In memory of Paul Bourke,
and in appreciation of shared friends

A Theory of Freedom


From the Psychology to the Politics of Agency

Philip Pettit


Copyright © Philip Pettit 2001

The right of Philip Pettit to be identified as author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988.

First published in 2001 by Polity Press in association with Blackwell Publishers Ltd

Editorial office:
Polity Press
65 Bridge Street
Cambridge CB2 1UR, UK

Marketing and production:
Blackwell Publishers Ltd
108 Cowley Road
Oxford OX4 1JF, UK

All rights reserved. Except for the quotation of short passages for the purposes of criticism and review, no part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the publisher.

Except in the United States of America, this book is sold subject to the condition that it shall not, by way of trade or otherwise, be lent, re-sold, hired out, or otherwise circulated without the publisher’s prior consent in any form of binding or cover other than that in which it is published and without a similar condition including this condition being imposed on the subsequent purchaser.

ISBN 0-7456-2093-0
ISBN 0-7456-2094-9 (pbk)
ISBN 978-0-7456-6815-4 (eBook)

A catalogue record for this book is available from the British Library.

Typeset in 10½ on 12pt Sabon
by Graphicraft Limited, Hong Kong
Printed in Great Britain by MPG Books Ltd, Bodmin, Cornwall

This book is printed on acid-free paper.



1    Conceptualizing Freedom

2    Freedom as Rational Control

3    Freedom as Volitional Control

4    Freedom as Discursive Control

5    Freedom and Collectivization

6    Freedom and Politicization

7    Freedom and Democratization





I amassed many debts in the course of writing this book. My thanks to David Held for proposing the project and to Victoria McGeer for talking through with me both the initial plan of argument, and many later versions. My thanks to Michael Smith with whom I share many of the ideas on responsibility and freedom that appear here and who has always been a great source of challenge and insight on the matters covered in the book (see Pettit and Smith 1996). My thanks to those who read through an early version and were most helpful with their comments. Michael Ridge and Jay Wallace were of incalculable help – Michael for an extended set of discussions and Jay for long and detailed observations – and I also found the comments of an anonymous reader very useful. And my thanks, finally, to John Braithwaite and Geoffrey Brennan with whom I have had a long-sustained discussion – a discussion sustained over nearly twenty years – of many themes that are central to the book. The text derives from a graduate seminar course that I taught as a Visiting Professor at Columbia University in 1999 but it was written for the most part at my permanent base in the Research School of Social Sciences, Australian National University. I thank all of those who took part in the New York seminar and made it so enjoyable and fruitful for me, as well as the many colleagues with whom I was able to discuss the material covered in the seminar.


This book differs from standard treatments of freedom in attempting to provide a connected discussion of free will issues and issues of political liberty. It looks for a theory of freedom in the classical, comprehensive mould exemplified by Thomas Hobbes in the seventeenth century, and Immanuel Kant in the eighteenth, as well as by their contemporaries and immediate successors. While Hobbes and Kant had distinguishable things to say about free will and political liberty, they clearly did not think of those topics as isolated and distinct. They derived their views in each area from deeper, common roots.

Why look for a theory in the classic, comprehensive mould? Why try to bring together themes that now belong to distinct bodies of literature, even distinct disciplines? I make two observations in explanation of the approach I have taken, one of them conceptual, the other methodological.

The conceptual observation is that the word ‘freedom’ as it is used in psychological and political contexts carries connected connotations and supports analogous implications. Thus, to mention the connotation that is prioritized in this book, the fact that someone is said to be free in either context normally means that they can be held responsible for what they do in exercise of that freedom. Suppose that someone is said to lack freedom of the will in a certain realm of activity. That implies straightaway that they should not be held responsible for what they do. Or suppose that someone is said, not to lack free will as such, but to lack some specific political liberty: say, the liberty to speak out against the government. Again that implies that the person cannot be held responsible – not at least fully responsible – for failing to speak out. In each case there is a tie between the ascription of freedom and the imputation of responsibility and no one can think that this is a mere accident. It testifies to a continuity of usage and meaning across the two domains of freedom talk.

But the fact that freedom of the will and political liberty are conceptually connected in this way – and connected, as we shall see, in many other ways too – does not argue in itself for treating them together. The compartmentalization that we find in the contemporary literature on freedom might still be useful and productive; it might still represent a profitable division of scholarly labour. It might. But I think it doesn’t. The reason turns on my second, methodological observation.

The sort of theory pursued in philosophy, whether in the psychological or political area, inevitably seeks to regiment various intuitions so as to fit them together in an appealing general structure. It looks for what John Rawls (1971) calls a ‘reflective equilibrium’ between particular intuitions or judgements – the data of the theory, if you like – and the more general, systematized claims that the theory defends. The methodological observation I want to make is that it will often make good sense, under this conception of philosophical theory, to seek a single theory of two conceptually connected domains rather than a different theory for each. I think that this makes good sense, in particular, with the domains of free will and political liberty.

The argument for this methodological claim is that the data in each of two connected domains may leave the choice of theory in that domain severely underdetermined, while the data in the domains taken together do much to constrain the choice of a single, comprehensive theory. The data in each of the separate domains may be consistent with multiple, more or less satisfactory equilibria, and yet the data in the domains taken together be consistent with only one equilibrium, or with a small family of equilibria. When we put the domains together, then we constrain the views to be defended in each domain, not just by the intuitions available in that area, but also by the intuitions available in the other domain. We ensure that there are more constraints on a satisfactory theory and we may thereby reduce the number of equally plausible candidates to a very few, or even to a singleton.

The lesson of this observation is that while domain-specific intuitions may be consistent with many theories of free will, and domain-specific intuitions consistent with many theories of political liberty, the combination of those sets of intuitions is capable of significantly constraining the choice of a single, unified theory of freedom. And this in fact is how I think it is.

There are many theories of free will, and many theories of political liberty, and little prospect in either area of definitively eliminating any of those options. The intuitions that guide theories of free will bear on what it means to say that an agent could have done otherwise, on what is involved in thinking that an agent authors an action in his or her own name, on what it is for an agent to be responsible for an action, and so on. And those intuitions have proven capable of being multiply interpreted and more or less satisfactorily systematized. The intuitions that guide theories of political liberty relate to whether raising the cost of an option inhibits an agent in the same way as the removal of an option, on whether we can say that natural obstacles as well as human beings interfere with choice, on whether a person can be unfree without suffering actual interference, and the like. And again those intuitions have lent themselves to a multiplicity of theoretical constructions.

In face of this underdetermination of theory in the two areas, it makes good sense to go back to the conceptual connection between free will and political liberty and to look into the prospects of a single, unified theory of freedom in general. And that is what I do in this book. I try to construct a theory that will bear at once on issues of free will and political liberty, and on the connections between the two. I seek out a theory that construes free will in such a way that it supports a defensible line on political liberty, and a theory that interprets political liberty in a way that fits with the line defended on free will. I seek a theory, in other words, that is constrained in each of its parts by the implications of that part across all the areas, psychological and political, in which we use the language of freedom.

I hope that the theory developed in the book will testify to the attractions of this holistic methodology. I do not think that the views defended here are so richly constrained that there are no plausible alternatives available; philosophy rarely works like that. But I do think that it is harder to see how the sort of unified theory presented here can be varied without significant loss than it is to see how to vary any of the familiar, compartmentalized positions that are defended in respect of free will and political liberty. I return to the issue in the Conclusion of the book.

In speaking of free will and political liberty, I have been using the terminology that has grown up under the very compartmentalization I reject. The language I prefer, and the language I use in this book, does not mark a distinction between psychological and political matters in such terms. I speak of freedom in the agent, rather than of free will, thereby avoiding any suggestion that it boils down to a psychological power of self-determination. And I speak, not of political liberty, but of the ideal that freedom in the agent would support as a target for political action; while I describe this as a political ideal of freedom, and even as an ideal of political freedom, I renounce the suggestion that it represents an autonomous domain of theory.

Freedom in the agent, as I think of it within my unifying project, has three aspects. It covers, first, the freedom of the action performed by an agent on this or that occasion; second, the freedom of the self implicit in the agent’s ability to identify with the things thereby done, rather than having to look on them as a bystander; and third, the freedom of the person involved in enjoying a social status that makes the action truly theirs, not an action produced under pressure from others. So construed, freedom in the agent has a social as well as a psychological aspect and the discussion of that freedom inevitably takes us beyond the realm of free will, traditionally conceived, and into politically relevant matters.

Freedom in an agent contrasts with freedom in the environment of an agent, where this is a function of how many and significant are the options made available by the impersonal parameters under which the agent exercises his or her freedom: say, the parameters dictated by a harsh natural order or a constraining social system. Questions to do with the environment of opportunity in which freedom is exercised come up only in the last couple of chapters of the book. In the earlier part I assume that the environment will make sufficient options available for people to have choices and I concentrate on the question of what it means, and what it is for them to enjoy agency freedom – freedom of action, self and person – in making those choices.

The book is organized in seven chapters and it may be useful to provide a brief overview of these. In the first I look at the concept of freedom, specifically as it is used on the side of agency, and I argue that it is unified by a connection with responsibility. To be free, in the most general sense, is to be fully fit to be held responsible; it is to be fully deserving of the sort of reactions, say those involving resentment or gratitude, that characterize face-to-face relations. The free action, the free self and the free person are nothing more or less than the sorts of action, self and person that are compatible with such fitness; they are, as I shall say, responsibility-compatible.

This characterization of the concept of freedom raises the question as to what it is – assuming that there is some single structure at work – that makes someone fit to be held responsible. Chapters 2, 3 and 4 explore this question in the context of individual as distinct from collective agents, looking at three theories of the capacity involved: these associate it, respectively, with rational control, volitional control and discursive control. The theory of freedom as rational control begins from an account of freedom in action, the theory of freedom as volitional control from an account of freedom in the self, and the theory of freedom as discursive control from an account of freedom in the person, though each serves to generate an overall view of free agency. I argue against the first two theories and defend the theory of freedom – specifically, the theory of freedom in the individual agent – that associates it with discursive control.

So far as the theory of freedom as discursive control involves a view of the free person it already has a social and political aspect to it. But the last three chapters of the book go on more explicitly to social and political questions. Chapter 5 argues that collective agencies, not just individual subjects, can possess freedom as discursive control; it thereby extends the theory developed in earlier, individual-centred chapters as well as connecting with the more social chapters that follow. Chapter 6 argues that one such collective agency, the state, should be given partial responsibility for furthering people’s enjoyment of freedom as discursive control and that the best conception of those requirements that the state can usefully monitor – the best political conception of freedom – is provided by the republican ideal of non-domination. Chapter 7 looks at the danger that any powerful state will itself represent for people’s enjoyment of non-domination, and ultimately of discursive control, and argues that the remedy lies in democratization, where this is represented as involving two dimensions, electoral and contestatory. The book ends with a short Conclusion in which I show how the main features of the position defended reflect the holistic methodology discussed in this Introduction.

The upshot is a treatment of freedom under which there is one single theme involved in all freedom talk – that of fitness for responsibility; there is one general theory of what constitutes such fitness in agents – discursive control; and this theory provides a standpoint from which we can see how issues of freedom go in the context of collectivization, politicization and democratization. The general line of argument can be gleaned from reading the concluding summaries that appear at the end of the different chapters.