. 126 Rawls, Political Liberalism, p. 108. 127 Rawls, Political Liberalism, p. 18. 128 Rawls, Political Liberalism, p. 108. 129 Rawls, Political Liberalism, p. 380. 130 Rawls, Political Liberalism, pp. 380–1. 131 Rawls, Political Liberalism, p. 110. 132 Rawls, The Law of Peoples, pp. 85–6. 133 Rawls, Political Liberalism, p. 109. 134 Rawls, Political Liberalism, pp. 109 and 110n.15. See H.L.A. Hart, The Concept of Law (Second Edition), Oxford: Oxford University Press, 1994, p. 193, where he claims that there must be a certain minimal content to natural law. He argues that

Notes

171

without such a content laws and morals could not forward the minimum purpose of survival which men have in associating with each other. In the absence of this content men, as they are, would have no reason for obeying voluntarily any rules. 135 Rawls, The Law of Peoples, p. 65. 136 Rawls, The Law of Peoples, p. 5 (emphasis added). 137 The notion of a thin universalism is explored in Haddock, Roberts and Sutch (eds), Principles and Political Order. 138 Rawls, The Law of Peoples, p. 68. 139 Rawls, The Law of Peoples, p. 80. 140 If this is the case, how do we make sense of Section 12.2 of The Law of Peoples, which Rawls calls ‘No Deduction from Practical Reason’ (pp. 86–7)? Here, Rawls seems almost to apologize for ‘misleading’ us in Political Liberalism when he admits that ‘I give the impression that the content of the reasonable and the rational is derived from the principles of practical reason’ (p. 86n.33). My thought is that we need to regard this ‘apology’ in the same way that we might regard his apology in Political Liberalism (p. 53n.7) for portraying justice as fairness as the outcome of a rational choice theory in A Theory of Justice (p. 15). Just as Rawls’s retraction is unnecessary in that case, A Theory of Justice is not a rational choice theory. So, his retreat away from a deduction from practical reasoning is also unnecessary. His project never was a deduction or derivation from practical reason; again, he is retracting a position he has never really advanced. 141 Rawls, Political Liberalism, p. 38n.41. 142 See G. Klosko, ‘Rawls’s “Political” Philosophy and American Democracy’, American Political Science Review, 1993, vol. 87, pp. 348–59, where he demonstrates that, for example, in a 1954 survey, more than 60 per cent of Americans actively supported preventing atheists and communists from making speeches and that a similar 1987 survey showed that only 33 per cent would favour allowing their leastliked political group to hold a rally (p. 352). This indicates strong reservations about liberal free speech. See also Klosko, ‘Political Constructivism in Rawls’s Political Liberalism’, American Political Science Review, 1997, vol. 91, pp. 635–46. 143 L. Wenar, ‘Political Liberalism: An Internal Critique’, Ethics, 1995, vol. 106, pp. 32–62; pp. 50–1. Wenar goes so far as to claim that ‘only those who affirm a Kantian conception of the person are eligible to be reasonable members of a society well-ordered by Rawls’s theory’ (p. 51). 144 G. Doppelt, ‘Rawls’ Kantian Ideal and the Viability of Modern Liberalism’, Inquiry, 1998, vol. 31, pp. 413–49; pp. 426 and 438–40. 145 W. Galston, ‘Moral Personality and Liberal Theory: John Rawls’s “Dewey Lectures” ’, Political Theory, 1982, vol. 10, pp. 492–519; pp. 515–16. 146 Walzer, Spheres of Justice, p. 318. 147 Wenar, ‘Political Liberalism: An Internal Critique’, p. 55. 148 B. Barry, ‘John Rawls and the Search for Stability’, Ethics, 1995, vol. 105, pp. 874–915; pp. 910 and 913. 149 I.M. Young, ‘Rawls’s Political Liberalism’, The Journal of Political Philosophy, 1995, vol. 3, pp. 181–90; pp. 189–90. 150 Martin, ‘Rawls’s New Theory of Justice’, p. 760. 151 Rawls, Political Liberalism, p. 97, where he says that ‘we can misdescribe our reason as we can anything else’. 152 Rawls, Political Liberalism, p. 112. 153 Rawls, Political Liberalism, p. 55. 154 Rawls, Political Liberalism, p. 227. This cannot simply refer to Rawls’s account of public reason and his account of the rules that should frame democratic political debate when applied at the level of fundamental ideas. Public reason limits the reasons advanced in disputes over constitutional essentials and matters of basic

172

155 156

157 158 159 160 161

Notes justice to those that have a properly ‘political’ character and so are not based on any particular comprehensive doctrine. It might be suggested that the ‘orderly contest’ between interpretations of the fundamental ideas takes place at this level, in a form of constitutional convention with arguments made in public reason. However, as we have seen, the fundamental ideas are required in order to lay out an original position, and a conception of public reason, just like the political conception, is understood by Rawls to be a product of the original position. As such, it already presupposes the successful recovery of appropriate fundamental ideas (Rawls, Political Liberalism, pp. 225–6). It is here that, once this is conceived of as a political process, public reason comes into its own. It may be that something like this is implied by Rawls’s coming to regard the content of the political conception as given by a family of generic liberalisms. See Rawls, Political Liberalism, p. 6, and ‘The Idea of Public Reason Revisited’, in Rawls, The Law of Peoples, pp. 129–80; pp. 140–41. Scheffler, ‘The Appeal of Political Liberalism’, p. 21. Rawls, ‘The Law of Peoples’, p. 52, and Rawls, The Law of Peoples, p. 5. Rawls, Political Liberalism, p. 110. Rawls, Political Liberalism, p. 109. Rawls, Political Liberalism, p. lxii.

4 Primary constructivism and O’Neill 1 O. O’Neill, Towards Justice and Virtue: A Constructive Account of Practical Reasoning, Cambridge: Cambridge University Press, 1996, p. 63. 2 O’Neill, Towards Justice and Virtue, p. 48. 3 O’Neill, Towards Justice and Virtue, p. 48. See also O. O’Neill, Constructions of Reason: Explorations of Kant’s Practical Philosophy, Cambridge: Cambridge University Press, 1989, pp. 208–11, and O. O’Neill, Bounds of Justice, Cambridge: Cambridge University Press, 2000, p. 72. 4 O. O’Neill, ‘Political Liberalism and Public Reason: A Critical Notice of John Rawls, Political Liberalism’, The Philosophical Review, 1997, vol. 106, pp. 411–28; p. 426. 5 O’Neill, Constructions of Reason, pp. 206 and 218. 6 O. O’Neill, ‘Constructivism vs. Contractualism’, Ratio, 2003, vol. XVI, pp. 319–31; p. 320. See also ‘Constructivism in Kant and Rawls’, in S. Freeman (ed.), The Cambridge Companion to Rawls, Cambridge: Cambridge University Press, 2003, pp. 347–67; p. 348. 7 O’Neill, ‘Constructivism vs. Contractualism’, p. 321. See also O’Neill, ‘Constructivism in Kant and Rawls’, pp. 348–9. 8 O’Neill, Towards Justice and Virtue, pp. 49–50. See also O’Neill, ‘Four Models of Practical Reasoning’, in Bounds of Justice, pp. 11–28. 9 O’Neill, Bounds of Justice, p. 13. 10 So far, this does not imply scepticism about Platonist reasons; rather, the claim is that they are no practical basis for general justification. However, as will become clear, O’Neill advances the further claim that arbitrary reasons cannot count as reasons at all nor will they have the authority of reasons. See O’Neill, Bounds of Justice, p. 12. 11 How could we reject instrumental reasoning entirely as we reason in this way (almost) whenever we act? No matter what our ends, we must identify the actions that would fulfil the end, whether this is selfish or selfless, moral or otherwise. 12 O’Neill, Bounds of Justice, p. 16. 13 O’Neill, Bounds of Justice, p. 23. 14 O’Neill, Towards Justice and Virtue, p. 51. See also O’Neill, Constructions of Reason, p. 23, and O’Neill, Bounds of Justice, p. 24.

Notes

173

15 Note that each of these methods of practical reasoning, and their problems, corresponds to a form of foundationalism outlined in the introduction: Platonic foundationalism, economic foundationalism and cultural foundationalism. 16 On the unavoidability of abstraction, see O’Neill, Towards Justice and Virtue, p. 40. Criticism of abstraction is widespread amongst communitarians and particularists of all sorts. For example, see M. Walzer, Spheres of Justice: A Defence of Pluralism and Equality, Oxford: Blackwell, 1983, p. xiv; M. Sandel, Liberalism and the Limits of Justice, Cambridge: Cambridge University Press, 1982; and C. Gilligan, In a Different Voice: Psychological Theory and Women’s Development, Cambridge: Harvard University Press, 1982. 17 See O’Neill, Towards Justice and Virtue, pp. 39–44. See also O’Neill, Constructions of Reason, pp. 208–10, and O. O’Neill, ‘Abstraction, Idealisation and Ideology in Ethics’, in J.D.G. Evans (ed.), Moral Philosophy and Contemporary Problems (Royal Institute of Philosophy Supplements), 1987, pp. 55–69. 18 O’Neill, Towards Justice and Virtue, p. 52. 19 This bears some obvious and not accidental affinities with Kant’s account of public and private reasons in I. Kant, ‘An Answer to the Question: What is Enlightenment?’ in T. Humphrey (ed.), Immanuel Kant: Perpetual Peace and Other Essays, Indianapolis: Hackett Publishing, 1983, pp. 41–8. See especially paragraphs 37–8 on pp. 42–3. 20 O’Neill, Towards Justice and Virtue, p. 53. 21 See O’Neill, Towards Justice and Virtue, pp. 53–5. 22 O’Neill contrasts modal questions with metrical ones (O’Neill, Constructions of Reason, p. 228). Here, there is no attempt to measure the comparative liberty/rightness that a right or obligation secures as there is on a metrical account. Instead, modal questions are concerned to understand what must or what might be the case in a given situation. Modal logic identifies two types of truth that could be attributed to a statement (as well as its falsity). A proposition may be true in one of two ways. First, a proposition may necessarily be true, as in the case that 2 + 2 = 4, this must be the case. Or the truth of a statement may be contingent, a matter of possibility not of necessity, as with the statement ‘I have come to work today’ (this is true, but it need not have been, I may have stayed at home). For O’Neill, modal questions are concerned with truth of the first sort. 23 O’Neill, Towards Justice and Virtue, p. 57. 24 O’Neill, Towards Justice and Virtue, pp. 57–8. 25 O’Neill, Towards Justice and Virtue, p. 62. 26 O’Neill, Towards Justice and Virtue, p. 63. See also footnotes at pp. 63–4. 27 O’Neill, Towards Justice and Virtue, p. 58. 28 ‘Open-ended questioning’ in a recursive process of critical reflection as the basis of justification is used by O’Neill to ground both reason and justice in her reading of Kant’s Critique of Pure Reason. See O’Neill, Constructions of Reason, p. 21. 29 For example, see O’Neill, Towards Justice and Virtue, pp. 47–8. 30 See, for example, J. Rawls, A Theory of Justice (Revised Edition), Oxford: Oxford University Press, 1999 (originally 1971), p. 581, where justification proceeds from ‘premises we both accept’ and ‘proofs become justification once the starting points are mutually recognized’ or actually ‘persuade us of the soundness of the conception expressed by their premises’. Also important is the general idea of justice as fairness, where the principles of justice are those ‘rational persons . . . would accept’ (p. 102). Cf. also the notion of overlapping consensus in J. Rawls, Political Liberalism (Paperback Edition), New York: Columbia University Press, 1996 (originally 1993), where we noted that the justification of a political conception in this way waited on the actual acceptance of citizens that this conception properly expressed the political commitments of their comprehensive doctrines and, therefore, on the actual development of such a consensus. But, see also O’Neill, Towards Justice and Virtue, p. 125n.1, where O’Neill distances herself from positions like Rawls’s.

174

Notes

31 O’Neill, Towards Justice and Virtue, p. 59. See also O’Neill, Constructions of Reason, p. 217. 32 O’Neill, Towards Justice and Virtue, pp. 54 and 77. 33 O’Neill, Towards Justice and Virtue, p. 93. The italics are in the original. See also p. 99. 34 O’Neill, Towards Justice and Virtue, p. 99. 35 O’Neill, Towards Justice and Virtue, pp. 100–1. 36 O’Neill, Towards Justice and Virtue, p. 101. 37 The qualifier ‘almost’ is added so as to make it clear that it may be logically possible that some action or other may not make these assumptions. However, and somewhat trivially, even when scratching my nose, I assume that I will not be prevented from doing so by existent others, many of whom may be in a position to attempt to do so, and also that I am the sort of being capable of suffering discomfort. 38 O’Neill, Towards Justice and Virtue, pp. 102–3. 39 O’Neill, Towards Justice and Virtue, p. 105. 40 See O. O’Neill’s ‘Justice and Boundaries’, in Chris Brown (ed.), Political Restructuring in Europe: Ethical Perspectives, London: Routledge, 1993, pp. 69–88. For example, see p. 83, where she writes, ‘Although the notions of “global community” or “global society” may be no more than a sentimental rhetoric of cosmopolitanism, the weaker relation of sharing a world with distant others is one that agents today cannot consistently deny.’ 41 O’Neill, Towards Justice and Virtue, p. 106. 42 O’Neill, Towards Justice and Virtue, p. 109. 43 O’Neill, Towards Justice and Virtue, p. 106 (italics original). 44 O’Neill, Towards Justice and Virtue, p. 113. 45 O’Neill, Towards Justice and Virtue, p. 114. 46 O’Neill, Towards Justice and Virtue, pp. 114–15. 47 O’Neill, Towards Justice and Virtue, p. 117. 48 See O’Neill, Towards Justice and Virtue, p. 113n.33, where she is clear that her account of scope does not require a cosmopolitan uniform treatment of all but ‘is simply an argument to show that, at least in our world, foreigners count’. 49 Accusations are often made that cosmopolitan claims to priority are apparently damaged by their abstracting from the ‘thick’ reasons for action that agents actually have. This is taken to provide justification for the priority of the particular. However, and as explained above, particularist accounts of reasoning also abstract from the ‘thick’ reasons of agents even if they do not idealize a particular account of a community and the agents within. Any general account of practical reasoning will look for some level of abstraction (what else could ‘general’ mean) and so will ‘omit’ something of an agent’s reasons. Priority between domains of discourse is not settled; all we can establish is that each of these domains or levels of reasoning is important. 50 O’Neill, Constructions of Reason, p. 212. 51 O’Neill, Constructions of Reason, p. 212. 52 O’Neill, Constructions of Reason, p. 213. 53 See, for example, O’Neill, Constructions of Reason, pp. 46–7, where she assumes ‘a plurality of parties . . . whose thinking and judging are to some extent independent’ and, again, ‘beings who are separate and whose coordination is not naturally given or preestablished’. 54 See O’Neill, ‘Justice and boundaries’, pp. 81–2. 55 O’Neill, Towards Justice and Virtue, p. 99. 56 Rawls, Political Liberalism, pp. 89–90. 57 O’Neill, Constructions of Reason, p. 213. 58 This question is dealt with in very much the same way in O’Neill, Towards Justice and Virtue, pp. 122–53, and in O’Neill, Constructions of Reason, pp. 187–205, 213–15 and 225–33.

Notes

175

59 O’Neill, Constructions of Reason, p. 189, and O’Neill, Towards Justice and Virtue, p. 147. 60 O’Neill, Constructions of Reason, p. 189, and O’Neill, Towards Justice and Virtue, p. 147. 61 For a clear and full statement of this argument, see O’Neill, Constructions of Reason, pp. 193–7. 62 This formulation crops up in very different sorts of theory. See, for example, Rawls, A Theory of Justice, p. 53, and J. Narveson, The Libertarian Idea, Philadelphia, Pa.: Temple University Press, 1988, p. 7. 63 O’Neill, Constructions of Reason, p. 197. 64 O’Neill, Towards Justice and Virtue, pp. 101 and 192–3, and O’Neill, Constructions of Reason, pp. 198–9 and 230. 65 O’Neill, Towards Justice and Virtue, pp. 134, 140–1, 143 and 146. 66 O’Neill, Towards Justice and Virtue, p. 144. 67 O. O’Neill, Faces of Hunger: An Essay on Poverty, Justice and Development, London: Allen and Unwin, 1986, p. 122. See also O’Neill, Constructions of Reason, p. 193, and O’Neill, Towards Justice and Virtue, p. 134. 68 O’Neill, Constructions of Reason, p. 213. 69 O’Neill, Towards Justice and Virtue, p. 163. 70 O’Neill, Constructions of Reason, pp. 13–14 and 19–20. 71 Whilst the construction of a complete set of obligations is a logical possibility, proving that we now possessed such a set would be impossible. 72 O’Neill, Towards Justice and Virtue, p. 164. 73 O’Neill, Constructions of Reason, p. 213. 74 O’Neill, Towards Justice and Virtue, p. 163, and O’Neill, Constructions of Reason, pp. 215–16. 75 O’Neill, Towards Justice and Virtue, pp. 174–5, and O’Neill, Constructions of Reason, p. 215. 76 O’Neill, Towards Justice and Virtue, pp. 176–8. 77 O’Neill, Towards Justice and Virtue, p. 184. 78 O’Neill refers to these assumptions as characterizing a meagre construal of the circumstances of justice in Constructions of Reason, p. 212. 79 Strictly speaking, there is a fifth step, constructing obligations of virtue, which should intervene between the third and fourth steps, and this is addressed in the following section. 80 O’Neill, Towards Justice and Virtue, p. 180. 81 O’Neill, Towards Justice and Virtue, pp. 168 and 182. 82 O’Neill, Towards Justice and Virtue, p. 183. 83 O’Neill, Towards Justice and Virtue, pp. 173–4. 84 This would seem to be so, despite several lapses by O’Neill. For example, ‘[This test] would rule out certain acts . . . as ethically unacceptable, and rule in others as ethically required’ (O’Neill, Faces of Hunger, pp. 48 and 136). In order for this test to identify obligatory actions over and above the obligation to reject forbidden principles, it would be necessary to ‘prove’ that an act/principle was the only possible non-forbidden alternative for action in a particular situation. This would be a daunting task. Ordinarily, we would regard this test as only demonstrating that in acting in such a manner, you would not be acting on a forbidden maxim. In order to go beyond this to designate the performance of that action/adoption of that principle as obligatory, we would have to be confident that we had successfully identified, tested and ruled out all other alternative courses of action so that there was only one nonrejected possibility. Could we ever be this confident? 85 O’Neill, Towards Justice and Virtue, pp. 184–5. 86 O’Neill, Towards Justice and Virtue, pp. 188–9. 87 O’Neill, Towards Justice and Virtue, pp. 191–3.

176 88 89 90 91 92 93

94 95

96 97 98

99

100 101

Notes O’Neill, Constructions of Reason, p. 230. O’Neill, Towards Justice and Virtue, p. 194. O’Neill, Towards Justice and Virtue, p. 194. O’Neill, Towards Justice and Virtue, pp. 195–6. O’Neill, Towards Justice and Virtue, pp. 200–5. O’Neill, ‘The Great Maxims of Justice and Charity’, Constructions of Reason, pp. 219–33. This chapter rehearses the arguments for imperfect obligations of virtue alongside justice and subsumes them under the general term charity. ‘Charity is concern for the needs of others’ (p. 219). Recall here the nature of the test. It can issue only one of two verdicts, rejected or permitted. There are obviously more ways of meeting needs than are outlined here. Philanthropic action, for example, meets needs in ways that are generally accepted to be above and beyond the call of duty. If this is the case, then philanthropy is not subsumed by, or synonymous with, the obligation to aid. The example can be found in O’Neill, Constructions of Reason, p. 219n.4. O’Neill uses the example to illustrate her point that many liberals have traditionally combined a stress on justice with a strong commitment to charity. Cited in O’Neill, Constructions of Reason, p. 219n.4. Here, I do not claim that Locke’s example can be read as denying an obligation to charity (after all Locke denies that the grain seller can legitimately just walk away). Instead, I use this example to illustrate the point that important needs can be met in ways other than those we would normally deem charitable. This also seems consistent with some of O’Neill’s presentations of the rejection of R. For examples, see O’Neill, Constructions of Reason, p. 199, where we must avoid only a ‘principled refusal to help’; also see O’Neill, Towards Justice and Virtue, p. 195, where we are told that we need only ‘take some care to sustain some others in some ways’ (italics original). Both these phrases would appear to be consistent with aR (mutual advantage) as well as with an obligation to charity. No state seriously questions whether it should at least attempt to help those who suffer because of earthquakes or floods. Indeed, it is often at just such times when the aid provided by the state is at its most obvious. O’Neill, Towards Justice and Virtue, p. 194.

5 Reasoning practically: a constructivist understanding of practical reasoning 1 This way of framing the issue owes much to illuminating, but often difficult, conversations with Bruce Haddock. 2 In this section, I draw very heavily on my ‘Why thin universalism needs conceptions of society and person’, in B. Haddock, P. Roberts and P. Sutch (eds), Principles and Political Order: The Challenge of Diversity, London: Routledge, 2006, pp. 111–27. 3 For example, see the classic statement of the circumstances of justice in David Hume, An Enquiry Concerning the Principles of Morals, Oxford: Oxford University Press, 1975, Sec. III, Part I, pp. 183–92, and A Treatise of Human Nature (2nd edition), Oxford: Oxford University Press, 1978, Bk III, Part II, Sec. II, pp. 484–501. See also J. Rawls, A Theory of Justice (Revised Edition), Oxford: Oxford University Press, 1999 (originally 1971), pp. 109–12. The account of the circumstances of justice offered by both Hume and Rawls is of a group of roughly equal and finite individuals, with a range of coinciding and conflicting interests coexisting in a situation of moderate scarcity. 4 See, for example, W. Kymlicka (ed.), The Rights of Minority Cultures, Oxford: Oxford University Press, 1995; C. Taylor (with Gutmann ed.), Multiculturalism: Examining the Politics of Recognition (with commentary from Appiah et al.), Prince-

Notes

5 6 7 8

9 10

11 12

13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

177

ton: Princeton University Press, 1994; and B. Haddock and P. Sutch (eds), Multiculturalism, Identity and Rights, London: Routledge, 2003. Rawls’s account of justification recognizes this in the claim that we must start from our ‘considered convictions’. These embody our basic assumptions about our contexts and environments. See Rawls, A Theory of Justice, p. 19. S. Chambers, Reasonable Democracy, London: Cornell University Press, 1996, p. 51. O. O’Neill, Towards Justice and Virtue: A Constructive Account of Practical Reasoning, Cambridge: Cambridge University Press, 1996, p. 41. This injunction is obviously similar to O’Neill’s claim that we should avoid the idealization involved in assuming false predicates of those to whom we offer reasons. For her distinction between idealization and abstraction, see, for example, Towards Justice and Virtue, pp. 39–44. See Rawls, A Theory of Justice, pp. 18 and 507. Something along these lines would seem to be what Rawls is referring to when he states that ‘justification proceeds from what all parties . . . hold in common’ and that ‘it is perfectly proper, then, that the argument . . . should proceed from some consensus. This is the nature of justification’ (A Theory of Justice, pp. 580–1). This form of expression owes much to O’Neill. See Towards Justice and Virtue, p. 57. Rawls, A Theory of Justice, p. 18. See also N. Daniels, ‘Wide Reflective Equilibrium and Theory Acceptance in Ethics’, Journal of Philosophy, 1979, vol. 76, pp. 256–82, and T.M. Scanlon, ‘The Aims and Authority of Moral Theory’, Oxford Journal of Legal Studies, 1992, vol. 12, pp. 1–23. J. Rawls, ‘The Independence of Moral Theory’, Proceedings of the American Philosophical Association, 1975, vol. 48, pp. 5–22; p. 8. This section draws on arguments in my ‘Identity, Reflection and Justification’, in B. Haddock and P. Sutch (eds), Multiculturalism, Identity and Rights, London: Routledge, 2003, pp. 142–57. M. Walzer, Thick and Thin: Moral Argument at Home and Abroad, Notre Dame, Ind.: University of Notre Dame Press, 1994, p. 99. A. MacIntyre, After Virtue: A Study in Moral Theory, Second Edition, London: Duckworth, 1985 (originally 1981), p. 220. M. Sandel, Liberalism and the Limits of Justice, Cambridge: Cambridge University Press, 1982, p. 179. B. Parekh, Rethinking Multiculturalism: Cultural Diversity and Political Theory, London: Macmillan, 2000, pp. 122 and 156. A. MacIntyre, ‘Intelligibility, Goods, and Rules’, Journal of Philosophy, 1982, vol. 79, pp. 663–5; p. 664. MacIntyre, After Virtue, pp. 209–10. R. Rorty, ‘Human Rights, Rationality and Sentimentality’, in S. Shute and S. Hurley (eds), On Human Rights, New York: Basic Books, 1993, pp. 112–34; p. 122. Parekh, Rethinking Multiculturalism, p. 156. Sandel, Liberalism and the Limitis of Justice, p. 150, and MacIntyre, After Virtue, p. 213. A. MacIntyre, ‘Is Patriotism a Virtue?’ in M. Rosen and J. Wolff (eds), Political Thought, Oxford: Oxford University Press, 1999, pp. 269–84; p. 282. This is implied by, for example, MacIntyre’s claims about unintelligibility, and therefore untranslatability, across cultures. S. Benhabib makes a similar point in argument against Rorty in The Claims of Culture: Equality and Diversity in the Global Era, Princeton: Princeton University Press, 2002, p. 32. K. Nielsen, ‘Relativism and Wide Reflective Equilibrium’, The Monist, 1993, vol. 76, pp. 316–32; p. 320. He says specifically that ‘no sense can be made of the idea that the conceptual

178

29 30 31 32 33 34 35 36

37

38

39 40 41 42 43 44 45 46

47 48 49 50 51 52

Notes

resources of different languages differ dramatically’. See D. Davidson, Inquiries into Truth and Interpretation, Oxford: Oxford University Press, 1984, p. xviii. D. Davidson, ‘On the Very Idea of a Conceptual Scheme’, in Inquiries into Truth and Interpretation, pp. 183–98; p. 192. D. Davidson, ‘The Method of Truth in Metaphysics’, in Inquiries into Truth and Interpretation, pp. 199–214; p. 200. Davidson, ‘On the Very Idea of a Conceptual Scheme’, p. 192. Davidson, ‘On the Very Idea of a Conceptual Scheme’, p. 197. Davidson, ‘On the Very Idea of a Conceptual Scheme’, p. 184. O’Neill, Towards Justice and Virtue, p. 105. We should be made particularly aware of this practical connectedness if we consider the global environmental impact of our activity, for example. This is not to say that there will not be contingent practical limits of information and time for reflection. I do not have much to say about these except that they do not necessarily prevent us from turning our attention to urgent challenges to the objectivity and justifiability of our principles, and as new information and time becomes available to us, we have the reason and the opportunity, respectively, to reflect further. For discussion of the relationship between contractarianism and constructivism, see, O. O’Neill, ‘Constructivism vs. Contractualism’, Ratio, 2003, vol. XVI, pp. 368–90. See also David Gauthier, ‘Political Contractarianism’, Journal of Political Philosophy, 1997, vol. 2, pp. 132–48, and T.K. Seung, Intuition and Construction: The Foundations of Normative Theory, New Haven: Yale University Press, 1993, where this sort of theory is regarded as constructivist. See, for example, the position of Glaucon in Plato’s Republic, Harmondsworth: Penguin, 1987, pp. 358e–60d; Hobbes, The Leviathan, Harmondsworth: Penguin, 1985, Chapter 17; and David Gauthier, Morals by Agreement, Oxford: Oxford University Press, 1986, pp. 8–20. David Hume, A Treatise of Human Nature (2nd edition), Oxford: Oxford University Press, 1978, Bk II, Part III, Sec. III, pp. 414–16. Of course, in what follows, I do not do justice to the range and power of Scanlon’s argument on these issues. T.M. Scanlon, What We Owe to Each Other, Cambridge, Mass.: Harvard University Press, 1998, p. 18. Scanlon, What We Owe to Each Other, p. 18. He calls these reasons, ones that really count in favour of something, reasons in the ‘standard normative sense’; What We Owe to Each Other, p. 18. Scanlon, What We Owe to Each Other, pp. 19–22. Scanlon, What We Owe to Each Other, pp. 33–4. I take it as read that the simplistic claim that we obviously have a desire to do whatever it is that we actually do is not a good response here. This tautologous claim cannot explain anything as it explains everything. It simply identifies our action and then treats this as evidence that we had a desire to do it. Nothing could possibly constitute a counter example to what is, in effect, the definitional identification of action and desire. I might just as well claim that all objects are being continually created and re-created by the flying spaghetti monster and then point to the fact that we continue to exist and to perceive the existence of other objects as evidence for the existence of such a pasta-based supreme being. Scanlon, What We Owe to Each Other, p. 40. Scanlon, What We Owe to Each Other, p. 41. Scanlon, What We Owe to Each Other, pp. 43–4. Scanlon, What We Owe to Each Other, p. 47. Scanlon, What We Owe to Each Other, p. 53. Of course, the geometric statements, for example, only hold in Euclidean or plane

Notes

53

54 55 56 57 58

59

179

geometry. In the geometry of curved surfaces, the internal angles of a triangle may add up to more or less than 180 degrees depending on whether the surface on which it is drawn is positively or negatively curved. However, this is to imply that the statement is not subjective but objective and applies only in clearly defined appropriate cases. No further claim is made here about the relationship between constructivist objectivity in practical reasoning and in mathematical reasoning beyond noting that they similarly do without a traditional account of objectivity. While there are constructivist accounts of mathematics, nothing in my argument refers to, or depends on, their plausibility. See F. D’Agostino, ‘Transcendence and Conversation: Two Conceptions of Objectivity’, American Philosophical Quarterly, 1993, vol. 30, pp. 87–108; p. 87. Rawls, A Theory of Justice, p. 543. T. Nagel, The View from Nowhere, Oxford: Oxford University Press, 1986, p. 5. These questions have been put to me most forcefully by Kerstin Budde and Bruce Haddock. See, for example, O. O’Neill, Constructions of Reason: Explorations of Kant’s Practical Philosophy, Cambridge: Cambridge University Press, 1989, p. 23. In Kantian terminology, appeal to an external ground for the authority of reason would be heteronomous. Regarding the authority of reason in this way is similar to O’Neill’s claims about reason being self-vindicating (for example, Constructions of Reason, p. 27). Likewise, I think that some argument along these lines is implicit in Rawls’s account of reflective equilibrium.

6 Making practice reasonable: political constructivism 1 J. Rawls, Political Liberalism (Paperback Edition), New York: Columbia University Press, 1996 (originally 1993), pp. 93–4. 2 Rawls, Political Liberalism, pp. 11–14. 3 Rawls, Political Liberalism, p. 14. 4 Rawls, Political Liberalism, p. 99. 5 In O’Neill’s case, this is clear from her focus on abstraction and rejection of idealization, and when she argues that constructivism gives ‘no place or weight to distinctively moral facts or properties, whether natural or non-natural, that can be discovered or intuited and do[es] not seek foundations for ethics in such facts’ (O. O’Neill, ‘Constructivism vs. Contractualism’, Ratio, 2003, vol. XVI, pp. 319–31; p. 320). Scanlon is clear when he says that the point of judgements about right and wrong is not to make claims about what the spatiotemporal world is like . . . [This] is not a concern about the metaphysical reality of moral facts. If we could characterize the method of reasoning through which we arrive at judgements of right and wrong, and could explain why there is good reason to give judgements arrived at in this way the kind of importance that moral judgements are normally thought to have, then . . . no interesting question would remain about the ontology of morals – for example, about the metaphysical status of moral facts (T.M. Scanlon, What We Owe to Each Other, Cambridge, Mass.: Harvard University Press, 1998, p. 2). 6 On the conception of society, see pp. 15–17, and on the conception of the person, see p. 19. 7 See, for example, O. O’Neill, Towards Justice and Virtue: A Constructive Account of Practical Reasoning, Cambridge: Cambridge University Press, 1996, pp. 50–1. 8 I have argued this case in ‘Why thin universalism needs conceptions of society and

180

9 10 11 12 13 14 15 16 17 18

19 20 21

22 23 24 25 26 27

28

Notes

person’, in B. Haddock, P. Roberts and P. Sutch (eds), Principles and Political Order: The Challenge of Diversity, London: Routledge, 2006, pp. 111–27, and draw on this in the argument here. Rawls, Political Liberalism, pp. 15–17. Rawls, Political Liberalism, pp. 18 and 108. Rawls, Political Liberalism, pp. 18–19. Rawls, Political Liberalism, pp. 109 and 380. O’Neill, Towards Justice and Virtue, pp. 100–7. O. O’Neill, Constructions of Reason: Explorations of Kant’s Practical Philosophy, Cambridge: Cambridge University Press, 1989, pp. 212–13. Rawls, Political Liberalism, p. 19 O’Neill, Constructions of Reason, p. 212. Rawls, Political Liberalism, pp. 108–10, and O’Neill, Constructions of Reason, pp. 212–13, and Towards Justice and Virtue, pp. 100–13. Even if the reasoning of the persons in this choice situation did take this basic form, then the overall justification of principles would still not have an instrumental character. This is because this process of decision-making is nested within a broader account of practical reasoning that has a non-instrumental character and that constrains the legitimate exercise of means-end reasoning. It is this broader account that underpins the constructivist understanding of the choice situation and its objectivity. F. D’Agostino, Free Public Reason: Making It up As We Go, Oxford: Oxford University Press, 1996, pp. 9–10. M. Walzer, Thick and Thin: Moral Argument at Home and Abroad, Notre Dame, Ind.: University of Notre Dame Press, 1994, p. 4. He describes how pictures of the suffering and struggle of others ‘resonate’ and ‘echo positively’ with our experience and foster feelings of solidarity (Walzer, Thick and Thin, pp. 3 and 10). Solidarity is built not on philosophical argument or on abstract reasoning but on our recognition that ‘we too remember, or we have listened to stories about, tyranny and oppression’ (Walzer, Thick and Thin, pp. 7–8). M. Walzer, ‘Nation and Universe’, The Tanner Lectures on Human Values, 1990, vol. XI, pp. 507–56; p. 527 [reprinted in Haddock, Roberts and Sutch (eds), Principles and Political Order], and Thick and Thin, p. 6. ‘Minimalism . . . is less the product of persuasion than of mutual recognition’ based on our capacity for sympathetic identification (Walzer, Thick and Thin, p. 17). R. Rorty, Contingency, Irony, and Solidarity, Cambridge: Cambridge University Press, 1989, pp. 190–2. The following paragraphs draw on the conclusions to my ‘Identity, Reflection and Justification’, in B. Haddock and P. Sutch (eds), Multiculturalism, Identity and Rights, London: Routledge, 2003, pp. 142–57. Rawls, Political Liberalism, p. 58, and C. Larmore, The Morals of Modernity, Cambridge: Cambridge University Press, 1996, pp. 152–5. See, for example, I. Berlin, ‘The Pursuit of the Ideal’, in I. Berlin, The Crooked Timber of Humanity: Chapters in the History of Ideas, London: John Murray, 1990, pp. 1–19; J. Gray, Enlightenment’s Wake: Politics and Culture at the Close of the Liberal Age, London: Routledge, 1995; W. Galston, Liberal Pluralism: The Importance of Value Pluralism for Political Theory and Practice, Cambridge: Cambridge University Press, 2002; W. Galston, The Practice of Liberal Pluralism, Cambridge: Cambridge University Press, 2005. To base constructivism on such an evaluative account of pluralism would be to build in a ‘thick’ moral claim from the outset, a strategy that constructivism is committed to avoiding. Recall the general account of political constructivism outlined above, whereby political constructivist position (1) draws only on general and abstract reasons that will include conceptions of society and person, (2) draws on an account of objectivity as impartiality or non-subjectivity and (3) involves an account of reflective and critical

Notes

29

30 31 32 33 34 35 36 37

38 39 40 41 42 43

181

practical reasoning that is concerned to continually hold the first to the standard of the second. Rawls, A Theory of Justice (Revised Edition), Oxford: Oxford University Press, 1999 (originally 1971), p. 10, and R. Milo, ‘Contractarian Constructivism’, The Journal of Philosophy, 1995, vol. 92, pp. 181–204; p. 184. L. Krassnoff in ‘How Kantian is Constructivism?’ Kant-Studien, 1999, vol. 90, pp. 385–409; p. 388, identifies this as one central understanding of constructivism. Milo, in‘Contractarian Constructivism’, pp. 184 and 186, says that moral truths are ‘truths about what norms and standards hypothetical contractors would have reason to choose’. As we have seen, Rorty and Walzer hold a position like this one. This appears to be the position of Gauthier, Morals by Agreement, Oxford: Oxford University Press, 1986, and Hobbes, The Leviathan, Harmondsworth: Penguin, 1985, for example. G.A. Cohen, ‘Facts and Principles’, Philosophy and Public Affairs, 2003, vol. 31, pp. 207–45. Cohen, ‘Facts and Principles’, p. 231. Cohen, ‘Facts and Principles’, pp. 218–19 and 221. That justice might be different if humans were invulnerable to physical attack is a notion also introduced by H.L.A. Hart, The Concept of Law (Second Edition), Oxford: Oxford University Press, 1994, pp. 194–5. Of course, some elements of practical reasoning do not seem to be dependent on context in this way, and these are often regarded as principles. However, these are not obviously normative principles in the usual sense but principles of logic and reasoning, such as non-contradiction and rules of deduction and inference. J. Rawls, ‘Kantian Constructivism in Moral Theory’, in S. Freeman (ed.), John Rawls: Collected Papers, Cambridge, Mass.: Harvard University Press, 1999, pp. 303–58; p. 351. Haddock has made this criticism in several conversations and draws on arguments in his ‘Thin Universalism as Weak Foundationalism’, in Haddock, Roberts and Sutch (eds), Principles and Political Order, pp. 60–75. Haddock, ‘Thin Universalism and Weak Foundationalism’, pp. 68 and 73. Although we cannot at present think that other such circumstances apply to us or we would have reason to revise our account of those circumstances along with the conceptions of society and person that constitute them. Milo, ‘Contractarian Constructivism’, p. 192. C. McKinnon, Liberalism and the Defence of Political Constructivism, Basingstoke: Palgrave, 2002, claims that even Rawls’s approach assumes too much and argues for any even thinner ‘many flowers’ account of the fact of pluralism and for self-respect as the basis of liberal democratic justice. The argument is that this is thinner and ‘less demanding’ than a straightforwardly Rawlsian account (pp. 53–6, for example).

Index

abstraction 85–6, 89, 90, 103, 133 Aquinas, Thomas 69 Archimedean point 132 Aristotelian principle 14–15, 36, 37–8, 41, 144 Aristotle 69 Barry, B. 76 Baynes, K. 37 Bentham, Jeremy 57, 75 Brink, D. 42 Brower, B. 61 burdens of judgement: political conception of justice 53–4; practical reasoning 115; primary and secondary constructivism 65, 66, 68, 71, 76 Campbell, Tom 32, 34, 37 charity 106, 107, 108 coercion 102 Cohen, G.A. 154, 155, 156 communitarianism: foundationalism 3, 4; intuitions 35; political constructivism 138, 150, 154; practical reasoning 86, 120, 135 connection: constructing justice 103; constructing virtue 108–9; political justice 143; practical reasoning 124; scope of practical reason 92, 93, 95, 98 Constructions of Reason (O’Neill) 6, 81, 96, 98, 111 constructivism 4, 5, 6, 8–10; see also political constructivism; Political Liberalism; practical reasoning; primary constructivism; secondary constructivism; A Theory of Justice contractarianism 125, 127, 128, 130, 154 critical reflection: Political Liberalism 67; practical reasoning 89, 118, 119, 128,

133; A Theory of Justice 24–5, 27, 29, 43, 45 culture 2–3, 120, 121, 123, 150, 151 D’Agostino, F. 148 Davidson, Donald 123 deception 102 desert 32, 34 desires 2, 14, 128, 129, 130 determinacy 9, 38–40, 45, 58, 144–5 disagreement 53, 54, 65, 66, 152, 157 diversity 68, 76, 82, 152, 153, 157 Doppelt, G. 75 Dworkin, R. 20 egoism 31 fact of pluralism 87, 153 finitude 92, 94, 98, 100, 103, 143 first principle of justice 18 foundationalism: justice as fairness 42, 43; justification 1–3; political constructivism 153, 155, 156; Political Liberalism 47, 59, 60; practical reasoning 127, 128, 131, 133, 135, 136; A Theory of Justice 9, 22, 23, 25–7, 44, 45 Galston, William 75 Gauthier, David 35, 75 Gray, J. 61 Habermas, Jürgen 72 Haddock, Bruce 155, 156 Hampshire, Stuart 61 Hart, H.L.A. 72, 80 higher and lower pleasures 37 Hobbes, Thomas 2, 55, 75, 143, 154 human goods 36, 52

Index 183 human rights 73, 149 Hume, David 75, 127, 128, 129, 143 ideal theory 73, 76, 80 idealism 113, 114 idealization: constructing justice 103; practical reasoning 85–6, 88, 89, 90, 133; scope of practical reason 97 imperfect obligations 99, 101, 105–7, 110, 145 indifference 105–6, 107, 109, 146 instrumental reasoning 83, 84, 86 intuitionism: justice as fairness 31–6, 39; political justice 144; Political Liberalism 47; A Theory of Justice 8, 9, 21–2, 23 Jones, P. 61 justice: political constructivism 143, 154, 155; practical reasoning 114, 125; primary constructivism and O’Neill 96, 99–105, 110; A Theory of Justice 9 justice as fairness: determinacy problems 38–40; foundational conception of the person 41–3; intuitions and the original position 31–6; political justice 144; Political Liberalism (outline of theory 48, 49, 50, 57; overview 46, 78, 80; primary and secondary constructivism 59, 66, 69–73, 75); reasoning in the original position 36–8; A Theory of Justice 9, 10, 19–21, 26, 27, 30 justification: justice as fairness 38, 40; overview 1, 2–3, 4, 6; political constructivism 148, 150, 153, 155, 157; Political Liberalism 54–5, 56–7, 59, 60; practical reasoning 113, 116–19, 126–7, 129–30, 136; primary constructivism and O’Neill 81–2; A Theory of Justice 8, 19, 25–30, 44 Kant, Immanuel: justice as fairness 42, 43; Political Liberalism 46, 47, 48, 57, 61, 72, 80; practical reasoning 83, 87, 134; primary constructivism and O’Neill 81, 82; A Theory of Justice 9, 44 Klosko, G. 75 Kukathas, C. 61 Larmore, C. 152 law of peoples 63, 65, 67–73, 79, 80 liberalism: justice as fairness 36–7, 40, 41, 42, 43; Political Liberalism 49, 61, 63, 73; A Theory of Justice 22

libertarianism 76 liberty 16, 18, 36, 43, 100 Locke, John 9, 108 Martin, Rex 69, 76 Mill, J.S. 37, 57, 76 Milo, R. 154, 156 monistic scheme 31, 32, 34, 39 moral facts/order: foundationalism 2, 3; justice as fairness 42; O’Neill 82; political constructivism 138, 139; Political Liberalism 47; practical reasoning 128, 131, 133, 136; A Theory of Justice 9, 23, 24, 26–7, 28, 30, 43 moral geometry 19, 20, 27, 39, 144 moral powers 50, 51, 52, 69, 72, 97 motivation 37, 38, 41, 124, 125–30, 144 multiculturalism 3, 4, 35, 120, 135, 150, 157 Nagel, T. 132 needs 14, 41, 105–9 Neurath’s boat 25, 122 Nielsen, Kai 30, 122 O’Neill, Onora: constructivism 5, 6; political constructivism 139, 141–7, 156–8; Political Liberalism 78; practical reasoning 83–91, 112, 116, 124, 132–4; primary constructivism 81–111 (constructing justice 99–105; constructing principles 98–9; constructing virtue 105–10; overview 81–3, 110–11; practical reasoning 83–91; scope of practical reason 91–8) objectivity: justice as fairness 40; overview 1, 2, 3, 4; political constructivism 139, 140, 147; Political Liberalism 66; practical reasoning 131, 132, 133, 134, 135, 136; primary constructivism and O’Neill 82, 84, 89, 90, 103, 110; A Theory of Justice 13, 27, 28, 29, 44 obligations: constructing justice 99–100, 101–3; constructing virtue 105–10; political constructivism 145, 150 overlapping consensus 54–7, 58, 66, 76, 78 particularist reasoning 83–4, 85, 86 perfect obligations 99, 103 perfectionism 32, 34 person, concept of: constructing substantive primary principles 98, 99, 103; political constructivism 154;

184

Index

person, concept of continued Political Liberalism 71, 72, 73, 78–9, 80; primary and secondary constructivism 151; primary constructivism and O’Neill 83, 110, 111; principles of political justice 140–6; scope of practical reason 95, 96, 97 Pettit, P. 61 Plato 2, 69, 72, 83, 84, 86 pluralism: foundationalism 3–4; justice as fairness 32, 33, 34, 38, 39, 40; political constructivism 141–2, 147, 148, 152, 153, 157; Political Liberalism (overview 46, 78; primary and secondary constructivism 65, 66, 68, 70; theory of justice 48, 49, 53, 54, 57); practical reasoning 6, 120, 135; primary constructivism and O’Neill 81, 82, 87; A Theory of Justice 8, 9, 12, 26, 29, 44 plurality: political justice 143; practical reasoning 115; primary constructivism and O’Neill 92–3, 94, 96, 98, 103 Pogge, Thomas 61 political constructivism 137–58; constructing principles 140–7; nature of 137–40; overview 1, 5, 46, 47, 153–8; Political Liberalism 59–62, 63, 64–5, 70, 78; primary and secondary constructivism 147–53; A Theory of Justice 17 political justice: justice as fairness 39; overview 2, 5; political constructivism 138, 140–7, 158; Political Liberalism 50, 54–8, 74–5, 78 political liberalism: overlapping consensus 56, 57; political conception 51; political constructivism 6, 157; primary and secondary constructivism 59, 61, 66, 67 Political Liberalism (Rawls) 46–80; outline of new theory of justice 48–59 (overlapping consensus 54–7; overview 48–50, 57–9; political conception of justice 50–4); overview 5, 6, 46–8, 77–80; political constructivism 137, 138, 141, 144, 145, 156; primary and secondary constructivism 59–77 (is political constructivism a secondary constructivism? 59–62; primary constructivism in Political Liberalism 62–4; primary constructivism in Rawlsian practical reasoning 64–7; primary constructivism in starting points 68–70; Rawls’s primary constructivism 70–4; Rawls’s secondary constructivism

74–7); primary constructivism and O’Neill 95, 110 positivism 131 practical reasoning 112–36; boundaries to reasons 119–30 (motivational limits on reasons 125–30; overview 119–20; practical limits on reasons 124–5; theoretical limits on reasons 120–3); constraints of practicality 113–19; justice as fairness 36–8; necessity of 112–13; objectivity and authority of reasons 130–5; overview 1–4, 6, 135–6; political constructivism 154; Political Liberalism 64–7, 68, 72, 74, 79; primary constructivism and O’Neill (constructing justice 103; O’Neill on practical reasoning 83–91; overview 81, 82, 110, 111; scope of practical reason 91–8); A Theory of Justice 10, 21 primary constructivism: constructivism 6; political constructivism 137, 138, 147–53, 154–8; Political Liberalism (is political constructivism a secondary constructivism? 60, 62; overview 47, 62–4, 79, 80; Rawls’s primary constructivism 70–4; Rawlsian practical reasoning 64–7; starting points 68–70); practical reasoning 131, 134; primary constructivism and O’Neill 81–111 (constructing justice 103, 104, 105; constructing virtue 109; overview 83, 110; practical reasoning 90; scope of practical reason 91–8) primary goods: justice as fairness 36, 37, 38, 40, 41, 42; political justice 144; Political Liberalism 48, 51, 52; A Theory of Justice 15, 16, 17, 18, 27, 31 principled pluralism 6, 137, 152, 153, 157 principles of indifference 105–6, 146 principles of injury 101–2, 104, 146, 151 principles of justice: constructing justice 104; justice as fairness 31, 32, 39, 41–2, 43; political constructivism 137, 140–7, 151, 154–6; Political Liberalism 51, 52, 62, 64, 68, 70, 71; practical reasoning 125; A Theory of Justice (original position, structure and parties 11, 12–14, 16; overview 8–10, 27, 29, 43, 44; reflective equilibrium/ourselves 20–6; well-ordered society 16–20) rational choice theory 19, 25, 26 Rawls, John: constructivism 5, 6; political constructivism (nature of political

Index 185 constructivism 137, 138, 139; overview 155–7; primary and secondary constructivism 148, 152; principles of political justice 140, 142–7); practical reasoning 112, 115, 118, 122, 132, 133, 135; primary constructivism and O’Neill 81–3, 87, 89–90, 95–9, 104–5, 110; see also Political Liberalism; A Theory of Justice reasonable pluralism 48, 49, 52 reasoning see practical reasoning reflective equilibrium: justice as fairness 32–3, 34, 35, 39, 40; Political Liberalism 60, 66–7, 78; practical reasoning 112, 118, 122, 127, 132, 133; primary constructivism and O’Neill 89, 97, 110; principles of political justice 145; A Theory of Justice 10, 20–6, 27–30, 43, 44 rights: constructing justice 99–100, 101, 102, 103; constructing virtue 105, 106, 109; primary and secondary constructivism 73, 149, 150; A Theory of Justice 16, 18 Rorty, R. 149 Rousseau, Jean-Jacques 9 Sandel, M. 36, 41, 43 Scanlon, T.M. 126, 128, 129–30, 139 Scheffler, S. 61, 80 second principle of justice 18–20 secondary constructivism: political constructivism 138, 145, 147–53, 154, 157; Political Liberalism (is political constructivism a secondary constructivism? 60, 61, 62; overview 6, 47, 79; primary constructivism 63, 67; Rawls’s secondary constructivism 74–7); practical reasoning 121, 131; primary constructivism and O’Neill 83, 87, 104, 105, 109, 110 self-respect 16, 18 Seung, T.K. 31, 32, 34 Sidgwick, Henry 57 slavery 21, 24, 28 social contract theory 9, 35 social primary goods 15, 16, 18, 48 society: political constructivism 140, 142–6, 151, 154; Political Liberalism 68, 69, 71–3, 75, 78–9, 80; primary constructivism and O’Neill 83, 95, 96–9, 103, 110, 111; A Theory of Justice 16–20 Theory of Justice, A (Rawls) 8–45; justice as fairness 31–43 (determinacy problems

38–40; foundational conception of the person 41–3; intuitions and the original position 31–6; reasoning in the original position 36–8); original position, structure and parties 11–16; overview 5, 6, 8–10, 43–5; political constructivism 144, 156; Political Liberalism (outline of theory 48, 49, 51–3, 54, 57, 58; overview 46, 76, 78, 80; primary and secondary constructivism 59, 60, 62, 63–4, 66, 67); principles of justice/well-ordered society 16–20; reflective equilibrium/ourselves 20–6; summary 26–31 thin theory of the good 14, 15, 16, 17, 36, 51 thin universalism: foundationalism 4; political constructivism 153, 157; political justice 147; primary and secondary constructivism 110, 148, 149, 152 Timmons, M. 2 Towards Justice and Virtue (O’Neill) 6, 81, 97, 98, 111 universalism: foundationalism 3, 4; political constructivism 147, 149, 152–4, 157, 158; Political Liberalism 62, 63, 80; primary constructivism 6, 83, 86, 90, 91 universalizability test: constructing justice 99, 101, 102, 103; constructing virtue 106, 107, 108 utilitarianism: justice as fairness 31, 32, 34, 35; Political Liberalism 57, 66, 76; A Theory of Justice 8, 9 veil of ignorance: justice as fairness 34, 39, 40, 41, 42; political justice 144, 145; Political Liberalism 51, 52, 68; practical reasoning 132; A Theory of Justice 13–17, 19, 22, 27, 31 virtue 105–10, 144–6, 158 Walzer, Michael: foundationalism 4; Political Liberalism 61, 75; primary and secondary constructivism 148, 149, 150; A Theory of Justice 22, 44 Warnke, G. 61 weak foundationalism 155, 156 Wenar, L. 75, 76 wide reflective equilibrium (WRE) 30, 110, 122–3 Young, I.M. 76