42,40 €
Mehr erfahren.
- Herausgeber: John Wiley & Sons
- Kategorie: Geisteswissenschaft
- Sprache: Englisch
This study uses techniques from economics to illuminate fundamental questions in ethics, particularly in the foundations of utilitarianism. Topics considered include the nature of teleological ethics, the foundations of decision theory, the value of equality and the moral significance of a person's continuing identity through time.
Sie lesen das E-Book in den Legimi-Apps auf:
Seitenzahl: 509
Veröffentlichungsjahr: 2017
Ähnliche
Table of Contents
Cover
Title
Copyright
Preface
Chapter 1: Introduction I: The Structure of Good
1.1 Acts versus consequences, and agent relativity
1.2 The working of ethical considerations
1.3 Teleological structure
1.4 Does good exist?
Notes
Chapter 2: Introduction II: Weighing Goods
2.1 Locations of good, and separability
2.2 Three dimensions of locations
2.3 Dimensions with varying lengths
2.4 What things are good?
2.5 An outline of the argument
Notes
Chapter 3: Similarity Arguments
3.1 Parfit, Section 1
3.2 Parfit, Section 2
3.3 Harsanyi
Notes
Chapter 4: The Separability Theorems
4.1 Two examples
4.2 Definitions and theorems
4.3 The significance of additive separability
4.4 The rectangular field assumption
Notes
Appendix to Chapter 4: Proofs
Chapter 5: Expected Utility and Rationality
5.1 Axiomatic expected utility theory
5.2 The sure-thing principle
5.3 Individuation of outcomes
5.4 Transitivity of preferences
5.5 Rational requirements of indifference
5.6 Individuation and the sure-thing principle
5.7 The dispersion of value amongst states of nature
5.8 The rectangular field assumption
Notes
Chapter 6: The Coherence of Good
6.1 The goodness of uncertain prospects
6.2 Betterness and rational preferences
6.3 The argument for coherence
6.4 Representing betterness by utility
6.5 Bernoulli’s hypothesis
Notes
Chapter 7: Coherence Against the Pareto Principle
7.1 The conflict
7.2 Two better principles
7.3 Welfare economics
Notes
Chapter 8: The Principle of Personal Good
8.1 Qualifications
8.2 The beginning of a defence
Notes
Chapter 9: Equality
9.1 A utilitarian case for equality
9.2 Types of egalitarianism
9.3 Equality under uncertainty
9.4 Fairness and equality
Notes
Chapter 10: The Interpersonal Addition Theorem
10.1 Proof
10.2 Connecting dimensions
10.3 The utilitarian principle of distribution
Notes
Chapter 11: Utilitarian Metaphysics?
11.1 The intertemporal addition theorem
11.2 The principle of temporal good
11.3 The metaphysical argument
11.4 Conclusion
Notes
Bibliography
Index
End User License Agreement
List of Tables
Chapter 1: Introduction I: The Structure of Good
Table 1
Chapter 2: Introduction II: Weighing Goods
Table 2
Table 3
Table 4
Table 5
Table 6
Table 7
Chapter 3: Similarity Arguments
Table 8
Table 9
Table 10
Chapter 4: The Separability Theorems
Table 11
Table 12
Chapter 5: Expected Utility and Rationality
Table 13
Table 14
Table 15
Table 16
Table 17
Table 18
Chapter 6: The Coherence of Good
Table 19
Chapter 7: Coherence Against the Pareto Principle
Table 20
Chapter 9: Equality
Table 22
Table 23
Chapter 11: Utilitarian Metaphysics?
Table 25
List of Illustrations
Chapter 2: Introduction II: Weighing Goods
Figure 1
Figure 2
Chapter 4: The Separability Theorems
Figure 3
Figure 4
Figure 5
Figure 6
Chapter 6: The Coherence of Good
Figure 7
Chapter 9: Equality
Figure 8
Figure 9
Chapter 10: The Interpersonal Addition Theorem
Figure 10
Guide
Cover
Table of Contents
Begin Reading
Pages
C1
iii
iv
v
ix
x
1
2
3
4
5
6
7
8
9
10
11
12
13
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
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
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
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
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
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
251
252
253
254
255
e1
WEIGHING GOODS
Equality, Uncertainty and Time
John Broome
Copyright © John Broome 1991, 1995
First published 1991
First published in paperback 1995
Transferred to Digital print 2003
Blackwell Publishers Ltd
108 Cowley Road, Oxford OX4 1JF, UK
Blackwell Publishers Inc.
238 Main Street
Cambridge, Massachusetts 02142, USA
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.
British Library Cataloguing in Publication Data
A CIP catalogue record for this book is available from the British Library.
Library of Congress Cataloging in Publication Data
Broome, John.
Weighing goods/John Broome.
p. cm.—(Economics and philosophy)
Includes bibliographical references and index.
ISBN 0–631–17199–1 (Hbk) — ISBN 0–631–19972–1 (Pbk)
1. Economics—Moral and ethical aspects. I. Title. II. Series:
Economics and philosophy (Cambridge, Mass.)
HB72.B76 1991
330′.01—dc20 90—26662 CIP
Preface
This is a book about ethics, which uses some of the methods of economics. I think it is already widely recognized that formal methods derived from economics can contribute to ethics. This book is concerned with some features of the structure of good, and in that area I believe these methods are especially fruitful.
Not only do many of my methods come from economics, but many of my conclusions have economic applications, too. I shall, for instance, be investigating the value of equality. I hope, therefore, that this book may attract readers from both economics and philosophy. Inevitably, each group will encounter some difficulties. For philosophers, many of the techniques may be unfamiliar, and perhaps in places daunting. I ask them to be patient with the symbols. Nowhere in the entire book is there any mathematics beyond elementary algebra, and I have done my best to explain the algebraic arguments carefully. (I have twice mentioned logarithms, but only in examples that may be skipped.) The Appendix to Chapter 4 and Section 10.1 may also be skipped. They contain informal proofs, where some readers may find the elementary algebra clusters too densely on the page.
Economists, on the other hand, will find much of the algebra tediously slow. I ask them to be patient too. What, I think, they may find difficult to understand is my focus on good rather than preferences. My techniques come from utility theory, which was developed for preferences, but I have redirected them towards the structure of good. Even to separate a person’s good from her preferences will make many economists uneasy; it will suggest to them a worrying paternalism. Chapter 7, particularly Section 7.3, is addressed to this concern. It explains why my argument cannot be built around preferences. I hope these economists will be willing to recognize that there is such a thing as a person’s good, which may or may not be determined by her preferences. In working with good rather than preferences I am being more general. I am allowing for cases where good is determined by preferences, and also for cases, which do exist, where it is not. It is not paternalistic simply to speak of a person’s good.
This book began life as a long paper I wrote while I was a Visiting Fellow at All Souls College, Oxford, in 1982–83. I thank the college for its kind hospitality, and also the Social Science Research Council, which supported me for that year. I developed various parts of the argument in articles during the succeeding years. They began to come together as a book during a course of seminars I gave at Princeton University in 1988. I am grateful to Princeton for this opportunity. Finally, the Economics Department at Bristol University, where I work, allowed me a term of study leave in 1989 to complete the writing. I am very grateful to the members of the department.
Over the years, I have been very fortunate in the many valuable comments I have received on parts of the book, and on earlier versions of the arguments. I am sure I have forgotten some of the people who have helped me, and to them I offer both apologies and thanks. Those I particularly remember include Jonathan Dancy, Nicholas Denyer, Peter Hammond, Sally Haslanger, Susan Hurley, Daniel Hausman, Donald Hubin, Richard Jeffrey, Douglas MacLean, Amartya Sen, John Skorupski, and Brian Skyrms. My thanks to all of them. I want to express my especial gratitude to John Harsanyi, Shelly Kagan, James Mirrlees, Adam Morton, Derek Parfit and Larry Temkin for the exceptionally helpful and detailed comments I received from them on sections of the book. Besides all these, Duncan Foley, Alan Hamlin, Frank Jackson, Philip Pettit and Paul Seabright were all generous enough to read drafts of the entire book. I am sure the finished product has benefited greatly from their excellent advice. I am very grateful to them.
Parts of Chapter 1 also appear in my contribution to Foundations of Decision Theory: Issues and Advances, edited by Michael Bacharach and Susan Hurley and published by Blackwell in 1991. Parts of Chapter 11 also appear in my contribution to Interpersonal Comparisons of Well-Being, edited by Jon Elster and John Roemer and published by Cambridge University Press in 1991.
I should also like to acknowledge the help of WordPerfect UK.
Chapter 1Introduction I: The Structure of Good
One part of ethics is concerned with good. This book is about that part. Specifically, it is about the structure of good.
More exactly, it is about betterness. Some things are better than others, so that a relation of betterness holds between them. This book is about the structure of this relation. Though I shall use the word ‘good’ as well as ‘better’, I am not referring to some metaphysically mysterious entity, but only to the homely matter of which things are better or worse than others.
I shall be particularly concerned with just one aspect of the structure of good. Good occurs, as I shall put it, at different ‘locations’. For instance, good comes to different people: there is your good and my good, and the publisher’s good. People, then, are locations for good. They constitute one ‘dimension’ of locations, as I shall say. Good comes at different times, too, so times are locations for good, and constitute a second dimension. The fact of uncertainty gives us a third dimension. We do not know whether our behaviour will cause temperatures to rise globally by one degree Celsius, or two, or five. Each of these possibilities is an example of a ‘state of nature’. What happens in one state of nature will be good to some degree, and may be better or worse than what happens in another. So we can think of states of nature as locations of good, forming a third dimension. This book investigates how good occurring at different locations goes together to make up overall good. How is good aggregated? When aggregated across the dimension of people, for instance, is overall good simply the total of people’s good, so that the better of two societies is always the one with the larger total? Or alternatively, might a society with a smaller total of good, evenly distributed across the population, sometimes be better than a less equal society with a larger total? I shall be concerned with questions like this.
The methods of economics can help with this project in ethics. I shall be making use of some formal theorems from economics and decision theory. In their original home, these theorems were intended to say something about the structure of preferences. But they can be redirected towards the structure of good instead. Specifically, they link together the different dimensions of good. For instance, they link the way good is aggregated across people with the way it is aggregated across states of nature. So they make a connection between the value of equality and the value of avoiding risk. This book explores connections of this sort.
But before coming to my more detailed argument, I need to talk generally about the idea of the structure of good. Moral philosophy these days is more often centred on rationality rather than good, so my interest in good needs explaining. That is the purpose of this first introductory chapter.
According to some ethical theories, the concern for good amounts to the whole of ethics, not just a part. How we should live our lives, these theories say, and how we should act on each particular occasion, are determined entirely by the pursuit of good. Once we know what is good, or more exactly what is better than what, we shall know the right way to live and the right way to act. Let us call a theory that says this teleological. A nonteleological ethical theory, on the other hand, gives a role in ethics to other considerations besides good. Most nonteleological theories, nevertheless, give some role to good too.1 So this book will have something to contribute to nonteleological theories too.
But what exactly is this distinction I am drawing when I pick out one part of ethics as concerned with good or betterness? What really distinguishes an ethical theory concerned only with good from any other ethical theory? How, indeed, could an ethical theory be concerned with anything other than good? Is not what an ethical theory is concerned with necessarily what the theory considers good? If a theory says we should live in such-and-such a way, or act in such-and-such a way, does that not mean it considers these ways of living and acting to be good? John Rawls defined an ethical theory as teleological if ‘the good is defined independently from the right, and then the right is defined as that which maximizes the good’.2 This is much the same as the definition I gave. But what distinction is Rawls making? If a theory claims that such-and-such is the right way to live or act, is it not also implying that it is a good way to live or act?
Evidently some work is needed to draw a line between the part of ethics that is concerned with good and other parts. That is the first object of this chapter. This book leaves the other parts aside, and I want first to separate them out clearly. Since teleology is concerned only with good, and with all of good, the boundaries of teleology coincide with the boundaries of good. So I shall proceed by asking what, exactly, distinguishes teleological from nonteleological ethics. This is a useful method for delimiting the concept of good. But it does not mean that the arguments of this book are relevant to teleologists only. As I say, good is important in nonteleological ethics too.
To start with, Section 1.1 describes two ways of making the distinction between teleological and nonteleological ethics, and explains why I think they are unsatisfactory. Sections 1.2 and 1.3 then describe the way I favour. Section 1.4 discusses what type of priority the concept of good has in ethics.
1.1 Acts versus consequences, and agent relativity
How, then, are we to draw the line between teleological and nonteleological ethics? The contrast between good and right suggests one answer. Rightness is commonly thought of as a property of acts, and goodness as a property of states of affairs. So we might make the distinction like this: teleological ethics first evaluates states of affairs, and then determines the value of an act from the value of the state of affairs it leads to – of its consequences, that is; nonteleological ethics, on the other hand, assigns intrinsic value to some acts, independently of their consequences. For instance, the view that breaking a promise is wrong in itself, quite apart from its consequences, would be nonteleological.
The idea is, then, that teleology insists acts are to be valued by their consequences alone, and that this is how it should be distinguished from other theories. This is the source of the term ‘consequentialism’. ‘Consequentialism’ is these days used more often than ‘teleology’, but it means the same, except that some authors narrow its meaning in a way I shall be explaining. I prefer the older term for two reasons that will soon appear.3
However, this way of making the distinction between teleological and nonteleological ethics turns out to fail. It relies on a division between an act and its consequences that cannot be maintained. It is not clear where an act leaves off and its consequences begin. If you perform an act, one consequence will be that you have performed it. If you break a promise, one consequence will be that you have broken a promise, and the wrongness of promise breaking can be taken as a bad feature of this consequence. Teleology, in evaluating the consequences of promise breaking, can therefore take account of the wrongness of promise breaking itself. In this way, the intrinsic values of acts can be absorbed into teleology.
This fact is generally recognized. In all the recent debate about teleology, nothing has been made to depend on separating the value of an act from the value of its consequences. The fact that an act has been done is generally counted amongst the consequences of that act, and the intrinsic value of an act is counted as a teleological consideration in the act’s favour. Samuel Scheffler, for instance, says:
When I speak of the act-consequentialist as requiring agents to produce the best overall outcomes or states of affairs, I do not mean that the act-consequentialist divides what happens into the act and the outcome, and evaluates only the latter with his overall ranking principle. Rather, the act itself is initially evaluated as part of the overall outcome or state of affairs. The act-consequentialist first ranks overall outcomes, which are understood, in this broad way, to include the acts necessary to produce them, and then directs the agent to produce the best available outcome so construed.4
I shall follow this practice. I shall not try to distinguish between the value of an act and the value of its consequences. I shall apply the notion of goodness to acts as well as to their consequences, and identify the goodness of an act with the goodness of its consequences. I shall take both to include any intrinsic value the act may have, as well as any good that may result from the act. Teleological ethics, then, in its pursuit of good, will take account of the intrinsic values of acts.
This is one reason I find the term ‘consequentialism’ unsatisfactory: whatever distinguishes teleological from nonteleological ethics, it is not that the former values only consequences, whereas the latter allows acts to possess intrinsic value.
Now teleology is interpreted so widely, it may seem to leave no room for a nonteleological ethics. Certainly, many values that have traditionally been classed as nonteleological can be brought under the umbrella of teleology taken this way. Of course, it means recognizing many sorts of good and bad besides the sorts recognized by the most traditional teleological theories – particularly by traditional utilitarianism. Breaking a promise must be taken as a bad thing in itself, quite apart from what anyone feels about it, and apart from the happiness, pleasure or pain that may result. Another example: the value of fairness can be recognized by teleology, provided unfairness is taken as a bad thing in itself, whether or not it makes the victim unhappy.5
Nevertheless, there are views that are incompatible with even this widely conceived teleology. A famous example is this.6 Suppose that by breaking a promise I can bring it about that in the future five promises are kept that otherwise would have been broken. Some people would think I ought to keep my promise nevertheless, unless the circumstances are exceptional. This opinion, however, seems inconsistent with teleology. If breaking promises is bad, then the breaking of five promises must be worse than the breaking of one. Other things being equal, then, keeping my promise will have worse consequences than breaking it, even taking account of the intrinsic badness of promise breaking. So teleology, it seems, must be in favour of breaking it.
How might one defend the view that I ought to keep my promise even at the expense of five broken promises? Here is one defence. In this moral dilemma, goes the defence, it is wrong for me to take up a neutral, impersonal point of view, from which everybody’s promise keeping counts the same. My own special position in the affair makes a difference to what I ought to do. Keeping my promise has a special value for me. It may count for me more than the keeping of five promises by other people. My valuations should be ‘agent relative’, as it is generally put. And why should that be? One possible answer is that I have a greater responsibility for my own acts than I do for the acts of other people, even if those acts are brought about by me.7 In this case, the five people who break their promises bear the responsibility for this wrongdoing of theirs more than I do, even though it is brought about by my act of promise keeping.
A lesson commonly drawn from this example and others like it is that nonteleological ethics must be agent relative in some way: if it is right for me sometimes to do something that will not have the best possible consequences, that must be because of my particular position as an agent. Conversely, agent neutrality is often included within the definition of consequentialism.8 Consequentialism has come to be distinguished by agent neutrality, and nonconsequentialism by agent relativity. Most of the recent discussion of consequentialism has really been a discussion of agent neutrality.9
But I think the debate between agent relativity and agent neutrality has distracted attention from other important questions. The issue between teleological and nonteleological ethics should be separated from the issue between agent-neutral and agent-relative ethics.10 It would be a mistake, therefore, to tie teleology down to agent neutrality. This is my second reason for using the term ‘teleology’ instead of ‘consequentialism’: to escape from the definition of consequentialism as agent neutral. There is a more natural way of characterizing teleological ethics, which allows for nonteleological theories that are agent neutral,11 and also teleological theories that are agent relative.12
1.2 The working of ethical considerations
When there is a decision to be made, there will often be ethical considerations on both sides. Against breaking my promise in the example is that this would be an intrinsically wrong act. In favour is that it will prevent five wrongs of promise breaking. The considerations must somehow come together to determine which act is right, all things considered. How do they do that? Teleological and nonteleological theories, as I prefer to distinguish them, give different answers to this question.
The answer of teleology is that each consideration contributes to the goodness or badness of the alternative acts. All the considerations together determine how good or bad the acts are. And which act is right is determined by the goodness of the alternatives. For instance, teleology takes the intrinsic wrongness of breaking a promise to be a bad feature of the act of promise breaking, which goes together with other good and bad features to determine the overall goodness of the act. The nonteleological answer, on the other hand, is that ethical considerations do not always work in this particular way. They may determine what ought to be done, not through determining the goodness of acts, but in some other way.
For the moment, I shall use this difference to define teleology informally: a teleological theory is one in which the rightness of acts is determined by their goodness. I shall tighten up the definition in Section 1.3.
A teleological theory will have to specify exactly how goodness determines rightness; it will have to supply a mapping from goodness to rightness. The standard sort of teleology is maximizing. The main elements of a maximizing mapping are these. If one act, out of those available, is better than the others, then that act ought to be done; only that act is right, and the others are wrong.
If several acts are equally good, and all of them better than any of the others available, then one of them ought to be done and it does not matter which. Any other act is wrong. Satisficing teleology has a different mapping.13 It says that any action is permissible – not wrong – provided it is ‘good enough’, even if a better act is available. So when Rawls says that in teleology ‘the right is defined as that which maximizes the good’,14 he is being a little too specific. He is referring to standard teleology only. My formula, that in teleology goodness determines rightness, is intended to be more general. But if we set aside nonstandard teleology including satisficing theory, Rawls’s formula is correct. In standard teleology, the right act is the best.
The metaphor of weighing often fits teleology. Each consideration for or against an act is a good or bad feature of the act. The act’s overall goodness is determined by putting together the goodness and badness of these features. This process is often analogous to weighing: good and bad features are weighed against each other.
Not all teleological judgement need be much like weighing, though. It is consistent with teleology to think that some considerations dominate others lexically. You might think, for instance, that any unjust act is worse than any just one, whatever other features the acts possess. There is nothing in literal weighing analogous to this, but your view is teleological because it is about the badness of acts.
You might also think that some considerations are incommensurable with others in a different way. Suppose some proposed social policy will make a community more affluent, but at a cost in social conflict. There is an alternative policy that will better promote harmony, but lead to less affluence. You might think that the benefits of these policies cannot be weighed against each other. You might think that neither policy is definitely better than the other, but that they are not equally good either. If they were equally good, you would be indifferent about which was chosen, but actually you think it matters very much. Your view is teleological. It takes the ethical considerations to be good or bad features of the alternative policies: affluence is good and so is harmony. But it supposes that some sorts of good cannot be weighed against others. Now, incommensurability may be a problem for teleology.15 In order to accommodate it, teleology will have to show how its mapping from goodness to rightness can cover cases like the one I described, where none of the alternative acts is better than the others, and yet the acts are not equally good either. Are all of these acts permissible, say? If so, what makes the case different from one where they are simply all equally good? But however teleology copes with them, goods that cannot be weighed against each other seem a genuine possibility. They cannot be ruled out at the start.
Furthermore, there may be complicated interactions amongst considerations, which also strain the metaphor of weighing. Chapter 2 contains some examples. In sum, weighing is typical but not definitive of teleology. This book concentrates on weighing, though, and examines in detail one aspect of it.
On page 5, I mentioned one opinion about the promising example: that I ought to keep my promise because my promise breaking counts more for me than other people’s. This is an agent-relative opinion. It is also teleological, as I have now defined teleology. It treats the wrongness of promise breaking as a bad feature of an act. It weighs the badness of my promise breaking against the badness of other people’s, to determine the overall badness of the alternative acts. For me, my promise breaking weighs more than other people’s. In the end, for me, breaking my promise comes out worse overall. Therefore I ought not to do it. This is a teleological argument – an agent-relative one.
One might doubt the validity of this argument, though. It claims that breaking my promise is worse for me than keeping it, and one might doubt that such a claim makes good sense. Certainly, there is one clear sense it could have. It could mean that breaking my promise is against my interest: that it would make me less well off than I would otherwise have been. But that is not what it is supposed to mean in the argument. The good in question is not my private wellbeing, and the argument is not appealing to my own self-interest. The good is general good, not mine; but general good evaluated from my own special position as an agent. And one might doubt that good can really be agent relative in this way. Agent-relative teleology appeals to agent-relative good, and one might be dubious of such a concept.
Not many authors have been willing to adopt explicitly an agent-relative concept of good; Amartya Sen is an exception.16 But anyone who uses an argument of the form I described is implicitly committed to one. On page 13 I shall specify more exactly what I mean by this form.
Now an example of a nonteleological opinion. It, too, is in favour of my keeping my promise in the example, but for a different reason. It also takes breaking a promise to be intrinsically wrong. But it does not treat this intrinsic wrongness as a consideration to be weighed against others in determining the overall goodness of the alternative acts. Instead, the wrongness of promise breaking simply determines that I ought not to do it. It is what Robert Nozick calls a ‘side constraint’.17 So the argument goes directly to what I ought or ought not to do, without first estimating the goodness of the alternatives.
I do not insist that the ethics of side constraints cannot, in the end, be reduced to a form of teleology. Perhaps it can, and Section 1.3 specifies exactly what would be required to make the reduction successful. But certainly, on the face of it, and as Nozick presents it, it is expressly nonteleological according to my definition. Compare, for instance, the view that promise keeping is a side constraint with the view that promise keeping is a good that dominates others lexically. The latter says that, whenever promise keeping is in conflict with any other consideration, promise keeping wins. This view is teleological. It says that breaking a promise is a very, very bad thing, so bad that its badness outweighs all other badness. In the example, it would imply that I ought to break my promise. Doing so will lead to only one instance of the great bad of promise breaking, whereas keeping the promise will lead to five. The side-constraint view, on the other hand, does not necessarily insist that promise breaking is very bad. It says only, and simply, that I ought not to do it.
Compare the side-constraint view, too, with the agent-relative teleological view that my promise keeping is a good that, for me, dominates others lexically. Imagine the five other promises in the example are actually mine: by breaking a promise now I can bring it about that I shall, in the future, keep five promises that otherwise I would have broken. Suppose, say, that if I break a promise now, the experience of guilt at my present tender age will stiffen my resolve in future life. In this case, the agent-relative teleological view would imply that I should break my promise now, since that would lead to five of my promise keepings instead of one. Yet a thoroughgoing side-constraint view about promises would say I ought to keep my present promise.
It is commonly believed that the ethics of side constraints is necessarily agent relative. But this is not so. Consider the view that whenever a miner is in mortal danger trapped in a mine, all available resources should be devoted to rescuing him. This will reduce the resources devoted to safety measures in mines, and so lead to the deaths of more miners in the future. Nevertheless, it is what ought to be done. This is a side-constraint view. But it is agent neutral. It says that all of us, equally, should contribute to saving the miner as far as we can.
Since side-constraint theory is an important example of nonteleological ethics, the popular belief that it is necessarily agent relative helps to sustain the popular association between nonteleological ethics and agent-relative ethics. But what makes side-constraint theory nonteleological is not agent relativity. It is the way it takes ethical considerations to work: side constraints determine what ought or ought not to be done directly, and not by determining goodness.
However, side-constraint theory might, perhaps, be called moment relative. Take the mining example. (The point is plausible in the promising example too.) At present we have a reason to rescue the miner who is now trapped, and this reason fully determines what we should do now. When the next miner is trapped, we shall then have a reason to rescue him. That reason will fully determine what we should do then. But at present it does not count at all; it gives us no reason to save resources for the future rescue. The reason, then, applies at one time but not at another. So it might be called a moment-relative reason.
It even seems possible to me that, with some work, side-constraint theory might be reduced to a moment-relative teleology, in which certain present reasons dominate other reasons lexically. Perhaps this is a way it might be brought within the fold of teleology. But at least until the work has been done, side-constraint theory stands as an example of nonteleological ethics that is not necessarily agent relative. I shall mention another example on page 15, and another later in the book, on page 159.
1.3 Teleological structure
I defined teleology as the theory that rightness is determined by goodness. In standard teleology, the right act is the best.
But, once again, the question arises: how does this simple definition exclude anything? How could any ethical theory deny that, when faced with a choice between acts, you should choose the best? I have already offered the ethics of side constraints as an example of a nonteleological theory, on the grounds that it does not work by assessing the goodness of the alternative acts. But whatever a theory says about the working of ethical considerations, once it has decided which act ought to be done, one might think there is nothing to stop it just calling this act the best. And surely this would be a very natural thing to do. So any theory could be made teleological in arrears.
Now, I did say that side-constraint theory might perhaps be reducible to teleology. But it would at least take some work, and it is certainly not true that any theory is automatically teleological according to the definition. That rightness is determined by goodness is genuinely a constraint on teleology. It means that teleology is constrained by the structure of good as I shall put it. A teleological theory implies that between acts there is a betterness relation:
(where the blanks are to be filled in with acts), and that this relation determines the rightness of acts. A betterness relation necessarily has a particular structure. When I speak of the structure of good, I mean more exactly the structure of this relation.
I can immediately say one thing about this structure: it is an ordering. Acts are ordered by their goodness. That is to say, the relation is transitive and reflexive. For the betterness relation to be transitive means that, for any acts A, B and C if A is at least as good as B, and B at least as good as C, then A is at least as good as C. For it to be reflexive means that for any act A, A is at least as good as A. These are both necessary truths.
Indeed, I take them to be truths of logic. The schema
denotes a transitive and reflexive relation, when the name of any property is substituted for ‘F’. Things are ordered by their Fness.18 The best way to make this clear is to examine an apparent counterexample. The relation
seems at first to be intransitive, because the corresponding strict comparative relation
seems to be intransitive. As you become more and more westerly, surely, you end up not westerly at all but easterly. But precisely what property is ‘more westerly than’ meant to be the comparative of? It might be the property of being to the west of somewhere else. In that case every place on earth has it equally, and nowhere on earth is more westerly than anywhere else. Alternatively, it might be some property of local westerliness, such as westerliness in Ireland. The comparative relation of local westerliness, however, is plainly transitive; think about the relation of being more westerly in Ireland. Or perhaps it might be the property of being to the west of here (Bath, that is). Does this property have an intransitive comparative? Boston is more to the west of Bath than Dublin is, and Papeete more than Boston is. Therefore, Papeete is more to the west of Bath than Dublin is. Is Sydney more to the west of Bath than Papeete is? We have a decision to make about this. We might decide that, being to the east of Bath, Sydney is not to the west of Bath at all. It is not, therefore, more to the west of Bath than Papeete is. No intransitivity here. Or we might decide that Sydney is indeed to the west of Bath, and more so than Papeete. We shall presumably draw the conclusion, eventually, that London is more to the west of Bath than Dublin is. Bath, presumably, is not to the west of Bath, and therefore not more to the west of Bath than London is. No intransitivity here either. Possibly we might decide that Bath is to the west of Bath. If we do, we shall be forced to conclude that everywhere on earth is equally as much to the west of Bath as everywhere else, and that I was wrong to say Boston is more to the west of Bath than Dublin is. This discussion shows that, whatever decisions we make, we shall always be guided by logic to preserve the transitivity of the comparative.
Strictly, a transitive and reflexive relation is only a partial ordering, or more strictly still a quasi-ordering.19 An ordering strictly needs also to be complete. For the betterness relation to be complete would mean that, for any acts A and B, either A is at least as good as B, or B at least as good as A. That is to say, of any pair of alternatives, one is better than the other, or the other than the one, or the two are equally good. But I see no reason to assume the betterness relation is complete; there may be different goods that are incommensurable. I mentioned an example on page 7. In this chapter I shall use the term ‘ordering’ loosely. When I say the betterness relation is an ordering, I do not mean to imply it is complete. I mean only that it is transitive and reflexive.
Teleological ethics, then, says there is an ordering of acts that determines the acts’ rightness. And this is actually enough to define teleological ethics. Any ethical theory with this implication is teleological. I am defining teleology, then, by its structure.Standard teleology has a maximizing structure: it says there is an ordering of acts such that, if one act out of those available comes higher in the ordering than the others, then that act ought to be done. My more general formula – that teleology says there is an ordering of acts that determines their rightness – is intended to allow for nonstandard theories such as satisficing. If a theory satisfies this condition, let us say it has a teleological structure.
Up to now, I have been defining teleology as the view that the rightness of acts is determined by their goodness. I am not now changing this definition; I am spelling out what it means. The reference to ‘goodness’ in the original definition simply indicates that rightness is determined by an ordering. It does not imply that teleology is defined in terms of a prior notion of good. Each teleological theory determines its own notion of good, constrained by the fact that betterness has to be an ordering. Each implies the existence of an ordering that determines rightness. The ordering is what the theory takes to be the betterness relation.* It defines the theory’s notion of good. Any ethical theory with a maximizing structure is teleological, and what it aims to maximize is what it takes to be good.
Look back to one of the opinions I mentioned about the promising example: the opinion that my promise breaking outweighs, for me, the promise breaking of others. This is agent relative. But I described it on page 8 as teleological also, because it takes the ethical consideration that breaking a promise is wrong to work in a particular way. It works by helping to determine the relative goodness of my alternative acts. Now I can say more precisely what this means. It is a matter of structure. It means that this ethical consideration goes, with others, to determine an ordering of the alternatives, which in turn determines what I ought to do. The ordering is a betterness relation, and it defines a notion of good. In this case, it will be an agent-relative notion. I said one might be dubious about agent-relative goodness. But to object to it, one would have to object to the whole ethical theory. The objection cannot simply be that one does not like to use the term ‘good’ in an agent-relative way. Any agent-relative theory with a teleological structure implies an agent-relative good.
I have by now departed from Rawls’s way of characterizing teleology. Rawls wishes to distinguish a moral theory as teleological, not simply by the structure of the judgements it makes, but also partly by the ground of its judgements. Even if a theory says the right act is the best, he would not count it as teleological unless it also says that this act is right because it is the best. As he puts it, teleology gives good priority over right. But once we have picked out theories that say the right act is the best, I doubt if there is really this extra distinction to be made. Suppose a theory identifies a particular act as right on the basis of some principles. And suppose these principles are such as to determine an ordering of acts, so that the right act may fairly be called the best according to the principles. How would we then decide whether or not the theory says the right act is right because it is best? Ultimately what makes it right is the principles. But these principles determine the goodness of acts as well as their rightness, and who is to say whether or not they determine rightness directly or by means of their determination of goodness? If someone says, ‘The right act is right because it is the best according to the principles’, surely it would be odd to disagree.
When it comes down to it, to make his distinction workable Rawls has to rely on a prior intuition about what principles are matters of right rather than good. He himself thinks of good as an impersonal or suprapersonal object for morality, so that the particular duties we owe to individuals count as principles of right. Here is an example of Rawls’s intuition at work:
If the distribution of goods is . . . counted as a good ..., and the theory directs us to produce the most good (including the good of distribution among others), we no longer have a teleological view in the classical sense. The problem of distribution falls under the concept of right as one intuitively understands it, and so the theory lacks an independent definition of good.20
But this sort of intuition does not seem to me secure enough to found a definition on. I, for one, have no intuitive grasp of why the problem of distribution should fall under the concept of right. In Chapter 9 I shall be examining several alternative ways of treating equality of distribution as a good.21 Furthermore, Will Kymlicka points out that even utilitarianism can be derived in a natural fashion from something Rawls would count as a principle of right: that each person’s interest be given equal respect.22 So on Rawls’s intuitive distinction, utilitarianism, arrived at this way, would come out nonteleological. But if it has this result, the distinction is pointless, and certainly not what Rawls has in mind. I think the only way to make a clear distinction is by structure, as I have made it.
That the betterness relation is an ordering is a serious constraint on teleological ethics. I said earlier that it may be possible to recast side-constraint theory as a moment-relative lexical teleology. We can now see what this job would entail. The theory must be given a transitive structure. I do not know whether that would be possible.
And consider this example. It is a cut-down version of Derek Parfit’s ‘Mere Addition Paradox’.23 A couple have a child, and are wondering whether to have another. If they do, they may either spread their resources evenly between the children, or concentrate them on the first. The alternatives are shown in Table 1; the numbers in the table stand for the children’s degrees of wellbeing.
Table 1
The couple might reason like this. In the comparison between A and B, A is better for the first child. And there is nobody for whom A is worse than B. Whom could it be worse for? – in A only the first child exists, and it is not worse for her. So A is better than B. Or at the very least, A is at least as good as B. In the comparison between C and A, C is better for the first child and no worse for the second. So C is better than A. Or at the very least, C is at least as good as A. But also, B is better than C, because B has a greater total of wellbeing, and has it equally distributed. So it is not true that C is at least as good as B. Transitivity, however, requires that it should be. This couple, then, might come to the conclusion that there is an intransitivity in the betterness relation. Larry Temkin has used examples like this, taken from Parfit, to argue powerfully that the betterness relation is indeed intransitive.24
But, despite its attractions, this is a conclusion that cannot logically be drawn. If there are intransitivities, they cannot be in the betterness relation. The situation must be described in other terms, however tempting it may be to use the terms of betterness. (It may seem puzzling that it should be so tempting to deny a law of logic. But any paradox is like that: it tempts one to assert both a proposition and its negation.) What Parfit’s examples show, I think, is the strong hold teleology has on us. We are strongly drawn to organize our moral thinking around the concept of good. But using that concept imposes a structure – a transitive one – on our thinking. If our thinking does not fit the structure, we shall have to give up teleology, or else restructure our thinking. Certainly, we must not fall into logical inconsistency.
That the betterness relation is an ordering is, by itself, a serious constraint on teleology. But there is much more to the structure of good than just that; there are many other structural conditions that a betterness relation must conform to. Any ethical theory is teleological if it implies the existence of an ordering that determines the rightness of acts. This one structural condition is enough to identify the theory as teleological; the ordering is the theory’s betterness relation; it fixes its notion of good. But a teleological theory cannot be correct unless its notion of good conforms to all the other structural conditions as well.
The purpose of this book is to examine in detail some of these other conditions. I shall mention just two examples now. First: I shall argue in Chapter 6 that the betterness relation must satisfy the axioms of expected utility theory. And second: this book is very much concerned with the way general good is made up of the good of individuals. I shall particularly be testing out the truth of this popular view about it:
The utilitarian principle of distribution. One alternative is at least as good as another if and only if it has at least as great a total of people’s good.
The utilitarian principle says it is best for good to be distributed in such a way as to maximize the total of people’s good. It values the total of good only, and not equality in the distribution of good. If the total would be maximized by making the society very unequal, then it says that is what should be done. Chapters 3, 9, 10 and 11 all discuss this principle.
1.4 Does good exist?
An ethical theory is teleological if it says one ought always to do the act that is best. A nonteleological theory must therefore on some occasions say one ought to do an act that is not the best. This may sound paradoxical, and may give nonteleological ethics a paradoxical air from the start. To counter this impression, Philippa Foot in her article ‘Utilitarianism and the virtues’ takes the radical step of denying that there is such a thing as ‘the goodness of states of affairs’. Since I am proposing in this book to examine the structure of the goodness of states of affairs, I need to respond to Foot’s denial.
What Foot really objects to is a notion of good that ‘stands outside morality as its foundation and arbiter’. She is happy with a notion that ‘appears within morality as the end of one of the virtues’ (namely benevolence).25 And she does not object to notions of good that are constructed within particular ethical theories.26 What she denies, if I understand her, is that there is such a thing as good existing in advance of ethical thinking, which ethical thinking takes as its object but does not itself determine. She argues that we simply do not have such a notion of good. She might have added, like J. L. Mackie,27 that this sort of good would be metaphysically mysterious. Let us, at least for the sake of argument, accept Foot’s objections to good conceived this way. Teleological ethics, she believes, sees good like this. Is she right?
This chapter suggests a different possibility. Certainly, teleology aims at good. But this need not mean it takes good as an externally determined objective. The pursuit of good may give to ethics, not an objective, but a structure. It may fix the way in which ethical considerations work and how they combine together. It may provide, not a foundation or an objective, but an organizing principle for ethics.
So far, this is only a suggestion. I have argued that teleology should be identified by its structure, but that is not enough to show that this new suggestion is correct. What matters is not how teleology is identified but how it is justified. How might one show that ethics does, indeed, have the structure teleology imputes to it? It needs to be shown, first, that rightness is determined by an ordering. But this is only the beginning of the task of justification. The structure of good, as I have said, includes many other conditions besides simply that betterness is an ordering. Some are described in this book. All of them will need to be justified too.
How? Roughly speaking, there are two styles of argument available. One is to say there is an object, good, that ethics aims at. This object determines what ought to be done. Acts will therefore be ordered by their goodness, and this ordering will determine their rightness. That would be the first part of the task of justification accomplished. It may also be possible to accomplish the other parts too – justify the other structural conditions – by arguments in this style. These arguments will appeal to a prior notion of good that stands outside ethics and gives it its object. If teleology is defended like this, it is indeed as Foot conceives it.
The alternative is to justify each of the structural requirements individually on other grounds. The most plausible are grounds of consistency or coherence.28 This second style of argument will almost certainly appeal to the idea of rationality. If an act is right, it must be rational, so the structure of rationality must tell us something about the structure of ethics. Rationality is constrained by structural conditions, which have been thoroughly examined and debated. These conditions are often defended on grounds of internal consistency rather than by an external object. For example, a commonly accepted structural condition is that a rational person must have preferences that constitute an ordering. That is to say, her preference relation
must be transitive and reflexive. If this condition could be established on grounds of internal consistency, then one might be able to argue from it to a conclusion about the structure of ethics: that rightness must also be determined by an ordering. If this argument was successful, the conclusion would have been derived from internal consistency rather than from an external object. And one might be able to do the same for all the other structural conditions of teleology.
Once the teleological structure of ethics had been justified in this second style, the notion of good would then be determined by the structure. It would arise out of ethical thinking, therefore. It would not stand outside morality as its foundation and arbiter. If teleology is defended by arguments in this second style, Foot should have no quarrel with it.
This book defends various claims about the structure of good. Does it use the external or the internal style of defence? Neither purely; both play a part. There is certainly a large proportion of the internal style. I shall explain in Chapter 6 that in one area it turns out to be unavoidable. Chapter 6 deals with acts whose results are uncertain. The goodness of such an act is a tricky notion. Take an act that is likely to have a good result, but suppose it turns out unluckily to have a bad one. It is hard to know whether this should count as a good or a bad act. We cannot expect to find a reliable notion of goodness here, prior to ethical examination, from which one might argue to structural conclusions. At that point in the book (particularly in Section 6.3), my arguments about the structure of good will be derived entirely from the structure of rationality. And that in turn will be derived partly from internal considerations of consistency. The argument in this book, then, is certainly not founded on a complete, prior, external, notion of good.
On the other hand, the notion of good cannot be determined entirely by structural conditions, either. A complete description of the structure of good does not constitute a complete analysis of the meaning of ‘good’. An ethical theory may have all the right structure, and so be teleological and possess a conception of good that meets all the structural conditions; yet nevertheless its conception of good may be wrong. An example is the theory that, when faced with a choice, one should always pick the act likely to produce the greatest total of pain. This is a teleological theory, with all the right structure, and its conception of good is pain. But good does not actually consist in pain. So this theory is wrong. It follows that teleological ethics cannot be fully justified on grounds of internal coherence. If it could, the theory I have just described would be fully justified. But it cannot be, because it is wrong.
It also follows that there are, actually, external criteria available for assessing the goodness of acts. The theory I mentioned is wrong because it recommends acts that are, actually, bad. It is therefore legitimate to appeal sometimes to external criteria in arguments about the structure of good, as I shall do. For instance: some things, such as happiness, are good, and others, such as pain, are bad. These are external facts about goodness. Foot may be right to doubt the existence of a complete conception of good, prior to ethics, on which ethics is founded. But there are certainly prior elements of a conception. Foot herself, indeed, recognizes enough of a prior conception of good for the virtue of benevolence to take it as its object.29 Furthermore, I shall argue on page 106 that internal conditions of consistency actually rely on external facts about goodness to give them a determinate meaning. I think our notion of good is formed out of external and internal conditions woven together.
I think the notion of good provides, not an object for ethics to aim at, but a valuable organizing principle in ethical argument. It is an accommodating principle: many ethical views can be made consistent with it. It can accommodate the value of fairness. According to Amartya Sen,30
