Desire as Belief: A Study of Desire, Motivation, and Rationality 2020952661, 9780198848172

Citation preview

Desire as Belief

Desire as Belief A Study of Desire, Motivation, and Rationality A L E X G R E G O RY

1

1 Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Alex Gregory 2021 The moral rights of the author have been asserted First Edition published in 2021 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2020952661 ISBN 978–0–19–884817–2 DOI: 10.1093/oso/9780198848172.001.0001 Printed and bound in the UK by TJ Books Limited Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

Acknowledgements I feel very lucky to belong to a profession where people are so generous with their feedback, and where even the most casual conversations at conferences are often highly intellectually productive. But this widespread generosity carries a cost, which is that it becomes impossible to remember and thank everyone: a list of everyone who has had a positive influence on this book would be a list of almost every philosopher I’ve spoken to over the last ten to fifteen years. I could try to prioritize these people by their degree of influence, but that feels somehow akin to trying to rank one’s friends. So rather than attempting such tasks, I’ll just record a more general thanks to everyone I’ve spoken to about these issues—departments, conference audiences, and particular individuals, young and old, world-­famous and comparatively unknown—and thank the following people who provided detailed and wide-­ ranging comments directly on this book manuscript: Giulia Felappi, St.John Lambert, Brian McElwee, Conor McHugh, Denis McManus, Michael Milona, Genia Schönbaumsfeld, Andrew Stephenson, Kurt Sylvan, Lee Walters, Jonathan Way, and Daniel Whiting. I also received extremely helpful and constructive comments from two anonymous readers for OUP: the book is definitely much improved for their input and I’m very grateful for the effort they put in. I’m also grateful to Peter Momtchiloff at OUP for seeing this book through to publication. Last, but not least: thanks to Jen. She models the virtues so well, and is especially well practised at patience. * Some small parts of this book borrow from prior publications of mine: • Some material from ‘Why Do Desires Rationalize Actions?’ in Ergo (2018) 5:40 is reused, primarily in Chapter 4. • Some material from ‘Are All Normative Judgements Desire-­Like?’ in The Journal of Ethics and Social Philosophy (2017) 12:1, pp. 29–55 is reused, primarily in Chapter 3. • I am grateful to The Thought Trust and Wiley Periodicals, Inc. for permission to reuse material from ‘How Verbal Reports of Desire May

x Acknowledgements Mislead’ in Thought (2017) 6:4, pp. 241–9, which primarily appears in §2.4.1 and §5.4. • I am grateful to The University of Illinois Press for permission to reuse material from ‘The Guise of Reasons’ in American Philosophical Quarterly (2013) 50:1, pp. 63–72 and (2017) 6:4, pp. 241–9, which primarily appears in §6.3. • Some work is also reused from my ‘Might Desires Be Beliefs About Normative Reasons for Action?’ in The Nature of Desire, edited by Julian Deonna and Federico Lauria, 2017, reproduced by permission of Oxford University Press—that material appears in a very diluted form primarily in §5.5 and §7.2.

List of Figures and Tables Figures 1.1. Taxonomy of normative beliefs

8

1.2. Desire-­as-­belief illustrated by analogy with disbelief

13

1.3. The two directions of fit of desire

15

3.1. Three views in metaethics

68

4.1. The Müller-­Lyer illusion

92

7.1. A simple model of the influence of emotion

133

7.2. A more complex model of the influence of emotion

134

7.3. How getting what you want might correlate with pleasure

145

7.4. How getting what you want might also cause pleasure

146

8.1. Uncertainty in desire

155

9.1. Conceptual role semantics

172

9.2. The roles of reasons beliefs

177

9.3. The roles of reasons beliefs again

179

9.4. The relations reasons stand in

179

A.1. An illustration of Lewis’ desire revision principle

204

Tables 3.1. The negation problem for conativism

71

8.1. Strength as certainty

152

8.2. Strength as a compound

152

8.3. Strength as importance

154

8.4. Strength as importance, and certainty as certainty about desire

158

Introduction It’s a sunny Wednesday morning, and I want to get lots of work done today—I want this book finished before Christmas. On the other hand, I also want to find time for a proper lunch break so I can do some exercise: I’m putting on weight and I’d prefer to reverse that trend sooner rather than later. Sadly, I put these thoughts to the back of mind as I notice that my youngest son has just been sick on the bed, and I want to get that cleaned up right away. I could easily continue this picture of my mental life by describing other things I want. We could construct a similar list for you, I’m sure. Such wants (desires, preferences) are an absolutely central part of our lives, in at least two ways. First, they are crucial for explaining our actions. For example, if we are trying to understand why you bought a sledgehammer, we will appeal (in part) to what you want. Since people’s actions are explained with reference to what they want, we can understand and predict their choices—what to buy, how to vote, how to parent, what career to pursue, and so on—only if we understand what they want. Second, our wants are central in our lives because they are crucial for determining what it’s rational for us to do. If we are trying to ascertain whether it’s rational for you to buy a sledgehammer, the answer will depend in part on what you want. We evaluate people’s choices with reference to what they want, and so we can critically engage with others only if we understand what they want. For example, your decision to refuse a promotion might strike me as foolish until I learn that you don’t want any more money, and desperately want to avoid further stress. In cases like this, we need to understand what people want if we are to critically evaluate their choices. The idea that we explain and evaluate actions with essential reference to what people want is compelling, and the popularity of decision theory is a testament to that: the field of decision theory aims to explain and evaluate actions with central reference to our wants. We might quibble about the details of decision theory, but the broad project is surely attractive, and is attractive precisely because it aims to systematize the common-­sense ideas that our wants lie at the heart of actual decision-­making and at the heart of good decision-­making. Desire as Belief: A Study of Desire, Motivation, and Rationality. Alex Gregory, Oxford University Press. © Alex Gregory 2021. DOI: 10.1093/oso/9780198848172.003.0001

2 Introduction In short, our wants are important. But all that said, when it comes to explaining and evaluating actions, our wants seem to have a competitor: our beliefs about what we ought to do, and other such normative beliefs. If we hope to understand why Peter gives money to charity, a good explanation will presumably appeal to such beliefs. And if we are trying to decide whether this action is rational, one important issue depends on Peter’s normative beliefs. For example, this act might be irrational if Peter himself believes he ought not do it. In short, it might seem possible to explain and rationally evaluate Peter’s actions without mentioning his wants after all. In short, it is tempting to assign a necessary role to our wants in explaining and evaluating everything we do. But it is also tempting to say that our normative beliefs sometimes suffice to explain or evaluate what we do. How can we resolve this tension? We might distinguish three broad possibilities: First, we might maintain that only our wants play these important roles, and try to explain away the attraction of saying that our normative beliefs can do so (e.g. Arpaly and Schroeder  2014; Haidt  2001; Sinhababu  2017). For example, we might insist that Peter’s philanthropy is really best explained and evaluated by his wants, perhaps including his desire to be moral. Second, we might maintain that our normative beliefs can play these roles, and try to explain away the attraction of saying that only our wants do so (e.g. Nagel  1970; Dancy  2002). For example, we might say that our wants are themselves always determined by our normative beliefs: you might buy a sledgehammer because you want to, but perhaps you want to do so only because you believe you ought to demolish the shed. Third, there is the conciliatory option, on which we explain away any appearance of conflict between our wants and our normative beliefs in explaining and evaluating actions. In this book, I pursue one such conciliatory strategy. I simply identify our wants with our normative beliefs: on my view, these are two different labels for a single thing.1 In my preferred terminology, I defend desire-­as-­belief, according to which our desires just are normative beliefs. According to this view, our actions are always explained by our wants, but this is consistent with our actions being explained by our normative beliefs, because our wants are our normative beliefs. And according to this view, a rational action is one that appropriately reflects what you want, but this is consistent with the rational importance of our normative beliefs,

1 In this respect my view is somewhat similar to non-­ cognitivism (e.g. Blackburn  1998; Gibbard 1990, 2003). But the two views are very different in numerous other ways, as will be obvious in what follows. I more directly contrast the two views in §3.3.

Introduction  3 because our wants are our normative beliefs. Desire-­as-­belief allows us to accept common sense about desires, and accept common sense about our normative beliefs. Why choose sides when everyone can win? Above I presented desire-­as-­belief as a conciliatory theory. But in a different respect, it is highly partisan. One long-­standing dispute in ethics is that between subjectivists and objectivists.2 Subjectivists believe that subjects project value onto the world, through their desires. On this view, our desires are the source of all value just as the sun is the source of all light. In contrast, objectivists believe that value is an objective part of the world, different in kind from other features of the world, but nonetheless equally real and mind-­ independent. Desire-­ as-­ belief finds a place for desire in an objectivist framework: objectivists say that value is out there already, and desire-­as-­belief adds that our desires aim to reflect those pre-­existing values. That is, on this view, desires are akin not to the sun, but instead to the moon, since they merely reflect pre-­ existing light. Desire-­ as-­ belief fits into the objectivist tradition, because it treats desires as beliefs about an objective normative landscape. Hume memorably writes, ‘’Tis not contrary to reason to prefer the destruction of the whole world to the scratching of my finger’ (T2.3.3.6). Many endorse similar claims, according to which in some fundamental sense our desires cannot be incorrect: they are just impulses you have, which might be outweighed by other desires, but can never be discounted as mistaken. But desire-­as-­belief takes the opposing line. If desires are beliefs, then desires can be incorrect just as beliefs can: they can fail to represent the normative landscape correctly. This, it seems to me, fits with common sense, according to which it is foolish to prefer the destruction of the whole world to the scratching of your finger, and it is foolish because it misrepresents the relative worth of these things. In short, this book shows how our wants and normative beliefs explain our actions, and how they can rationally justify our actions. I claim that all human actions can be explained by appeal to our wants, which is to say our normative beliefs. I also show how the rationality of our actions is explained by our wants, which is to say our normative beliefs. I explore these issues because 2  Subjectivism, or a nearby view, is endorsed by Hobbes (L I.VI.7); Goldman (2009); Schroeder (2007a); Sobel (2016). Objectivism, or a nearby view, is endorsed by Parfit (2011a); Raz (2000); Scanlon (1998,  2014); Shafer-­Landau (2003), and plausibly Plato (Euthyphro 9e–11b) and Aristotle (Metaphysics 1072a29). There is also the error theory, according to which values are supposed to be objective but in fact don’t exist at all (Olson  2014; Streumer  2017). Such a view is consistent with desire-­as-­belief—in combination they tell us that all our desires are in error. But I set this radical view aside.

4 Introduction they are important for making decisions, for understanding the actions of those around us, and for criticizing or excusing others for what they do. But I also explore these issues in service of a broader objectivist view about ethics. * Here is an overview of the book. The first chapter serves to put desire-­as-­belief on the table so we have the theory clearly in view for the remainder of the book. Along the way I say a little about the direction of fit of belief and desire, and about the distinction between reason and passion. The next chunk of the book appears across Chapters 2 to 4, where I present two arguments for desire-­as-­belief. Chapters 2 and 3 combine to produce the first argument: Chapter 2 defends the view that all motivation is explained by desire. Chapter 3 defends the view that some normative beliefs can motivate, and then combines these two claims to argue for desire-­as-­belief. Chapter 4 presents the second main argument for desire-­as-­belief: desire-­as-­belief is best placed to explain why desires make a difference to what it’s rational to do. Many object to desire-­as-­belief by claiming that it has dubious implications for various topics. So Chapters 5 to 9 of the book aim to show how desire-­as-­ belief can escape those objections. But these chapters also have more positive elements, since I often say that desire-­as-­belief is not only defensible but also positively attractive for the light it casts on the relevant topics. For this reason, these chapters are structured by theme rather than by (say) a count of different objections. Chapter 5 discusses irrationality, especially akrasia, and shows how desire-­ as-­belief is not only consistent with such irrationality but also positively predicts it. One key claim is recycled from earlier chapters: not all desires succeed in motivating us. Chapter 6 discusses the general idea that our desires are constrained by our normative beliefs, and shows how desire-­as-­belief, as I formulate it, can permit that our desires can float free of our value beliefs. Chapter 7 turns to the relationship between desires and various feelings, and I show how desire-­as-­belief illuminates such connections. Chapter 8 discusses two further topics: degrees of desire, and reasoning with desire. I show how desire-­as-­belief makes sense of these things. Chapter 9 considers the mental lives of animals and whether they cast doubt on desire-­as-­belief. But to get there, the majority of the chapter sketches a theory of what it takes to have the concept of a reason: I hope that discussion is illuminating independently of the primary use it serves.

Introduction  5 Finally, with everything else settled, Chapter  10 returns to the broader choice between objectivism and subjectivism, above. I rehearse some classic arguments against subjectivism, and also present some debunking explanations of its appeal—I point to some ways in which desires are normatively relevant even on an objectivist view. Chapter 11 briefly concludes. There is also an appendix on David Lewis’ famous objection to desire-­as-­ belief (1988, 1996). For some readers this appendix might seem crucial, since Lewis’ objection is well known and somewhat influential in some circles. But I have nonetheless placed it in an appendix, since the relevant issues are somewhat technical, and I don’t think there are any interesting wider lessons to be learnt from my discussion: Lewis is wrong for reasons that add little to the rest of the book. So I present it as optional reading. If you do read it, do read §8.1 first, which the appendix draws upon. On a similar note, §2.3.1, §5.7, and §8.2.3—all titled ‘An Aside’—can be skipped if you prefer.

1 What is Desire-­as-­Belief? 1.1  Desire and Belief Let’s start by contrasting our wants with our preferences. When you have a preference, you prefer one thing over another. But often we talk more simply about what people want, without comparing that thing with some alternative. I will almost always do the same. Many of my claims are much simpler to state when expressed in terms of what people want, and how strongly, rather than in terms of people’s preferences between alternatives. By stating my claims more simply, I hope to make them easier to evaluate. What I say could later be extended, in a natural manner, to make sense of preference. In fact, this emphasis might reflect the underlying reality. Preferences seem to be more complex attitudes than wants, because they involve comparing two things, rather than assessing only one alone. As a result, it is tempting to think of preferences as complex states, composed of wants: your preference for one thing over another is explained by your wanting the former more than you want the latter (for an argument, see Pollock 2006, 22–7). If our wants are more basic than our preferences in this way, it makes good sense to investigate wanting first, and then later use that theory to develop a corresponding theory of preference. This might provide a further reason for focusing on wants, since preferences are plausibly composed of such wants. Talk of what we ‘want’ is somewhat ambiguous: sometimes we mean to refer to what someone wants most, and sometimes we mean to refer to what they want to some extent. The former is implicitly comparative, like preferences are. Unless I say otherwise, I will always have the latter sense in mind, the sense in which it is clearly true that all of us have many conflicting wants. If I mean to talk about what someone most wants, I will explicitly say so. I will sometimes refer to our wants using the word ‘desire’. The word ‘desire’ in English is often associated with particularly strong desires, and sometimes associated more narrowly still with sexual desires. But the notion I have in mind need not have these associations: in my sense, it is true that I desire to get a good night’s sleep tonight. To this extent, the word ‘want’ is probably a better fit than ‘desire’ as a label for the state of mind I focus on. But the word Desire as Belief: A Study of Desire, Motivation, and Rationality. Alex Gregory, Oxford University Press. © Alex Gregory 2021. DOI: 10.1093/oso/9780198848172.003.0002

Desire and Belief  7 ‘desire’ is useful since it can more naturally be used as a noun, and so I will make use of both ‘want’ and ‘desire’, meaning the same by each. Another important mental state is belief. To believe p is to represent p as true. Again, in ordinary English the word ‘belief ’ has distracting associations: it is often used to refer solely to our religious commitments. But as is now standard in philosophy, I will use the word more broadly: we have beliefs about the population of China, about the best theory of gravity, about whether it will rain later today, and so on. In everyday English we use phrases like ‘he thinks’ and ‘in her opinion’ to talk about beliefs, but the word ‘belief ’ is less ambiguous than words like ‘think’ and ‘opinion’, and is anyway firmly entrenched in the philosophical literature. States of mind like beliefs and desires are attitudes, and those attitudes have contents. The content of a belief is the thing you believe (e.g. that the population of China is large), and the content of a desire is the thing you desire (e.g. that you buy a sledgehammer). For clarity, I will often place the contents of our attitudes in square brackets, so that you might believe [that the population of China is large] and desire [that you buy a sledgehammer]. I normally treat the contents of beliefs and desires as propositions: things that are expressed in English by ‘that’ clauses, and which can be grammatically preceded by ‘it is true’ or ‘it is false’. But I don’t think very much hangs on this assumption, other than ease of presentation.1 (A little more on this shortly.) Our beliefs can have many different contents, about just any subject matter. But an especially important subset of our beliefs is the set of beliefs with normative contents, such as beliefs about what we ought to do, or about what is good. I’ll refer to beliefs with normative contents as ‘normative beliefs’. You have very many normative beliefs. Perhaps some of those are moral beliefs, as when you believe you ought to keep your promise to Ahmed, believe that no-­ one should be cruel, believe that it’s good to be a vegetarian, or believe that it’s bad to be at war. But you also have large numbers of nonmoral normative beliefs, as when you believe that you ought to keep doing exercise, that no-­ one should wear a bowtie, that pie for dinner would be good, or that it’s bad to have no pension. Understood in this extramoral way, we have very many normative beliefs indeed. 1  For example, I rely on no assumptions about the metaphysical nature of propositions. A distinct worry is that we have some desires for objects, not propositions, such as when you desire chocolate. I agree with those who think that there are no such desires, and that sentences that appear to ascribe such desires are really elliptical claims about propositional desires (Sinhababu 2015; for opposition, see Brewer 2006; Thagard 2006). But note that this is consistent with thinking that many other attitudes are not propositional, including attitudes such as likings (Grzankowski 2015; Montague 2007). For the contrast between desiring and liking, see §7.3.

8  What Is Desire-as-Belief?

1.2  Reasons and Reasons Beliefs Our normative beliefs fall into various subcategories. For example, there are beliefs about what you ought to do, beliefs about what you have reason to do, beliefs about things being good overall, and beliefs about things being good in certain respects. See Figure 1.1 for a simple taxonomy. Most of my discussion will focus on beliefs about just one of the nodes in Figure 1.1: beliefs about reasons. For example, if you are debating whether to opt for surgery, you might well form beliefs about the reasons for, or against, surgery. Or you might believe you have reason to do more exercise, to help your sister, or to drink some tea. I will very often refer to such beliefs, and I will refer to them as ‘reasons beliefs’. In fact, by ‘reasons beliefs’, I have something relatively narrow in mind. You might believe [that Sarah has a reason to help you], believe [that you don’t have reason to jump through the window], or believe [that if dragon fruit is tasty then you have reason to eat it]. Though these are beliefs about reasons, I won’t include them as ‘reasons beliefs’, in my sense. In my sense, ‘reasons beliefs’ are atomic beliefs about single reasons you yourself have: beliefs with the content [I have reason to v].2 This shouldn’t be too confusing: such reasons beliefs are the canonical kind. And don’t worry: I’ll remind you of this restriction at the most crucial points. What, exactly, are reasons? Two clarificatory remarks are crucial. First, by ‘reason’, I mean normative reason. Normative reasons are facts that count in favour of actions, or (equivalently) that contribute towards justifying actions. I will never use the word ‘reason’ to refer to motivating reasons (or any other purely explanatory reasons). Motivating reasons are the reasons Ought Directive Reason Normative Good overall Evaluative Good in a respect

Figure 1.1  Taxonomy of normative beliefs. 2  To be clear, ‘v’ here might be logically complex, as when you believe [that you have a reason to (either A or B)]. But any kind of wider logical complexity ensures that the belief in question is not a ‘reasons belief ’, in my sense.

Reasons and Reasons Beliefs  9 why people act, and merely explain, rather than contribute to justifying, their actions. For example, we might describe Hitler’s reasons for invading Poland: such motivating reasons explain what he did rather than justify it. In contrast, normative reasons justify actions, or show that they are appropriate. We will talk a lot about motivation, but since other phrases are available to refer to motivating reasons, I’ll keep things clearer by reserving the word ‘reason’ for normative reasons only. Similarly, by ‘reasons belief ’, I mean beliefs about normative reasons, so that your beliefs above amount to beliefs that there is something to be said for doing exercise, helping your sister, or drinking tea. Second, if I leave it unqualified, by ‘reason’ I will always mean ‘reason to act’, where an act (action) is something you intentionally do, such as cutting your own hair, or buying a toupee. Really, normative reasons can favour attitudes as well as actions—you might have some reason to believe that the butler did it, or to feel sad. But since I focus on reasons for action, and say little about reasons for attitudes, I will just use ‘reason’ to mean ‘reason to act’. Similarly, by ‘reasons belief ’, I mean only beliefs about normative reasons for actions, never beliefs about reasons for attitudes. In short, by ‘reason’ I mean the things that count in favour of certain actions—things that go in the ‘pros’ column for an act. And when I talk about ‘reasons beliefs’, I mean our beliefs about such things. Can we say anything more about such normative reasons for action? Two broad points are worth mentioning. First, as I said in the Introduction, desire-­as-­belief fits most neatly into the objectivist tradition according to which normative reasons are relatively independent of your own state of mind.3 On this view, you might have various reasons to have surgery, donate money to charity, or keep your promises, and those reasons are independent of what you think or feel on these issues. Since I assume this kind of objectivist view, I assume that there are real facts of the matter about whether you have certain reasons or not, and your reasons beliefs aim to track objective facts about those reasons, just as your beliefs about planets aim to track certain objective facts about those planets. But though I assume some kind of objectivist view, I will stay silent on exactly which kind of objectivist view we should endorse.4 There are many

3  I say ‘relatively’ independent because of course some reasons can depend on your own state of mind: if you are in pain, that might be a reason to take a painkiller. But objectivists think that these are special cases and certainly that not all reasons are like this. 4  In my (2016), I defended some claims about reasons that are in some ways awkward partners for my claims about desire in this book. I think the two views could be reconciled, but I won’t undertake that task here, and I shall proceed to ignore my claims from that paper.

10  What Is Desire-as-Belief? kinds of objectivism, and desire-­as-­belief should be compatible with the vast majority of those. For example, it is compatible with pluralist views like Ross’, as well as more monistic views like utilitarianism. It is also consistent with metaethically naturalist views as well as metaethically non-­naturalist views. Desire-­as-­belief tells us that desires are beliefs about reasons, and these different objectivist theories will disagree about which such beliefs are true, and what makes them true. These seem like relatively independent issues. Just about any view that makes the truth of these beliefs independent of our desires is likely to fit perfectly well with desire-­as-­belief. For that reason, I won’t express my own views about the respective merits of different objectivist theories, since almost all are compatible with desire-­as-­belief. Whatever your preferred objectivist theory, you can combine it with desire-­as-­belief. (One notable exception are objectivist theories that incorporate desire-­ based theories of wellbeing—I discuss those in §10.3.) But second, I will make one small assumption: I shall assume that reasons connect in a systematic way with what you ought to do. More specifically, I assume that you ought to do something just when you have most reason to do it. The idea is that our reasons weigh against one another, and the balance of reasons determines what you ought to do overall. As a result, at some points I shall make claims about what you ought to do, trusting you to understand that these relate in an obvious and systematic way to what you have reason to do. I hope this assumption seems as natural to you as it does to me. To summarize: I will use the label ‘reason’ to refer to normative reasons for action. In turn, I will use the label ‘reasons beliefs’ to refer to our beliefs about such normative reasons for action—beliefs about justifications for various acts. Most crucially, you should never get misled by the alternative use of ‘reason’ where it means a motivating reason: a mere explanation of what moved someone to act. I will talk about such things, but never using the word ‘reason’.

1.3  Desire-­as-­Belief In this book, I will argue for Desire-­As-­Belief.5 This view identifies our desires with some of our normative beliefs, and more specifically with our reasons beliefs:

5  Similar views can be found in Campbell (2018), McNaughton (1988), and Little (1997).

Desire-­a s-­B elief 

11

Desire-­As-­Belief:  To desire [to v] just is to believe [that you have ­reason to v]. Desire-­as-­belief says ‘desire’ is really just a label that refers to a certain subset of our beliefs. It says that when a person desires to do something, that is just the very same thing as their believing that they have reason to do it. For example, desire-­as-­belief says that to desire to start work early is to believe that you have reason to do so (perhaps you work best then). Vice versa, desire-­as-­belief says that every reasons belief qualifies as a desire, so that if you believe you have reason to buy some wellies, you thereby desire to buy some. As stated above, desire-­as-­belief made a claim only about desires [to __]. But it seems that some of our desires are desires [that __], rather than desires [to __], as when I desire [that Spurs win]. What should defenders of desire-­ as-­belief say about such propositional desires? One option here is to insist that desire-­as-­belief is a theory only about desires [to __], and to treat propositional ‘desires’ as distinct states—perhaps as hopes, or wishes, rather than desires (cf. §6.5). But a second more ecumenical option is to instead extend desire-­as-­belief and make a further related claim about desires [that __]: Desire-­As-­Belief:  To desire [that p] just is to believe [that you have reason to bring about p]. We could separately accept both this analysis of desiring [that __] and the above analysis of desiring [to __]. Or, more neatly, we could simply treat the original view as a mere consequence of this second one.6 We could treat the desire [to v] as the desire [that you v], and could treat the belief [that you have reason to v] as the belief [that you have reason to bring it about that you v]. If that were right, the initial analysis of desiring [to __] is really just a special case of the above analysis of desiring [that __]. In what follows, I proceed on those assumptions. That is, I treat desire-­as-­ belief as a claim about desiring [that __], and assume that claims about desiring [to __] are just shorthand for related claims about desiring [that __]. This way of proceeding is certainly cleanest, since it allows me to largely stay with the orthodoxy and treat desires as propositional attitudes, and because it allows me to deploy examples without much care as to whether they involve desires [to __] or instead desires [that __]: again, I assume that desire-­as-­belief 6  For potentially relevant discussion, see Schroeder (2011).

12  What Is Desire-as-Belief? covers both and in fact that the former are just a special case of the latter. But I don’t believe I rely on this assumption at any point, and it really serves only to keep my presentation of the issues nice and clean. A related thing I should emphasize is that we might distinguish a more general formulation of desire-­as-­belief which merely identifies desires with normative beliefs of some kind, and my more specific version of desire-­as-­ belief which identifies desires with reasons beliefs in particular.7 The choices here are interesting—I discuss related matters in Chapters 3 and 6, where I argue for my more specific formulation of desire-­as-­belief. Much of what I say would remain relevant if we instead pursued some other nearby view. Still, for ease, by ‘desire-­as-­belief ’ I will always mean my specific version of that view. To help us understand desire-­as-­belief, let’s turn our attention to another state of mind: disbelief (cf. Price  1989, 120–1). As I said above, we believe many things. Some of these things involve negation (¬). For example, you might believe [that it is not Christmas today]. Indeed, it seems that we have many beliefs in negations: you presumably believe [that it’s not the case that grass is tasty], [that it’s not the case that sheep wear top hats], [that it’s not the case that 2 is larger than 10], and so on. Given how common beliefs with negated contents are, it might in some contexts be useful to use the word ‘disbelief ’ to refer to such states of mind. Disbelief is not some new state of mind over and above belief: it is just a belief with a negated content. With this word in place, we can talk about the things you disbelieve: [that grass is tasty], [that sheep wear top hats], [that 2 is larger than 10], and so on. When we talk about these disbeliefs of yours, that is just another way of talking about the above beliefs of yours. In this way, we might put the word ‘disbelieve’ to good use, where ‘disbelieving [that p]’ is just shorthand for ‘believing [that ¬p]’. The introduction of the word ‘disbelief ’ allows us to describe beliefs with negated contents in a more concise manner, where we move the negation out of the content of a belief and into our description of the attitude itself. Of course, we are not actually moving anything around: this is just a convenient way of talking. When you use the label ‘disbelief ’, you aren’t committing yourself to the existence of a new state of mind beyond belief. Rather, you are just using a convenient label that lets you talk about some of our beliefs slightly more concisely. We can now get a better understanding of desire-­as-­belief. Remember, it says: 7  A slightly different option is to identify desires not with beliefs, but with perceptual states, or similar. I discuss and reject such views in §4.4.

Desire-­a s-­B elief 

13

Desire-­As-­Belief:  To desire [that p] just is to believe [that you have reason to bring about p]. Desire-­as-­belief says that ‘desire’ functions exactly like ‘disbelief ’. Desire-­ as-­belief says that to ‘desire [that p]’ is just to believe [that you have reason to bring about p]. Just as the word ‘disbelief ’ allows us to move negation out of the content of a belief and into our description of the attitude, the word ‘desire’ allows us to move reason-­to-­bring-­about out of the content of a belief and into our description of the attitude. And just as the word ‘disbelief ’ might allow us to talk more concisely about beliefs with negated contents, by using this label ‘desire’, we might talk more concisely about our reasons beliefs. Desire-­as-­belief literally identifies desires and beliefs with a particular normative content, in just the same way that we identified disbeliefs and beliefs with a particular negated content. It says that when we talk about desires, this is just a particular way of talking about a particular subset of our beliefs. That is the view I will defend.8 For a simple illustration, see Figure 1.2. It’s not the case that

Belief

P

... is the same as ... Disbelief

Belief

I have reason to bring about

P

P

... is the same as ... Desire

P

Figure 1.2  Desire-­as-­belief illustrated by analogy with disbelief.

8  Are there other words like ‘disbelief ’ and ‘desire’? If so, they are probably words for attitudes with contents governed by monadic propositional operators, where attitudes with such contents are sufficiently common that greater parsimony of expression is helpful. There are some candidates: perhaps someone doubts [p] just when they believe [probably not p], or someone expects [p] just when they believe [p will happen in the future] (Searle  1983, 31). Another possibility is that ­someone suspends judgement whether p when they believe [I have insufficient reason to believe p or its negation] (Raleigh Forthcoming; see also Friedman  2013). On this broad topic, see also Campbell (2018).

14  What Is Desire-as-Belief?

1.4  Direction of Fit With desire-­as-­belief clarified in this way, this is a good place to address the direction-­of-­fit metaphor (Humberstone  1992; Gregory  2012; Smith  1994, 111–16; Platts 1997, 256–7). The idea is that desires and beliefs have different ‘directions of fit’. Whilst in each case something goes right if the content of the attitude matches up to the world, the thought is that the direction of this fit travels in opposing directions: whereas beliefs aim to fit the world, desires aim to have the world fit them. That is, we try to revise our beliefs to bring them into line with the way the world is, whereas we try to bring the world into line with how we want it to be. (If you like, think of beliefs as soft wax that press against the world and have the world imprint the truth upon them, and desires as stamps that press against the world and imprint their contents upon it.) This metaphor seems to suggest that when we think about beliefs and desires, we are thinking about states of mind that are in some fundamental sense the opposite of one another. That might seem to count against desire-­ as-­belief, which identifies desires with a particular subset of our beliefs (Smith 1994, 116–25).9 But with desire-­as-­belief clarified via the comparison with disbelief, we can see why this line of thought is mistaken. According to desire-­as-­belief, we can describe desires in two different ways, and depending on how we describe them, we describe their content in a different way.10 In turn, when we talk about the direction of fit of a mental state like desire that can be described in more than one way, we need to be careful about which content the relevant direction of fit is being ascribed to. Desire-­as-­belief says that to desire [that p] is to believe [that you have reason to bring about p]. The former content— [that p]—is plausibly one that you are supposed to impose on the world. The latter content—[that you have reason to bring about p]—is plausibly one that 9  Smith presents the problem as a problem for the ‘besire’ theory, rather than desire-­as-­belief, though he surely thinks it would undermine both. I discuss the besire theory in §6.1 and reject it for reasons that are independent of the present issue. 10  This claim also has some bearing on views according to which desires are not states with normative contents, but instead have non-­normative contents that they represent with normative force (Schafer  2013; Tenenbaum  2008). These views are sometimes pitched as alternatives to a view like desire-­as-­belief. But I agree with these views insofar as reasons rarely feature in the contents of our desires: I rarely desire [that I have reason to v]. This fact is consistent with desire-­as-­belief: according to desire-­as-­belief, that desire would be an unusual belief—one about a higher-­order reason: a reason to bring it about that I have reason to v. Those defending these views also tend to say that desires relate to the good in the same way that beliefs relate to the true. Let’s set aside the fact that I identify desires with beliefs about reasons rather than beliefs about goodness. Even so, this clam isn’t right. Since claims about goodness are themselves truth-­apt, it would be more accurate to say that desires relate to the truly good as beliefs relate to the true. But so understood, this claim fits perfectly well with the claim that desires are a subset of our beliefs. In these ways, I think these views are either consistent with desire-­as-­belief or else implausible.

Direction of Fit  15 Belief

I have reason to bring about

P

Should fit world P

‘Desire’ World should fit

Figure 1.3  The two directions of fit of desire.

you are supposed to make fit the world as it already is. This state of mind has two directions of fit, each with respect to a different content: it is both supposed to bring about p, and also supposed to be responsive to whether it’s true that you have reason to bring about p (Little  1997, 63–4; Price  1989, 120–1).11 See Figure 1.3. Understanding this may be easier if we step away from the metaphor. Really, the idea behind the direction-­of-­fit metaphor is that whereas beliefs are, or should be, responsive to evidence, desires do, or should, influence us to act in ways that are productive of their content. Desire-­as-­belief says that both of these things are true of desires, which is to say, both of these things are true of reasons beliefs. Clearly, your beliefs about what you have reason to do should be sensitive to evidence about whether you really have reason to do those things. But such beliefs should also rationally influence your actions in appropriate ways. And—I here repeat the previous claims in different words— your desires should be sensitive to what you have reason to do, and ought to rationally influence your actions in appropriate ways. Desire-­as-­belief is perfectly consistent with the distinction between directions of fit: it merely says that this state of mind has both directions of fit at once, each with respect to a different content. It might also help if we return briefly to our comparison with disbelief. Imagine someone reasoning as follows: ‘Disbelief cannot be a belief of any kind, because those states of mind have opposing directions of commitment. Whereas the belief [that p] commits

11  Some might deny that that one state of mind could have two contents. Perhaps it would be better to say that this state of mind has just one content—that [I have reason to bring about p]—but also has [p] as a part of its content, and the desire-­like direction of fit applies to a content-­part rather than a content. I am not so clear about this—too much hinges on how we think of mental contents—and at any rate the analogy with disbelief ensures that there must be some appropriate way to understand states of mind which can be described in different ways and which seem to get ascribed different contents depending on how they are described.

16  What Is Desire-as-Belief? one to [p], the disbelief [that p] commits one to [¬p]. Beliefs commit one in favour of their contents, whereas disbeliefs commit one against their contents. So disbelieving cannot be a kind of believing’.

This is bad reasoning. The very idea is that disbelieving [p] just is believing [¬p], and as such disbelieving [p] both commits you in favour of [¬p] and commits you against [p]. Similarly, according to desire-­as-­belief, desiring [that p] just is believing [that you have reason to bring about p], and as such the desire [that p] both aims to fit whether [you have reason to bring about p] and aims to have the world fit [p]. A final point: desire-­as-­belief is consistent with the idea that describing a state of mind as a belief, or as a desire, makes one of its directions of fit more salient. For example, when we describe a state of mind as a belief [that you have reason to bring about p], by referring to it as a ‘belief ’ we might thereby emphasize the way in which this state should be responsive to the truth about whether you in fact have reason to bring about p. Vice versa, when we describe this very same state of mind as a desire [that p], by referring to it as a ‘desire’ we might thereby emphasize the motivational role it plays in affecting your actions. So it is possible that by describing states of mind as ‘beliefs’ or as ‘desires’, we thereby emphasize different features they have, and in turn we might emphasize one direction of fit by describing the state in a certain way. To this extent, one direction of fit is more closely associated with beliefs, and another is more closely associated with desires. But this is all consistent with desire-­as-­belief, which says that ultimately, there is just one state of mind here with both directions of fit. Our ability to draw attention to one direction of fit at the expense of the other is consistent with this state of mind ultimately always having both. I conclude that the direction-­of-­fit metaphor is perfectly fine, but does absolutely nothing to undermine desire-­as-­belief. Before we move on, I can briefly respond to another simple objection to desire-­as-­belief. Whereas we refer to beliefs as ‘true’ or ‘false’, we never attach these labels to desires. Isn’t that a simple but effective argument against desire-­ as-­belief? It’s again helpful to start by thinking about disbelief. To the extent that we might successfully make use of the word ‘disbelief ’, we would be unlikely to describe disbeliefs as ‘true’ or ‘false’. Why? Because it would be confusing: if someone described the disbelief [that pigs fly] as ‘true’, it would be unclear whether they meant that the content of the disbelief was true or that the content of the belief was true: whether they meant that pigs do fly or  that they don’t. To this extent, it would be far clearer to communicate

Some Broad Attractions of Desire-­a s-­B elief  17 the relevant facts by describing the state of disbelief as ‘appropriate’ or ‘inappropriate’, or else by describing the relevant belief as ‘true’, ‘false’, ‘appropriate’, or ‘inappropriate’. Given that we have these helpful ways of communicating the relevant facts, it would be bizarre to try to communicate those facts with the highly unclear assertion that the relevant disbelief is ‘true’.12 Similar reasoning applies to desires, given desire-­ as-­ belief. You could accurately describe desires as ‘true’ or ‘false’. But it would be extremely confusing. If someone called the desire [that pigs fly] ‘true’, it would be unclear whether they meant that the content of the desire was true, or that the content of the belief was true: whether they meant that pigs do fly, or instead that there is a reason to make them fly (i.e. whether they meant that the desire is satisfied, or that it is appropriate). As a result, it’s clearest to make the intended claim by saying that the relevant desire is ‘appropriate’ or ‘inappropriate’, or else by saying that the relevant belief is ‘true’, ‘false’, ‘appropriate’, or ‘inappropriate’. To this extent, desire-­as-­belief positively predicts that we don’t refer to desires as ‘true’ or ‘false’. Doing so would be unnecessarily confusing, and that is a good explanation of why we do not talk in that way.

1.5  Some Broad Attractions of Desire-­as-­Belief In the chapters ahead I’ll present some detailed arguments for desire-­as-­ belief. But before we get to those details, we can here note two more simple attractions of the view. The simplest attraction of desire-­ as-­ belief is that the view is highly parsimonious, reducing desires to beliefs. According to other views, we have two distinct and important states of mind: beliefs and desires. But according to desire-­as-­belief, we have just one important state of mind: belief. On this view, talk of ‘desire’ is just another way to talk about some of these beliefs. I take it that theories are more plausible if they can explain the data while positing the existence of fewer kinds of entity. So if desire-­as-­belief can explain the data, it is a more plausible theory than many rivals. This is clearly an attraction of the view. True, a lot hinges on whether desire-­as-­belief can explain the data, but as we might put it, desire-­as-­belief is a good starting hypothesis that we should abandon only if it can’t explain the data; other more positive arguments for the view might be superfluous. 12  Similar reasoning applies to ‘deny’ and ‘reject’: denials and rejections are truth-­apt, and yet we don’t call them ‘true’ or ‘false’. This is surely just because it would be deeply confusing to talk that way.

18  What Is Desire-as-Belief? A second simple attraction of desire-­as-­belief comes from noting how much there is in common between desiring something and believing there’s a reason to bring it about (we will investigate many of these in greater depth in the following chapters). Plausibly, reasons beliefs and desires are both capable of motivating you to do things. Reasons beliefs and desires both play important roles in practical deliberation. Reasons beliefs and desires both seem capable of rendering your actions (ir)rational. Reasons beliefs and desires both come in degrees: you can think you have a weak reason, or a strong reason, and you can want things a little or a lot. You can believe that you have conflicting reasons, and so too you can have conflicting desires. Reasons can be believed to be instrumental, when they favour a means to an end, or ultimate, when they favour an end for its own sake, and this same distinction holds for our desires. Ultimate reasons beliefs and ultimate desires seem relatively stable over time, whereas instrumental reasons beliefs and instrumental desires do not. Reasons beliefs and desires come in both non-­ comparative and comparative forms: just as you can believe [that you have reason to bring about p] and desire [that p], you can believe [that you have more reason to bring about p than q], and you can prefer [p to q]. Reasons beliefs and desires can both have their demands met: you can comply with the reason you believe you have, or satisfy the desire. Reasons beliefs and desires are both evaluated in a manner that is agent-­relative: some reasons beliefs and desires might be more appropriate for me than you, or vice versa. This list could go on, but the basic and simple point should be clear: it looks as though desires and reasons beliefs are made for each other, in that they have many properties in common. This isomorphism needs explaining, and one obvious explanation of it is that the two are in fact one and the same. To this extent, desire-­as-­belief seems in a good position to explain some obvious facts about reasons beliefs and desires.

1.6  Some Broad Defences of Desire-­as-­Belief Some might think that desire-­as-­belief is sufficiently implausible that we should just reject it out of hand. Again, in the following chapters, I defend desire-­as-­belief in detail. But again, before we get to those details, we can note some general reasons for optimism about desire-­ as-­ belief. David Lewis claimed that incredulous stares can’t be answered (1986a, 133), but I hope they might nonetheless be prevented by casting a view in a different light. We can then move onto more articulate arguments.

Some Broad Defences of Desire-­a s-­B elief  19 Sometimes, when we talk about our desires, what we say seems inconsistent. For example, as you reluctantly drag yourself out of the house, filled with dread, and head towards the dentist, it would be odd to insist that you want to go to the dentist. But at the same time, when the bus driver asks where you’re heading, you might quite truthfully say that you want to go to the dentist. So which is it: do you want to go, or not? Our thoughts here might seem inconsistent. In a case like this, I think we should try to find resources that explain the apparent variation in the kinds of claims we make about desire. Rather than treating such variation as demonstrating inconsistency in our thoughts, I shall try to find theories that predict the relevant variation. More specifically, I will make sense of what we say about desire in part by showing how some of what we say about our desires is misleading. We might have thought that a theory of desire is most plausible if it vindicates the truth of everything we say about our desires. But, in fact, we need only vindicate the truth of those claims that ought to be read at face value. We might then argue that the other claims that we tend to make are misleading and ought not be read at face value. By aiming to vindicate only some claims about desire, we make a difference to the range of phenomena that a theory of desire ought to explain. For example, in the case above I think you really do desire to go to the dentist, and that we can explain away our tendency to say that you do not want to go (§2.4.1). If this is right, we can maintain that although our theory of desire must make sense of your desire to go to the dentist, it need not vindicate the appearance that you don’t desire to go. Instead, we explain that latter appearance away by appeal to independent linguistic theories that explain why we make such misleading claims. With the above point in hand, when I say that desires are reasons beliefs, the plausibility of that claim might depend on our prior decisions about which ordinary language claims about desire we take seriously and try to vindicate, and which we instead treat as misleading. Part of my job in what follows is to show that we should understand ordinary desire-­talk in a way that makes desire-­as-­belief more plausible. For example, hunger is not a belief of any sort, but this doesn’t threaten desire-­as-­belief so long as I can defend the claim that hunger is not really a desire, whatever some ordinary language might suggest (§7.2). Or, for another example, I agree that we are sometimes weak-­willed, and fail to be motivated to do things that we think we ought to do. But again, I don’t think this threatens desire-­as-­belief, so long as I can defend the claim that motivation is distinct from desire, whatever some ordinary language might suggest (§2.3, §5.5). As Austin said, in philosophy

20  What Is Desire-as-Belief? there’s the bit where you say it, and the bit where you take it back (1965, 2): though desire-­as-­belief is substantial and interesting, it nonetheless survives some kinds of criticism in virtue of having more modest implications than its objectors assume it must have. So we should not reject desire-­as-­belief until we have gotten clear about exactly what it does commit us to. I used to think that philosophers should be interested only in the world itself, and not interested in contingent facts about the arbitrary language we use to describe it. But Williamson (2008, 284–5) rightly points out that this is like an astronomer insisting that astronomers ought to be interested in the stars and not their telescopes: though it is true, their understanding of the stars will improve if they come to understand how their telescopes might distort their vision. In this book I defend desire-­as-­belief in part by applying the parallel lesson to our theorizing about desires and desire-­talk.

1.7  Reason and Passion Many of the details of desire-­as-­belief are new. But in many ways, desire-­as-­ belief is an old view. The underlying picture on which our desires are entwined with our normative or ethical beliefs is hinted at, or outright endorsed, by many of those who form the canon of at least Western philosophy. So the view is not an aberrant flight of fancy, but instead a development of a long-­ standing tradition of thinking of our desires and ethical views as deeply connected. The oldest and most well-­known figures in this tradition are probably the Stoics, who apparently claimed that ‘impulses are acts of assent’ (Long and Sedley 1987, sec. 33I1; see also 53R, 53S, 60F, 65A4, 65G1, 65I4). But many other figures also seem to endorse some view in this vicinity. For example, Aristotle wrote that ‘The object of desire always moves, but this is either the good or the apparent good’ (DA 433a27–9; see also EN III.3 1113a23–4, EE VII.2 1235b25–7), and Aquinas wrote that ‘there is appetite only for a good which is proposed to it by a cognitive power’ (DV 24.2; see also ST I–II.1.1). Some of the rationalists may have also held views of this sort: Spinoza wrote that ‘The will and the intellect are one and the same’ (Ethics II P49 Cor.; for some discussion see Youpa  2007), and Leibniz claimed that ‘volition is the effort or endeavour (conatus) to move towards what one finds good and away from what one finds bad, the endeavour arising immediately out of one’s awareness of these things’ (1982, 172; see also 1989b, sec. 13 and 1989a, 279). More recent influential authors endorse something in the vicinity of

Reason and Passion  21 desire-­as-­belief as well. For example, Davidson writes, ‘If an agent judges that it would be better to do x than to do y, then he wants to do x more than he wants to do y’ (2001b, 23), Murdoch writes that ‘Will and reason then are not entirely separate faculties in the moral agent’ (2013, 39), and Scanlon writes that ‘desire, in order to play the explanatory and justificatory roles commonly assigned to it, needs to be understood in terms of the idea of taking something to be a reason’ (1998, 7–8). Still, not everyone has been happy to subscribe to a view of this kind. Even the Stoics were not united in their views: most notably, Posidonius appears to have explicitly rejected this kind of view.13 Posidonius seems to have been concerned with cases where we act irrationally (see e.g. Long and Sedley 1987, sec. 65K3), and this concern is shared by others. For example, Locke resists the ‘established’ and ‘settled’ view that we desire the good on the same basis (Essay, II.XXI.31–8). I address such objections in Chapter 5. Another famous objector to desire-­as-­belief is Plato. In The Republic, Plato presents an argument that Reason and desire form different parts of the soul (436b–439d). (‘Reason’ here has a capital R, and thereby refers to the faculty of the mind, rather than a normative reason.) If correct, this would presumably entail that desire-­as-­belief is false. Plato’s basic idea is that we can be conflicted about what to do, and this requires that our conflicting impulses come from different parts of the soul. That is, one and the same part of the soul could not simultaneously be F and not-­F: that would be a contradiction. And so, one and the same part of the soul could not simultaneously be in favour of, and not in favour of, drinking. Now imagine a man who is thirsty, but who is unwilling to drink. Plato argues that his thirst—a desire—must originate in a different part of his soul from whatever ethical principles lead him to avoid drink.14 Otherwise, we would have to say that he both wants to drink and doesn’t want to drink, and that would be a contradiction. So desire and belief must originate in different parts of the soul. But Plato’s argument can’t be right (Annas  1981, 137–8). We should all allow that conflicts can occur within individual parts of the soul: your desires might conflict with one another, and your reasons beliefs might conflict with one another, as when you think that there is something to be said for going to 13 See e.g. Galen’s remarks in Long and Sedley (1987, sec. 65K2); for detailed discussion, see Cooper (1999). Some of the other Stoics were apparently equally unsure—see Long and Sedley (1987, sec. 65K1). 14  Later (§7.2), I argue that thirst is not even a desire, but the general points here could easily be made with other examples.

22  What Is Desire-as-Belief? the pub and something to be said for staying in. To this extent, everyone needs to make sense of conflicting attitudes in a manner that doesn’t require us to locate each conflicting attitude in a different part of the soul. So we should reject Plato’s argument. And that is easily done, since it involves a scope fallacy: Plato conflates desiring [not p] (D¬p) and failing to desire [p] (¬Dp).15 Plato seeks to avoid the conclusion that the thirsty man both wants [to drink] and fails to want [to drink], since that would be a contradiction. But even if the soul is unitary, we can avoid that conclusion. We should say that the thirsty man who is unwilling to drink is someone who wants [to drink] and also wants [not to drink]. That combination of attitudes is perfectly consistent with the law of non-­contradiction, and is enough to explain his conflicted psychology. So we should not accept Plato’s argument that Reason and desire form different parts of the soul. There is a further—better—objection to desire-­as-­belief, related to Plato’s. The idea is that it is perfectly rational to have conflicting desires, but definitely irrational to have conflicting beliefs. It seems to follow that desires could not be beliefs: they do not stand in the right consistency relations with one another (Archer 2016, 3–4; Tenenbaum 2007, 38–9). But this objection also fails. Desire-­as-­belief says desiring [that p] is believing [that you have reason to bring about p]. So now imagine that you desire [that p], but that you also have a conflicting desire [that ¬p]. If desiring [that p] is believing [that you have reason to bring about p], your desire [that ¬p] should be understood as the belief [that you have reason to bring about ¬p]. It should be clear that these beliefs are perfectly consistent: they can both be true, since you might well have reasons to bring about p and competing reasons to bring about ¬p. Here, your conflicting desires might accurately represent a genuine normative conflict. As I have formulated it, desire-­ as-­ belief entails that conflicting desires are beliefs about conflicting reasons, and not conflicting beliefs about reasons. So desire-­as-­belief permits that conflicting desires are rational, even though conflicting beliefs are not. That is, desire-­as-­belief does not analyse conflicts between desires as conflicts between beliefs (i.e. B[Rp] & B[¬Rp]), but rather as conflicts between reasons you believe you have (i.e. B[Rp] & B[R¬p]). It thereby entails that conflicts between desires are totally unlike conflicts between beliefs: conflicts between desires can accurately represent conflicting normative pressures, whereas conflicts between beliefs can never accurately represent anything (cf. De Sousa  1974; Williams  1976). So this further objection to desire-­as-­belief also fails. 15  This crucial distinction features again in §2.4.1 and §5.4.

Reason and Passion  23 Plato’s objection to desire-­as-­belief, and its descendant, both fail. But might we still want to maintain some distinction between Reason and passion? And if so, might that distinction undermine desire-­as-­belief? No. Though we should draw some useful distinctions in this vicinity, none of them conflict with desire-­as-­belief. For example, we should definitely distinguish between beliefs and emotions (see also §2.4.4, §7.1). But this is consistent with desire-­as-­belief, which is a theory of desire, not emotion. To the extent that the distinction between Reason and passion tells us something only about emotion, it is consistent with desire-­as-­belief. Or for another example, we should definitely distinguish between those states of mind that are produced by (conscious?) reasoning, and those that are not. But that distinction is consistent with desire-­as-­belief, since it cuts across the belief/desire distinction: many beliefs are not produced by reasoning (but instead, say, by socialization), and many desires are produced by reasoning (such as our political preferences).16 Or for a final example, we should definitely distinguish between states of mind that are rational and those that are not. But again, that distinction is consistent with desire-­as-­belief, since it also cuts across the belief/desire distinction: many desires may be irrational, but so too are many of our beliefs. We have always known that our beliefs can go awry in various ways, but the catalogue of our failures is constantly growing (e.g. Kahneman  2011). Moreover, our normative beliefs are especially prone to irrationality: more hangs on them, and so they are more liable to distortion from incentives such as self-­interest. To this extent, desire-­as-­belief is not only consistent with the claim that our desires are often irrational, but in fact positively explains such irrationality (for more, see §5.3). So again, the distinction between rational and irrational states of mind does nothing to undermine desire-­as-­belief. In these ways, we should be wary of rejecting desire-­as-­belief out of hand because it conflicts with the distinction between Reason and passion. That distinction is ambiguous, and on obvious disambiguations it is perfectly consistent with desire-­as-­belief. After Plato, Hume is the main canonical figure who cemented the division between Reason and passion.17 He writes: 16  I discuss our ability to change our desires by reasoning more thoroughly in §8.2. 17  That said, Hume’s views are not perfectly straightforward. For example, Hume seems to suggest some sympathy for something like desire-­as-­belief when he says things like ‘Desire arises from good consider’d simply’ (T2.3.9.7; see also T2.1.1.4, T2.3.9.2), though interpreting such claims is difficult given that Hume seems to use the words ‘good’ and ‘pleasant’ interchangeably (T2.3.9.8).

24  What Is Desire-as-Belief? A passion is an original existence, or if you will, modification of existence, and contains not any representative quality, which renders it a copy of any other existence or modification. When I am angry, I am actually possest with the passion, and in that emotion have no more a reference to any other object, than when I am thirsty, or sick, or more than five foot high. ’Tis impossible, therefore, that this passion can be oppos’d by, or be contradictory to truth and reason; since this contradiction consists in the disagreement of ideas, consider’d as copies, with those objects, which they represent. (T2.3.3.5)

Hume here suggests that desire-­as-­belief is false. One concern is that Hume gains unfair rhetorical force by switching between distinct states of mind: he moves back and forth between desires and emotions, even though these are clearly distinct states of mind. Defenders of desire-­as-­belief should not be too concerned if emotions like anger do not represent the world, given that they have a theory of desire and not emotion. But even if we set this issue aside, Hume doesn’t actually give any argument in this passage against desire-­as-­ belief. Rather, he simply asserts that passions do not represent the world: he asserts that our passions make no reference to any other object (see also T2.3.3.6). But this is just the very claim that is in question when we assess desire-­as-­belief. And so Hume’s only real argument for that division is the argument that Reason alone cannot motivate us: I address that across the next two chapters.

1.8 Summary In this chapter I outlined some basic terminology and assumptions that I make, and described desire-­as-­belief, according to which desiring something just is believing that you have reason to bring it about. I showed that the view is consistent with the direction-­of-­fit metaphor, has some simple attractive features, and can avoid some initial objections, including those relating to inconsistency in desire. Our next chapter discusses the role that desires play in motivating people to act, and defends the view that everything anyone has ever done can be explained in a systematic manner by appealing to their desires. That view will be important for Chapter 3, since it is a premise in an argument there for desire-­as-­belief.

2 Desire and Motivation Between this chapter and the next, I develop an argument for desire-­as-­belief. The basic idea is that only desires can motivate us, and yet reasons beliefs can motivate us. It follows that reasons beliefs must be desires. From there, it is a short step to desire-­as-­belief. This chapter focuses solely on the first of these claims. Only in Chapter 3 do I turn to the remainder of the argument. As a result, in this chapter, desire-­as-­belief steps backstage: here, I focus solely on the relationship between desire and motivation. A brief reminder from §1.1: I treat ‘desire’ and ‘want’ interchangeably. We might instead talk about preferences, which tell you how much someone wants one thing relative to another. But I will focus on our wants, setting preferences aside. And when I talk about what you want, I mean what you want to some extent, not what you want most.

2.1  Only Desires Motivate One grey winter’s day, I trudged through the cold rain into my office. A parcel had arrived! This in itself was puzzling: I wasn’t expecting anything, and at any rate the parcel was too small to be a book. I’d hoped that opening the parcel would resolve my confusion, but instead it only baffled me further: inside was a plastic model of a snail. This meant nothing to me, and there was no note to indicate who had sent it, or why. Why on earth would someone send me a plastic snail?! I had no idea what was going on. But even in my ignorance, I knew something: that it had been sent in order to achieve some goal. What that goal was, I didn’t know—perhaps someone hoped to amuse their friends with a quirky prank, or hoped to bring a smile to my face with a surreal joke. But I did know that whoever sent it had some goal, whatever it was. My inference was based on the following simple and attractive theory, Only Desires Motivate: ODM:  You can be motivated to act only by the combination of a desire [that p] and a belief [that this act will bring about p]. Desire as Belief: A Study of Desire, Motivation, and Rationality. Alex Gregory, Oxford University Press. © Alex Gregory 2021. DOI: 10.1093/oso/9780198848172.003.0003

26  Desire and Motivation ODM applied to my snail conundrum: I didn’t know what someone hoped to achieve by sending me the snail, but I did at least know that this act was motivated by some goal—a desire—and the instrumental belief that sending me the snail would help achieve that goal. ODM can be illustrated by any number of other examples. If you are motivated to wear a bobble hat, perhaps that’s because you want to look stylish, and believe that wearing a bobble hat will help you do so. And if you are motivated to cross the road, perhaps that’s because you want to buy a toy dinosaur, and believe you can buy one over there. And if you are motivated to overthrow the king, perhaps that’s because you want to improve the world and believe that overthrowing the king will do so. Vice versa, ODM tells us that if you desire p, but have no idea how to bring it about, you won’t be motivated to do anything with respect to that desire. This seems plausible: I’m not motivated to do anything about my desire for magic wellies because I have no idea where to find some. And ODM tells us that if you believe that doing something would lead only to p, but have no desire that p, you won’t be motivated to do it. Again, that seems plausible: I’m not motivated to eat my chair because I can’t see that it would achieve anything that I want. We can think of ODM in terms of navigation. If you are navigating your way around an area, you need two things: a map and a goal. A map tells you about the lay of the land, informing you about which things can be found in which direction. A goal is your end point: what you are trying to make your way towards. If you had a map, but no goal, you wouldn’t travel in any particular direction: you’d have nothing to aim for. If you had a goal, but no map, at best you’d wander randomly: you’d have no idea how to head towards, rather than away from, your goal. What you need is both: some idea of where you are headed and some idea of how to get there. We can think of our beliefs as sophisticated maps of the world around us, and our desires as goals that tell us what we are aiming for. I formulated ODM as the view that we are motivated by desires combined with instrumental beliefs about what our acts will bring about. In fact, this is not quite right. For example, you can be motivated by desires combined with beliefs about what your acts might bring about, as when you act on a best guess. Perhaps your desires can even motivate you by combining with representations that somehow fall short of belief, such as credences, so long as they help you to see how you might satisfy your desires. And perhaps, in some cases, a belief is not needed at all: perhaps I can perform basic actions like

Why Accept ODM?  27 blinking without the need for mediating beliefs about how to perform them.1 For my purposes, I needn’t fuss about these issues: everything that follows hinges on the role of desire in motivation, not the role that instrumental beliefs play. As a result, I shall often talk roughly, as though ODM just says that only desires motivate, omitting the reference to belief. My label for the view indicates this focus. ODM fits neatly not only with common sense, but also with decision ­theory, which models actions in terms of ‘utilities’ and ‘credences’, which roughly correspond to the desires and instrumental beliefs in ODM. If we needed to, we could supplement ODM with more specific claims about exactly how degrees of desire and confidence in our beliefs make a difference to our levels of motivation, as decision theory does. But again, for most of my purposes, this level of detail isn’t needed, and I set these details aside (though see §8.1 and Appendix A). ODM is not exactly the consensus view, but it’s nonetheless dominant. Nearby views are expressed by various authors, such as Davidson (2001a), Smith (1987,  1994, 92), Stalnaker (1987, 15), and Sinhababu (2009,  2017). The view is implicit in some older figures too, as for example when Aristotle says, ‘that which produces movement is a single thing, the faculty of desire’ (DA433b; see also DM700b15–24). Probably the most famous defender of ODM is Hume. In fact, so great is Hume’s influence that ODM is sometimes labelled ‘the Humean theory of motivation’. But I avoid this label: Hume committed not only to ODM but also to some other claims that I wish to distinguish from ODM, such as about the rationality of desire, and the difference between belief and desire. Unifying all of Hume’s views all under one label can wrongly encourage us to think of them as forming a unified package rather than as the mere conjunction of the various views Hume happened to endorse. I will return to the connection between desire-­as-­belief and the larger Humean picture at the end of the next chapter (§3.5).

2.2  Why Accept ODM? Why is ODM so popular? There might be some sophisticated argument for it—for example, Smith gives one such controversial argument for the view 1  For some discussion, see the exchange between Smith (2004, 101–4) and Harcourt (2004, 124–8).

28  Desire and Motivation (1994, 116). If you are moved by his argument, great. But it seems to me to appeal to premises that are less obvious than the truth of ODM itself, and so  I’m not clear that it will really do much to move anyone not already attracted by ODM. Instead, I think the central attractions of ODM are very straightforward. First, most importantly, for any action you care to name, it is plausible that some belief/desire pair explains it. Any plausible theory of motivation needs to be able to explain the ubiquitous relevance of our beliefs and desires to our actions, and ODM does so in the most obvious way. Imagine, for example, that I learn that you regularly give money to Oxfam. Why do you do that? It seems perfectly appropriate for me to infer that you have some desire— perhaps to help others?—and believe that giving to Oxfam will achieve that goal. But why would this kind of inference be warranted unless a theory like ODM is true? If rival views simply deny that our actions can always be explained by our desires, then those views are radically sceptical about our ordinary capacity to explain behaviour, since they tell us that we are wrong to always so quickly hunt for some belief/desire pair in such contexts. On the other hand, if rival views agree that it is always appropriate to explain actions by appeal to desires, then it’s not at all clear how they can square that with the denial of ODM.2 Second, relatedly, ODM is attractive because it promises to explain so much (see also Sinhababu  2017). Humans act in enormously varied ways, and a theory that promises to explain all those actions in the very same way has enormous explanatory power. ODM promises to explain, in a systematic manner, everything that anyone has ever done, from my tying my shoelaces to Caesar crossing the Rubicon. Given the extremely wide variation in human action, a theory that promises to explain it all in the same way is a theory that promises to be extraordinarily powerful. Moreover, ODM explains all those actions in a manner that is extremely simple: it appeals to just two states of mind. In this way, ODM is quite brilliant: it promises to explain a very wide range of phenomena in a very simple and systematic way. These arguments for ODM are clearly defeasible, because they are arguments from success: if we can find some act that doesn’t fit with ODM, then ODM would no longer be attractive in these ways.3 But equally, as we investigate various possible counterexamples to ODM and find that none succeed, we should be enormously impressed at just how successful ODM is: 2  Nagel might be understood as attempting something like this task; I discuss his view in §2.4.2. 3  In §2.4, I address some hard cases for the view.

Why Accept ODM?  29 it seems to succeed at a task which we might have thought impossible. In short, I think we should endorse ODM just because the theory fits with ordinary thought and talk which makes ubiquitous reference to belief/desire pairs as explanations of action, and because ODM has extraordinary explanatory power.4 With these thoughts in mind, we should briefly return to Hume. As I said, some think of ODM as the ‘Humean’ theory of motivation. Above I sketched why I resist this label, but there is a further important reason for doing so. Hume focuses not on the claim that we are motivated by desire/belief pairs, but rather on the negative claim that our beliefs alone cannot motivate us. As Hume puts it, he aims to establish that ‘reason alone can never be a motive to any action of the will’ (T2.3.3.1). But we should be careful about the relationship between this claim and ODM. For clarity, here they are again: ODM, roughly:  Only desires motivate. Hume’s negative claim:  Beliefs alone can never motivate. People sometimes talk as though Hume’s negative claim is entailed by ODM. But this is correct only if desire-­as-­belief is false. If desire-­as-­belief is true, then desires are beliefs, and as such it might be true that only desires motivate and yet also true that beliefs alone can motivate. So if we want to take desire-­as-­belief seriously, we should keep the above claims apart. Moreover, once we keep these claims apart, we can see that the two arguments above are definitely arguments only for ODM, not Hume’s negative claim. The first argument was that ODM fits with ordinary claims about the relevance of desires for explaining actions. The second argument was that ODM is plausible because it is parsimonious and has great explanatory power. These are not arguments that beliefs do not motivate. After all, Hume’s negative claim is not a theory that promises to explain actions at all: it merely tells us how not to explain actions. So the attractive arguments above are arguments for ODM, and not for Hume’s negative claim. To put that another way, we will ultimately arrive at the view that ODM and desire-­as-­belief are both true, and therefore that Hume’s negative claim is false. That combination 4  Some generalize this explanatory ambition further and try to explain other things by appeal to just our beliefs and desires. For example, some try to reduce other states of mind such as intentions, emotions, hopes, fears, and wishes to combinations of beliefs and desires (e.g. Davis 1981; Hobbes L  I.VI; Hume T2.3.9.6; Marks  1982; Ridge  1998; Searle  1983, 29–36; Sinhababu  2017, 100–12; Smith 1994, 117). I discuss some related issues in §2.4.3, §2.4.4, §6.5, and §7.1, but for independent reasons: I here leave it as an open question whether belief/desire psychology can explain more than just motivation.

30  Desire and Motivation of views definitely gets to capture the above attractive features of ODM, since it does indeed explain all action by appeal to belief/desire pairs. Indeed, if anything, that combination of views makes ODM all the more explanatorily powerful, explaining all human action by appeal to only systematic combinations of a single state of mind: beliefs. Since defenders of desire-­as-­ belief can capture the attractive features of ODM whilst denying Hume’s negative claim, the central arguments for ODM are arguments for ODM, and not arguments for Hume’s negative claim. Did Hume have any arguments for his negative claim? He argues that beliefs are either about ‘relations of ideas’ or ‘matters of fact’, and then asserts that neither is motivationally relevant (T2.3.3.2–3). But he focuses wholly on non-­normative beliefs when he makes these claims. It is highly plausible that beliefs with non-­normative contents cannot motivate alone. If you believe that triangles have three sides, or that peanuts are brown, these beliefs alone cannot motivate you to do anything. But this leaves open whether beliefs with normative contents might motivate us. (I leave open whether these are beliefs about relations of ideas, about matters of fact, or neither.) When Hume later turns to his discussion of the influence of moral judgement on the will, he simply refers back to his previous discussion as having shown that beliefs alone cannot motivate us (T3.1.1.6). But since the earlier discussion ignored beliefs with normative contents, this is unconvincing. It is akin to first asserting that plants are never taller than 6ft but failing to say anything about trees, and then later claiming that trees can’t possibly be plants because you have already shown that no plant is taller than 6ft: Hume first asserts a universal claim whilst ignoring an obvious candidate counterexample, and later replies to that counterexample merely by appealing to his earlier discussion. We should see Hume’s important insight as ODM, not his negative claim that beliefs alone cannot motivate.5 Again, I return to ‘Humeanism’ in §3.5.

2.3  Desires Sometimes Motivate ODM is a theory of motivation. To be motivated to do something is to have a variety of dispositions towards that action: most obviously, to do it, but also to 5  Smith’s argument for ODM is equally consistent with denying Hume’s negative claim. Smith does also argue for Hume’s negative claim, but his argument relies on the claim that no state of mind can have both directions of fit at once (Smith 1994, 117–25). That claim is independent of ODM, and is false (§1.4).

Desires Sometimes Motivate  31 think about ways of doing it, and to think about how other actions might impede your doing it (see also Johnson King  2020, 412). As I understand motivation, we can only be motivated to perform actions: we might have various inclinations to form new mental states, but such processes don’t motivate those mental states. For this reason, as I use the word, beliefs and desires themselves are never motivated. Moreover, as I understand motivation, it is a special kind of cause for an action. We might explain some action with reference to your genes, or give some purely physical story about the prior movements of atoms in your body, but in neither case would we thereby be giving a motivational explanation of your act. ODM says that desires are the only things that bring motivation about. I said that by ‘desire’, I mean what we desire to some extent. Similarly, by ‘motivation’, I mean what you are motivated to do to some extent. On this use of the word, you might be motivated to do more than one thing at once. For example, perhaps you are somewhat motivated to finish reading this chapter today, but I imagine that you might be distracted by competing motivations (perhaps the pub beckons). When ‘motivated’ is understood in this way, it’s clear that we need not actually do everything we are motivated to do: you can’t act on all of your competing motivations. But what you do is nonetheless sensitive to what you are motivated to do: plausibly, your motivations determine your actions in the sense that you will do whatever you are most motivated to do. (Or, at least, you’ll try: perhaps goblins will kidnap you at just that moment.) To this extent, our actions can be fully explained by our levels of motivation. If our actions can be fully explained by our levels of motivation, and motivation is always explained by desire, does it follow that we always do whatever we most want to do? That is a basic assumption of revealed preference theory, and a common assumption in general.6 But it does not follow from the claims above. ODM places a necessary condition on motivation, and not a sufficient condition. ODM says that we are only motivated by desire, and this is consistent with our sometimes failing to be motivated by desires that we have.7 And even if some desire does motivate us, ODM permits that we might be motivated out of proportion to the strength of that desire. These claims are absolutely crucial in what follows, and so I repeat the bottom line: ODM says that the only source of motivation is desire.

6  On revealed preference theory, see e.g. Samuelson (1947, 97–8). 7  See also Armstrong (1993, 157–8); Arpaly and Schroeder (2014); Foot (1972, 306); Sinhababu (2017); Swartzer (2015); cf. Dancy (2002, 85–90).

32  Desire and Motivation But that is consistent with our sometimes being motivated to do things by weaker, rather than stronger, desires. Agreed, the default expectation is that desires will motivate us, and that stronger desires will motivate us to a greater degree. But such expectations are defeasible, and so ODM permits that our actions may not perfectly reflect our desires. An analogy may be helpful: perhaps the only source of electricity for your house is a single cable, and that cable is supposed to provide 240 volts. Saying this is perfectly consistent with the cable sometimes providing no electricity at all (perhaps there is a fault up the line somewhere) and also perfectly consistent with the cable sometimes providing more, or less, than 240 volts (perhaps there is a partial fault further up the line). Similarly, the only source of motivation is desire, but this is consistent with your desires sometimes failing to motivate you altogether, and consistent with your desires sometimes motivating you out of proportion to their strengths. Mismatches between desire and motivation can occur in various ways. One I shall set aside: you might have some desire but have no idea how to satisfy it. Plausibly, such a desire might generate no motivation at all. If you accept this possibility, then it’s clear that desires need not result in motivation. But I won’t (yet) say much about this kind of case, or rely on it, since it raises possible complications given a possible distinction between desires, on the one hand, and wishes and hopes on the other. We might think that if you have no idea how to achieve your goal, it is a hope or wish rather than a desire. I will defer that topic until §6.5.8 So in what follows I instead describe four other kinds of mismatch between desire and motivation. (Note also that we return to some of the issues below in §5.5, which addresses akrasia.) First, you might fail to respond to some of your desires altogether. Just as you might know the answer to a question in a quiz and yet not be able to bring it to mind, you might have a desire relevant to a choice you face and yet fail to bring it to bear. For example, you might want to buy some milk on your way home today, but absent-­mindedly fail to be motivated by this desire as you head home. An objector might suggest that in such a case you have simply lost the desire to buy the milk for a period of time. But more plausibly, when you later chide yourself, ‘Damn, I wanted to buy some milk on the way home!’, you mean that you failed to act on this desire, not that you ceased to have it. 8  A nearby possibility is that you might want something to come about without your interference, as when you desire that Jeff will fall in love with you, of his own accord (Arpaly and Schroeder 2014, 114; Schroeder 2004, 17). It might seem that such a desire cannot motivate you to do anything, since that would be self-­defeating. But as David Wall explains, if this is what you want, then you might well be motivated to do those things that you think will have the best chance of bringing this about: namely, not interfering with Jeff (Wall 2009).

Desires Sometimes Motivate  33 A second kind of mismatch between desire and motivation is that you might rationally be unmoved by some desires because you are focused on other more important goals instead. For example, I am not at all motivated to exercise for four hours a day, because I know that the benefits will be so enormously outweighed by the costs. This is not the same as my lacking the desire to exercise for that amount of time: I definitely want to do so, to some limited extent (I’d be so fit!). Similarly, though I have some desire to learn about astronomy, I am not presently motivated to do so because this aim is not urgent and I have many more important things to be doing. In circumstances like these we can rationally ignore options that would on balance satisfy our desires only to a small degree, and by ignoring an option we can lack the motivation to pursue it. Trivial goals may be put on the backburner, where the relevant desires persist but play no present role in our motivational economy. Whereas it is easy to desire many things at once—a list of all your desires would be enormous—motivation is a more limited resource that we spend more sparingly. As a result, our desires need not all be reflected in present motivation. A third kind of potential mismatch between desire and motivation is that you might irrationally have too little motivation to pursue some options, given your desires. For example, you might want to get your marking done today, but be moved more strongly by your desire to watch Game of Thrones, though you would prefer to get your marking done and watch the show later than to organize your time in any other way. Or for another example, though I want to, I am (sadly) not appropriately motivated to exercise one hour a day, because I struggle to summon the energy and willpower. It’s crucial to see that in such cases the problem is not merely that the relevant desire is weak: I might really want to get fit, know that to do that I need to exercise for an hour a day, and yet fail to be sufficiently motivated to do so. This, sadly, is a familiar phenomenon. Perhaps in extreme cases we completely lack the motivation to pursue some of our goals, but probably more commonly we are motivated to pursue those goals, but too weakly. Finally, a fourth kind of potential mismatch between desire and motivation is that you might irrationally have too much motivation to pursue some goals, given your desires. After all, if your desires can undermotivate you, why couldn’t they overmotivate you?9 (Given our original analogy, this would be 9  Admittedly, I sometimes talk about desires being disposed to motivate you, and those who discuss dispositions tend to focus on cases of underfiring rather than overfiring. But the basic point is that desires play a certain role, which they can play too forcefully. This is attractive even if it should not be couched in dispositional terms. It parallels plausible claims about other functional objects, such as when your coffee machine overfills your cup.

34  Desire and Motivation akin to a power surge, rather than a power failure.) For example, imagine that you want to rearrange the books in your office so they are ordered more sensibly. Imagine that this desire is present, but weak: you aren’t particularly fussed. You have some free time this morning, and so, moved by your mild desire, you begin to tidy your books. But before you know it you get carried away and you’ve embarked on a massive all-­day project of devising the most perfect book arrangement possible, at the cost of achieving several other goals that you care about more. Here, you are strongly motivated to achieve something you only weakly desire. Or, for another example, imagine that you want to have a successful career, but more strongly want to have time with your family. You might nonetheless find yourself getting carried away by your pursuit of your career, working late at night and at weekends, and so on. In such a situation you might feel foolish, and feel foolish precisely because you recognize that your endeavours to further your career are coming at the cost of something you want more. In cases like these, we are motivated too strongly by our desires: we are motivated to pursue our goals in ways that are disproportionate to our commitment to those very goals. Such cases of fixation mirror weakness of will in that they involve overmotivation, rather than undermotivation, by our goals. In short, we can distinguish four key kinds of mismatch between desire and motivation: i. You might simply fail to respond to some desires. ii. You might rationally be undermotivated by some desires, and focus on more important goals. iii. You might irrationally be undermotivated by some desires, and focus on less important goals. iv. You might irrationally be overmotivated by some desires, and focus on them at the expense of other equally or more important goals. These claims are all consistent with ODM, which says that we are only motivated by belief/desire pairs, not that every such pairing motivates us appropriately. According to ODM, our beliefs and desires are the only generators of motivation, but they need not always fire at the correct power. To help cement the claim that there can be mismatches between desire and motivation, it’s helpful to think about other mental states such as our beliefs, and the effects that they (alone) can have. For example, we might think that our beliefs dispose us to make appropriate inferences with their contents: plausibly, if you believe [p → q], you are thereby disposed to believe [q] if you

Desires Sometimes Motivate  35 come to believe [p]. Does it follow that if you come to believe [p], you will in fact also come to believe [q]? Absolutely not: we fail to make at least some such inferences. Though believing [p → q] involves a disposition to believe [q] if you believe [p], that disposition may not manifest itself, even if you do indeed come to believe [p].10 The above point about desire and motivation is exactly the same. A desire [that p] is disposed to motivate you to do whatever you believe will bring about p. But even if there is something you believe will bring about p, your desire is only disposed to motivate you, and this disposition might fail to become manifest under some circumstances. In short, though we should accept ODM, we should deny that we are always motivated in exact proportion to our desires. You might strongly desire to finish your paper today—indeed, desire to do so more than anything else in the world—and yet irrationally get tempted into pursuing other more tangible goals that you have. Part of what is frustrating about common lapses of rationality—overeating, overdrinking, under-­exercising—is precisely that we know exactly what we want and yet we cannot summon the willpower to pursue those goals. If you say that we are always motivated in proportion to our desires, then you are thereby forced to conclude that people don’t very strongly want to avoid obesity, hangovers, or ill health. Our more plausible view is that people are strongly averse to these things, but struggle to summon the willpower to execute those desires.

2.3.1  An Aside: Dispositional vs. Occurrent Desires Some philosophers distinguish between ‘occurrent’ and ‘dispositional’ states of mind, such as between occurrent and dispositional beliefs, and between occurrent and dispositional desires (e.g. Price 1969). Intuitively, some desires are presently alive to you, whereas others exist but play no role in your present mental life (perhaps your desire to keep reading is presently in the former category, and your desire to have children one day in the latter). We might use this terminology to mark the contrast between desire and motivation: perhaps an occurrent desire is one that is presently motivating you, and a dispositional desire is one that is not. But in fact, I think this terminology is somewhat unhelpful.

10  What about the reverse possibility: can this disposition overfire? It can: you might end up believing [q] with too much confidence relative to your confidence in [p] and [p → q].

36  Desire and Motivation First, talk of whether a mental state is ‘occurrent’ is unhelpfully ambiguous. We might mean that the state of mind is present to consciousness, or we might instead mean that it is having some behavioural effect. Our desires can be occurrent in one of these senses without being occurrent in the other (Mele 1995, 396–7; Smith 1994, 106–16). I am rarely consciously aware of my desire to avoid death, though it often plays a role in determining my actions (e.g. when driving on busy roads). In contrast, I am often consciously aware of my desire that there be parsley on the moon (Nagel  1970, 45), but only because I like thinking about the example, not because I ever act on it. We should keep these two possibilities separate in our minds, and not run them together in our terminology (perhaps; Strandberg  2012). Certainly, when I say that a desire might motivate to a greater or lesser degree, this could be translated into talk about ‘occurrent desires’ only if ‘occurrent’ is used in its behavioural sense, not in its present-­to-­consciousness sense. More problematically, talk of whether a desire is ‘occurrent’ is most naturally heard as binary, so that a desire is either fully occurrent or else not occurrent at all. But I am thinking of motivation as coming in degrees, since you might be more or less strongly motivated to do something. Of course, an occurrent desire can have a certain strength, but I am claiming that a desire can have motivational effects that are disproportionate to its strength. It is not clear how we could translate that into a claim about whether a desire is occurrent or not. Finally, I argued that just as desires can undergenerate motivation, they can also overgenerate motivation. If that point is granted, it is hard to see how my claims could be described using the distinction between occurrent and dispositional desires, unless we are happy to say that some desires are ‘too occurrent’. For these reasons, I prefer to avoid the ‘occurrent’ vs. ‘dispositional’ terminology, and instead just talk about the degree to which desires motivate.

2.4  Objections to ODM In §2.2 I claimed that ODM is attractive because it fits with ordinary thought and talk, and because it explains all human action in a way that is simple and systematic. Especially with these claims in mind, I had better address possible counterexamples to ODM. One key objection to ODM is that our normative beliefs can motivate us, and so ODM must be false. I address this separately in the next chapter and

Objections to ODM  37 beyond. In the remainder of this chapter, I focus on other objections to ODM.11

2.4.1  Misleading Desire Ascriptions Our first objection to ODM is refreshingly simple: people sometimes do things that they don’t want to do (Platts 1997, 256; Searle  2001, 169–70; Shafer-­Landau  2003, 123). For example, as we ordinarily talk, it seems that many of us don’t want to get out of bed in the mornings, don’t want to do chores around the house, and don’t want to fulfil some of our promises. Nonetheless, we are motivated to do these things. One possible response to such examples is to say that ODM uses the word ‘desire’ in a very broad stipulative sense, so it is true by definition that we do want to do these things (Davidson 2001a, 3–4; cf. Schueler 1995, 29–30). That is, we might suggest that ‘desire’ is just a broad label for any motivating state, so by definition whatever moves you to act—whatever it might be—is a desire. But we should not defend ODM in this way: it threatens to trivialize ODM by leaving the nature of desire too unconstrained (e.g. Platts 1997, 256; see  also the discussion of Nagel in §2.4.2 below). Moreover, part of the attraction of ODM is that it fits with common-­sense thoughts about the importance of our desires for explaining our actions. If we insist that ODM uses ‘desire’ in some technical sense that is independent of its ordinary usage, we thereby abandon part of the initial attraction of ODM. As a result, I will not pursue this strategy: I take ‘desire’, as used in ODM, to have something very close to its ordinary meaning (or, at least, the ordinary meaning that ‘want’ has—see §1.1). Instead, we should defend ODM by appeal to more general linguistic principles: principles that are independently plausible, and which apply in the 11  There is one broader objection to ODM I don’t address below: Jonathan Dancy argues that people are never motivated by any mental states whatsoever (Dancy 2002; see also Alvarez 2010). The thought is that people are moved by features of the world (as they see them), such as the prospect of helping someone, or of avoiding poverty. Since facts like these are not mental states, they are clearly not desires, beliefs, or combinations of the two. There is some truth in what Dancy says. People are indeed moved by features of the world when they act. But for all that, Dancy should acknowledge that our mental states clearly play an important role in our mental economy, and stand in some important explanatory relationship to our actions. Whatever disagreement remains is presumably just about which of these explanatory chains to think of as ‘motivation’, and I’m not sure what hangs on that. So it’s best to think that Dancy is absolutely correct that one explanation of our actions makes no reference to our mental states, but to insist that another parallel explanation of our actions does. ODM concerns the latter topic, and so is consistent with the truth in what Dancy says (for nearby thoughts, see Fogal 2018; Smith 2012, 392; and see also Pettit and Smith 1990).

38  Desire and Motivation examples above to vindicate ODM. This way of defending ODM allows us to maintain ODM without trivializing it. I will discuss two such general linguistic principles in turn. First, sentences with negations are frequently (syntactically) ambiguous with respect to what exactly is being negated. For example, if I say, ‘I didn’t take the job because of the money’, I might mean that the money explains why I didn’t take the job, or else mean that though I took the job, it wasn’t because of the money. Similarly, if I say, ‘he doesn’t believe p’, that might mean that it’s not the case that he does believe p (¬Bp), but might alternatively mean that it is the case that he believes not-p (B¬p): this is why both atheism and agnosticism can be expressed by ‘I don’t believe in God’.12 Before I apply this point to our desire-­talk, it’s crucial to first remember something I said earlier: By ‘want’, I mean ‘want to some extent’. In this sense, you might have conflicting desires: you might want to go on the water slide (for the thrill), but also want to not go on the slide (in order to stay dry). To this extent, whether you desire [that p] is independent of whether you desire [that ¬p]: you can see something attractive about one of these, the other, both (conflict), or neither (disinterest). Now, take some negated assertion about desire, such as ‘Zara doesn’t want to paint my office’. Such an assertion might mean that Zara fails to want [to paint my office] (¬Dp), or might instead mean that Zara does want [to not paint my office] (D¬p) (cf. Broome 2013, 274). In the first case, Zara might have nothing against painting my office, but just have not thought about it either way. In contrast, in the second case, Zara might have thought about it long and hard, and decided that she’s positively against the idea. This same contrast can be made in other ways: you don’t want Zara to paint my office (until now, the issue had presumably not crossed your mind), but only my office-­mate has the positive desire that Zara not paint my office (she is known for her lurid choice of colours). One way to check which reading is intended is to imagine supplementing the assertion with ‘really’: ‘Zara really doesn’t want to paint my office’ surely conveys the presence of a strong desire not to paint my office, since it would be senseless to assert the strong absence of a desire. This reasoning suggests that if ‘really’ can sensibly be added to the relevant sentence, it ascribes a strong negative desire, not the absence of desire. The important upshot of all this is that we should acknowledge that someone might truthfully assert ‘Zara doesn’t want p’ even though Zara does want 12  It’s easy to see how this happens in English: Both ‘do not __ __’ and ‘do __ not __’ get contracted to ‘don’t __ __’, and as a result the last is ambiguous between the former two. Some discuss nearby phenomena under the label ‘neg-­raising’ (see e.g. Horn 2001, 308–29).

Objections to ODM  39 p: if this assertion indicates that Zara desires [that ¬p], it is consistent with Zara also desiring [that p]. Zara might really want not to paint my office but also have some desire to do so (perhaps she hates painting, but does want to make me happy). Under those conditions I can truthfully assert ‘Zara doesn’t want to paint my office’. This sentence is confusing, because Zara does want to paint my office. But the assertion is nonetheless truthful because it correctly says that she has a (competing) desire not to paint my office. With this point in hand, we can respond to our original supposed counterexamples to ODM. We might say that we (really) don’t want to get out of bed in the mornings, but plausibly this expresses a desire not to do so, and that is consistent with our also desiring to do so, at least for instrumental reasons. As a result, ODM (surprisingly) permits that you might be motivated to get out of bed, and yet truly assert that you don’t want to do so. Similar remarks apply to the other examples with which we began. In this way, defenders of ODM can accommodate the thought that we sometimes do things whilst appropriately saying that we ‘don’t want’ to do them. Such assertions are consistent with ODM because they convey the existence of desires not to do those things, not the non-­existence of desires to do them. I now turn to a second (complementary) defence of ODM. The basic idea is that there are pragmatic grounds for failing to assert the existence of certain desires. Such pragmatic factors can distort our assertions in ways that are misleading.13 To see this, begin by noting that our desires are very numerous. You might simultaneously desire to become a singer (for the fame), to become a gardener (for the sunshine), and to become a philosopher (for the pedantry). Probably, you have some desire to pursue a very large number of different paths in life, including even very remote options: you might well have some desire to become a hermit, in order to avoid your emails. Parallel reasoning can be applied to almost any other choice you might face. Insofar as you have very weak desires even for fairly remote options, it follows that your desires are extremely numerous. With that in mind, let us turn to our second more general linguistic principle: Grice’s maxim of quantity, which says to make assertions only when they are informative (Grice  1989, 26). Because we have some desire for so many things, merely learning that someone has some desire that p is rarely informative. As a result, Grice’s maxim tells us not to assert ‘Zara desires that p’ if we mean only to convey the bare fact that Zara has some desire that 13 The discussion below parallels Mark Schroeder’s discussion of negative reason existentials (2007a, 92–7).

40  Desire and Motivation p: often, that modest fact will be too trivial to mention. Further, with this expectation in place, if someone does assert ‘Zara desires that p’, we are in a position to infer more than the mere fact that Zara has some desire that p. We should infer that this assertion has been made because this desire is especially worth noting, most probably because it is very strong, or else because it is of some notable kind. For example, if you tell me that you want to go to the pub, I might reasonably infer that you don’t merely have some desire to go (who doesn’t?), but rather have a fairly strong desire to go. Or if I ask about your new job and you happily tell me that you now want to go to work, I might reasonably infer that you don’t merely have some desire to go for the money (who doesn’t?), but rather desire to go not merely because of the money. This point provides a second way to defend ODM from the counterexamples above. I hardly need to be told that you have some desire to get out of bed in the mornings when it is common knowledge that doing so is instrumental to many things you want (e.g. food). As a result, you might be disinclined to assert that you want to get out of bed in the mornings. But this disinclination can be explained by Grice’s maxim in a way that permits that you do nonetheless have this desire. As a result, ODM (surprisingly) permits that you might be motivated to get up and yet be disinclined to assert that you want to do so. Similar remarks apply to the other examples with which I began.14 This story is obviously compatible with what I said above about negation: Grice’s maxim of quantity inclines us to not assert that we do want to get out of bed, and the existence of a desire not to get out of bed inclines us to assert that we don’t want to get out of bed. But such assertions are consistent with the fact that your motivation to get out of bed is explained by your desires. In summary, our assertions might suggest that we are sometimes motivated to do unpleasant things that we have no desire to do, but we can insist, for principled reasons, that such assertions are misleading. Assertions about what we ‘don’t’ want are ambiguous with respect to what we are negating, and Grice’s maxim of quantity will incline us not to assert the obvious. These phenomena make our desire-­talk misleading in various ways. Once we clarify these effects, ODM escapes many putative counterexamples.

14  The claim in the main text is effectively that we refrain from making certain assertions about our desires because, although true, they would have misleading conversational implicatures. Such implicatures are normally thought to be cancellable (Grice 1989, 39; and see Zakkou 2018 for a helpful overview). If our disinclination to make the relevant assertions about our desires is explained by their misleading implicatures, we should be happier to make those assertions when the relevant implicature is cancelled. This prediction is accurate: you might be disinclined to assert that you want to go to work, but happier to assert that if you qualify it: ‘well, I mean I do want to go, but only for the money’.

Objections to ODM  41 There might well be other linguistic effects that can also give rise to illusory ‘counterexamples’ to ODM. For example, some assertions are implicitly restricted in scope, so that if you are talking about your baby and you say, ‘I’ve run out of milk’, you might really mean ‘I’ve run out of baby milk’: your original claim might have been implicitly consistent with your allowing that you have some milk left, just nothing suitable for the baby. Similarly, you might assert, ‘I have no desire to keep this promise’, but given the context really mean only that you have no self-­interested desire to keep this promise. For that reason, ODM permits that you might reasonably make this assertion, and nonetheless be motivated to keep your promise: your assertion conveys the absence of a certain kind of desire, not the complete absence of any desire. More generally, ODM is consistent with claimed contrasts between what you want to do and what you know you ought to do: we might understand those as contrasts between some subset of your desires and what you want to do on balance.

2.4.2 Nagel I now turn to Thomas Nagel’s rejection of ODM (1970). Nagel’s remarks, though brief, have strongly influenced many main opponents of ODM (e.g. McDowell  1998; Parfit  2011b, 381; Schueler  1995). What was influential in Nagel was not so much an objection to ODM, but instead a rival theory that nonetheless promises to capture everything that is attractive about ODM. To explain Nagel’s view, it will be easiest to work with the simplistic version of ODM, the one that omits explicit reference to the role of beliefs in motivating us to act. So understood, ODM says: ODM, roughly: Only desires motivate. Nagel is keen to distinguish that claim from the following weaker one: If you are motivated to act, then you desire to act in that way. These two claims look very similar, but whereas the first says that your desires explain what you are motivated to do, the second only makes a weaker claim about entailment: wherever there’s motivation, there’s desire. With this distinction in hand, Nagel’s basic thought is that only the second claim is true, and that the first claim looks attractive only insofar as we confuse it with the second.

42  Desire and Motivation How could the second claim be true but the first false? That is, why would you expect desire wherever you get motivation, unless there was some explanatory link between the two? Nagel could be a little clearer on this point, but I take it that his basic answer is that ‘he wanted to . . .’ sometimes merely says that the relevant action was motivated (cf. Dancy  2002, 86–8). For example, he writes: ‘whatever may be the motivation for someone’s intentional pursuit of a goal, it becomes in virtue of his pursuit ipso facto appropriate to ascribe to him a desire for that goal’ (Nagel 1970, 29). That is, on Nagel’s view, it is because you intentionally do something that it becomes appropriate to say that you ‘wanted’ to do it. Or, again, as he puts it: ‘that I have the appropriate desire simply follows from the fact that [the relevant] considerations motivate me’ (1970, 29). Given this view, Nagel is in a position to agree that wherever there is motivation, there is desire. But he thinks that ODM is false, and has a misleading veneer of plausibility only because it can be confused with this nearby truth. Again, this is not so much an objection to ODM as a rival, and even then it’s not quite clear whether Nagel argues for this rival so much as asserts it. At any rate, Nagel’s view is implausible, as I now explain. But before I present my objections, I should clarify that Nagel’s view is actually slightly more complicated than I presented it above, since he claims that desires come in two kinds: ‘motivated’ and ‘unmotivated’ (1970, 29).15 The claims above are claims about motivated desires: they are really just shadows of independent choices we make. In contrast, Nagel allows that unmotivated desires—such as hunger and thirst—are more substantial entities which can genuinely explain those choices. Unmotivated desires are somewhat primitive urges that merely ‘assail’ us (1970, 29). These desires have no special role to play in debunking ODM, above, but are relevant insofar as they make a difference to my objections to Nagel. Here is a first (flat-­footed) objection to Nagel. I said that ODM is attractive because it fits with the way in which we very commonly explain actions with reference to people’s desires, and because ODM promises to explain so much (§2.2). Nagel’s rival to ODM, on which motivated desires explain nothing at all, obviously fails to share these advantages. In turn, at first pass, Nagel’s view has the surprising consequence that no-­one ever does anything because they 15  Properly speaking, I would prefer to say that only actions, and never mental states, are ‘motivated’ (§2.3). But I’ll set this aside to focus on the content of Nagel’s view, rather than the labels he uses to express it. Equally, I set aside the question of whether Nagel really means to say that there are two kinds of desire rather than that the word ‘desire’ has two meanings.

Objections to ODM  43 want to. Really, this is the very point of the view. But rather than undercutting ODM, this instead seems to highlight quite why ODM alone is attractive: only it captures the commonplace thought that desires explain actions. Now Nagel could here appeal to his distinction between motivated and unmotivated desires, and claim that to the extent it is appropriate to positively explain actions by appeal to desire, we are explaining them by appeal to unmotivated desires. But this is highly implausible: consider all of the everyday actions that we would normally explain by appeal to desire: my voting choices, my pursuit of my chosen career, the choice of books I read, and so on. In every case, we would ordinarily be tempted to explain these actions by appeal to my desires. On Nagel’s view this is inappropriate. Adding that this might be appropriate because these actions might be explained by unmotivated desires hardly helps, since it is vastly implausible that this is so: in all these cases, it seems that my actions are explained by desires that arose through some kind of reasoning (however brief). So in these cases my desires are clearly not unmotivated desires. But the only alternative for Nagel is to say that these desires are motivated desires, which don’t explain my behaviour at all. Again, that claim is not plausible. Here is a second objection to Nagel. Motivated desires are ascribed to people in virtue of how they intentionally act. But we have many desires which do not move us to act. For example, imagine that I am deciding which career to pursue, and after some investigation and deliberation, I find myself with two conflicting ambitions: I want to be a singer, but also want to be a gardener. Eventually, after much soul-­searching, I decide to become a gardener. Here my outweighed desire to become a singer cannot be classed as one of Nagel’s motivated desires, because I am not pursuing this goal. But at the same time, it is clearly not one of Nagel’s unmotivated desires, since this desire did not simply assail me but was rather produced by a process of reflection. Nagel’s view entails that all desires are either acted on or else unmotivated, and this is false. Let me summarize this section. Nagel aims to offer a view on which ODM is false, but nonetheless close to something true. He thereby aims to undercut its appeal. But his view faces two serious problems. First, he is forced to deny the truism that many actions are explained by desires, including desires that we reach after deliberation. To the extent that Nagel cannot agree with this, he hardly captures the attractive elements of ODM. Second, on Nagel’s view any desire that you fail to act on must be an unmotivated desire. But this is clearly false.

44  Desire and Motivation

2.4.3 Emotions It might seem that we can be motivated by emotions such as fear, as when my fear leads me to back away from a horse. But this thought can easily be accommodated by ODM. If fear involves aversive desires, then such desires explain the motivational capacity of fear. On the other hand, if it is possible to fear something without desiring to avoid it—perhaps having the relevant bodily responses, and/or conscious feelings, suffices—then we should ask whether fear can motivate in those cases where it does not involve desire. Since the most obvious answer is not, we can again be reassured that fear motivates only insofar as it involves a desire. To this extent, ODM is not threatened by the idea that our emotions can motivate us: they do so only to the extent that they involve desire. We can say a little more about the second option above, on which emotions are only contingently associated with desire. At a more general level, we might think that emotions primarily serve to make us especially responsive to certain features of our environments: fear is designed to make you more alert to dangers and quicker to respond to them, anger makes you more alert to wrongs and quicker to respond to them, and so on. In this way, emotions might make us more effective agents by making us more responsive to more important features of our environment. Emotions might do this in a variety of ways, such as by directing our attention, by preparing our bodies for certain actions, and by disposing us to form certain desires. If we think of emotions in this way, your fear of horses doesn’t necessarily involve any desire to back away, but nonetheless does strongly dispose you to form such a desire. So on this view, whilst ODM is true, our emotions play an important role in determining how we act by often affecting what we come to desire.16 I now turn to a second objection to ODM, which also appeals to the capacity of emotions to explain our actions. The objection above was undermined because it appealed to cases where emotions motivate us in a goal-­directed manner. That makes it plausible to suppose that they motivate via desire after all. But in other cases we seem to act from emotion but not in  a goal-­directed manner. For example, Rosalind Hursthouse discusses ‘arational’ actions, such as throwing an uncooperative tin opener on the ground, or rolling in the clothes of one’s dead wife (Hursthouse 1991, 58; see also Döring 2003). She claims that such actions constitute counterexamples to

16  I say more about emotion, especially in relation to desire-­as-­belief, in §7.1.

Objections to ODM  45 ODM. Such cases seem potentially more problematic for ODM, since, unlike the examples above, such actions are not goal-­directed in any way: they serve to express emotion rather than secure any outcome. That looks more obviously inconsistent with ODM, both in letter and in spirit. But we should be careful with these cases. Insofar as such actions are possible, they are plausibly not motivated at all: again, they are expressive acts, and not motivated.17 That is, Hursthouse’s argument focuses on whether we can act independently of desire. Perhaps that is true, but it doesn’t undermine ODM which says that we can only be motivated to act by desire. Insofar as some intentional actions are not motivated at all, they are not relevant to our assessment of ODM. (In a note, I give another reason for being relaxed about such expressive actions.18) In summary, we can act on our emotions in two different ways. When we act on them in a way that is goal-­directed, it is plausible that we act out of emotion only insofar as those emotions involve or produce desires. When we act on emotions in a way that is merely expressive, ODM should be understood as avoiding such counterexamples because it is a thesis about only motivated action, not all action. In what follows, I shall usually gloss over this later complication, and loosely talk as though desires explain all actions: expressive actions won’t feature again.

2.4.4 Intentions We clearly have various intentions (plans) for what we might do in the future: you might intend to go clubbing tonight, intend to buy a round of Tequila, and later, intend to never drink again. Our intentions clearly play an important role in our lives, since they explain many actions. They may also play other roles: famously, Michael Bratman argued that they play a crucial role in enabling us to coordinate our actions over time and with others (Bratman 1987). For 17 Perhaps equivalently: such actions are not done for motivating reasons (Ginet  1990, 130; Hursthouse 1991, 59–60). 18  Hursthouse sometimes hints that expressive actions also constitute a counterexample to the claim that all actions are explained by our normative beliefs (see e.g. Hursthouse 1991, 58–9; see also Döring  2003, 215). Let’s imagine that she is right, and her examples are counterexamples both to ODM and to the view that all actions are explained by our normative beliefs. Then I could simply modify my arguments in what follows so that they are restricted only to non-­expressive actions: since the relevant claims about desires and about normative beliefs go hand in hand, this would do nothing to undermine my overall argument for desire-­as-­belief, which will appeal to the very fact that desires and normative beliefs play the same role in motivation. Indeed, if anything, Hursthouse’s argument promises to support desire-­as-­belief: desire-­as-­belief would explain why the only actions not explained by desire are equally not explained by normative belief.

46  Desire and Motivation example, if I intend to take paternity leave, I can rearrange my other commitments around the fact that I’ll be away from work during that period, and I  can inform others of this intention so that they can also rearrange their plans around me. These roles of intention presuppose that intentions are not merely desires. After all, we can desire all sorts of things that we know we will not do, as when I idly fantasize about taking a holiday during term-­time: neither I nor others should rearrange our plans around this possibility even though I have some desire for it. What is distinctive about intentions is precisely that we commit, in some way, to acting in a certain manner, and it is this level of commitment that makes intentions useful for coordinating our actions over time and with others: they serve as fixed points which we plan around. The way in which intentions allow us to commit to courses of action also allows intentions to serve as the practical equivalent of memories, storing our decisions for later. This is useful not only for planning ahead, but also if we predict that our later reasoning may be substandard so that it is better to deliberate now (again, see Bratman 1987; also Holton 2009). Do intentions motivate us in a manner that undermines ODM? No. The obvious thing to say is that intending is a state that goes beyond desiring, but which is nonetheless always partly constituted by desire. And so we can agree that intentions are not merely desires, but also insist that they motivate only because they have a desire as one component. The idea that intentions motivate independently of desire would clearly be an overreaction to arguments which aim to show that intentions are not merely desires. We might instead think of intentions as desires+: states that involve desire but which also involve a degree of commitment and stability over time. We might slightly more precisely say that intending to v is desiring to v, where this desire (a) is strongly disposed to motivate you, (b) is relatively stable over time, and (c) is normally known to have these features, so that it is useful for facilitating planning for oneself and others. Roughly, we should think of intentions as desires that have been locked into the firing chamber, akin to beliefs that are settled and that are fresh in the mind, ready for use in inference. No doubt we could fuss about the exact way to formulate clauses (a), (b), and (c), and present a more detailed reductive theory of intention (see e.g. Ridge 1998; Sinhababu 2017, chap.  6), but so long as you allow that there is some way to describe what intentions do, the important point is just that we can define an intention by appending those conditions, whatever they are, onto a desire. Any such view enables us to agree that intentions can motivate but only because they involve desires, just as ODM says.

Summary  47

2.5 Summary In this chapter I have explained and defended ODM: ODM:  You can be motivated to act only by the combination of a desire [that p] and a belief [that this act will bring about p]. ODM says that our desires explain what we are motivated to do. It says that you are never motivated to do anything by any state of mind that is not a desire. But I also emphasized that ODM does not say that our desires always motivate us: it says that they are like an intermittent spring, which is the only source of water, but nonetheless one that is variable in output. I also showed how ODM can overcome various objections. But by far the strongest objection to ODM is that our normative beliefs can motivate us, and that this is inconsistent with ODM. In the next chapter, I will discuss this objection, and explain how our actions are affected by our normative beliefs, and how our desires and normative beliefs relate to one another. Desire-­as-­belief will make its promised return. I never did find out where that snail came from.

3 Normative Belief and Motivation The previous chapter defended ODM (Only Desires Motivate), the claim that we are motivated only by desires, combined with instrumental beliefs. This chapter makes use of this claim as part of my first argument for desire-­as-­ belief. §3.1 defends the claim that reasons beliefs can motivate us, and §3.2 combines this claim with ODM to give us the motivation argument. That argument is sometimes thought to favour non-­cognitivism rather than desire-­ as-­belief, but across §3.3 and §3.4 I argue that non-­cognitivism is implausible, and that desire-­as-­belief is the more appropriate inference to draw from the motivation argument.

3.1  Normative Beliefs Motivate We very often explain the things people do by appealing to their normative beliefs. For example, I might explain why you order a vegetarian pizza by appealing to your moral beliefs, I might explain why you vote the way you do by appealing to your political beliefs, and I might explain why you invest in your pension by appeal to your beliefs about what is in your interest. Such common-­sense thoughts suggest that normative beliefs have an important role in motivating our actions. Moreover, it seems like the very point of normative beliefs is to provide us with guidance about how to act: we deliberate about what to do by thinking about what we ought to do, and we give guidance to others using phrases like ‘you ought to . . .’. If normative beliefs didn’t play this action-­guiding role, it would be unclear why we have them at all: our interest in such matters is not idle curiosity, in the way that we might wonder about the air on Venus. We are interested in normative truths precisely because they help determine what we’ll do. Some claim that normative beliefs have no special power to motivate us, but instead only play the role of ordinary beliefs as stated by ODM (e.g. Brink  1989, chap.  3; Railton  1986, 170). On such views, normative beliefs motivate us only in combination with desires to do what we ought to do. On that view, normative beliefs are motivationally relevant only in exactly the Desire as Belief: A Study of Desire, Motivation, and Rationality. Alex Gregory, Oxford University Press. © Alex Gregory 2021. DOI: 10.1093/oso/9780198848172.003.0004

Normative Beliefs Motivate  49 same way any other beliefs are, in that you might, or might not, have desires regarding the objects of those beliefs. But plausibly, this understates the special action-­guiding role of normative beliefs. It is not as though we might meet an alien species who have firm views about what they ought to do, but who treat such questions as matters of purely theoretical interest, utterly irrelevant to any choice they might face (see also Dreier 1990, 9–14; Tresan 2009). To put the above points in a different way, when we think about the role that normative beliefs play in explaining actions, it is that they set us goals: they tell us what to aim for. Returning to the map analogy that I used to illustrate ODM, it seems that normative beliefs mark a point on a map to head for, rather than merely representing the layout of the terrain. In this respect, normative beliefs don’t merely provide information about how to pursue other goals that we independently have. This can be further corroborated by noting that normative beliefs seem to motivate only in combination with other beliefs that do merely provide information about how to pursue goals. Just as my desire that I win at snail racing only motivates me insofar as I have some ordinary non-­normative belief about how to bring that outcome about, so too my belief that I ought to win at snail racing only motivates me insofar as I have some ordinary non-­normative beliefs about how to bring that outcome about. In this way, normative beliefs seem to supply us with goals, and their motivational influence towards those goals is directed by independent beliefs about the best means to those ends. It’s easy to see how the above parallels between normative beliefs and desires might favour desire-­as-­belief. But let’s hold off on that inference for now. For now, let’s stick with the more modest thought that normative beliefs motivate in a manner unlike other beliefs, and that they motivate by combining with such ordinary beliefs.

3.1.1  Motivation Internalism Views on which normative beliefs have some special role in motivating us to act tend to get labelled as versions of ‘judgement internalism’. I find that label misleading, for minor reasons, and so instead use the label ‘motivation internalism’.1 But previous discussion of motivation internalism has sometimes been awry in three more serious ways. 1  ‘Judgement’ most naturally names a conscious mental act, rather than a persisting attitude which may be nonconscious. Moreover, ‘judgement’ sounds cognitive, and is therefore less neutral than intended on non-­cognitivism, below.

50  Normative Belief and Motivation First, many of those who discuss motivation internalism focus on moral motivation internalism: a view about the motivational powers of our moral beliefs (see e.g. Björklund et al.  2012; Smith  1994). But this is at best too narrow. I am interested in the motivational role that all of our normative beliefs play, whether beliefs about morality, about prudence, or about any other practical norm. Various normative beliefs play an important role in determining your actions, such as your beliefs about which pension you ought to invest in, or about whether you ought to do your shopping on the way home today, or about the order in which you ought to reply to your emails. Motivation internalism is best understood as making a claim about all of these beliefs, not just about moral beliefs. Another way to put this is that we should not isolate morality from other kinds of normativity as though it has nothing in common with them. Plausibly, the point of any norm is to guide us in appropriate ways, and to this extent all normative beliefs play the same kind of guiding role as one another. Moral norms might be interestingly unique in some respects—e.g. their connection to practices of blame—but not with respect to the role that such norms play in guiding our actions. Moral norms are very much like other norms in that they all employ concepts like ought and reason. Plausibly, those concepts are characterized in part by the role they play in guiding us, and to this extent we should think of moral beliefs as merely one instance of the broader class of beliefs that play a motivational role. Focusing solely on the motivational role of moral beliefs is artificially narrow, and thereby threatens to unnecessarily limit our remit and/or to distract us with idiosyncrasies of distinctively moral beliefs. So far I have suggested that focusing on moral beliefs wrongly narrows our discussion, since we are interested in the motivational role of normative beliefs in general, and not merely moral beliefs. But in fact my initial claim above was that focusing on moral beliefs is too narrow at best. I say this because perhaps someone could have various moral beliefs but have no corresponding normative beliefs: they might think that they ‘morally ought’ to do various things but think this has no bearing on what they really have reason to do (cf. Foot  1972). That is, we might believe that amoralists are possible, who think that they have no reason to do what morality requires of them (see e.g. Brink 1989, 46–50; Nichols 2002; Roskies 2003). Such people accept that they morally ought to do all sorts of things, but deny that this is of any normative significance to them. Plausibly, such figures would be utterly unmoved by their moral beliefs. As such, amoralists are sometimes treated as counterexamples to motivation internalism: that view is often understood as

Normative Beliefs Motivate  51 making a claim about the motivational power of moral beliefs, and amoralists will plausibly be left cold by their moral beliefs. But I need not worry about the possibility of amoralists. Perhaps such figures aren’t even possible: perhaps you don’t count as having the relevant moral beliefs unless you accept that the relevant considerations are normative.2 Then my claims about normative beliefs straightforwardly extend to moral beliefs. On the other hand, perhaps amoralists are possible. Plausibly, they would be left unmotivated by their moral beliefs. But that would not undermine motivation internalism, as I understand it. Instead, their lack of motivation would be positively predicted by my version of motivation internalism: surely, amoralists would be left cold by their moral beliefs precisely because they don’t think that morality is normative. So whilst discussion of motivation internalism sometimes treats it as a claim about moral beliefs, and focuses on amoralists as a potential counterexample to that claim, I will instead focus on motivation internalism as a claim about normative beliefs, and I thereby need not worry about the possible existence of amoralists.3 In short, motivation internalism is often taken to be a claim about the motivational role of moral beliefs. But I will instead think of it as a claim about the role of normative beliefs. Focusing on moral beliefs would be at best too narrow and at worst irrelevant for my purposes. I now turn to a second way in which some previous discussion of motivation internalism has been awry. Motivation internalism is often thought to posit a necessary connection between the presence of a normative belief and motivation (Björklund et al.  2012, 125; Gibbard  2003, 153; Hare  1952, 163–9). Such a view faces straightforward counterexamples involving akrasia (‘weakness of will’) (Shafer-­Landau 2000; Svavarsdóttir 1999, 176–83). For example, imagine a soldier who thinks she ought to shortly storm the opposing trenches. Perhaps as the time approaches, she gradually loses her resolve, and so her motivation to storm the trenches gradually dissipates until she has no motivation left at all. She might thereby lack the motivation to storm the trenches despite thinking that she ought to do so. 2  See e.g. Smith (1994, chap. 3) on ‘moral rationalism’. 3  Perhaps I should be worried by the possibility of ‘anormativists’—people who accept that they have reasons to do things, but treat that as motivationally irrelevant (cf. Archer 2020, sec. 3). However, it’s far from clear that such figures are possible—we might think that someone who made assertions along these lines would not sufficiently understand the word ‘reason’ to count as deploying that concept at all. A related option for me in response to anormativists would be to restrict motivation internalism, and desire-­as-­belief, so that they are views only about normative beliefs held with conceptual competence.

52  Normative Belief and Motivation And this lack of motivation hardly shows that she no longer really thinks she ought to storm the trenches: we can well imagine her feeling guilty for her inaction, and such feelings would be best explained by her continuing belief that she ought to act differently. More generally, it would be surprising if there were a necessary connection between normative belief and motivation. It is a familiar claim that mental states have no effects by themselves but only in the context of other states of mind (e.g. Block 1981, 11–12). Furthermore, it is a familiar claim that laws in (at least) the special sciences have exceptions: at best only the laws of fundamental physics are exceptionless. Each of these claims suggests that all states of mind have their effects only contingently: it is tempting to think that states of mind dispose us to do things, but that such dispositions will not become manifest under various circumstances. For these reasons, motivation internalism is surely too bold if it posits a necessary connection between normative belief and motivation. I will instead focus on the more modest claim that such beliefs dispose us to be motivated (cf. Broome 1997, 139; Dancy 1993, 22–3; Ridge 2014, 50). By formulating the view in this way, we render it consistent with examples like the soldier above: since dispositions need not always become manifest, it’s possible for people to believe they ought to do things but lack any motivation to do them. And by formulating the view in this way, we also render it consistent with the broader thought that mental states in general have their effects only contingently.4 And yet despite its modesty, this claim is not without content. It says that normative beliefs themselves have the capacity to motivate us. Ascribing a disposition to a state of mind tells us something about the nature of that state of mind—something about the role that it plays in our mental economy. It is a special feature of normative beliefs that they have this power to motivate us. They are in this respect different from other mental states which are not motivational, such as the belief that peanut butter is brown, or the visual perception of a carrot: those states of mind are not even capable of moving us to act. Normative beliefs are special states of mind in that they have a defeasible action-­guiding function: they can motivate us to do things, by setting goals for us to pursue. By understanding their motivational power in this defeasible way, we maintain the interest in normative beliefs whilst 4  Michael Smith reformulates motivation internalism in a different direction: ‘If an agent judges that it is right for her to φ in circumstances C, then either she is motivated to φ in C or she is prac­tic­ al­ly irrational’ (Smith 1994, 61; see also Korsgaard 1986, 15). But this is no longer a claim about the descriptive nature of a certain state of mind, but instead the statement of a rational norm governing that state. Such a normative claim might well be true (cf. Chapter  4), but we should also maintain some purely descriptive clam about the motivational effects of normative beliefs, as I do here.

The Motivation Argument  53 nonetheless avoiding traditional objections to motivation internalism that appeal to cases of akrasia.5 The third way in which my discussion diverges from some historical discussion of motivation internalism is that motivation internalism is often thought to make a claim about the motivational powers of either all normative beliefs, or else only about beliefs regarding what we ourselves ought to do. In the remainder of this chapter one task is precisely to adjudicate exactly which beliefs fall under the remit of motivation internalism. But this is a convenient point at which to make one relevant claim: the idea that we can be motivated only by beliefs about what we ought to do is too restrictive. If you think that you ought to do one thing, but also think that you have very strong reasons to do something else, it seems clear that you might be motivated to do the latter. This might well be irrational on your part, since you believe those reasons are outweighed. But it nonetheless seems clear that beliefs about reasons can motivate us to do things, and to that extent it is not only beliefs about what we ought to do that have motivational powers. This oversight is easy to explain: many have assumed that motivation internalism makes the claim that some normative beliefs necessarily motivate (above). But it would be very implausible to claim that beliefs about reasons necessarily motivate: beliefs about very strongly outweighed reasons, for example, may not motivate. But once we correct this assumption and think of motivation internalism as making a claim about the defeasible powers of the relevant states of mind, rather than their necessary effects, it becomes more plausible to think that motivation internalism applies not only to beliefs about what we ought to do but also more generally to beliefs about what we have reason to do. To summarize, normative beliefs can motivate us to act. This claim is highly plausible. But it should not be confused with the claim that moral beliefs can motivate us to act (which is either unduly narrow or possibly just plain false), with the claim that normative beliefs necessarily motivate us to act (which is false), or with the claim that beliefs about what we ought to do can motivate us to act (which is unduly narrow).

3.2  The Motivation Argument In the previous section, I defended a modest version of motivation internalism: 5  I return to akrasia again in Chapter 5.

54  Normative Belief and Motivation Motivation internalism:  Your normative beliefs can motivate you to act, by combining with relevant non-­normative beliefs about how to achieve the rele­vant ends. And in Chapter 2, I defended ODM: ODM:  You can be motivated to act only by the combination of a desire [that p] and a belief [that this act will bring about p]. It is easy to see that these two claims together suggest that normative beliefs are desires. ODM tells us that you can be motivated only by combinations of desire and beliefs about how to achieve those ends, and motivation internalism tells us that normative beliefs motivate by combining with beliefs about how to achieve those ends. These claims can be reconciled only by claiming that normative beliefs are desires. Let’s write out this argument—the motivation argument—more succinctly as follows: Motivation internalism, roughly:  Normative beliefs can motivate. ODM, roughly:  Only desires motivate. So, C:  Normative beliefs are desires. This way of expressing the argument isn’t perfectly precise, since it ignores the role that ordinary beliefs play in combining with normative beliefs, and with desires, to enable them to motivate. But it nonetheless captures the basic line of reasoning well enough, and this simplification will make the following discussion easier to follow. The motivation argument, as expressed above, is hard to reject. When discussing this argument, objectors sometimes reject the second premise, ODM, which I defended in Chapter 2 (e.g. Shafer-­Landau 2003, chap. 5). As I said above, rejecting ODM is costly. It is our best model of human action, a model that fits with ordinary practices of explaining actions, and which successfully explains a wide variety of events—all human action!—by appeal to just two states of mind, combining in systematic ways. ODM is highly plaus­ible, and we should not reject it unless we have to. Objectors to the mo­tiv­ation argument sometimes instead object to the first premise, motivation internalism (e.g. Brink 1989, 44–5). But these objections focus on the unnecessarily bold version of motivation internalism according to which certain normative beliefs necessarily motivate. But if we reformulate motivation internalism so that it permits that normative beliefs can fail to move us, it is much harder to deny.

Against Non-­C ognitivism 

55

I have argued that we should accept motivation internalism, which c­ aptures the natural thought that normative beliefs have a special role to play in motivating us to act. And I have argued that we should accept ODM. As a result, it seems that we should accept the conclusion of the argument above. So far as this reasoning goes, it seems that we might accept either desire-­as-­belief, or else its rival, non-­cognitivism—both accept some kind of identity claim between normative beliefs and desires. But in the remainder of this chapter I show that desire-­as-­belief is the more appropriate conclusion to draw from the motivation argument above.

3.3  Against Non-­Cognitivism Non-­cognitivism says that normative beliefs are really just desires. Many take the motivation argument to be an argument for non-­cognitivism (Blackburn  1984, 187–9; 1998, 70; Gibbard  2003, 11–13; Hare  1952, 1; Stevenson 1937, 16), and you might even think that Hume introduced ODM precisely in order to argue for non-­cognitivism (Hume T3.1.1.6). But I will argue that non-­cognitivism sits poorly with the motivation argument, and that desire-­ as-­ belief fits much better. Here is the basic objection to non-­cognitivism: if I come to think that I ought to go truffle hunting, we might think that this attitude involves a kind of inclination—if I really think this, I must be somewhat inclined to go. To this extent, motivation internalism seems plausible. But in contrast, if I think that you ought to go truffle hunting, it is far less clear that this involves any kind of inclination on my part: I might really think that you ought to go, but need not be at all in favour of your doing so (indeed, perhaps I would much prefer you to act wrongly, and join me in the pub). Cases like this suggest that normative beliefs about others are a matter of mere recognition, not inclination. In turn, this casts doubt on any universalized formulation of motivation internalism. And in turn again, this casts doubt on non-­cognitivism, which claims that normative beliefs are all desires. I will defend these claims in the following sections.

3.3.1  Non-­Cognitivism and Conativism To explain this in greater detail, I first need to say a little more about how I understand non-­cognitivism.

56  Normative Belief and Motivation The label ‘non-­cognitivism’ indicates the central commitment of non-­ cognitivists: that when we adopt some normative point of view, that is not fundamentally a cognitive state—it is not belief-­like, and is incapable of truth or falsity. That is, non-­cognitivists say that our points of view on normativity don’t aim to represent normative aspects of the world around us. This non-­ cognitivist view runs against ordinary ways of speaking, and as a result a small piece of jargon will be useful. I have made frequent reference to our ‘normative beliefs’. But non-­cognitivists deny that we have any of those, strictly speaking. So let us instead talk about our normative ‘attitudes’. As I see it, this is just another label for our normative beliefs. But using the label allows me to discuss non-­cognitivism in more neutral terms, and in turn to reject the view on neutral ground. Despite the label, non-­cognitivists don’t merely make negative claims about what normative attitudes are not, but also make positive claims about what normative attitudes are. In particular, non-­cognitivists tend to accept: Conativism:  All normative attitudes are desires. Non-­cognitivists should accept conativism. Why? If you are moved by the motivation argument to accept non-­cognitivism, you should be moved to accept the conativist kind of non-­cognitivism, since the motivation argument aims to establish the conclusion that normative attitudes are desires, not any other non-­cognitive state of mind. That is, a non-­cognitivist who rejected conativism would have to reject either ODM or else reject motivation internalism. That would be a cost since these views are so plausible, and doubly a cost insofar as the rejection of either of these would undercut the non-­cognitivist’s best argument against those who claim that all normative attitudes are just regular beliefs. In short, non-­cognitivists should endorse not only the negative claim made by non-­cognitivism, but also the positive claim made by conativism. In the remainder of this chapter, I will most directly argue that desire-­as-­belief is superior to conativism. But I thereby demonstrate the broader conclusion that desire-­as-­belief is superior to non-­cognitivism. My objection to conativism will also extend to nearby views in two further important ways.6 6  A third: some non-­cognitivists claim that normative attitudes are not desires, strictly speaking, but instead desire-like states, with the same motivational profile as desires (for this combination of claims, see Blackburn  1998, 9–10, 13–14, 66; and plausibly Gibbard  1990, 55–6, 75). That is close enough for my purposes. When I object to conativism, it is this aspect of it that I focus on, and it is therefore unimportant whether some non-­cognitivist denies that normative attitudes are desires for reasons that are independent of their motivational profile.

Against Non-­C ognitivism 

57

First, conativism is a view about the nature of certain mental states: normative attitudes. By itself, conativism says nothing about normative language. Most conativists defend expressivism, which combines conativism about normative attitudes with a further claim about normative language: expressivists say that the meaning of normative utterances is determined by the state of mind they express. In what follows I continue to focus on the nature of various states of mind rather than the meanings of utterances, but my objection nonetheless undermines expressivism because expressivists incorporate conativism into their view. Second, many non-­cognitivists are quasi-­realists. Surprisingly, quasi-­realist non-­cognitivists accept that normative attitudes are beliefs (and are truth-­ apt). So in what sense are quasi-­realists non-­cognitivists at all? The answer to this question is somewhat unclear (Dreier  2004), but the basic quasi-­realist idea is that normative attitudes are beliefs, but only in a comparatively thin sense of ‘belief ’ (again, see Dreier  2004, especially 26–9). Quasi-­ realist conativists might therefore claim that normative attitudes are both desires and beliefs, though the latter only in some minimalist sense. I need not worry about the details of the view beyond this: my objection to conativism applies equally to quasi-­ realist conativism, since I object to the claim that all normative attitudes are desires, not the claim that they are not beliefs. In short, in the remainder of this chapter I focus only on conativism, but since conativism is part of any plausible non-­cognitivist theory, my arguments undermine non-­cognitivism more broadly.

3.3.2  Other-­Regarding Attitudes To evaluate conativism, we first need to clarify the motivation argument. To remind you, here it is again: Motivation internalism, roughly:   Normative attitudes can motivate. ODM, roughly:  Only desires motivate. So, C:  Normative attitudes are desires. In this formulation of the argument, I have left the scope of motivation internalism and (C) unstated: is this an argument about all normative attitudes, or just some of them? Conativism is the view that all normative attitudes are desires, and so is most directly supported by the argument if these claims are universalized. My key claims in what follows are (a) that this way of

58  Normative Belief and Motivation formulating the argument renders motivational internalism false, and (b) that the relevant counterexamples also undermine conativism itself. I will make these points by focusing on other-­regarding normative attitudes: attitudes about what other people ought to do. Since ‘other-­ regarding normative attitudes’ is quite a mouthful, let’s take the ‘normative’ bit as a given, and just label them ‘other-­regarding attitudes’. I claim that other-­regarding attitudes have no motivational powers of their own. For example, if Tanya thinks that Mo ought to sing Happy Birthday to his son, it seems that this attitude has no power to motivate Tanya to do anything. It is an attitude about what Mo should be doing, and by itself has no bearing at all on what Tanya herself will do. When Tanya thinks that Mo ought to do something, that is a matter of recognition, not inclination. As a result, motivation internalism should be understood as making a claim about just first-­personal normative attitudes, not about all normative attitudes. This point can be obscured by the fact that other-­regarding attitudes can play a role in inference to further normative attitudes which do have motivational powers. For example, if Tanya also thinks that she ought to assist others in doing their duty, she might infer that she ought to help Mo sing, and this attitude might motivate her. But here the motivational power is injected only by the addition of a further first-­personal normative attitude: without it, the original other-­regarding attitude is motivationally inert. So this kind of example is consistent with other-­regarding attitudes having no motivational powers of their own. Similar reasoning undermines other possible reasons for thinking that other-­regarding attitudes can motivate us. For example, you might say that if Tanya thinks that Mo ought to keep his promises, but finds that he doesn’t, this might motivate her to avoid him. Or you might say that if Tanya thinks that Mo doesn’t invest his money as he ought, this might motivate her to avoid lending him money. But plausibly what really motivates Tanya in the first case is the attitude that she ought to avoid people who don’t keep their promises, and what really motivates Tanya in the second case is the attitude that she ought not lend her money to people who are bad with money. So, again, it’s doubtful that her other-­regarding attitudes have any motivational powers of their own. The claim that other-­regarding attitudes are motivationally impotent is starkest in cases where one person thinks that another ought to do something that conflicts with the goals and values of the first. Most extremely, imagine that Sara is a consistent egoist, who thinks that everyone ought to only

Against Non-­C ognitivism 

59

promote their own wellbeing. Sara might thereby think that Kumar, in his dealings with her, ought to use and abuse her. This attitude, it seems clear, would have no motivational power over Sara. The point can be seen equally well in less extreme cases: in a prisoner’s dilemma, Ursula might think that her partner ought to rat, but it’s doubtful that this alone can motivate Ursula to do anything. Or for a final example, I might think that you ought to save your mother at the expense of mine, but not be remotely inclined to help you do your duty. In general, we will find stark illustrations of the motivational impotence of other-­regarding attitudes whenever the relevant norms are believed to be agent-­ relative, as many prudential and moral norms are believed to be. In such cases, someone can think that some norm applies to another but not to themselves, and as such be left cold by their recognition of that norm (see also Chrisman 2015, 180–1; Ridge 2018; Scanlon 2014, 58–9). Someone might reply that these cases show only that other-­regarding attitudes sometimes fail to motivate, and that is consistent with motivation internalism as I defined it: I said that motivation internalism is most plausibly understood as the view that normative attitudes can motivate us, not the view that they necessarily do so. But the worry is that in cases like those above it is not even plausible that the relevant attitudes could motivate us: other-­ regarding attitudes can’t motivate us under any circumstances. It just doesn’t seem intelligible for someone to act in some way because they recognize that someone else ought to do something. Moreover, the most obvious cases where our normative attitudes fail to motivate us are when we are being weak-­willed. But clearly, that is not the issue in the cases above: it is not as though egoists, or prisoners in the prisoner’s dilemma, are weak-­willed when they refuse to help others do what they ought to do. Indeed, if we are rationally required not to be weak-­willed, then the proposal would imply that (e.g.) the prisoner who doesn’t try to persuade her partner to rat is being irrational, and this is highly implausible. Nothing need be irrational or even abnormal about someone who thinks that they have no reason to comply with a norm that they think governs someone else but not themselves. In light of this, it seems reasonable to deny that other-­regarding attitudes have motivational powers, and in turn reasonable to think that motivation internalism is best formulated as a claim about only first-­personal normative attitudes. And so, the proper conclusion of the motivation argument is equally a claim about only first-­ personal normative attitudes. It follows that conativism goes well beyond the conclusion of the motivation argument, since it commits us to the view that all normative attitudes are desires, but the

60  Normative Belief and Motivation argument only demonstrates that this is true of first-­personal normative attitudes. But worse, these examples cast doubt not only on the inference from the motivation argument to conativism, but also directly on conativism itself (cf. Dreier 1996). Indeed, this would straightforwardly follow if we accepted the popular theory which analyses desires precisely in terms of their capacity to motivate (Smith 1994, 92–129; Stalnaker 1987, 15; see also §4.2). It is true that desires motivate only when combined with suitable means-­ends beliefs (§2.1), but this does nothing to help the conativist: it’s clear that we do often have the relevant means-­ends beliefs and still lack the corresponding mo­tiv­ ations. For example, Sara might think that Kumar ought to use and abuse her, and believe that she can get him to do so by anonymously sending him the works of Ayn Rand, but still not have any motivation to do that. It seems to follow that her attitude that Kumar ought to use and abuse her is not a desire that he use and abuse her, and it is not clear what other desire it might be. Since other-­regarding attitudes cannot motivate, there is a simple but forceful case for thinking that they cannot be desires. Moreover, it is just independently implausible that other-­regarding attitudes bear any necessary connection to desire. If Tanya thinks that Mo ought to sing Happy Birthday to his son, that seems to leave completely open whether she altruistically hopes he does, spitefully hopes he doesn’t, or (more likely) just doesn’t care either way about what she sees as Mo’s business. And again more starkly, a committed egoist surely need not desire that others do what they ought to do, and someone in a prisoner’s dilemma can recognize that their opponent ought to rat without desiring that they do so. Even independently of anything we say about motivation, it is not plausible that we necessarily desire that others do what they ought to do. In short, when we reflect on other-­regarding attitudes, it seems as though they serve as counterexamples to the claim that all normative attitudes have motivational powers, and as counterexamples to the claim that all normative attitudes are desires. Reflection on such attitudes thereby (a) encourages us to formulate motivation internalism as a claim about only first-­personal normative attitudes, (b) casts doubt on the motivation argument if it is supposed to favour conativism, and (c) casts doubt on conativism itself. When we adopt attitudes about what others ought to be doing, that seems to be a matter of mere recognition, and need not involve any inclination on our own part. In the following section, I rebut two possible replies open to the conativist.

Against Non-­C ognitivism 

61

3.3.3  Objections and Replies We might think that certain reactive attitudes such as blame are important to morality (e.g. Blackburn  1998, 8–14; Gibbard  1990, 41–5). This might encourage the conativist to claim that other-­regarding attitudes are desires after all: they might say that other-­regarding attitudes are desires to respond to others in certain ways. In what follows, I focus on our desires to criticize others for failing to act as they ought, but the relevant arguments extend in obvious ways to nearby alternatives such as our desires to shun or to punish wrongdoing, or to praise or reward virtue. It’s crucial to note that the non-­cognitivist pursuing this line would need to defend the view that other-­regarding attitudes are literally identical to desires to criticize others, and not merely defend a weaker view, such as the view that other-­regarding attitudes dispose us to have such desires. The conativist says that every normative attitude is a desire, and so the present suggestion must be that other-­regarding attitudes are desires to criticize. Once we recognize that the proposal must come in this bold form, its plausi­bil­ity plummets. Clearly, there is some connection between normative attitudes and our inclinations to criticize, but to save their view the conativist must maintain that for every other-­regarding attitude, there is a way of describing it as a desire, and that is too far-­fetched. For example, you might think some act to be wrong but blameless, and under those circumstances it seems that you might adopt an other-­regarding attitude but have no corresponding desire to criticize (cf. D’Arms and Jacobson  1994; Nichols  2004, chap.  4; Ridge  2014, 142–3). Moreover, even if there were some necessary connection between moral attitudes and desires to criticize, it is still not plausible that any similar necessary connection obtains between other, nonmoral, normative attitudes and desires to criticize. Perhaps morality is a kind of social phenomenon, and so it might look initially plausible to claim that moral attitudes are accompanied by desires to criticize. But no similar claim seems plausible for attitudes about others’ prudence, rationality, or aesthetic choices. Take our prudential attitudes. Imagine that Jane thinks that prudence requires Jeff to buy new running trainers. It is doubtful that she must thereby want him to buy those t­ rainers: she might adopt this attitude and yet not really care whether he does what he ought. If she doesn’t care whether Jeff does what he ought to do, it seems she might equally not care what happens to Jeff if he doesn’t do what he ought to do (again, cf. Ridge 2014, 143). If she doesn’t care much about Jeff

62  Normative Belief and Motivation at all, she might care neither whether he is prudent nor whether he is chastised when he isn’t.7 Note that I need not claim that such attitudes are common. The point is simply that it is not a necessary condition on thinking that someone else ought to v that you desire to criticize them if they don’t. If we find some alien culture where people have no concept of criticism, this wouldn’t conclusively demonstrate that they have no normative concepts at all. Consistent utilitarians might have various views about what people ought to do, but have no corresponding desires to criticize anyone for wrongdoing, because they consider it counterproductive. Such possibilities seem perfectly coherent, but would be impossible if other-­regarding attitudes were identical to desires to criticize others. In short, it may be plausible that some normative attitudes have some connection to our desires to criticize others. But it is not plausible that all other-­regarding attitudes are desires to criticize others, and this undermines the idea that conativists can understand other-­regarding attitudes as such desires. I now turn to a second conativist reply to my objection, as presented primarily by Allan Gibbard (2003, 49–53; see also 2012, 174–7; Ridge 2014, 19, 177–8).8 (Note: the discussion of this reply is relatively long, gets relatively involved, and has no further bearing on the rest of the book. I won’t be offended if you skip it.) On Gibbard’s view, other-­regarding attitudes are desires to do certain things if you were in the other person’s place. That is, Gibbard claims that Tanya’s attitude [that Mo ought to sing] is a conditional desire [to sing if she were in Mo’s place]. Such a view rightly permits that Tanya might think that Mo ought to sing but not want him to (at least, not unconditionally). But such a view also nonetheless treats Tanya’s attitude as a desire, as conativists must. In this way, Gibbard’s view might seem to give the conativist everything they need. We should distinguish two more precise ways we might develop Gibbard’s proposal. First, the simpler option:

7  In his (2017), Brian McElwee defends the claim that kinds of normativity other than moral normativity—such as prudential and aesthetic normativity—are associated with various practices of interpersonal criticism. But even if this is right, McElwee claims that the relevant criticisms apply only when agents fail to do what they ‘must’ do, where ‘must’ is a unique normative concept, which contrasts with other concepts such as ‘ought’ and ‘reason’. To that extent, even McElwee will agree that very many normative beliefs have no straightforward connection to interpersonal criticism. 8  Gibbard treats normative attitudes as plans rather than desires, but as I said in footnote 6, this small kind of difference is unimportant for our purposes, since Gibbard nonetheless treats plans as being desire-­like in the way that they motivate. For simplicity, I will continue to talk in terms of desire.

Against Non-­C ognitivism 

63

Gibbard-­simple:  When Tanya thinks [that Mo ought to sing], that consists in Tanya’s conditionally desiring [to sing if she, Tanya, were in Mo’s circumstances].9 Second, the more sophisticated option: Gibbard-­sophisticated:  When Tanya thinks [that Mo ought to sing], that consists in Tanya’s conditionally desiring [to sing if she were Mo, and in Mo’s circumstances]. The difference between the proposals is that the sophisticated proposal, unlike the simple, understands other-­regarding attitudes to consist in desires for circumstances where we have different identities (haecceities) than we in fact have. So far as I can tell, Gibbard himself endorses the second option (Gibbard 2003, 50–1), but for completeness I will object to both, in turn. So first, there is Gibbard-­simple. Gibbard-­simple is highly counterintuitive. It is counterintuitive to say that when Tanya adopts a normative attitude about Mo, she is adopting a conditional desire for the eventuality that she end up in his circumstances. The view is all the more surprising once we remember that ‘circumstances’ here has to include not only Mo’s external environment, but also anything that might be relevant to what he ought to do: his ignorance, his character traits, his emotions, and so on (Gibbard 2003, 50–1; 2012, 174–5). The suggestion had better be that when Tanya thinks that Mo ought to sing, this is a desire of hers to sing if she had all of his properties. But that is a desire about a very strange counterfactual possibility, and a correspondingly strange desire to have. Moreover, it seems to have the wrong object. We were trying to analyse other-­regarding attitudes, and it is surprising that they turn out to be attitudes about ourselves under different circumstances, and not really about other people at all. To consolidate these doubts, think about how children adopt normative attitudes about others: it is highly doubtful that they do so by thinking about various highly remote possibilities. If a young boy thinks that girls aren’t supposed to have short hair, they do not come to that thought by imagining 9  This sort of view may also have been held by Hare (see e.g. Hare 1981, chap. 7). But matters are not so clear because Hare often talks about what a person is ‘committed to’, where this more often sounds like a normative claim rather than a descriptive one. Gibbard-­simple says that other-­regarding attitudes literally are desires of the relevant sort, and Hare may have meant to endorse only the thesis that other-­regarding attitudes rationally require desires of the relevant sort. I say nothing here against this latter thesis, which might well be true, but fails to provide the conativist with what they need. See also the discussion of Supervenience, below.

64  Normative Belief and Motivation themselves as a girl. Perhaps it might be conceded that children sometimes adopt normative attitudes about others by considering the relevant counterfactuals, but it is highly doubtful that they always do so, especially when the relevant counterfactual possibilities are (or are seen to be) very remote. Indeed, we seem to appeal to such counterfactuals precisely to teach children facts they wouldn’t otherwise accept: this practice presupposes that their present normative attitudes float free of their counterfactual thoughts about themselves. So the first concern is that Gibbard-­simple is counterintuitive and to that extent ad hoc. Gibbard-­simple also faces a second objection. Gibbard-­simple analyses Tanya’s attitude [that Mo ought to sing] as Tanya’s desire [to sing if she were in Mo’s circumstances]. Presumably, Gibbard-­simple would also have us analyse Tanya’s more explicitly conditional attitude [that she ought to sing if she were in Mo’s circumstances] as Tanya’s desire [to sing if she were in Mo’s circumstances]. Since these two attitudes receive the same analysis, Gibbard-­ simple forces us to identify them. That is, according to Gibbard-­simple, Tanya’s attitude [that Mo ought to sing] is just the very same as Tanya’s attitude [that she ought to sing if she were in Mo’s circumstances]. But these two attitudes are distinct. For example, Tanya might have concluded that Mo ought to sing without having extended the conclusion to her own case. The reverse is also possible: Tanya might have made a plan for herself in Mo’s circumstances without having actually considered what Mo himself ought to do. By analysing other-­regarding attitudes as conditional desires, we lose the ability to give an independent analysis of genuinely conditional normative attitudes. Might Gibbard reply that other-­ regarding attitudes and genuinely conditional normative attitudes are in fact the same state of mind, and claim that this truth is merely non-­obvious to us? But that reply suggests that there are no inferences between these states of mind, and that is implausible. Plausibly, we can sometimes get people to change their minds about how they judge others precisely by having them consider the relevant counterfactual claims about themselves, and vice versa (‘How would you like it if . . .?’). This would not be possible if the relevant states of mind were literally identical. On these issues, Gibbard-­ simple gains illusory plausibility because of Supervenience, the claim (roughly) that the same norms apply to all people and vary only insofar as their circumstances differ. Given Supervenience, if Mo ought to sing, then Tanya ought to sing if she were in Mo’s exact circumstances. But even if Supervenience is true, that doesn’t tell us much about Tanya’s attitudes. Tanya might fail to accept Supervenience, or more likely, she

Against Non-­C ognitivism 

65

might fail to accept every single implication of that truth. So Supervenience does not show that Tanya’s thinking that Mo ought to sing is the very same thing as Tanya’s thinking that she ought to sing if she were in Mo’s circumstances. Some claim that Supervenience is not a metaphysical truth, but instead a conceptual one (e.g. Blackburn  1993). But even this will not help. Even if Tanya’s attitudes embody failures to accept conceptual truths, this is no bar to her having those attitudes (cf. Williamson  2008, 73–133). Tanya might be conceptually confused and deny Supervenience. Or again, more likely, she might accept Supervenience but fail to accept various truths that are entailed by combining Supervenience with other beliefs that she holds—she might fail to combine her very abstract belief in Supervenience with her attitude that Mo ought to sing. It is surely possible to fail to endorse every implication of a conceptual truth one recognizes. In short, Gibbard-­simple is counterintuitive, and moreover it collapses the distinction between Tanya’s thinking [that Mo ought to sing] and Tanya’s thinking [that she ought to sing if she were in Mo’s position]. Those attitudes are distinct, and Supervenience does not show otherwise: it is a claim about the world, not our mental states. I now turn to Gibbard-­sophisticated. To remind you, it says: Gibbard-­sophisticated:  When Tanya thinks [that Mo ought to sing], that consists in Tanya’s conditionally desiring [to sing if she were Mo, and in Mo’s circumstances]. This differs from Gibbard-­simple. It analyses other-­regarding attitudes as desires for circumstances where your identity is different, not just your circumstances. But it’s not obvious that Gibbard-­sophisticated really improves on Gibbard-­simple. It does allow us to distinguish Tanya’s attitude [that Mo ought to sing] and Tanya’s conditional attitude [that she ought to sing if she were in Mo’s circumstances]. Gibbard-­sophisticated entails that the former but not the latter consists in a desire that is conditional on Tanya’s being Mo. But this is not obviously progress, since we might now worry that Gibbard-­ sophisticated fails to distinguish a different pair of attitudes: Tanya’s attitude [that Mo ought to sing] and Tanya’s attitude [that she ought to sing if she were Mo and in Mo’s circumstances]. Gibbard-­sophisticated will presumably give these two attitudes the same analysis, and this might seem mistaken for just the same reasons as those given above. Again, by analysing other-­regarding attitudes as conditional desires, we lose the ability to give an independent analysis of genuinely conditional normative attitudes.

66  Normative Belief and Motivation Perhaps an objector will say that here it is less clear that these two attitudes really are distinct. But whatever plausibility Gibbard-­sophisticated gains here, it loses with respect to the first worry above. To whatever extent Gibbard-­simple was counterintuitive, Gibbard-­ sophisticated is worse. Gibbard-­ sophisticated says that people who adopt attitudes about what other people ought to be doing are adopting desires for circumstances where their identity differs. It is highly counterintuitive to suppose that we have many conditional desires for such impossible circumstances: even if such conditional desires are possible, it does not seem that they are commonly occurring parts of our mental lives. And again, it is deeply implausible that we learn to adopt other-­regarding attitudes by learning to think about how our own identity and circumstances might differ: it is not clear that the average child is even capable of thinking about such matters. I conclude that Gibbard’s strategy is at best ad hoc and counterintuitive, and at worst inconsistent with clear distinctions between different normative attitudes. I also argued above that conativists should not try to rescue their view by appeal to the connection between normative attitudes and desires to criticize. With no other obvious option on the table, I conclude that the objection in §3.3.2 stands, and conativism is implausible because we should not identify other-­regarding attitudes with desires.

3.4  In Favour of Desire-­as-­Belief Where are we? Back in §3.2 I described the motivation argument. Its premises are ODM, and motivation internalism: the claims that we are motivated only by desire, and that normative attitudes can motivate us. Since each of these claims is highly plausible, we must accept that some normative attitudes are desires. Conativists accept that claim in a very bold form: they claim that all normative attitudes are desires. But it seems that other-­regarding attitudes, at least, are not. In light of what I have said about other-­regarding attitudes, we might formulate the motivation argument in a more restricted form, as follows: Motivation internalism, roughly:  First-­personal normative attitudes can motivate. ODM, roughly:  Only desires motivate. So, C:  First-­personal normative attitudes are desires. Despite qualifying the argument in this way, it is still significant. For example, it rules out any view on which no normative attitudes are desires. Many

In Favour of Desire-­a s-­B elief  67 standard cognitivist views are of this kind: they claim that all normative attitudes are beliefs, and that beliefs and desires are distinct states of mind (e.g. Boyd  1988; Brink  1989, chap.  3; Railton  1986). We should reject such views: they either imply that normative attitudes motivate only as regular beliefs do, which is implausible given the special action-­guiding role of such attitudes (§3.1), or else imply that ODM is false, which is a cost given just how persuasive and powerful that theory is (Chapter 2). To this extent, this restricted motivation argument rules out standard cognitivist views on which normative attitudes are beliefs and not desires. My arguments above (§3.3) ruled out conativist views on which normative attitudes are all desires. What we need is a theory on which some normative attitudes are desires, and some are not. That theory is desire-­as-­belief: Desire-­as-­belief:  To desire [that p] just is to believe [that you have reason to bring about p]. Desire-­as-­belief says precisely that some normative attitudes are desires, and that some are not. It says that all and only reasons beliefs are desires, and that all other normative attitudes are not desires. It identifies desires with only reasons beliefs—which, to remind you, are beliefs about what you yourself have reason to do (§1.2). It thereby says that beliefs about others’ reasons are not desires. This is the first of my central arguments for desire-­as-­belief: it is the perfect theory with respect to the motivational powers of normative attitudes. It is better than standard cognitivist theories because it explains how some normative attitudes can motivate us, and better than conativist theories because it explains how other normative attitudes cannot mo­tiv­ate us.10 It might be helpful to illustrate the three candidate views as Euler diagrams. See Figure 3.1. The first Euler diagram represents standard cognitivism. On that view, there is an exclusive distinction between beliefs and desires, and all normative attitudes are a subset of our beliefs. The motivation argument undermines this view, since it shows that some normative attitudes are desires. The second Euler diagram represents conativism. On that view, there is an exclusive distinction between beliefs and desires, and all normative attitudes are a

10  What about beliefs about what we ought to do? I claim that such beliefs motivate only insofar as they are partially constituted by relevant reasons beliefs: for example, perhaps such beliefs are really just beliefs about what we have decisive reason to do or most reason to do. See also §1.2.

68  Normative Belief and Motivation Belief

Desire Standard Cognitivism

Normative Attitude

Belief

Desire Normative Attitude

Belief

Normative Attitude

Desire

Conativism

Desire-as-belief

Figure 3.1  Three views in metaethics.

subset of our desires.11 Other-­regarding attitudes undermine this view, since they are not plausibly analysed as desires. The third Euler diagram represents desire-­as-­belief. According to desire-­ as-­belief, there is no exclusive distinction between beliefs and desires. Rather, our desires are a subset of our beliefs. In particular, they are beliefs with a particular normative content. Since these beliefs are desires, desire-­as-­belief fits well with the motivation argument: these beliefs can motivate even though only desires can motivate. But not just any belief with a normative content is a desire; that is how the view permits that other-­regarding attitudes are not desires. Let me summarize. Desire-­as-­belief identifies our desires and our reasons beliefs. This view is attractive because it allows us to accept ODM, which attractively explains all action in an elegant manner, as stemming from our desires, at the same time as accepting motivation internalism, which attractively claims that some normative attitudes can motivate us. According 11  Quasi-­realist conativists might say that the distinction between beliefs and desires is not exclusive, because normative attitudes are both—they are beliefs in some minimalist sense (see §3.3.1). But this makes no important difference here, since such a view is still committed to the claim that all normative attitudes are desires.

In Favour of Desire-­a s-­B elief  69 to desire-­as-­belief, these claims are compatible because desires and reasons beliefs are the same state of mind. That is, desire-­as-­belief allows us to maintain the dominant theory about how all human action is explained— ODM—whilst also permitting space for our normative views to motivate us, as they surely do. Desire-­as-­belief has all these attractions without committing us to the unattractive claim that all normative attitudes can motivate, and are desires, as conativists must say. In these respects, desire-­as-­belief is highly attractive, and uniquely so.

3.4.1  Extending the Argument We can approach our overall topic from another direction. Conativism and desire-­ as-­ belief might look like similar theories: conativism reduces normative attitudes to desires, and desire-­ as-­ belief reduces desires to normative attitudes. So the two views might seem similar, in that both identify normative attitudes and desires with one another. One issue that divides the two views is whether both normative attitudes and desires are beliefs, as desire-­as-­belief says, or whether neither normative attitudes nor desires are beliefs, as conativism says. This issue may be difficult to resolve, especially once conativists adopt quasi-­realism and claim that normative attitudes are beliefs, in some minimalist sense (again, see Dreier  2004). But there is a further and clearer contrast between the two views, and that is the scope of the identity that they posit. Conativism identifies all normative attitudes with desires, whereas desire-­as-­belief identifies only some normative attitudes with desires. In this respect, desire-­ as-­ belief clearly has the upper hand. For example, one central difference between desire-­as-­belief and conativism is that desire-­as-­belief does not treat other-­regarding attitudes as desires. It thereby permits that such attitudes can’t motivate, whilst nonetheless retaining the central attraction of conativism: it explains the motivational power of first-­personal normative attitudes. Once we see the issue in this way, it is tempting to think that the relevant point will generalize beyond other-­ regarding attitudes (cf. Finlay  2014, 130–4). Whereas conativism says that every normative attitude is a desire, desire-­as-­belief says that only reasons beliefs are desires. So these two theories make different predictions about all normative attitudes which are not reasons beliefs: desire-­as-­belief will say that none of those other normative attitudes are desires, and conativism will say that they all are. In all such cases, desire-­ as-­belief looks more plausible. For example, conativism has to claim that

70  Normative Belief and Motivation attitudes about epistemic normativity are desires of some kind, whereas desire-­as-­belief need not, since it identifies only beliefs about reasons for action with desires. On the face of it, this favours desire-­as-­belief: thinking that the evidence justifies believing p is not the same as wanting to believe p. Or for another example, conativism implies that attitudes about goodness correspond to desires even when they run free of attitudes about reasons, whereas desire-­as-­belief does not imply this. Again, on the face of it, this favours desire-­as-­belief: if you think that something is good but that you have no reason to bring it about, you need not desire it (see also Chapter  6). In cases like these, desire-­ as-­ belief has more plausible implications than conativism. But the starkest examples of this kind are given by logically complex normative attitudes. Desire-­as-­belief identifies only beliefs with the atomic content [I have reason to bring about p] with desires. In contrast, conativism identifies all normative attitudes with desires, no matter how logically complex their contents. For example, it says that there are desires corresponding to the thoughts [that Chen has been fired and Patel ought to replace him], [that either Chen has been fired or else he should resign], and [that everyone should eat before 11pm]. But in each of these cases, it is hard to see what desire the relevant attitude could possibly be. The simplest and best illustration of this point is provided by normative attitudes with negated contents (Schroeder 2008, 39–55; Unwin 1999, 2001). Imagine that Patel thinks [that it’s not the case that she ought to drink tea]: she believes she lacks that obligation. This attitude seems impossible to analyse as any desire. It is not a desire [to drink tea], since she’s not in favour of that. Nor is it a desire [not to drink tea], since she need not be against it either: she might be neutral on the matter. But nor is it a failure to desire [to drink tea], since that’s the absence of an attitude, not the presence of one. Worse, even if we could find some way to analyse this attitude as one of these desires, that would only move the bump in the rug: the attitude more naturally associated with the relevant desire would itself now lack a suitable analysis. We can present this problem as in Table 3.1.12 This reasoning favours the same general conclusion as above. Identifying some normative attitudes with desires seems plausible, and to that extent desire-­ as-­ belief is plausible. But it’s not plausible that all normative

12  I’ve lined the attitudes up in a manner that seems natural, so that the problem is that we lack a candidate desire for the third thought, but as above, the broader problem is simply that there are four attitudes on the left and only three candidate desires to reduce them to.

In Favour of Desire-­a s-­B elief  71 Table 3.1  The negation problem for conativism Thinking that she ought to drink Thinking that she ought not drink Thinking that it’s not the case that she ought to drink Not thinking that she ought to drink

T[Ov] T[O¬v] T[¬Ov]

Desire to drink Desire not to drink ???

D[v] D[¬v] ???

¬T[Ov]

Failure to desire to drink

¬D[v]

attitudes—including logically complex attitudes—are desires. In this way, conativism is implausible. The objection that conativism cannot offer an adequate theory of logically complex normative attitudes is in fact the best-­known objection to conativism: it is the so-­called Frege-­Geach objection.13 It is worth pausing briefly to dwell on the convergence we have reached with the Frege-­Geach objection. Whilst the motivation argument is by far the most influential argument for non-­cognitivism, the Frege-­Geach objection is by far the most influential argument against non-­cognitivism. If we want a theory that has the former payoff without the latter cost, desire-­as-­belief is exactly the right theory for the job. Conativism has the former advantage because it says that reasons beliefs are desires, and has the latter disadvantage because it says that all other normative attitudes are also desires. Desire-­as-­belief adopts the obvious position once this is clarified: it identifies only reasons beliefs with desires, and denies that any other normative attitudes are desires. That is just the obvious thing to say if we want the best of conativism without the worst. Again, desire-­as-­belief is the perfect theory for making sense of the mo­tiv­ ation­al powers of normative attitudes. Conativism is attractive insofar as it coincides with desire-­as-­belief, and it faces problems insofar as it diverges from desire-­as-­belief.

13  See Mark Schroeder (2008, 2010) for good places to start on the vast literature. Often, the Frege-­ Geach objection is expressed as an objection to the linguistic theory of expressivism (see §3.3.1), as the problem of accounting for the meaning of logically complex sentences that employ normative predicates. But since I have defined conativism as a theory about the nature of certain states of mind, rather than the meanings of certain sentences, the Frege-­Geach objection applies to conativism only if we express it as an objection that makes reference to the nature of states of mind rather than to the meanings of sentences (this is not wholly unusual: see e.g. Unwin 1999, 2001). This is plausibly the best way to think about the fundamental source of the Frege-­Geach problem. Expressivists claim that sentences’ meanings are inherited from the states of mind those sentences express. As a result, if we can find states of mind that constitute logically complex normative attitudes, it seems likely that we thereby allow expressivists to explain the meanings of logically complex normative sentences. Vice versa, if there are no such states of mind, no fiddling with anything else is going to fix the problem for expressivism. See also Eklund (2009) for helpful related discussion.

72  Normative Belief and Motivation

3.5  Summary and Humeanism We began this chapter by exploring motivation internalism, according to which normative attitudes can motivate us. I contrasted this claim with some nearby less plausible claims. I then combined motivation internalism with ODM to give us the motivation argument: motivation internalism says that normative attitudes can motivate us, but ODM says that only desires can motivate us. We are thereby forced to conclude that normative attitudes are desires. Both conativism and desire-­as-­belief aim to accommodate this truth. Conativism does it by claiming that all normative attitudes are desires. But this is problematic: if you think that someone else ought to do something, that is no desire of yours. So we should not accept conativism. Desire-­as-­ belief, in contrast, says that only some normative attitudes—reasons beliefs— are desires. That claim allows us to accept the motivation argument, but without overcommitting to the claim that all normative attitudes are desires. In this way, desire-­as-­belief is the perfect theory for capturing the motivational powers of normative attitudes. This is my first argument for desire-­as-­belief. Having rejected conativism, from here onwards I will cease using the technical label ‘normative attitudes’ and instead revert to talking about our normative beliefs. * Before we move on, it might be helpful to locate myself in the broader philosophical terrain. When philosophers think of ‘Humean’ views about desire, they often package together a variety of distinct claims. But desire-­as-­belief tells us to accept some of those claims but not others. Let’s distinguish five Humean claims. (I leave it open which, if any, Hume himself actually endorsed.) I’ll take the first two together: (1) ODM:  You can be motivated to act only by the combination of a desire [that p] and a belief [that this act will bring about p]. (See Chapter 2.) (2) Hume’s negative claim:  You can never be motivated to act by beliefs alone. (See §2.2.) Clearly, ODM is compatible with desire-­ as-­ belief: I defended ODM in Chapter 2, and it played a crucial role in the argument for desire-­as-­belief in this chapter. One key objection to ODM is the claim that normative beliefs can motivate. But by accepting desire-­as-­belief, we completely defuse

Summary and Humeanism  73 this objection, and ODM is thereby still more attractive if we accept desire-­as-­belief. Desire-­as-­belief also entails that ODM is equivalent to the following claim: ODM+:  You can be motivated to act only by the combination of a belief [that you have reason to bring about p] and a belief [that this act will bring about p]. According to ODM+, we are motivated only by reasons beliefs. This obviously follows from ODM and desire-­as-­belief. ODM+ is inconsistent with Hume’s negative claim. So when I defend desire-­as-­belief and ODM, I thereby reject Hume’s negative claim. Hume’s negative claim is sometimes treated as equivalent to ODM, but it is not, as desire-­as-­belief illustrates. I now turn to a third Humean claim: (3) The distinct existences claim:  Desires and beliefs are different types of mental state. Desire-­as-­belief is obviously inconsistent with the distinct existences claim. If we were to assume the distinct existences claim, we could move without comment between ODM and Hume’s negative claim, treating ODM, Hume’s negative claim, and the distinct existences claim as a single package of views. But we should not treat these views as a single package in this manner. Desire-­ as-­belief is Humean to the extent it fits well with ODM, but anti-­Humean to the extent it forces us to deny the other two claims. There are two other Humean claims: (4) Instrumentalism:  All desires are either ultimate or instrumental. Ultimate desires cannot be changed by reasoning. Instrumental desires can be changed by reasoning only insofar as we can change them by inference from other beliefs and desires we have. (5) Subjectivism:  What you have reason to do always depends on your desires. I discuss these claims in §8.2 and Chapter  10, respectively, and reject them both. In these ways, desire-­as-­belief is best described as ‘somewhat’ Humean. It fits neatly with ODM, which might be considered the most central plank in

74  Normative Belief and Motivation the Humean tradition. But desire-­as-­belief fits poorly with the other four Humean claims. To this extent, we might want to reconsider some standard ways of laying out the options in this area: the best theory is one that takes one central element from the Humean tradition but rejects many others. * Our next chapter turns to rationality. What does it take to make rational choices? A natural thought is that rational choices are responsive to our desires. But why should this be true? I will argue that desire-­as-­belief provides a plausible answer to this question, and as such this provides a second argument for the view.  

4 Desire and Rationality In Chapter  2, I discussed the explanatory role that desires play: only they motivate us to act. In Chapter  3, I then presented a related argument for desire-­as-­belief. In this chapter I present a different argument for desire-­as-­ belief, focusing on the normative role of desires. The argument in this chapter for desire-­as-­belief emerges over several sections, and so it will be useful to give a brief overview here. The argument relies on distinguishing what we rationally ought to do from what we ought to do in some more objective sense. Roughly, the idea is that you rationally ought to do something just when you can be charged with inconsistency if you don’t, whereas the more objective ought binds you not on the grounds of personal consistency, but just because the facts themselves speak in favour of the relevant option. With that thought in mind, I say that desires make a difference to what we rationally ought to do: you are somewhat inconsistent if you most want to A, and yet do something else instead. But this rational force of desire needs explaining: why should there be rational pressure on people to act on their desires? Echoing Warren Quinn, I argue that many theories of desire will struggle to explain the relevance of desires to rationality. But desire-­as-­belief can explain this relevance, since a widely accepted constraint on rationality is that you live up to your own normative beliefs: to do otherwise is to be akratic, a paradigm case of irrationality. As a result, desire-­as-­belief should be endorsed as the best explanation of why desires make a rational difference. The argument above is presented across §4.1 to §4.4. §4.1 defends the claim that desires make a difference to what it is rational to do. §4.2 argues that dominant rivals to desire-­as-­belief cannot explain this truth. §4.3 argues that desire-­as-­belief can explain this truth. And §4.4 shows how desire-­as-­belief is superior to a nearby rival view—presentationalism—which tries to explain why desires make a rational difference by treating desires as perception-­ like states.

Desire as Belief: A Study of Desire, Motivation, and Rationality. Alex Gregory, Oxford University Press. © Alex Gregory 2021. DOI: 10.1093/oso/9780198848172.003.0005

76  Desire and Rationality

4.1  Desires Rationalize Actions Imagine that you have cancer, and falsely believe that a herbal tea will cure you. In one clear sense, you ought not drink the herbal tea under these circumstances: it will do absolutely nothing to cure your cancer. In this sense of ‘ought’, what you ought to do is largely independent of your own state of mind.1 But given your belief, it’s clear that you are behaving somewhat confusedly if you don’t drink the tea. As we might put it, a rational response to your belief is to drink the tea—that reaction makes perfect sense, given your belief. In cases like this, we can identify what you rationally ought to do, where what you rationally ought to do is highly dependent on your own state of mind, and contrast it with what you non-­rationally ought to do, where that is largely independent of your own state of mind. To ensure this distinction is clear, let me give some further examples. Imagine someone who holds a moral theory that you take to be mistaken. In one clear sense, if the theory is mistaken, they ought not live up to that theory. But in another clear sense, something is going wrong with them if they fail to abide by their theory: they are inconsistent. Or imagine that Donald is a horrible person. We might think that is a clear failing of his. But it is a further failing of his if he pursues his horrible goals inefficiently: as we might put it, he’s not only horrible, but also stupid. I label this distinction as the distinction between ‘non-­ rational’ and ‘rational’ oughts. But terminology here is partly a matter for stipulation: if you don’t like my use of the word ‘rational’, I’m happy for you to use other words to mark the distinction I have in mind.2 The important thing is just to distinguish between oughts that don’t hinge on facts about internal consistency and oughts that do. As I see it, we can assess people’s actions in both of these ways: as conforming with the relevant facts, or else as conforming with their own state of mind. Often the former is our focus, but we also sometimes criticize people on the latter ground, as when someone is inconsistent, weak-­willed, hypocritical, or otherwise fails to live up to their own views. I think the word ‘rational’ is natural in relation to the latter topic, 1  I say ‘largely independent’ because your own state of mind can sometimes be relevant to what you ought to do even in this sense, as when you ought to avoid something because it will be painful, or when you ought to see a psychologist because of your intrusive thoughts. I won’t try to make the idea any more precise—the basic contrast in what follows should be clear enough. 2  For example, following Scanlon (2007), some use the label ‘structural rationality’ for what I call the ‘rational ought’, and ‘substantive rationality’ for what I label the ‘non-­rational ought’.

Desires Rationalize Actions  77 but need not insist on this choice of terminology. In a note, I briefly connect these claims to some recent discussion of rationality.3 With this distinction clarified, I will now focus on desire and rationality for a while. We return to non-­rational oughts in Chapter 10. Given my use of the word ‘rational’, it’s plausible that our desires make a difference to the rationality of our actions. My neighbour might well have described me as foolish when they saw me climbing onto the roof with a dishcloth, but this act was in fact perfectly rational, given that I wanted to glue the guttering (don’t ask), and believed that cleaning the guttering would help the glue take hold. In cases like this, knowing about someone’s desires can help us assess their rationality. Other examples make the same point. If you want to go dancing, and I don’t, that makes some difference to the rationality of our going to Club De Fromage this evening: perhaps going makes sense for you, but not for me. If I go despite my hatred of dancing and have little to say in my defence, you might reasonably criticize me for going irrationally. Or, imagine that Nala regularly takes efficient steps to satisfying her desires. She wants to stay dry, to drink some wine later, and to see the new Sharknado film, and she responds to these desires by carrying an umbrella to the cinema and walking home via Bargain Booze. Or perhaps she has some desires that she satisfies in a more complex manner—perhaps she manages, through a cunning series of currency exchanges over the course of a year, to make a large profit. It seems to me that there is something admirable about Nala. Depending on the wider facts, she might be a failure in some other respects, might be morally bankrupt, and these desires might themselves be misguided in various ways. But such wider failings would leave intact the fact that there is still something

3  The view I adopt here fits most neatly with the views of Dan Fogal and Alex Worsnip (Fogal 2019; Worsnip 2018). It fits less neatly with other views on which there is only one kind of assessment, by the evidence (Kiesewetter 2017; Lord 2018; Weatherson 2019). I can only report that I find the defences of those views unsatisfactory (for some objections, see McHugh 2018; Way 2018), and that we should maintain some distinction along the lines of the one I draw in the main text. Even so, we might want to allow that we can assess people by the available evidence. I might allow this as a third kind of evaluation beyond the two in the main text, or else try to somehow combine it with one of them. For example, I might claim that evaluation by the facts is really evaluation via just the available facts, or else adopt a theory on which evidential evaluation is really evaluation via mental states such as perceptual states. For my purposes in the main text, I need not decide between these options: so long as we can rationally evaluate actions by appeal to people’s mental states, and so long as there is also a further kind of evaluation that is more objective, that is all I need. There are some other disputes about rationality (see e.g. Broome 2013; Parfit 2011a; Scanlon 2007), but I largely set those disputes aside: I can’t see that anything hangs on them for my purposes here. See also footnote 10.

78  Desire and Rationality she is doing well at. As I use the word, the thing Nala succeeds at is rationality: she responds in a rational manner to her desires. Thoughts like those above are commonplace. They stem partly from the irresistible thoughts that rational actions respond appropriately to your mindset, and that your mindset is affected by your desires. The attractiveness of thoughts like these also surfaces in decision theory. Decision theory says precisely that the rationality of your actions depends on your desires. Even if we don’t accept every claim made by decision theory, we should surely accept that the core of the view is onto something: rational action is sensitive to your own desires. In short, our desires make a rational difference. To keep things simple, I will summarize this claim by saying that desires ‘rationalize’ actions, and I will sometimes (somewhat loosely) say that desires make actions rational.4 Our theorizing about rationality and about desire had better make good sense of this fact. (In a note, I briefly explain why, unlike others, I ignore the power of desires to make actions ‘intelligible’.5) Before we move on, I should mention an alternative way of running the argument in this chapter. I aim to explain how our desires make a difference to the rationality of our actions. But some might worry that our actions ­cannot be assessed as rational or irrational, because rationality supervenes on the mind (e.g. Broome  2013, 88–89, 151–2). But there is a simple fix to this. I could instead aim to explain how our desires make a difference to the rationality of our intentions. The resulting argument would run exactly as I  present it here: nothing I say hinges on the idea that desires rationalize actions rather than intentions. If you prefer to read the argument in this ­alternative manner, feel free.

4  I say that this is a loose claim because it needs to be qualified by some ‘pro tanto’ or ‘all else being equal’ clause—cf. §4.3. 5  Some authors discuss the normative significance of desire and focus on their capacity to make actions ‘intelligible’ (e.g. Johnston 2001). Perhaps desires do make actions intelligible, but this is not what I have in mind when I say that they make actions rational. More is required to make actions rational than to make them intelligible: actions might be intelligible and yet irrational, if, for example, they exemplify an understandable kind of irrationality (cf. Johnston 2001, 190–1; see also Yao 2019). When you are weak-­willed, your actions might be perfectly intelligible in the sense that we can understand the relevant temptations and can understand how you succumbed to them, but such actions are nonetheless not rational. Vice versa, if you manage to act in ways that maximize the expected degree of desire satisfaction that you will get, your actions are not merely intelligible but instead have a higher status: you have accomplished something difficult, and succeeded in a feat of rationality. In short, I claim that desires, unlike common biases, can confer some positive normative status on our actions, not merely make them comprehendible or predictable. Decision theory, for example, is not a theory of how to act in ways that people will understand: it positively recommends courses of action. To this extent, our interest is rationality, not mere intelligibility.

Against the Dispositional Theory of Desire  79

4.2  Against the Dispositional Theory of Desire So far I have claimed that desires make actions rational. Can we explain why this is true? Before we get to desire-­as-­belief, let’s examine a rival theory of desire: doing so will allow us to reject that rival theory, and to clarify the explanatory challenge at hand. According to the dispositional theory of desire, to desire that p just is to be disposed to do things that you believe will secure p (Smith  1987,  1994; Stalnaker 1987, 15). For example, this theory says that desiring a pig amounts to being in a state which would lead you go to the shop, if you thought it would sell you one. On this view, desires are really just tendencies to act in certain ways. Warren Quinn famously objected to this theory of desire, precisely on the grounds that it fails to explain why desires rationalize actions (Quinn 1998). He worries that if we understand desires as mere tendencies to do certain things, that might explain why we do those things, but would not explain the normative import of desires. He asks: ‘How can the fact that we are set up to go in a certain direction make it (even prima facie) rational to decide to go in that direction?’ (Quinn 1998, 189).6 Quinn’s worry is compelling. If desires are merely dispositions to act in certain ways, then acting on a desire is no more rational than manifesting any other disposition you have. Perhaps you are disposed to get sleepy in the afternoons, to get bags under your eyes when tired, to squint in bright light, and to trip over when you’re distracted. These dispositions are nonrational dispositions: it’s not at all rational to manifest these dispositions, or irrational to fail to manifest them. What is different about the dispositions that constitute your desires? You may well have other dispositions which are positively irrational. Perhaps you are disposed to affirm the consequent, to be biased towards the status quo, and to be overconfident in your own views. Manifesting these dispositions does not make you rational, but instead the reverse. Again, why are desiderative dispositions different? The dispositional theory of desire has no answer to this question. Only some dispositions are rational dispositions, and so far as the dispositional theory of desire goes, we have no explanation of why desires belong to the privileged kind. So we

6 Quinn’s worry should be distinguished from the reverse worry that the dispositional theory makes no room for irrationality, since it seems to suggest that no-­one ever fails to act on their strongest desires (Korsgaard 1997). That is not a strong objection to the dispositional theory: according to the dispositional theory, desires are dispositions, and dispositions are not always manifested. So the defender of the dispositional theory can say that irrationality consists in having desires that fail to manifest their latent motivational potential (§2.3; see also Sinhababu 2011).

80  Desire and Rationality should reject the dispositional theory of desire (even if we nonetheless allow, with desire-­as-­belief, that one thing desires do is dispose us to act). Quinn supplements his argument against the dispositional theory with an example (Quinn  1998, 189–95). He imagines someone having a brute disposition to turn on radios—a disposition that is independent of any normative beliefs they hold. He asks whether this state would make it rational for them to turn on radios. His plausible answer is that it wouldn’t. Such a state would cause them to turn on radios, but it is doubtful that we would find this rational: it would be explicable but not sensible or otherwise deserving of any positive normative status. Let’s call Quinn’s imaginary figure Radiohead. Given that desires do make behaviour rational, we should conclude that Radiohead doesn’t really have a desire to turn on radios (but instead, say, a compulsion), and thereby conclude that the dispositional theory of desire is false. Quinn presents his wider argument, and his Radiohead case, not only as a negative argument against the dispositional theory of desire but also as a positive argument in favour of the view that desires represent their objects as having some favourable normative property (see also Stampe  1987; Schafer 2013). I will discuss the positive version of Quinn’s argument in the next section. In the remainder of this section, I will continue to focus on his negative argument against the dispositional theory of desire. Later discussion of Quinn often focuses on Radiohead (e.g. Scanlon 1998, 37–49). Objectors commonly claim that the dispositional theory can be qualified in ways that make Radiohead’s urge fail to qualify as a desire. In particular, objectors claim that desires dispose us not only to act, but also to  do other things such as feel pleasure (Copp and Sobel  2002, 261; Sinhababu 2017, 185–6; on a similar note, see Smith 2011, 93; 2012, 394). For example, if you want to go on a date with Khan, you will presumably feel excited if he agrees to go. On this revised version of the dispositional view, desires are not merely tendencies to act in certain ways, but also tendencies to do other things, such as have certain feelings. (I discuss related claims in §7.6.) With this revised dispositional theory of desire in hand, we can insist that Radiohead does not have all of the dispositions associated with desire: he has no disposition to feel pleasure when he turns radios on. We can thereby infer that he does not desire to turn on radios. In turn, this revised theory of desire is consistent with the claim that there is nothing rational about Radiohead turning on radios. To this extent, objectors claim that Quinn’s Radiohead case undermines the simple dispositional theory of desire, but not this nearby view.

Against the Dispositional Theory of Desire  81 This reply to Quinn is compelling as far as it goes, but it doesn’t go far enough. The essential point is that Quinn makes his case with his broader argument, and illustrates it with Radiohead. But the reply above focuses only on Radiohead, and thereby fails to address the main thrust of Quinn’s argument. Let us simply concede that Radiohead can be dealt with. Quinn’s argument was that the dispositional theory of desire has no explanation of why desires rationalize actions: it cannot explain why desires, unlike other dispositional states, have rational consequences. That argument need not appeal to Radiohead. Another way of putting this is that the Radiohead case focuses on a mental state that fails to rationalize actions, and asks why that is so. The worry was that the dispositional theory of desire cannot explain this. Our objectors aim to show that even on their dispositional view, they can agree that Radiohead is not rational and has no desire. That is fine so far as it goes, but is not a positive explanation of why things that are desires do rationalize actions, but rather at best a partial and negative explanation of why Radiohead’s state of mind fails to rationalize his actions. Again, let us simply accept that Quinn’s objectors are right about Radiohead. They still need a positive explanation of why, in general, desires do rationalize actions. So our objectors above have yet to reply to Quinn’s broader argument. Might they simply redeploy the same resources to do so? I don’t think so. They could say that desires rationalize actions because desires dispose us to feel pleasure when we get what we want. But, first, at best it seems that rationality might hinge on the pleasure you expect to get from the options open to you, not the amount of pleasure you are disposed to actually get from them, whether you know it or not. At least, this seems right if we are thinking of rationality as hinging on your perspective in an important sense. If this is so, then the hedonic aspect of desire fails to bear directly on their capacity to rationalize actions, since the hedonic aspect of desire concerns the pleasure they dispose us to feel, rather than our beliefs about pleasures we might feel. Second, Quinn himself gives a compelling reply to this suggestion (Quinn 1998, 196; see also Smithies and Weiss 2019, 44).7 If desires rationalize actions just because of the pleasure they promise, then the very same rationalizing force should be present when the pleasure is available but in the absence of desire: the present suggestion is that the rationalizing force of desire comes purely from their hedonic side. But that seems wrong. If I don’t want any cake, but would nonetheless get X pleasure from it, whereas you do 7 Quinn also offers two further replies to this suggestion (1998, 196–7), but I find neither compelling and so set them aside.

82  Desire and Rationality want cake and would get X pleasure from it, it seems that you are under greater rational pressure than me to eat the cake. But on the present suggestion, since the rationalizing force of desire is exhausted by their hedonic effects, we should both be equally rational in eating the cake. That seems incorrect. A more dramatic example is death: everyone would get the same amount of pleasure from death (i.e. none), but it doesn’t follow that it’s equally rational for anyone to choose it, no matter their desires.8 Or, for another example, it doesn’t seem that the rationality of being moral hinges entirely on whether you get pleasure from doing your duty. In short, the present suggestion commits us to a hedonistic theory of rationality, but the common-­ sense thought behind decision theory is that desires, not prospective pleasures, are the crucial thing for rationality. And that thought seems right.9 I conclude that Quinn’s argument succeeds. It undermines dispositional theories of desire, and nearby variants, since they fail to explain why desires rationalize actions. I now turn to the positive interpretation of Quinn’s argument, and explain how it favours desire-­as-­belief.

4.3  In Favour of Desire-­as-­Belief In contrast to the dispositional theory of desire, desire-­as-­belief provides a simple explanation of why desires rationalize actions. One central principle of rationality is that you should live up to your own normative beliefs—you should follow your conscience. To do otherwise is to be akratic, and akrasia is the paradigm case of irrationality: it involves incoherence between your normative beliefs and your actions. Since desire-­as-­belief tells us that desires are normative beliefs, it combines with the injunction against akrasia—the Enkratic Requirement—to explain why it is rational to act on your desires. This line of thought is simple and compelling: desire-­as-­belief promises to explain why desires rationalize actions, and it does so via a principle of rationality that absolutely everyone ought to accept. So we should accept

8  An objector might reply that desires also dispose us to feel frustrated when they are unsatisfied, and that this explains the death case, since some people but not others would feel frustrated if they don’t die. But this only moves the bump in the rug: imagine someone who doesn’t want to die but who will feel frustrated if they don’t (perhaps they are just eternally grumpy). Again, it’s doubtful that such a person is just as rational in pursuing death as someone who wants it. 9  A further worry with the present suggestion is that we should allow that we desire some things out of proportion to the pleasure we would feel if those desires were satisfied (Sinhababu 2017, 29; see also Arpaly and Schroeder 2014, 126 and onwards). This provides further grounds for doubting that desires rationalize in proportion to the amount of pleasure we would feel if those desires were satisfied.

In Favour of Desire-­a s-­B elief  83 desire-­as-­belief on the grounds that it best explains why desires rationalize actions. This is my second argument for desire-­as-­belief. But there is a slight hitch with this argument. Remember desire-­as-­belief: Desire-­as-­belief:  To desire [that p] just is to believe [that you have reason to bring about p]. Desire-­as-­belief says that desires are beliefs about reasons. But the Enkratic Requirement of rationality is standardly formulated in (roughly) the following way (e.g. Broome 2013, 90; Ewing 1948, 120–1): The Enkratic Requirement:  If you believe you ought to v, rationality requires v-­ing.10 These two views don’t obviously combine to explain why desires rationalize actions. Desire-­as-­belief tells us that desires are beliefs about reasons, whereas the Enkratic Requirement refers only to beliefs about what we ought to do. So how could the Enkratic Requirement combine with desire-­as-­belief to explain why desires rationalize actions? This problem can be fixed: we should broaden the Enkratic Requirement. The Enkratic Requirement above says that rationality requires us to do whatever we believe we ought to do. But it’s also more broadly true that actions are rationally favoured if we believe we have reasons to do them.11 Imagine you believe that you have conclusive reason to do Acrobatics, less reason to do Ballet, and still less reason to go Climbing. Rationality might require you to do Acrobatics, but if you do Ballet instead, that is at least more rational than Climbing. So although it is true that you are rationally required to do whatever you believe you ought to do, we should treat this as just one specific implication of the more general truth that actions are rationally

10  I leave open whether this should be understood with ‘requires’ taking wide scope over the whole conditional, or narrow scope over just doing v (see e.g. Broome 2004 and Kiesewetter 2017 for discussion of this distinction). With minor modifications, my claims would apply either way (Gregory 2018, 1075). Moreover, as I said before (at the end of §4.1), I am happy if you prefer to read my arguments in this chapter as concerning the rationality of intentions rather than of actions. So understood, the Enkratic Requirement makes the plausible claim that if you believe you ought to v, rationality requires you to intend to v, and similar adjustments could be made to my other claims below. 11  For some discussion of rational pressures that fall short of requirements, see e.g. Brunero (2010, 33–4), Fogal (2019), and also the literature on ‘subjective reasons’ (e.g. Parfit  2011a, 33–5; Schroeder 2007a, 14–15; Whiting 2014).

84  Desire and Rationality favoured to whatever extent you believe you have reason to do them. We should reformulate the Enkratic Requirement as follows: The Best Enkratic Requirement:  If you believe you have reason to v, rationality favours v-­ing.12 With this version of the Enkratic Requirement in hand, desire-­as-­belief does explain why desires rationalize actions. Let’s take stock. I began with the basic thought that if desire-­as-­belief is true, then the injunction against akrasia explains why your desires make a difference to the rationality of your actions. But this basic line of thought faces a stumbling block, because desire-­ as-­ belief identifies desires with reasons beliefs, whereas the Enkratic Requirement is normally stated as a claim about beliefs about oughts. But this problem can be overcome: the Enkratic Requirement should really be stated in a broader form, as a claim about reasons beliefs. With this revised Enkratic Principle in hand, the original argument succeeds. To ensure the resulting argument for desire-­as-­belief is clear, here it is, one more time: (P1)  If you desire to v, rationality favours v-­ing. (P2)  If you believe you have reason to v, rationality favours v-­ing. By inference to best explanation, (C) To desire to v is to believe you have reason to v. Here, P1 is the claim with which I began: that desires rationalize actions. As I said in §4.1, that claim is highly plausible. P2 is the Enkratic Requirement, stated in the best way. That requirement is rightly widely accepted. C is desire-­ as-­belief. It is favoured by the premises as an inference to best explanation: Without C, it is hard to see how P1 might be explained. But given P2, C provides an excellent explanation of P1. This is my second argument for desire-­ as-­belief. Quinn was right to insist that we need an explanation of why desires rationalize actions, and desire-­as-­belief supplies an attractive explanation. 12  We should probably qualify this view so that the strength of the reason determines quite how strongly rationality favours v-­ing, and perhaps qualify it further to take into account any residual uncertainty about whether your belief is correct (see also §8.1.4). But I’ll pass over such details here. With this view in hand, we should also be clear about how to define ‘akrasia’: as failing to do what is most rational. Otherwise we might be forced to conclude that even the best choice is often akratic since you often defy some competing reasons you believe you have.

In Favour of Desire-­a s-­B elief  85 Here is a worry: does the explanation above merely kick the can down the road? I’ve explained why desires rationalize actions by appeal to the Enkratic Requirement. But do we need an explanation of the Enkratic Requirement itself? And if we can explain why reasons beliefs rationalize actions, might that same explanation be deployed directly to desire, without the need for desire-­as-­belief? Vice versa, if we instead insist that we don’t need any further explanation of why reasons beliefs rationalize actions, why insist that there need be any explanation of why desires rationalize actions? Might we say in both cases that these are just brute facts about rationality? The argument is an inference to the best explanation, and to that extent I am happy to agree that my argument is defeasible, and that we will have to assess rival explanations of why desires rationalize actions as and when they are proposed. But I can nonetheless say two short and general things in reply to this worry. First, most importantly, it’s worth repeating that the Enkratic Requirement is widely accepted, and that it is widely accepted for good reason: it is clearly intelligible how normative representations can bear on what it is rational for us to do. If you think you have good reason to do something, that thought clearly bears on the rationality of your doing it: the transition from that thought to action is perfectly appropriate in that you have the thought just when you think of that action as appropriate. In contrast, if desire-­as-­belief is false, we can say nothing like this about our desires. Unless we have some special story to tell about the nature of desire, there is just no positive reason to expect desires to have some comparatively unique input to rationality. In this way, there could be some norm bearing on desire without the need for desire-­as-­belief, but its existence would be mysterious and inexplicable. In contrast, desire-­as-­belief explains, rather than merely asserts, the fact that desires have rational consequences. Second, more speculatively, there might be some broader picture of rationality that more strongly favours desire-­as-­belief in particular as the explanation of why desires rationalize actions. A tempting thought is that what a mental state can rationalize is determined by the inputs to that mental state (cf. Fogal 2019, 1055-­6; Millgram 1997, chap. 2). For example, perceptual states can rationalize beliefs, and they surely do so precisely because those perceptual states are sensitive to the state of the world: the inputs of perceptual states are uniquely placed to bear on the appropriateness of the resulting beliefs. We might tell a similar story about the power of beliefs to rationalize other beliefs: a belief has the power to rationalize another just when the characteristic inputs of the former—the evidence to which it is supposed to be

86  Desire and Rationality sensitive to—also bear on the appropriateness of the latter. If we generalize these thoughts, perhaps desires can rationalize actions only if those desires have characteristic inputs which bear on the appropriateness of those actions. Desire-­ as-­ belief says just this. According to desire-­ as-­ belief, desires are influenced by evidence just as other beliefs are, and in turn that is a good explanation of why they rationalize actions: because they have the function of being sensitive to evidence about what you have reason to do. More could be said if we wanted to develop this thought into a comprehensive theory of rationality, but the basic idea is surely attractive that rational norms are never brute, but instead explained in some way by the nature of the state of mind in question, and more specifically by the characteristic inputs to those states (cf. Kauppinen Forthcoming). That is just what desire-­as-­belief promises, and just what the dispositional theory rules out.13 Before we move on, we can address a nearby objection to desire-­as-­belief. Neil Sinhababu worries that desire-­as-­belief must say that our beliefs come in two different flavours, with almost all our beliefs playing one motivational role, but reasons beliefs playing a different motivational role. This might make it sound like desire-­as-­belief succeeds in reducing desires to beliefs only by  having an extraordinarily disunified account of belief. He writes: ‘Understanding desires as beliefs with some properties unusual for beliefs is almost like understanding desires as cheeses with some properties unusual for cheeses’ (Sinhababu 2017, 43).14 With the Best Enkratic Requirement in hand, we can now respond to this worry. Desire-­as-­belief identifies only reasons beliefs with desires. This makes sense, given the Best Enkratic Requirement: the Best Enkratic Requirement privileges our reasons beliefs as beliefs that have a special power to rationalize actions. In turn, that might explain why these beliefs alone have special ­motivational powers. Just about any state of mind could, in principle, cause us  to make certain physical movements. But those movements caused by our  desires are special: those movements are motivated. The Enkratic Requirement, in combination with desire-­as-­belief, explains why this is so: such movements uniquely count as motivated because they comply with a 13  To make this clear, it is helpful to distinguish functionalist and dispositionalist theories. Both views understand mental states in terms of their relations to other things. Functionalism understands mental states in terms of both their inputs and outputs, and as such I am broadly happy with such views. Dispositionalism, in contrast, understands mental states purely in terms of their outputs: what they dispose us to do. It is this restriction to onward effects which seems to me to render dispositional theories worse at explaining the rational significance of desire. 14  In fact, Sinhababu worries not only about the unique motivational role of desires but also about other roles they play, such as in directing attention or producing pleasure. But it’s less obvious to me in these cases that no other beliefs play the relevant roles. On pleasure, see §7.6.

Against Presentationalism  87 rational norm. So although desire-­as-­belief must say that reasons beliefs can do things that other beliefs cannot, that does not amount to a disunified theory of belief, but rather follows from the localized focus of the Best Enkratic Requirement. That reasons beliefs are unique amongst beliefs in their power to rationalize actions should not raise eyebrows any more than parallel claims such as that altruistic desires are unique amongst desires in their power to make actions kind, or that desires for future wellbeing are unique amongst desires in their power to make actions prudent.

4.4  Against Presentationalism So far I have focused on explaining how a version of Quinn’s argument ­succeeds in favouring desire-­as-­belief. But Quinn doesn’t focus on desire-­as-­ belief, and instead treats his argument just as favouring some view in this general vicinity (see e.g. 1998, 200). In this section, I explain why Quinn’s argument favours desire-­as-­belief, and not its nearby competitor, presentationalism. After all, I said that the argument above was an inference to the best explanation, and perhaps this other theory can offer a better explanation than desire-­as-­belief of why desires rationalize actions. Presentationalism is the view that desires are appearances of the good (Oddie  2005; Schafer  2013; Stampe  1987; Tenenbaum  2007; see also Moss 2012).15 The idea is that just as perceptual states present things as being a certain way (e.g. the wall as red), and just as intellectual ‘seemings’ present things as being a certain way (e.g. 0.999 . . . as less than 1), desires present their objects as good.16 Such presentational states are independent of your beliefs: you can distrust how things seem to you, as when you think you are subject to a visual illusion, or when you know that the intellectual appearances are deceptive. Presentationalism says that when we desire something, it seems good to us, though we might believe those appearances to be misleading. (A quick aside: I treat desire-­as-­belief as the view that desires are beliefs about reasons. Presentationalists typically take desires to be appearances of the good. In this section I focus on the difference between treating desires as beliefs and treating them as appearances: the difference between views on which desires represent the good, and those on which they represent reasons, 15  Some endorse a version of presentationalism not about desire, but instead about emotion (e.g. Döring  2003; Tappolet  2016). My concern here is purely with presentationalism about desire. See also §7.1. 16  For more on intellectual seemings, see Huemer (2007b).

88  Desire and Rationality is the focus of Chapter 6, and independent of the issues here. For ease, I will address my presentationalist opponents in their own preferred terms: as ­taking desires to be appearances of the good.) Presentationalism might seem to explain why desires rationalize actions. Presentational states can clearly rationalize belief (see e.g. Huemer  2007a). For example, if you perceive the wall as red, that might make it rational to believe the wall is red. So if we think that desires are presentations of the good, it might seem plausible that such states can rationalize actions (Schafer 2013; Stampe 1987, especially 362). If p is presented to you as good, isn’t it rational to do things that you think will bring about p? In this way, presentationalists hope to explain the rationalizing force of desire on the model of the rationalizing force of presentational states. As such, we might think that Quinn’s argument also favours presentationalism because it also explains the rationalizing power of desire. But we should not accept presentationalism. One worry for the view is that it conflicts with the motivation argument for desire-­as-­belief that I offered in the last chapter. Some defend presentationalism with the hope of holding onto ODM whilst also allowing that our normative attitudes can motivate us (e.g. Oddie  2005, chap.  2; Tenenbaum  2007, chaps 1–2; see also Sinhababu  2017, chap.  4). But this hope is misguided. Given presentationalism, if we are only motivated by desire, then it follows that we are only ever motivated by appearances of normative properties, never our beliefs about such properties. And so according to presentationalism, our normative beliefs are motivationally inert. That is implausible. For example, my beliefs about the merits of heading home after a couple of beers can motivate me to do that. And in some cases, these beliefs can motivate us even when the appearances are silent or else point the other way. For example, perhaps your beliefs about the best investment strategy can motivate you even when the relevant issues are too complex for any option to seem right. Or for another example, perhaps your moral beliefs about the need to protect the environment might motivate you to eat insects, even if it unreflectively seems wrong. More generally, we might think that virtuous people often manage to act against the normative appearances. Presentationalists must abandon such thoughts, or else abandon ODM, and neither option is attractive. So this provides one simple worry for presentationalism. But I will present three further objections to it, and then independently show that it provides a poor explanation of why desires rationalize actions. First, presentationalism understates the rational responsiveness of desire.17 If I ask whether you want to visit Tonga with me this summer, you might 17  See also Gregory (2017, 214). For more on reasoning with desire, see §8.2.

Against Presentationalism  89 investigate the issue and deliberate at length in order to decide whether you want to go (and if so, how strongly). The claim that our desires are sensitive to deliberation may be controversial with respect to our non-­instrumental desires, but it is undeniable with respect to our instrumental desires. We form instrumental desires by rationally combining our prior desires with our beliefs, and so it obviously follows that we form instrumental desires in light of the beliefs we hold and the inferences we have made about how our beliefs interact with our desires. So our instrumental desires, at least, are rationally responsive. But if such desires are presentational states, they should not be rationally responsive in this manner. How things are presented to us is not normally something we can change by acquiring further information, or by reasoning. For example, no amount of reasoning is likely to change the colours things perceptually appear to have. So it is not plausible that all desires are presentational states. Someone might object that presentational states can be influenced by our beliefs. For example, if you tell me that snakes live nearby, this might make nearby vines look more snake-­like. But first, the evidence for such ‘cognitive penetration’ is far from decisive—see Firestone and Scholl (2016) for scepticism. Second, even if our beliefs do affect our presentational states in this manner, this falls short of explaining the rational responsiveness of desire. Our instrumental desires can quite clearly be changed by inference. I can deliberate about my goals, and settle on appropriate means to them, and thereby come to desire those means. But cognitive penetration is a merely causal process, not an inferential one. One way to see this is to note that such cognitive penetration is rationally unconstrained: you might make nearby vines look more snake-­like by telling me that snakes live nearby, but you might achieve the very same effect by telling me that no snakes live nearby: the mere mention of snakes might be enough to put me on edge. But our formation of instrumental desires is not a merely causal process like this, but instead a rational process of inference. The second (related) problem for presentationalism is that it cannot permit that desires are ever irrational (cf. Stampe 1987, 359).18 This criticism is somewhat ironic, since one of the main attractions of presentationalism is supposed to be that it makes room for desiderative failures (e.g. Tenenbaum 2007, chaps 6–8). For example, the view is supposed to be attractive because it permits that your normative beliefs might fail to align with your desires.19 According to presentationalism, this can occur just because you can believe the appearances to be illusory. 18  This mirrors a common objection to presentational theories of emotion—see Brady (2009, 2013, 112–16); Helm (2015). 19  I discuss related matters in Chapter 5.

90  Desire and Rationality But in this respect, presentationalism is less attractive than it appears. Presentationalism does permit that you might deny that what you want is good. And since presentational states can be inaccurate, it also permits that desires can be incorrect, as when you prefer the destruction of the world to a scratch on your finger. But presentationalism does not permit that desires can be irrational. Presentational states cannot be irrational. No matter the circumstances, you cannot be irrational because it perceptually appears to you that a wall is red. Perhaps under some circumstances you might be irrational to believe this appearance, and perhaps we can say that your perceptual faculties are malfunctioning or inaccurate. But these claims fall short of saying that your perceptual state is itself irrational: crucially, these are not criticisms of you, as charges of irrationality are. Since we do criticize people for their desires, it follows that desires are not presentational states. Some might protest that some desires—such as your liking of chocolate— are not subject to rational assessment. I discuss some cases like this in Chapter 7. But we need not wait for that discussion to make the point here. Even if some desires are rationally unassessable, other desires clearly are. Since presentationalism treats all desires as presentational states, it cannot permit even this modest claim. Let me give some more concrete examples to illustrate the problem for presentationalism. Intransitive preferences are clearly irrational. Imagine that you overall prefer to go to the conference by train than by plane (because it is better for the environment), overall prefer to go to the conference by plane than not go at all (because the conference will be worth the short journey), but also overall prefer not to go to the conference at all than to go by train (because the conference will not be worth the long journey). Someone with these preferences seems criticizable for being irrational (see e.g. Broome 2006). But presentationalism cannot allow this. If you are subject to some illusion whereby A looks better than B, B looks better than C, and C looks better than A, you are not thereby irrational, any more than you would be irrational for being subject to a similar illusion about length. So presentationalism cannot explain the irrationality of intransitive preferences. There are other ways in which desires are clearly subject to rational assessment. If you desire E, and believe that M leads to E, you should rationally desire M (see e.g. Broome 1997; Smith 2004, 97–100). Someone who fails to desire the known means to their ends seems irrational.20 But again, presenta20 Perhaps this needs some qualification so that, for example, the desire in question is strong enough to be worth thinking about in this context (cf. Harcourt  2004). But this and other

Against Presentationalism  91 tionalism cannot allow this. If it appears to you that E is good, and you believe that M causes E, you are not irrational if M does not seem good to you. We are never rationally required to form new presentational states. So again, presentationalism cannot explain the irrationality of failing to desire the means to our ends. The above two objections to presentationalism may seem local to our instrumental desires: instrumental desires are most obviously sensitive to information and reasoning, and instrumental desires are most obviously vulnerable to criticisms of irrationality. So in response, the presentationalist might restrict their theory, offering it as a theory of only non-­instrumental desires. But this move is unattractive for three reasons. First, it is doubtful that it would avoid the objections above. Intransitive preferences, for example, seem to be irrational even when non-­instrumental. Second, it forces us to accept a disjointed theory of instrumental and non-­instrumental desires, and it seems that a more unified theory would be better if one is available: desire-­ as-­belief does not pay this same cost.21 Third, remember that presentationalism is supposed to explain the rationalizing power of desire. But if it does not extend to instrumental desires, it is not clear how it might explain the rationalizing power of such desires. Since instrumental desires do seem to rationalize actions—they rationalize the pursuit of money, for example—the presentationalist who constrains their theory in this way will fail to fully explain the rationalizing power of our desires as originally advertised. The objections above show that presentationalism is false: it offers an implausible theory of desire on which desires are never sensitive to reasoning and are never irrational. The other problem with presentationalism is that it fails to explain why desires rationalize actions—it fails to benefit from Quinn’s argument with which I began. Presentationalists hope to explain why desires rationalize actions by appeal to their theory that desires are appearances of the good, in combination with the idea that presentational states have the power to rationalize belief. But when stated like this it should be obvious that presentationalism would not explain how desires rationalize actions, but

qualifications will leave intact the basic thought that there are some rational requirements to form instrumental desires. 21  In reply, some might say that instrumental desires don’t exist at all, and that talk of such desires is really just shorthand for talk about combinations of non-­instrumental desires and beliefs (for such claims, see Chan 2004; Sinhababu 2017, 27–8). But this view seems implausible: it is a feat of rationality to see how your beliefs and desires interact. Insofar as we need to store the results of such reasoning— which we surely do, so as to speed up future deliberation—we surely store it as a new, instrumental, desire. This argument runs parallel to a natural argument for thinking that humans sometimes believe the consequents of conditionals.

92  Desire and Rationality instead explain how desires rationalize beliefs, and that is not what we hoped to explain. That is, according to presentationalism, the desire that p is a presentational state in which p appears good. But the standard view is that the appearance as of p rationalizes the belief that p. So according to presentationalism, the primary rational upshot of a desire [that p] should be the belief [that p is good]. But this is wrong. It is doubtful that desires rationalize beliefs in this manner. Imagine you ask me why I think that giving money to Oxfam is a good thing to do, and I reply that I think this because I want to give them money. This seems like an inappropriate answer. Since the desire that p fails to rationalize the belief that it’s good that p, it’s doubtful that desires are appearances of the good. Vice versa, whereas desires rationalize actions, presentational states do not. What it is rational for you to do depends on what you want. But what it is rational for you to do does not depend on your presentational states at all, except insofar as those states influence your beliefs. For example, imagine that you are presented with the Müller-­Lyer lines (Figure 4.1), and some decision (e.g. whether to accept a bet) hinges on their respective lengths. Though the first line seems longer, let’s imagine that you know about the illusion, and have even measured the lines in this case and confirmed that they are really equally long. Under such circumstances, it seems that the appearances are irrelevant to what it is rational for you to do: what it is rational for you to do depends only on what you believe about the lines, and not how they appear. Parallel reasoning suggests that if A seems to you to be better than B, that alone tells us nothing about whether you ought to pursue A: everything depends on whether you accept these appearances. When we rationally evaluate your actions, we do not take into account your presentational states, only your beliefs and desires. It is useful to think of decision theory here, which aims to analyse rationality. It presents a plausible theory of rationality that appeals only to a person’s beliefs and desires, not at all to their presentational states. Since appearances do not bear on the rationality of actions, but desires do, desires are not appearances of the good. In summary, we should reject presentationalism. Presentationalism cannot permit that our desires are sensitive to information and reasoning.

Figure 4.1  The Müller-­Lyer illusion.

Summary  93 Presentationalism cannot permit that desires can be irrational. And most crucially for our purposes, presentationalism cannot adequately explain why desires rationalize actions: presentationalism suggests that desires rationalize beliefs, rather than actions, and this is not plausible. Pace presentationalism, the primary rational output of desire is action, not belief. And these problems, remember, are in addition to the fact that presentationalism conflicts with the motivation argument. In the face of these objections, presentationalists might weaken their claims. They might agree that desires are not literally appearances of the good, but maintain that they are nonetheless similar to such states. But in this form the view is too vague to be able to explain why desires rationalize actions, as originally advertised. After all, if the presentationalist says that desires are similar to presentational states only in that they have the power to rationalize something, that at best asserts, rather than explains, the capacity of desires to rationalize actions. But if the presentationalist says that desires are similar to presentational states in other respects which genuinely explain why they rationalize things, they need to say in which respects they are similar, so as to show that they can be similar enough to rationalize something without being so similar as to rationalize beliefs rather than actions. Without such details, presentationalism fails to explain why desires rationalize actions, and so fails to capitalize on Quinn’s argument with which we began.

4.5 Summary In this chapter I have provided a second argument for desire-­as-­belief. Quinn was right to claim that we need some positive explanation of why desires rationalize actions. Objectors to Quinn have undersold his worry, and failed to address it in its strongest form. And in that form, it favours desire-­as-­belief: desire-­as-­belief provides a compelling explanation of why desires rationalize actions, via the Best Enkratic Requirement, which tells us to live up to our own normative beliefs. Other theories of desire, such as the dispositional theory of desire, and presentationalism, cannot explain why desires rationalize actions. Our next chapter turns away from the lofty ideals of rationality, and towards the sad existence of irrationality. Does desire-­as-­belief leave enough room for stupidity, in all its spectacular forms?

5 Irrationality 5.1 Recap Let’s recap where we are. We’ve seen that desire-­as-­belief says that our desires are just a subset of our beliefs. This view is easiest to understand if we remember the analogy with disbelief: someone disbelieves [p] just when they believe [¬p]. Talk of disbelief is just shorthand for talk about beliefs with negated contents. Desire-­as-­belief says the same thing about ‘desire’: someone desires [p] just when they believe [I have reason to bring about p]. Talk of desire is just shorthand for talk about beliefs with reason-­to-­bring-­about contents. We’ve seen two arguments for desire-­as-­belief. The first is that it lets us hold onto ODM (Only Desires Motivate)—the claim that we are motivated only by desires, combined with suitable beliefs—at the same time as holding onto motivation internalism—the claim that we can be motivated by our normative beliefs. Desire-­as-­belief says that our desires are normative beliefs. It thereby says that these two states of mind are not competitors in our motivational system, but one and the same motivational state of mind. That is, desire-­as-­belief allows us to maintain the dominant theory about how all human actions are explained—ODM—whilst also permitting that our ethical views motivate us, as they surely do. The second argument for desire-­as-­belief is that it explains why desires make a rational difference. Our desires clearly have implications for the rationality of our actions, and rival theories of desire cannot explain that truth. But desire-­as-­belief can, by appeal to the widely accepted Enkratic Requirement, according to which you rationally ought to live up to your own normative beliefs. Desire-­as-­belief thereby makes good sense of the way in which desires have normative relevance. In these ways, desire-­ as-­ belief allows us to maintain and explain various elements of common sense, and that is a strike in its favour. Some will accept that desire-­as-­belief would have these merits if it were true, but will claim that the theory faces decisive objections of one sort or another. So this chapter, and those following, often focus on various objections to desire-­ as-­ belief. This chapter focuses on one specific and influential Desire as Belief: A Study of Desire, Motivation, and Rationality. Alex Gregory, Oxford University Press. © Alex Gregory 2021. DOI: 10.1093/oso/9780198848172.003.0006

Desire-­a s Belief and Failures of Rationality  95 objection to desire-­as-­belief: that it cannot make sense of irrationality such as  akrasia. §5.2 describes the objection, and §5.3 to §5.6 present a series of complementary replies. Finally, §5.7 discusses a slight aside: some think that reference to second-­order desires is key to understanding akrasia. I argue not.

5.2  Desire-­as Belief and Failures of Rationality Sometimes people act badly because they are ignorant, as when you mis­ taken­ly lock your neighbour’s cat in your shed. But sometimes, people act badly despite knowing what they ought to do: perhaps you know that you shouldn’t smoke, but do so regardless, and perhaps you know that you should exercise, but fail nonetheless. It might seem that desire-­as-­belief is inconsistent with such possibilities: it might seem that desire-­as-­belief implies that you’ll always do whatever you believe you have most reason to do. In this way, desire-­as-­belief might seem inconsistent with the existence of weakness of will, when people act against their better judgement.1 I’ll call this the objection from akrasia. In fact, the objection from akrasia rests on a more fundamental objection to desire-­as-­belief: that desire-­as-­belief is wrong to claim that people always want to do things to whatever extent they believe they have reason to do them. The objection from akrasia is that our reasons beliefs can fail to align with our actions, but it arises because desire-­as-­belief entails that our reasons beliefs always align with our desires. Desire-­as-­belief says that whenever someone believes that they have reason to bring about p, they desire p, and also says that whenever someone desires p, they believe that they have reason to bring about p. Some find these claims implausible. In this way, the objection from akrasia really rests on two prior objections: the objections from underwanting and overwanting. The objection from underwanting is that you might not desire enough: you might believe you have reason to bring about p, and yet not desire p. The objection from overwanting is that you might desire too much: you might not believe that there’s any reason to bring about p, but desire it nonetheless. In the remainder of this chapter I address the objections from underwanting and overwanting, and in turn the objection from akrasia.

1  The literature on akrasia is comparatively vast: for an overview, see Stroud (2014). Richard Holton (1999) discusses a further kind of weakness of will involving an over-­readiness to modify your intentions: I set this aside, since it raises no obvious problem for desire-­as-­belief.

96 Irrationality The objections from underwanting and overwanting are key objections to desire-­as-­belief. Many worry that we can be irrational in what we desire, so that our desires are at least somewhat independent of our normative beliefs (e.g. Smith  1994, 134; Stocker  1979). For example, imagine that Sally is a smoker. She might think she has good reasons to quit, and yet not want to do so. This kind of underwanting might seem quite common: can’t you think that you have good reasons to go for a run, to volunteer your weekends for charity, to give up your money for good causes, to stick to your diet, or to be brave and step out to face danger, but not want to do these things? Vice versa, imagine that Gary is a glutton. He might think he has no good reasons to eat more cake, and yet want to. This kind of overwanting might also seem quite common: can’t you want to stay in bed despite believing it would be a bad idea, or want to shout at your boss despite believing that it would be a bad idea? I will address these objections in three stages. First, I will briefly point to some relatively simple ways in which desire-­ as-­ belief is consistent with irrationality in desire. Second, I will show how desire-­as-­belief can explain away some objections in this vicinity by appeal to some linguistic resources. And third, I will show how desire-­as-­belief can explain away other objections in this vicinity by appeal to the distinction between desire and motivation. These replies are not quite exhaustive. Supposed counterexamples to desire-­ as-­belief can go wrong in any number of ways, and the responses I offer in this chapter don’t exhaust the possibilities, even if they deal with the central possibilities. Much of what I say in Chapters 6 and 7 will also help us to see some further ways in which putative counterexamples to desire-­as-­belief might go wrong. But the claims in this chapter are nonetheless central to my defence of desire-­as-­belief.

5.3  Irrationality in Belief Desire-­as-­belief certainly does not entail that our desires are always perfectly rational. To that extent, it can permit that ‘overwanting’ occurs if that simply means that we are irrational for having some of our desires, and can permit that ‘underwanting’ occurs in the sense that we are irrational for lacking some desires. Because our beliefs are often irrational, desire-­as-­belief entails that our desires are too. And, to the extent that our normative beliefs are especially irrational compared with other beliefs, desire-­as-­belief entails that our desires

Irrationality in Belief  97 are especially irrational compared with other beliefs.2 I will illustrate irrationality in belief with four examples, but these are just illustrations. In principle, for every possible rational failure in belief—for every fallacy or bias we are prone to—desire-­as-­belief tells us that our desires can be irrational in that same way. A first way in which our beliefs can be irrational is that we can easily draw mistaken conclusions when we deliberate too quickly. For example, if you are splitting the bill in a restaurant, and try to swiftly calculate everyone’s share in your head, you might make some trivial mathematical mistake. Similarly, we form some reasons beliefs quickly and without adequately checking them for accuracy. For example, you might look at the clouds and think that they provide you with a reason to put an umbrella in your bag, forgetting that you are staying inside all day. Under such circumstances, you form the irrational desire to put an umbrella in your bag, even though you yourself would realize on reflection that you have no reason to do so. More generally, desire-­as-­belief permits that we have some irrationally ‘impulsive’ desires, insofar as they are best understood as reasons beliefs that were formed too hastily.3 Perhaps such irrationality explains some dubious decisions, such as impulsive purchases. A second related way in which our beliefs can be irrational is that they can be oversensitive to the most readily available facts (cf. Tversky and Kahneman  1973). For example, if you try to estimate the proportion of the population who are younger than twenty-­five, you might more easily overestimate it when you are at university and surrounded by students than when you are visiting your grandparents and surrounded by older people. Similarly, many of our reasons beliefs are oversensitive to the nearby facts. You might generally believe that you ought to maintain your diet, but when the strudel is right in front of your eyes, you might find it easy to believe that the tastiness of the strudel outweighs the benefits of your diet, which are far less salient. Under such circumstances, you might end up irrationally preferring the strudel to staying thin, even though this desire is irrational. This exemplifies a second way in which desire-­as-­belief positively predicts irrationality in desire: some desires might be overly influenced by the things in front of you.4 2  For discussion of irrationality in normative belief, see Enoch (2009); Greene (2013); Haidt (2001). 3  Some of the Stoics, such as Plutarch, may have said something similar in defence of their view: see e.g. Long and Sedley (1987, sec. 65G3). 4 If you want to avoid this kind of irrationality, Benjamin Franklin provided helpful advice: ‘When . . . difficult Cases occur, they are difficult chiefly because while we have them under Consideration all the Reasons pro and con are not present to the Mind at the same time; but

98 Irrationality A third way in which our beliefs are often irrational is that they can be biased by wishful thinking (see also Enoch 2009, 25–7; May 2018, chap. 7). For example, if you are assessing who will win the next election, you might attend more closely to the evidence that suggests your preferred candidate will win, and less closely to the evidence that suggests they will lose. You might thereby end up believing that your preferred candidate will win even though the evidence, impartially considered, suggests otherwise. Similarly, many of our normative beliefs are subject to wishful thinking. For example, you might wonder whether you are obliged to have a difficult conversation with Jack about his behaviour in your classes, and attend more closely to the evidence that suggests you shouldn’t (e.g. he will graduate soon anyway) and less closely to the evidence that suggests you should (e.g. his behaviour is getting worse). Under such circumstances, you might end up irrationally preferring not to speak to Jack, even though an impartial spectator would rightly conclude that you ought to do so. In this way, desire-­as-­belief permits that some desires are irrational, such as when they are biased by wishful thinking driven by self-­ interest. And if normative beliefs are especially susceptible to wishful thinking—say, because a lot hangs on them—desire-­as-­ belief predicts that desires may be especially irrational. (It might be objected that wishful thinking is precisely when our desires interfere with the formation of our beliefs, and so you might think that the very possibility of wishful thinking is flatly inconsistent with desire-­as-­belief, since it presupposes a separation between our desires and our beliefs. But this is not right. For one thing, wishful thinking can occur when your normative beliefs—even normative beliefs that are not desires—irrationally influence your other beliefs, as when you wishfully believe that world peace will arrive as a result of thinking that it would be good. Of course, wishful thinking can also occur when your desire [that p] influences you into believing [that p], as when you come to believe that you will pass the exam just because you want to do so. But this would be inconsistent with desire-­ as-­belief only if it identified the desire [that p] and the belief [that p], which it doesn’t do. Desire-­as-­belief permits that wishful thinking can occur because some desires interfere with some other beliefs: you might come to believe [that p] because sometimes one Set present themselves, and at other times another, the first being out of Sight. [. . .] To get over this, my Way is, to divide half a Sheet of Paper by a Line into two Columns, writing over the one Pro, and over the other Con. Then during three or four Days Consideration I put down under the different Heads short Hints of the different Motives that at different Times occur to me for or against the Measure. When I have thus got them all together in one View, I endeavour to estimate their respective Weights [. . .] when [. . .] the whole lies before me, I think I can judge better, and am less likely to make a rash Step’ (https://founders.archives.gov/documents/Franklin/01-­19-­02-­0200).

Irrationality in Belief  99 you believe [that you have reason to bring about p]. It can even happen at a higher-­order level, when you come to believe [that you have reason to bring about p] because you believe [that you have reason to bring it about that you have reason to bring about p]: for example, you might think you have prudential reasons to ensure that you don’t have decisive reasons to speak to Jack, and wishful thinking might lead you to thereby conclude that you don’t have such reasons. That kind of wishful thinking in your reasons beliefs is entirely consistent with desire-­as-­belief.) A fourth way in which our beliefs can be irrational is that they can be recalcitrant. By this, I mean that they can persist over time even if conscious reflection concludes in their rejection. For example, you might consciously avow that women and men are intellectual equals, and yet still find yourself acting and making inferences as though they are not, as when you treat intelligent women as surprising outliers. Or for another example, imagine that you have been going to yoga on Wednesday nights for many years, but then give up. If you have been going for long enough, you might find yourself habitually believing that you have yoga on Wednesday, even though you consciously know this is mistaken. Such recalcitrance is obviously also possible with our normative beliefs. For example, you might consciously judge that planes are 100 per cent safe and yet still find yourself acting and making inferences as though they are not, such as by making the effort to check whether the wings look wobbly. Or, for another example, you might consciously judge that you have no significant reason to meet your ex-­ boyfriend, but persistently find yourself thinking that you should. In such cases, you might have desires that you consciously reject (see also §7.5, and Enoch 2011, 228–9). In summary, our beliefs can be irrational because we form them quickly, because they are oversensitive to particularly salient facts, because they are biased by wishful thinking, and because they are recalcitrant. Desire-­as-­belief predicts that our desires can be irrational in these very same ways: they can be impulsive, oversensitive to the near, biased, and recalcitrant. To this extent, desire-­as-­belief predicts that our desires can be irrational. And to repeat, these were just illustrations of the more general point: according to desire-­as-­ belief, any common failure with respect to belief formation is bound to also be a common failure with respect to desire formation. In this way, desire-­as-­ belief is certainly not a view on which our desires are always rational. If the objections from underwanting and overwanting merely amount to the thought that we can be irrational in what we desire, then they are not objections to desire-­as-­belief at all.

100 Irrationality

5.4  Misleading Desire Ascriptions, Again I now turn to a second (complementary) response to the objections from underwanting and overwanting. The response relies on claims I made about ascriptions of desire back in §2.4.1. Some of what we say about our desires is misleading. I argue that this makes the objections from underwanting and overwanting seem stronger than they are. I start with the objection from underwanting, and then turn to the objection from overwanting. In §2.4.1 I said that there were two notable ways in which our desire ascriptions can be misleading. The first was that assertions about what we ‘don’t want’ are ambiguous: they can express what we fail to desire (¬Dp), and can also express what we desire not to happen (D¬p). For example, when you say that you don’t want to eat blancmange, that might mean that you lack the desire to eat it (perhaps you don’t know what it is), or might instead mean that you have a desire to avoid eating it. With this distinction in hand, we can see that the latter is consistent with your also wanting to eat it: you might have conflicting desires, and want to avoid it for your health but nonetheless also want to eat it for the taste. This kind of ambiguity is clearly relevant to the objection from underwanting. There are many cases where underwanting appears to occur: you might think that you have reasons to do all sorts of things that you don’t want to do, such as to go for a run, to volunteer your weekends for charity, or to give up your money for good causes. But clearly, the ambiguity above is relevant in these cases. It is undeniable that we say things like, ‘I should go for a run today, but I don’t want to’. But given the ambiguity above, this might not be a case of underwanting at all, but instead a case of desire conflict: perhaps this assertion says that you think you should go for a run, and yet have some conflicting desire not to do so. In fact, this way of disambiguating such assertions is mandatory, since you might more emphatically say, ‘I should go for a run today, but I really don’t want to’, and this surely indicates the presence of a strong desire not to go for a run, not (impossibly) the strong failure of a desire to go for a run. As a result, sentences like, ‘I should go for a run today, but really don’t want to’ must really mean ‘to some extent, I want to go for a run today, but to some extent, I want to stay in’, or equivalently, ‘I think I have some reason to go for a run today, but some reason to stay in’. Clashes like these, between different desires—between different reasons beliefs—are entirely consistent with desire-­as-­belief. I now turn to the second point from §2.4.1 about expressions of desire. I said that we might be hesitant to express some desires because those desires

Misleading Desire Ascriptions, Again  101 are too obvious to be worth mentioning. Since we have so many desires for so many things, we must have some special grounds for asserting that we want something: often, that this desire is especially strong or of a notable kind. For example, you might be hesitant to say that you want to go to the dentist, not because you have absolutely no desire to go, but because saying this might imply that this desire is of a non-­obvious kind—say, grounded in the pleasure you take at being at the dentist, rather than a merely instrumental desire to  go for your future health. To this extent, it is surely true that we might be strongly disinclined to say that we want to do everything we believe we have reason to do. But such facts about our linguistic practices might in some cases be better explained by Gricean considerations rather than the falsity of desire-­as-­belief.5 I’ll illustrate both points above with an example of Scanlon’s. Scanlon might be thought to subscribe to a view like desire-­as-­belief (see, e.g., 1998, 7–8), but in fact he rejects the view. For example, he writes: I might see something good about drinking a glass of foul-­tasting medicine, but would not therefore be said to have a desire to do so.  (1998, 39)

This is the objection from underwanting, and it is unconvincing for the ­reasons given above. Let’s take our second point first: assuming that the bene­ fi­cial potential of the medicine is common knowledge, Scanlon might well be disinclined to assert that he wants to take the medicine. But his disinclination is best explained by the fact that this assertion is needless: if it’s common knowledge that the medicine will benefit him, the desire in question is too obvious to mention. Alternatively, imagine that the beneficial effects of the medicine are not obvious to Scanlon’s audience. If we imagine the case in that way, then it might well be appropriate for Scanlon to say that he wants to drink it. Imagine that I find what appears to be some foul-­smelling sludge, and go to throw it in the bin. Scanlon might reasonably interject: ‘No! I wanted to drink that!’. In this way, Scanlon’s example gets false traction by presupposing that it’s common knowledge that the drink is medicine: this 5  The claim in the main text is that we refrain from making certain assertions about our desires because, although true, they would have misleading conversational implicatures. Such implicatures are normally thought to be cancellable, in the sense that you can explicitly disown them (Grice 1989, 39; and see Zakkou  2018 for a helpful overview). If our disinclination to make the relevant assertions about our desires is explained by their misleading implicatures, we should be happier to make those assertions when the relevant implicature is cancelled. This prediction is accurate: you might be disinclined to assert that you want to go to the dentist, but happier to assert that if you qualify it with ‘well, I mean I do want to go, but only for the sake of my future health’. See also Chapter 2, footnote 14.

102 Irrationality renders the relevant desire too obvious to be worth mentioning. Vice versa, if we imagine it’s not common knowledge that it’s medicine, it’s clear that Scanlon could correctly be said to desire to drink it. Our first point above may also apply in this case. Scanlon claims that it wouldn’t be appropriate to say that he does want to drink the medicine, but this point gains rhetorical force by the natural thought that he could positively assert that he doesn’t want to drink it. But such an assertion would be ambiguous, and on a natural reading, would be appropriate insofar as it conveyed only the presence of a desire not to drink the medicine, rather than the absence of a desire to drink it. At least, this is the natural way of hearing that assertion if the foul taste of the medicine is remotely relevant: such a taste might compete with, but would not eliminate, a desire to drink the medicine, and so bears on whether you desire not to drink it, not on whether you desire to drink it. If Scanlon could assert that he doesn’t want to drink the medicine only as a way of conveying that he desires not to drink it, then his assertion doesn’t bear on whether he does have some positive desire to drink it. And in that case, the example fails to undermine desire-­as-­belief. In short, Scanlon’s example fails to threaten desire-­as-­belief. It is true that under some circumstances we might be disinclined to assert that we want to drink a glass of foul-­tasting medicine, but this can be explained by the Gricean point that we shouldn’t assert the obvious. And if we imagine that this desire is non-­obvious, it seems perfectly assertable. Scanlon’s example gets additional rhetorical force from the appropriateness of saying that we don’t want to drink foul-­tasting medicine. But that kind of assertion is consistent with desire-­as-­ belief, given that it surely conveys the presence of a desire not to drink it, not the failure to desire to drink it. In these ways, Scanlon’s example of underwanting fails to threaten desire-­as-­belief. The wider point is that these claims about Scanlon generalize. Many examples that are supposed to illustrate underwanting can be explained away by our two linguistic effects above. You might say that you don’t want to go for a run, to volunteer your weekends for charity, to give up your money for good causes, to stick to your diet, or to be brave and step out to face danger, despite believing that you have good reasons to do these things. But when you say that you don’t want to do these things you really mean that you want not to do them, which is consistent with desire-­as-­belief. And any remaining inclination you might have to not assert that you do want to do these things can be explained by the obviousness of such desires. The points above all pertained to the objection from underwanting; I now turn to the objection from overwanting. Our linguistic effects above have

Failures of Motivation  103 little direct relevance for assessing cases of this kind: those effects show that we want more than our verbal reports suggest, but the objection from overwanting is that we want too much for desire-­as-­belief to be true, not that we want too little. But a similar thought is nearby. In many putative cases of overwanting, it is doubtful that the relevant person really does think that they have no reason for doing the thing in question. Even in cases where we are intuitively irrational in our desires—desires to smoke, or drink—it seems that we do believe there is some reason to do the thing in question: it would be relaxing, fun, tasty, or whatever. It is surely true that you can desire something whilst thinking that it is on balance not what you ought to do, but this does not threaten desire-­as-­belief in the least. Just as we need to be careful with the objection from underwanting to ensure that supposed counterexamples definitely involve no desire, we need to be careful with the objection from overwanting to ensure that supposed counterexamples definitely involve no reasons beliefs. Let’s summarize the key claims in this section. To remind you, underwanting is when someone believes they have reason to do something but without wanting do it, and overwanting is when someone wants to do something without believing they have reason to do it. In many apparent cases of these kinds, the appearances are misleading. In particular, in many apparent cases of underwanting, the relevant desires exist but are just too obvious to be worth stating, and moreover, we can easily misinterpret assertions about negative desires as assertions about failures to desire. Similarly, in many apparent cases of overwanting, we can wrongly be misled into thinking that someone lacks a reason belief when in fact they have one but only think the reason is outweighed. In these ways, many putative counterexamples to desire-­as-­belief can be avoided.

5.5  Failures of Motivation I now turn to my third complementary response to the objections from underwanting and overwanting, and this response also serves as my primary response to the objection from akrasia. Recall the claims I made in §2.3: we should distinguish between desire and motivation. We may fail to be moved by our desires: desires are disposed to motivate us, but do not necessarily manifest this disposition. This distinction between desire and motivation provides further resources for desire-­ as-­ belief to avoid objections from underwanting, overwanting, and akrasia. In this section I explain the point

104 Irrationality with respect to the objections from underwanting and overwanting. Then, in §5.6, I apply the same points to make sense of akrasia. First, the objection from underwanting. Again, imagine Sally, who smokes but does so akratically, knowing she shouldn’t. Is Sally’s problem that she lacks a desire to quit smoking, or that she lacks the motivation? I submit that the latter is much more plausible. Imagine asking Sally: ‘Would you prefer to stop smoking or continue doing so?’ It would be incredible if Sally responded that she preferred to smoke, when she herself maintains that she has much stronger reasons to quit. She clearly does want to stop smoking: this is precisely why finding yourself in a situation like Sally’s can be frustrating. Precisely her problem is that although she wants to quit, she cannot motivate herself sufficiently to achieve this goal (her desire is failing to exercise its motivational power). This seems like a much more natural way to describe her problem, as well as the similar problems faced by those who think they should eat less, exercise more, and so on. In cases like Sally’s, the problem is that we lack the motivation to achieve our goals.6 To this extent, cases like Sally’s fail to undermine desire-­as-­belief: they illustrate mismatches between desire and motivation, not between reasons beliefs and desires. Similar reasoning permits desire-­ as-­ belief to handle numerous other putative counterexamples to the view, such as counterexamples involving people with depression (Stocker 1979, 744–6; Smith 1994, 119–21).7 Someone with depression might believe that they have reason to do something and yet not be remotely motivated to do it. But this is irrelevant to the truth of desire-­ as-­belief. Desire-­as-­belief only entails that reasons beliefs are sufficient for desire, not that reasons beliefs are sufficient for motivation, and as such is perfectly consistent with many cases that supposedly illustrate underwanting. To help cement these thoughts, it might be helpful to return to the distinction between desire and motivation. I said that motivation consists in having a set of dispositions: most obviously, to do the thing in question, but also to think about ways of doing it, and to think about how other actions might impede your doing it (§2.3). Desires, in contrast, consist in a slightly different set of dispositions: most centrally a disposition to be motivated (§2.3), but also dispositions to feel pleasure under certain circumstances (§7.6), to make various inferences (§8.2), and to make various assertions (e.g. about what you want). These two sets of dispositions might well come apart, 6  Again, the Stoics may have said something similar in defence of their view: see Galen in Long and Sedley (1987, secs 65J5–9 and 65T). 7 For claims similar to mine here, see Coleman (2008); Garrard and McNaughton (1998, 49); Swartzer (2015); Toppinen (2017).

Failures of Motivation  105 and indeed do so in cases of akrasia. I might have a desire that fails to manifest its potential to motivate me, but nonetheless clearly have that desire insofar as I have the more general disposition to be motivated, and have dispositions to feel pleasure under certain circumstances, to make various assertions, and so on. This explains how desire-­as-­belief permits that reasons beliefs can fail to motivate us. I now turn to the objection from overwanting. For example, if Gary is a bit of a glutton, couldn’t he want some cake whilst believing he has no reason to eat it? Cases like this may appear to threaten desire-­as-­belief. But that is not so. One initial point worth repeating is that Gary probably does believe that he has some reason to eat the cake: it would be tasty, after all. But this by itself does not wholly undermine the objection: perhaps Gary’s desire to eat the cake could be too strong given the reasons he believes he has to eat it. Most obviously, perhaps Gary could believe that he has much stronger reasons to diet than to eat the cake, but nonetheless more strongly desires to eat the cake than to diet. But again, with the distinction between desire and motivation in hand, this is not so clear. Above I said that desire and motivation can fail to align, so that you might fail to be motivated (might be undermotivated) to pursue something that you want. If we allow for that possibility, it seems as though we should also allow for the reverse possibility, where you are overmotivated by your desires (again, see §2.3). Sometimes, your desires may generate more motivation than they ought to. Plausibly, this is the situation that Gary is in: although he has only a comparatively weak desire to eat the cake—he thinks he only has a weak reason to do so—this desire is strongly motivating him to eat it. Desires dispose us to be motivated, but this disposition might overfire, so that we are motivated to a degree that is disproportionate, given the desire. This possibility is familiar with the functional roles of other mental states, as when you too confidently infer other claims from a belief you hold only tentatively, or when you too readily express a certain belief even when it is not conversationally helpful. Similarly, Gary might have some (perfectly reasonable) desire for the cake, but find this desire moves him too firmly.8

8  The comments above bear on comparatively everyday examples. A more extreme illustration of overwanting might be drug addiction. Finding a precise understanding of what is going on in real-­life cases of addiction is difficult, and so I am wary of tying desire-­as-­belief to any particular understanding of such a complicated phenomenon. But one option worth mentioning is a hybrid picture, on which addicts’ behaviour is driven in part by desires that are overmotivating them, and in part by compulsions which don’t even count as motivating them at all. Such a view might suggest that addicts are partially but not fully responsible for their behaviour.

106 Irrationality

5.6 Akrasia In summary, desire-­as-­belief is consistent with various putative examples of underwanting and overwanting. Some putative cases of these kinds might just involve irrationality in belief (§5.3). Moreover, we can be misled into thinking that our reasons beliefs fail to align with our desires by certain linguistic effects: by confusion about ascriptions of desire involving negations, and by the Gricean maxim that tells us not to state the obvious (§5.4). And desire-­as-­belief can accommodate other putative counterexamples by appeal to the distinction between desire and motivation: our motivations can diverge from our reasons beliefs, but that is consistent with desire-­as-­belief since our motivations can diverge from our desires (§5.5). With all these resources, desire-­as-­belief has plenty of scope for making sense of rational failures in desire and action, and in turn can emerge unscathed from the objections from underwanting and overwanting. These points all combine to allow us to make sense of akrasia, where your actions fail to align with your normative beliefs. The claims in §5.3 suggest that some apparent cases of akrasia might not be akratic at all, because we have irrational reasons beliefs that support our choice. For example, you might not speak to a troublesome student, even though you clearly should, because you are biased by self-­interest into believing that you are in fact not obliged to speak to them at all (see also Hare 1963, 76; Holton 2009, 98–103). The claims in §5.4 suggest that some other apparent cases of akrasia might not really be akratic at all because the person is merely reporting that they have some normative beliefs that militate against the action, not that they think it is wrong all-­things-­considered. For example, you might say that eating more cake is a bad idea, and yet continue to eat, because you think that eating is only somewhat bad, and think that the merits of eating more outweigh those costs. But §5.5 gives us our central explanation of akrasia, given desire-­as-­belief. Desire-­as-­belief says that you’ll most want to do things that you think you have most reason to do, and that is perfectly consistent with your failing to actually do them. Desires are disposed to motivate us, but they can fail to manifest that disposition. When they do so, we might find that our resulting actions fail to reflect our desires, and in such cases, we are akratic. This is what happens, for example, when you fail to maintain an exercise regime: you want to lose weight, but cannot summon sufficient willpower to put that desire into action.

Akrasia  107 To ensure these claims are clear, consider another way of presenting the objection from akrasia: Akrasia occurs when you fail to live up to your own normative beliefs, and so it therefore must occur because some other mental state motivates you to act. So akrasia occurs precisely when something that is not a normative belief motivates you to act, and the most obvious possibility is that it is your desires that do so. So we cannot simply identify our desires and our normative beliefs. Posidonius apparently criticized his fellow Stoics along these lines, and many other authors, ranging from Plato to Gary Watson, present similar arguments (Long and Sedley  1987, 65K3; Plato The Republic 436–9; Watson  1982; see also §1.7). But this line of argument is mistaken. Akrasia occurs only when you fail to live up to your all-­things-­considered normative beliefs: akrasia occurs when you fail to act as you think you ought. But you might do that as a result of acting on other normative beliefs that you have. In this way, we can maintain that akratic actions, despite being akratic, are nonetheless motivated by your normative beliefs. When you akratically eat another slice of chocolate cake, you are motivated by your belief that the taste is a good reason to eat it. This is irrational given that you think you have stronger reasons to diet, but your irrationality consists in your being motivated by the wrong normative belief, not in your being motivated by some other kind of motive. In this way, we can deny that akrasia consists in your being motivated by desire rather than by normative belief. Even akratic actions are undertaken in light of our beliefs about their merits. To put that in yet another way, consider one way to express the general puzzle raised by akrasia (e.g. Smith and Kennett 2004, 56–8; Watson 1977). The general puzzle about akrasia is the puzzle of explaining how akratic actions are both actions and also how they are akratic. If supposedly akratic acts are wholly driven by unendorsed and uncontrollable urges, they look more like bodily spasms than voluntary actions. On the other hand, if supposedly akratic acts are not driven by such urges but instead by our normative beliefs, it is unclear how they could count as akratic. Our view gives us the answer: akratic actions count as actions because they are motivated by reasons beliefs (=desires). And yet such actions are nonetheless akratic because they are motivated by the wrong reasons beliefs (=desires).

108 Irrationality That is, when we are akratic, we are motivated by beliefs which, if they were our only relevant beliefs, would fully justify what we do. The problem is not the local problem that we act on beliefs that fail to favour that action, but instead the global problem that those beliefs should have been overridden by other normative beliefs we had. For this reason, desire-­as-­belief is perfectly consistent with the existence of akrasia. There is still another worry I should address. How can we be motivated by the wrong normative beliefs? How do some beliefs about weak reasons nonetheless strongly motivate us? Perhaps this explanation will need to make reference to independent desires, and so desire-­as-­belief is hopeless after all. But that explanation would only postpone the problem, since we would still need to explain why some weaker desires strongly motivate, as I claim they can (§2.3). This claim can be further supported by the phenomenon of picking, as when you pick one specific tin of baked beans off the shelf, despite it being obviously no better than nearby identical tins (Ullmann-­Margalit and Morgenbesser 1977). In such a case you are moved to take a certain tin despite believing it no better than rivals. Must we therefore say that you wanted that tin more than the others? No. Rather, we should think of this as a case where you are uniquely motivated to do something despite having no uniquely strong desire for it. Moreover, there are very general reasons for expecting reasons beliefs to sometimes motivationally underfire or overfire. It’s quite generally true that dispositions can be masked so that they don’t manifest themselves—as when a fragile vase fails to break because it is padded—and clearly such masking is not always due to some intervening desire. Similar remarks apply within the realm of the mental: the laws of psychology are generally laws that have exceptions, and there is no general reason to think that this defeasibility will always be down to the interference of desires. For example, failures of inference are failures to manifest the dispositional tendencies of your beliefs, and such failures are rarely driven by desire. Equally, there are plenty of candidate explanations of why reasons beliefs might undermotivate or overmotivate that don’t appeal to our desires: for example, perhaps we could explain such mismatches by appeal to systematic biases in thinking, temporary failures of memory or reasoning, interfering hormones, failures of synapses to transmit information appropriately, or whatever else: take your pick. So long as you allow that mental states have their normal effects only contingently, and can see that there are numerous possible explanations for the exceptions that don’t appeal to our desires, there are no grounds for insisting that reasons beliefs could fail to motivate only

An Aside: Second-­O rder Desires  109 because of the interference of desire. It is surely true that when we explain your actions, we need to explain them with reference to your mental states. But when we come to explain why you were motivated by those mental states rather than by others, that explanation need not appeal to further mental states you have, and even when it does, that explanation need not appeal to desires that you have.

5.7  An Aside: Second-­Order Desires In this section, I briefly explore a different way that some people have tried to make sense of akrasia. They do so by appeal to second-­order desires. Second-­ order desires are desires about your own desires. The idea is that just as you might want psychological states such as pleasure or virtue, you might want psychological states that are themselves desires. For example, perhaps you desire [to desire exercise]. With this thought in mind, some have claimed that akrasia occurs when you don’t have the desires you desire to have, or else do have desires you desire not to have (Frankfurt  1988a,  1988b; Jeffrey  1974; Lewis 1989). For example, imagine that you don’t want [to exercise], but do want [to want to exercise]. Since your first- and second-­order desires fail to align, you will end up being akratic: you will act on the first-­order desire, but will wish you didn’t. I agree with the basic idea that akrasia occurs when there is a gap between what a person ‘really’ cares about, and what they actually do. This view makes sense of this contrast by claiming that what you really care about is a matter of your second-­order desires, whereas your actions are determined by your first-­ order desires. In contrast, I think it is better to make sense of this contrast by claiming that what you really care about is a matter of what you (first-­order) desire, whereas your actions are determined by what you are motivated to do. If we accept the distinction between desire and motivation, we can see that the appeal to second-­order desires is gratuitous: we can explain gaps between what a person really cares about and their actions, without appealing to second-­order desires. But worse, the appeal to second-­order desires is problematic.9 On the view in question, our second-­order desires track what we ‘really’ care about. But

9  Some people discuss not only second-­order desires, but also higher-­order desires more generally, since you might have desires about your second-­order desires, and so on. The arguments that follow could easily be generalized to cover views about akrasia that appeal to the agent’s highest-­order desires

110 Irrationality this is mistaken: our second-­order desires are unsuitable candidates for representing what we really care about, since they are sensitive to the merits of having the relevant first-­order desires, rather than to the merits of the objects of those first-­order desires (cf. Harman 2000, 130–1). To illustrate the point, imagine that a powerful and vain demon threatens to kill you unless you desire his company. He thereby gives you an incentive for having that desire, and so you might desire [to desire the demon’s company]. This is a clear example of a second-­order desire. But it is doubtful that this constitutes your ‘really’ caring about the demon’s company. If you want to have a desire merely because of the merits of having that desire, this shows nothing about your attitude towards the object of that desire. To this extent, it is deeply im­plaus­ ible to suppose that you really care about something just when you desire to desire it.10 In turn, it is implausible to suppose that akrasia consists in mismatches between your first- and second-­order desires: if you desire to desire the demon’s company, but don’t actually desire his company, that is unfortunate for you given his threat, but not akratic on your part. Indeed, it might be perfectly appropriate, given his terrible threat and terrible company. When outlining his view on free will, Frankfurt notes that we might want a desire merely because of some incentive for having that desire. He writes: Suppose that a physician engaged in psychotherapy with narcotics addicts believes that his ability to help his patients would be enhanced if he understood better what it is like for them to desire the drug to which they are addicted. Suppose that he is led in this way to want to have a desire for the drug.  (Frankfurt 1988a, 14)

Again, Frankfurt’s physician does not really care about the drug even though he desires to desire it. In light of this case, Frankfurt draws a distinction between second-­order desires and second-­order volitions. A second-­order volition is a desire that some desire be your will—a desire that some desire move you to act. It follows that not all second-­order desires are second-­order rather than just their second-­order desires: just stipulate that in the problematic cases in question, the relevant agents have only first- and second-­order desires, and no others. 10  One response to the objection in the text is to restrict our attention to only non-­instrumental second-­order desires (Lewis 1989, 115, footnote 4). But understood this way, the original suggestion loses much of its appeal. Though we have many instrumental second-­order desires—you might desire to desire to exercise, because this desire would indirectly improve your health—it is not remotely plausible that we have many non-­instrumental second-­order desires, if any at all. Certainly, none of the authors cited in the main text give a single example of such a desire, and it’s difficult to think of plausible examples. It seems unlikely, for example, that many people non-­instrumentally want to want to exercise; more likely they only instrumentally want this desire as a means to health.

An Aside: Second-­O rder Desires  111 volitions, because you might want to have some desire without wanting it to be your will: perhaps you want to desire Alfred but only because having this desire would make Bella jealous, and not because you want to actually act on this desire. Or, as with Frankfurt’s physician, perhaps you want to have a desire for a drug but only because having this desire will give you information, not because you want to be moved to consume the drug. Or, as with our demon, perhaps you want to desire the demon’s company, but only because having this desire would placate him, not because you want to actually act on this desire. With this thought in mind, we might say that what you really care about is determined by your second-­order volitions, rather than your second-­order desires. In turn, we might try and make sense of akrasia by appeal to second-­ order volitions, not all second-­order desires. But appealing to second-­order volitions rather than second-­order desires simply postpones the problem. Imagine that a demon threatens to kill you unless you are moved by a desire for his company. You might thereby want to [be moved by a desire for his company], and that still wouldn’t show that you really cared about his company, as Frankfurt seems to suggest, or that you are akratic if you fail to be so moved. Just as rewards and threats can change what you desire to desire, they can also change what desires you desire to be moved by, and thereby change your second-­order volitions. But in doing so, they needn’t make you care about the objects of those first-­order desires, or akratic if your first-­order desires fail to respond to these second-­ order volitions. So appealing to second-­order volitions does not solve the problem. In short, it is not plausible that we can make sense of akrasia by appeal to second-­order desires, because our second-­order desires are responsive to incentives in ways that what we really care about is not. The very same argument undermines the idea that we can make sense of akrasia by appeal to second-­order volitions. (In a note, I briefly connect these remarks with wrong kinds of reasons.11) 11  The reason given by the above threatening demon is known as a wrong kind of reason (see Gertken and Kiesewetter 2017 for an overview and further references). It is known by that name since it is a reason for having a certain attitude, but a reason given by the pragmatic benefits of forming that attitude. As such, it seems to be unlike standard reasons for attitudes, which normally focus on the intrinsic appropriateness of those attitudes, not their instrumental benefits. Indeed, I agree with those who say that wrong kinds of reasons are not reasons for attitudes at all, but instead reasons for acting: for acting in whatever ways might result in your having the attitude in question (Parfit 2011a, appendix A; Rowland 2019, sec. 6.1–6.5; Skorupski 2007a, 10–12; Way 2012). Desire-­as-­belief explains why second-­order desires track wrong kinds of reason: it says that desires are beliefs about reasons for actions (§1.2): that to desire p is to believe that you have reason to bring about p. So if you desire some attitude, such as a belief or a desire, desire-­as-­belief tells us that you think you have reason to act in such a way as to acquire that attitude. That just is the paradigmatic case of a wrong kind of reason: it is

112 Irrationality

5.8 Summary This chapter focused on the objections from underwanting, overwanting, and akrasia. I offered three complementary responses. First, desire-­ as-­ belief permits that our desires are often irrational. Second, many putative counterexamples to desire-­as-­belief are misleading insofar as they rely on false assumptions about which desires or reasons beliefs people have. These assumptions are generated by misleading assertions, and once we clarify those assertions, many putative counterexamples to desire-­ as-­ belief disappear. Third, we should distinguish between desire and motivation, and desire-­as-­ belief is certainly consistent with irrational shortages and gluts of motivation. This last claim is perhaps the most important one for making sense of akrasia. It is worth noting that it cuts the link between desire-­as-­belief and the view that knowing the good suffices for being a good person. Some might have thought that the driving force behind desire-­as-­belief is that all moral failures are failures of knowledge (cf. McDowell 1998; see also Socrates’ claims in Plato’s Protagoras 358b–d). But this isn’t true, and desire-­as-­belief doesn’t imply that it is. There certainly are many moral failures which are failures of knowledge, and certainly improving our normative beliefs is a good way of improving our choices. But we sometimes exhibit pure failures of motivation, and being a good person consists partly in avoiding those by displaying willpower and executing your desires correctly. Desire-­as-­belief is consistent with this truth. Our next chapter discusses two nearby rivals to desire-­as-­belief: the besire theory and the guise of the good. I aim to show that desire-­as-­belief is superior to both, and thereby how desire-­as-­belief escapes some objections to other views in the same vicinity.

not a reason to have the attitude, but instead a reason to cause yourself to acquire it. In turn, desire-­as-­ belief tells us that second-­order desires will always be desires that recognize wrong kinds of reason.

6 The Guise of the Normative This chapter contrasts desire-­as-­belief with two nearby rivals: the besire theory and the guise of the good. I do this in part to show that defenders of those views ought to instead adopt desire-­as-­belief. But primarily I do this to show how desire-­as-­belief can overcome some further objections. Many will think desire-­as-­belief belongs to the same broad tradition as these two views, and might think that it fails for similar reasons. But I instead show that desire-­as-­ belief can escape objections that undermine these views. §6.1 presents some simple reasons to prefer desire-­as-­belief to the besire theory. One further issue that divides the two views is that desire-­as-­belief, unlike the besire theory, commits to the guise of the normative, according to which you can desire p only if you have some favourable normative belief about p. The remainder of the chapter assesses this view. §6.2 briefly clarifies the landscape in this area by distinguishing two more specific versions of the guise of the normative: the guise of reasons and the guise of the good. Desire-­ as-­belief commits us only to the former. Then, §6.3 discusses cases where people are motivated by value-­independent duties, §6.4 discusses appropriately partial desires, and §6.5 discusses the contrasts between desiring, hoping, and wishing. In each case, I argue that the guise of the good comes off worse than the guise of reasons. I conclude that the besire theory is in some ways worse than desire-­as-­belief and in no way better, and that desire-­as-­belief, as I formulate it, should be dissociated with the guise of the good: it is not committed to the guise of the good, but instead to the guise of reasons, and that is a superior view.

6.1  Against Besires The besire theory is a little unclear. Some of the best-­known discussion— including the label—comes from critics who don’t fuss about every detail of the view (Altham 1986; Smith 1994, 118–25). Positive defence of the view is probably most strongly associated with John McDowell (1998), who doesn’t

Desire as Belief: A Study of Desire, Motivation, and Rationality. Alex Gregory, Oxford University Press. © Alex Gregory 2021. DOI: 10.1093/oso/9780198848172.003.0007

114  The Guise of the Normative use the label at all, and whose discussion is superb but somewhat difficult to pin down with precision. Let’s distinguish two possibilities. First, perhaps the besire theory says that ODM (Only Desires Motivate) is false, and that not only desires but also normative beliefs can motivate us. I think this is perhaps the best way to understand McDowell. In one sense, this view is comparatively orthodox since it permits that the distinction between beliefs and desires is useful and exhaustive, and it categorizes normative attitudes as beliefs. But in another sense, this view is ambitious, because it commits us against ODM. As I said in Chapter 2, that view is highly attractive, and widely accepted. To this extent, rejecting it is a serious cost that we ought to avoid if possible. For example, ODM fits with ordinary thought and talk in that it explains why it is always appropriate to explain actions by appeal to what people wanted, and offers an explanatorily powerful theory of all human action.1 So if this is the besire theory, we have implicitly discussed it already. So, second, if the besire theory is of any further interest, I think we can helpfully understand it in the following way. It says that in addition to beliefs, and desires, there is a third state of mind, with elements of each: besires. Besires are states with both directions of fit: they ought to be responsive to evidence, and also have motivational capacities (for the notion of direction of fit, see §1.4). With this idea in hand, the besire theory says that normative attitudes are besires, and that this permits them to be belief-­like in some ways, and desire-­like in others. As such, the view promises to treat a middle way between views on which normative attitudes are beliefs with no motivational capacities, and conativist views which face other objections. In this respect, the besire theory might share some of the advantages of desire-­as-­belief that I described in Chapter 3. But the besire theory and desire-­as-­belief nonetheless differ in significant ways. The most central difference is that the two theories deviate from orthodoxy in precisely opposite directions. To make this point, it will be useful to distinguish between type and token states of mind, where a type is a general repeatable kind, whereas a token is a single specific instance of some kind. For example, the word ‘bee’ has three token letters but only two types of letter. The orthodox view is that our actions are explained by two types of mental state: belief and desire (§2.1). Desire-­as-­belief agrees with the claim that our actions are explained by combinations of belief and desire, but since it treats 1  McDowell tries to accommodate at least the first of these thoughts by appeal to Nagel’s theory about desire (McDowell 1998, 79). But again, we should reject that: see §2.4.2.

Against Besires  115 desires as reasons beliefs, it thereby says that our actions are explained by combinations of beliefs alone (see also §3.5). That is, desire-­as-­belief agrees that our actions are explained by two token states of mind, but it also claims, against the orthodoxy, that those tokens are of the same type. In this way, desire-­as-­belief is more parsimonious than the orthodoxy. Desire-­as-­belief explains everything that the orthodoxy explains, whilst positing fewer types of thing in the world. Now contrast this with the besire theory. It deviates from the orthodoxy in precisely the opposite direction. It says that our actions are explained by three types of mental state. For according to the besire theory, some actions are explained by our desires combined with our beliefs, and other actions are explained by our besires combined with our beliefs. The besire theory treats these three as all distinct types: all beliefs have one direction of fit, all desires another, and all besires both at once. In this way, whereas desire-­as-­belief is more parsimonious than the orthodox theory, the besire theory is less parsimonious than the orthodox theory. It says that to explain the full range of human action, we need to appeal to beliefs, desires, and besires. Not only is this unparsimonious, but worse, besires are no part of ordinary thought and talk. When philosophers refer to our beliefs and desires, that has a clear basis in everyday thought and talk about what people think or want. But it is far less clear that besires have any place in everyday thought and talk about our mental lives. To this extent, the besire theory is committed to the idea that ordinary thought and talk is terribly under-­resourced, since we lack a word for one of our most central states of mind. That is obviously a possibility, but its truth would be surprising. In short, the besire theory differs from desire-­as-­belief because it is less parsimonious and posits states of mind that are no part of ordinary thought and talk. In these respects, desire-­as-­belief seems superior to the besire theory. But the two theories also differ in a second way. Desire-­as-­belief says that all desires are beliefs. In contrast, the besire theory says that desires—unlike besires—are not belief-­like at all. We can express this difference between the two views by saying that desire-­ as-­ belief, unlike the besire theory, is committed to the guise of the normative, the view that you can desire p only if you have some favourable normative belief about p. In this respect, the besire theory seems unlikely to benefit from the argument that I presented in Chapter 4, since that argument supported views on which all desires represent their objects in some favourable way. Still, the guise of the normative is sometimes thought to be implausible. The remainder of this chapter discusses this implication of desire-­as-­belief in greater detail.

116  The Guise of the Normative

6.2  The Guise of the Normative In fact, this chapter is not a comprehensive discussion of the guise of the normative, because some other chapters of this book discuss related issues. Some objections to the guise of the normative appeal to cases of overwanting of the sort I discussed in Chapter  5. Other objections to the guise of the normative focus on appetites, such as hunger, which I discuss in §7.2 (see also §7.5). Still other objections appeal to the desires of animals, which I discuss in Chapter 9. In the remainder of this chapter, I set these issues aside, and focus solely on three other issues surrounding the guise of the normative. I said that desire-­as-­belief commits us to the guise of the normative. To remind you, here is that view again: The guise of the normative:  You can desire p only if you have some ­favourable normative belief about p.2 The basic idea behind the guise of the normative is that no matter how foolish our desires might be, their objects must seem at least somewhat worthwhile to you at the time. For example, imagine that you want to date Sam. Perhaps onlookers all agree that your relationship would be disastrous for both of you. Still, if you want to date Sam, you must think that there is at least something positive about that state of affairs. This chapter largely focuses on a choice between two more specific versions of the guise of the normative: The guise of the good:  You can desire p only if you believe p is good. The guise of reasons:  You can desire p only if you believe that you have reason to bring about p. Many historical defenders of the guise of the normative have defended the guise of the good, or something close enough (Anscombe 1963, 75; Aquinas ST I–II.1.1, DV 24.2; Davidson 2001b, 22–3; Kant G5:59–60; Raz 2000, 2010). But desire-­as-­belief commits us to the guise of reasons, not the guise of the 2  The guise of the normative says that our desires are constrained by our beliefs. Some other versions of the guise of the normative say that our desires are constrained by other representational states, such as by perceptual states. But I rejected such views in §4.4. Another option is to say that our desires represent their objects with normative force, rather than having normative properties as part of their content (Schafer 2013; Tenenbaum 2008). On that view, see Chapter 1, footnote 10.

The Guise of the Normative  117 good. This might seem like a minor difference—after all, reasons and goodness correlate to a large degree: in normal cases, reasons are reasons to secure certain goods. To that extent, the guise of reasons and the guise of the good might seem very close. But reasons and goodness don’t perfectly correlate, and anyway, what matters for the guise of the normative is what agents think, not what is true. I say that in cases where agents believe— however mistakenly—that what is good differs from what they have reason to do, their desires follow their beliefs about reasons, not goodness. Many will associate desire-­as-­belief with the guise of good, and be sceptical of it for that reason. But I say that the failings of the guise of the good are not shared by the guise of reasons, or in turn, by desire-­as-­belief. To this extent, when we say that desire-­ as-­ belief commits us to the guise of the normative, this is potentially misleading since it nonetheless allows us to reject the guise of the good. Before we move on to present the relevant arguments, two final clarifications: First, those who defend the guise of the normative often defend it as a standalone constraint on desire, rather than as a downstream implication of a view like desire-­as-­belief. This might seem to rest on mere optimism about human nature. In contrast, I defend the guise of the normative as an implication of desire-­as-­belief. This is potentially more satisfying: desire-­as-­ belief explains the guise of the normative. If desires are normative beliefs, it is no surprise that whenever you have a desire, you have a normative belief. Second, some treat the guise of the normative as a constraint on intentions or actions rather than on desires (e.g. Raz 2000, 23; 2010, 111; Setiya 2010, 16; Sussman  2009, 614). I instead accept the guise of the normative most fundamentally as a claim about desire. Still, I said that intentions and actions are constrained by our desires (§2.4.4, §2.1). As a result, I accept that our intentions and actions are also constrained by our normative beliefs.3

3  Still, the order of explanation here is important. Kieran Setiya has argued against the guise of the normative by claiming that it lacks explanatory power (Setiya 2010, pt. 1). He argues that it fails to explain self-­knowledge of our actions (see especially 2010, 50–1). But if we understand the guise of the normative as primarily a constraint on desire, we should not expect it to explain that. We can say that although the guise of the normative constrains actions, it does so only because it constrains desire, and in turn say that we should not expect the guise of the normative to directly explain facts that are distinctive of action but not desire. We might then explain the relevant facts about action in wholly independent ways. Setiya seems to suggest that if the guise of the normative lacks explanatory power in this respect, then it should be rejected. But to undermine one argument for a view is not to undermine the view itself.

118  The Guise of the Normative

6.3  Reasons without Goodness In this section I present a series of counterexamples to the guise of the good. I am not sure that any of them is decisive by itself, but they nonetheless collectively make a case for rejecting the guise of the good. In contrast, none undermines the guise of reasons. Take the following two examples: Suppose that I promise my children that I will tie my right shoelaces before my left shoelaces on alternate days of the week if they will do their homework without fuss. One can imagine arguing that though I ought to tie my right shoelaces before my left shoelaces today, since I did the opposite yesterday, my doing so has no value of any form.  (Dancy 2000, 168) Althea is a lovely person, who lives in ways that can only be described as admirable. As a theoretical matter, she believes that the only legitimate sense of ‘good’ is the attributive sense, where it means that something meets the internal standards for things of its kind (e.g. Geach 1956). In fact, she further believes that actions cannot be good in this sense, because there are no such standards (Thomson  2015, 25).4 Thankfully, she still thinks there are some acts she has reasons to perform: she just believes that reasons are independent of goodness.

It seems that Dancy might desire to tie his shoelaces in a way that he does not believe good, and that Althea might desire to perform various acts even if she doesn’t believe they are good. These are not the standard sorts of counterexample to the guise of the good, to be sure. It is tempting to think that counterexamples to the guise of the good, like those we examine below, will appeal to people with bad desires. But in these examples, we have people who are committed to doing the right thing. The problem for the guise of the good is that these people believe that the right thing to do is not a good thing to do. The guise of the good implies that no-­one committed to such a view could desire to do the right thing. But this looks hopelessly implausible. Regardless of whether such views are true, they are at least views which people can hold and be motivated by. But clearly, this same objection does not undermine the guise of reasons. The guise of reasons says that people must 4  A similar example could be described in which Althea accepts that actions can be good in this sense, but just thinks that those internal standards are irrelevant to what she has reason to do (cf. Enoch 2006).

Reasons without Goodness  119 believe they have some reason to do the things they desire to do, and that is true in these cases. With this point in mind, let’s turn to a better-­known objection to the guise of the good. David Velleman writes: Consider the figure of Satan in Paradise Lost, who responds to his defeat with the cry, ‘Evil be thou my Good.’ Satan is here resolving to desire and pursue evil, and hence—as he himself puts it—to regard evil as good. But he cannot reasonably be interpreted as adopting new estimates of what’s valuable—that is, as resolving to cease judging evil to be evil and to start judging it to be good. If Satan ever loses sight of the evil in what he now desires, if he ever comes to think of what he desires as really good, he will no longer be at all satanic; he’ll be just another well-­intentioned fool. The ruler of Hell doesn’t desire what he wrongly thinks is worthy of approval; he desires what he rightly thinks isn’t. (Velleman 1992, 18; see also Dreier 1990, 10–11)

This case seems to undermine the guise of the good: Satan desires to do things whilst failing to believe that they are good. Defenders of the guise of the good might respond as follows. They might say that things can be good in a number of different ways, and so claim that even if Satan sees no moral good in what he desires to do, he must still see some other good in it (Alvarez 2010, 218–19; Anscombe  1963, 75; Tenenbaum  2007, 252–3). Perhaps he sees something good in demonstrating his liberty, in being a rebel, or something similar. But it is far from clear that this response can render the guise of the good compatible with a plausible interpretation of Satan’s beliefs and desires. We are being offered a view according to which Satan wants to act in ways that he believes are morally bad but good in some other way. So the overall picture being presented here is that Satan is in the common position of believing that certain goods conflict—being moral, and being free—and believes that the latter kind of good outweighs the former. But this seems to understate the divergence between us good folk and Satan. After all, we might agree that liberty is good, that rebellion can sometimes be worthwhile, and so on. Is the only remaining difference between us and him the respective weights we assign to these kinds of goodness? This seems implausible. In contrast, the guise of reasons has a much more plausible story to tell about Satan. It seems plausible that Satan desires to act in ways that he believes he has reason to act in. The kernel of truth in the above response is that Satan has a rationale for his desires, and to this extent it is true that he believes that

120  The Guise of the Normative he has some reason to act in these ways. But we can say this and agree that he is pursuing badness - not that he is merely prioritizing one kind of goodness over another, but that he is prioritizing the bad, in general, over the good, in general.5 I now turn to yet another counterexample to the guise of the good. Stocker writes: Just as there are desires and appetites directed at harming others, there are desires and appetites directed at harming oneself. In certain moods, such as the self-­directed modes of disgust, hatred, guilt, shame, I may seek to humble, abase, or harm myself. (Stocker 1979, 748; see also Velleman 1992, 17–18)

It is again doubtful that the guise of the good can make sense of such cases: such desires seem precisely aimed at making things worse. But again, the guise of reasons is not impugned by these examples. It is plausible to suppose that if you genuinely desire to harm yourself in this way, this must be because you believe that you have reason to do so: because you think this is the appropriate action in these circumstances (or at least, this is what you think whilst you have the desire). It is worth comparing this case with the quote from Dancy, above. One natural way of understanding Stocker’s case has the person desiring to harm themselves because they think it is somehow fitting that their life goes worse. In that respect, this attitude is just like the deontologist who believes that some action is right independently of its goodness. This thought also extends to Gary Watson’s squash player who wants to hit their victorious opponent with the racquet (Watson  1982, 100–1). Watson persuasively suggests that they might want to do this without thinking that it would be good for their opponent to suffer. But again, it is tempting to think that the squash player believes (however briefly) that their opponent deserves 5  A superficially similar example: in Saint Augustine’s Confessions he seemingly says that he stole a pear precisely because he saw it as wrong (Book II, especially §4 and §6; see also §8 and §9). This might seem like a problem for the guise of the normative, and perhaps the guise of reasons. But the exact interpretation of the case is a little unclear. Augustine himself seems to endorse the guise of the normative in some form when he writes, ‘I have never seen a man purposely throw out the good when he clears away the bad’ (Book I, §7). And when he describes the pear case, he actually writes that ‘our real pleasure consisted in doing something that was forbidden’ (Book II, §4, my emphasis). That is consistent with the guise of the normative (in either form) given that it is a view about desire, not pleasure: it permits that someone might find wrongdoing pleasurable and do wrong for the pleasure of it (see also §7.3). At any rate, it is not obvious to me that we should treat Augustine’s description of events as theory-­neutral data that any theory must accommodate, given that it is clearly driven by his religious convictions on related matters.

Correct Variation in Desire  121 to be hit in the face, and a natural interpretation of this is that they believe they have reason to harm their opponent. Here, too, it seems that this is a counterexample to the guise of the good but not the guise of reasons. In short, the cases above undermine the guise of the good. But on the other hand, they are all consistent with the guise of reasons. In this way, we should be careful with objections to desire-­as-­belief. Some objections may look forceful only if we confuse reasons and goodness, and wrongly presuppose that desire-­as-­belief commits us to the guise of the good.

6.4  Correct Variation in Desire I now turn to a second advantage of the guise of reasons over the guise of the good, and in turn a second way in which some objections to desire-­as-­belief might miss their target. Desire-­as-­belief clearly implies that our desires can be correct or incorrect, depending on whether their objects are in fact things we have reason to bring about.6 The guise of the good and the guise of reasons have similar implications, depending on whether the objects of our desires are good, or are things we have reason to bring about. Both views say that if you desire p, you must have a corresponding belief which may or may not be correct. With this thought in mind, we can ask a simple question: are the same desires correct for everyone, or is the correctness of a desire relative to the person whose desire it is? Imagine that I would enjoy chocolate, but you wouldn’t. Under these circumstances, it seems that it would be correct for me to want it, but not you. This might suggest that different desires are correct for different people. But this is unclear, since we might think there are actually two different desires here: It’s correct for me to desire that I (AG) eat some chocolate, and incorrect for you to desire that you (NN) eat some. So it’s not obvious that we here have an example of a single desire which it’s correct for me to have but not for you to have. But other examples do show that different desires are correct for different people. Imagine that you ought not be concerned with my wellbeing, or else that you should, but to a lesser degree than me. Then it would follow that your 6  To be clear, by ‘correct’ I don’t mean rational. I think that our desires are rationally assessable, but that is not the claim I need here. Some prefer to describe beliefs as ‘true’ rather than ‘correct’. But the real bearers of truth and falsity are propositions, and so I think it clearer to say that beliefs are correct depending on whether their contents are true. (On the objection to desire-­as-­belief that we never describe our desires as ‘true’, see §1.4.)

122  The Guise of the Normative desire that I (AG) eat chocolate should be weaker than my desire for that chocolate. Similarly, it’s correct for me to desire to buy my girlfriend an anniversary present, but it seems like a further independent question whether it’s correct for you to desire that I do so: perhaps you are a rival for her love and so might correctly hope that I forget (cf. §3.3.2). Or for a final example, imagine that my mother and your mother are each drowning but just one can be saved. It seems that it might be correct for me to prefer that my mother gets saved, and correct for you to prefer that your mother does so. Cases like these show that the correctness of a desire is relative to the person whose desire it is.7 If different desires are correct for different people, that favours the guise of reasons over the guise of the good (see also Schroeder and Milona 2019). This is because reasons are themselves agent-­relative, whereas goodness is not. That is, different people have reasons to bring about different things. The guise of reasons therefore predicts that some desires might be correct for you but incorrect for me, because only you have reasons to bring about those things. In contrast, what is good is not relative to different people (more shortly). The guise of the good thereby wrongly entails that the same desires are correct for everyone. To illustrate these claims, let’s return to the final example above. When my mother and your mother are each drowning, it seems that we each have a stronger reason to save our own. So the guise of reasons tells us that you and I might correctly desire different things, since we have different reasons. In contrast, it is doubtful that the guise of the good entitles us to the same conclusion: it is either best that my mother be saved, or best that yours is. In the former case, the guise of the good says we should both desire that my mother be saved, and in the latter case, the guise of the good says we should both desire that your mother is. In neither case can it be correct for us to have different desires. In this way, such cases show that the guise of reasons is more plausible than the guise of the good. We should address possible replies to this argument for the guise of reasons. But before we do so, let’s note two nearby thoughts. First, reasons can be time-­ relative as well as person-­relative. After all, if I have some reason that you

7  It’s important not to confuse this agent-­relativity with the claim that desires are not subject to any standard of correctness whatsoever. I say that some desires really are correct, but it’s just that what is correct for you might not be correct for me. In this way they are like beliefs about where you live: there is a perfectly objective fact of the matter, but that fact differs for each person. That different desires are correct for different people is perfectly consistent with the objectivism that underwrites desire-­as-­belief.

Correct Variation in Desire  123 don’t because Kim is my friend but not yours, then I can also gain or lose that reason over time depending on whether Kim is my friend at that time. Insofar as this is true, the guise of reasons rightly entails that the correctness of a desire is time-­relative. This seems plausible: you might be correct to desire your child’s existence to a greater degree after they arrive.8 (Indeed, your earl­ ier weak desire might be correct even though you know you will in future be correct to have the stronger one, just as you might correctly think that you have no obligations to read the work of a student over the summer even though you know you will have that obligation once the course begins.) Again, these are thoughts that the guise of the good will struggle to make sense of. Second, the guise of the normative gives a necessary condition on desiring, not a sufficient condition. But beyond this chapter, our broader focus is desire-­as-­belief, and desire-­as-­belief commits us to the stronger claim that certain normative beliefs are not only necessary but also sufficient for desire. Further examples, parallel to those above, show that beliefs about goodness are not sufficient for desire. I have in mind cases where we recognize that something is good but not in a way that is relevant for us personally (cf. Baker  2017, 590). For example, I can see that it’s good to understand chemistry. But this project has insufficient bearing on my life to give me significant reasons to pursue this good, and as a result I might acknowledge the goodness of chemistry but be left relatively cold by it. Similarly, you might think that it would be good to donate money to charity. But you might think that this is not your responsibility, and so be not very inclined to give any money away. Cases like these undermine the view that beliefs about goodness suffice for desire, but are consistent with the claim that beliefs about reasons suffice for desire. They thereby favour my formulation of desire-­as-­belief over nearby formulations that identify desires with beliefs about goodness. Insofar as we do not have reasons to respond to everything good in the world, it seems that our desires ought to track our partial reasons rather than impartial goodness. Let us now turn to replies to the main argument above: I said that the guise of reasons is superior to the guise of the good because only the former can permit that different people can be correct despite having different desires. How might defenders of the guise of the good respond? Their best option is to maintain that goodness is also agent-­relative: that what is good can also vary 8  See also Harman (2009) and Wallace (2013, especially chapter 3.1), and for other nearby issues see Hedden (2015, chap. 4); Parfit (1984, pt. 2).

124  The Guise of the Normative from person to person.9 Then the guise of the good could explain why it’s correct for different people to have different desires, just as the guise of reasons can. But whereas it is highly plausible that different people have reasons to do different things, it is not plausible that different things are good relative to different people. (In a long note, I address two other possible defences of the guise of the good.10) It’s easy to get distracted into thinking that goodness can vary from person to person by two thoughts, neither of which is to the point (cf. Schroeder  2007b). First, something can clearly be good according to a particular person (but not others), as when I say that the Holocaust was good, according to Hitler. But this really means only that Hitler believed the 9  One possible source of commitment to agent-­relative goodness is the fitting-­attitudes account of goodness (most famously, Scanlon  1998, 97; for an overview, see Jacobson  2011). Such views have many defenders, but also face serious worries (e.g. Bykvist  2009; Gregory  2014; Olson  2009; Rabinowicz and Rønnow-­Rasmussen  2004, 2006). Moreover, if the fitting attitude in question is desire, such views cannot be combined with the guise of the good without circularity. Further, even if such accounts allow us to construct a theoretically respectable notion of agent-­ relative value, the guise of the good says that desires are constrained by ordinary thoughts about goodness, had by all creatures with desires. As a result, fitting-­attitude accounts supply a notion of agent-­relative goodness that is too removed from ordinary thought to be of use for defenders of the guise of the good. 10  First: Graham Oddie says that desires somehow present things as good from a certain perspective (Oddie 2005, 60–3, 218–26). Oddie pursues this as part of his view on which desires are more akin to perceptual states than beliefs, which I argued against in §4.4. But the relevant idea here can stand independently of that (perhaps desires are beliefs about how good things are, from your perspective). On Oddie’s view, your perspective is determined by your ‘distance’ from a good, where distance is determined by the ‘location’ of those goods, in terms of who they affect, and when (Oddie  2005, 220–1). But goods have no literal location and we have no literal distance from them. As a result, it is not clear that Oddie has made progress on the problem rather than merely labelled it: the guise of the good can’t say that desires represent goodness (since that is agent-­neutral), and must instead say that desires represent some agent-­relative thing related to goodness. Labelling that as ‘goodness from a certain perspective’ doesn’t seem like a way of making progress on the question of what this thing is. Of course, we might mean that desires are beliefs about goods insofar as they are relevant for what we have reason to do, but this just is the guise of reasons, or near enough. Second: Allan Hazlett argues that we should adopt the guise of the good, and just allow that the same desires would be correct for everyone (Hazlett 2018; for a similar strategy in another context, see Olson 2009). But he softens this by claiming that a virtuous person would have many incorrect desires. On his view, virtuous people have incorrect desires, in just the same way that a virtuous parent might have an inaccurate estimation of their child’s abilities: on some matters, the virtuous thing to do is to have attitudes which are (clearly) incorrect. Hazlett hopes to thereby show that desires represent the good, and claim that the same desires are correct for everyone, but nonetheless permit that in a different sense, different people ought to have different desires. But this cannot be right either. When we form desires, we do so only by appeal to their (supposed) correctness, not other considerations, such as financial payoffs (cf. Shah  2003,  2006). For example, I cannot get you to desire to drink something by offering you money to form that desire, but only by making it seem to you that you have good reason to actually drink it (see also §5.7, §8.2; Kavka 1983; Millgram 1997, chap. 2). But when we deliberate about what to desire, we clearly can take into account our partial commitments: you might reflect on what you want and come to desire new trainers, because they would benefit you. According to the guise of reasons, this is perfectly legitimate and consists in your aiming to form correct desires. In contrast, on Hazlett’s view, this is akin to your forming a desire in order to secure some pragmatic benefit, rather than because it is correct. But again, we should doubt that considerations other than (supposed) correctness can influence deliberation in this manner.

Wanting, Wishing, Hoping  125 Holocaust to be good, not that it was actually good relative to him. So this can’t really give a standard of correctness for desire. A second thing that can mislead us into thinking that goodness can vary from person to person is the concept of something being good for someone, as when my being your boss is good for me but not good for you. In one obvious sense, this kind of goodness varies from person to person. But the sense in which this is true is also not helpful for defenders of the guise of the good. For then their view would be that when we desire things we must believe them to be good for ourselves. But this kind of egoism is rightly widely rejected. When I desire that my mother gets saved, I don’t do so because I believe it is good for me. So this thought cannot explain why it is correct for me but not you to desire that my mother gets saved. In summary, the guise of reasons is much better placed than the guise of the good to explain why different desires are correct for different people. The guise of reasons can explain this because different people have reasons to do different things. In contrast, it seems as though goodness is agent-­neutral. Obvious responses on behalf of the guise of the good fail. Again, the more general lesson is that whereas there are grounds for rejecting the guise of the good, those objections do not undermine the guise of reasons, or desire-­as-­belief.

6.5  Wanting, Wishing, Hoping I now turn to a final difference between the guise of the good and the guise of reasons. Take the following principle (Streumer 2007; see also Lord 2015): Reason-­implies-­can:  You have a reason to v only if you can v. Reason-­implies-­can captures attractive thoughts such as that we have no ­reasons to change the past, that inanimate objects have no reasons at all, and that you cannot be criticized for failing to do things you could not do. But for our purposes, the most important thing is not the truth of the reason-­implies-­ can principle, but instead the psychological fact that most of us ordinarily accept some principle like this, and as a result at least normally believe that we have reasons to v only if we can v. The guise of reasons thereby implies that most of us do not desire things that we believe are impossible or else out of our control, since most of us deny that we could have reasons to bring such things about.

126  The Guise of the Normative In contrast, there seems to be no parallel constraint on goodness. Options that are closed to you might nonetheless hypothetically be good, or else have good consequences. I might have no reasons to change the past, but we can nonetheless agree that it would be good if I did. In this way, the guise of the good, unlike the guise of reasons, permits that we might desire things that we believe are impossible or else out of our control. In short, the guise of reasons, unlike the guise of the good, seems to entail that we have no desires for things we believe are impossible or out of our control. Is that plausible? If so, this might be a further mark in favour of the guise of the reasons. If not, this might be an objection to the guise of reasons and in turn to desire-­as-­belief. To settle this, I first need to take a slight detour, and explain the distinction between wanting, wishing, and hoping. We wish for things when we want them, but believe them to be impossible. That is the natural way of understanding my wishes to own magic wellies, to see a dinosaur, and to visit space. Many have then claimed that hoping is a kind of halfway house to wishing, in that hoping for something consists in wanting it but believing it unlikely (Anscombe  1963, 67; Hume T2.3.9.6; Searle 1983, 32; see also Hobbes L I.VI.14). But though this account of wishing seems right, the account of hoping seems false: you can hope for things that you don’t think are improbable. For example, I think it probably will be dry tomorrow, but I nonetheless hope that my expectations are fulfilled. Equally, I hope I get some good presents for Christmas this year, even though I don’t find that especially improbable. And a moment’s reflection makes me hope that you, the reader, are happy, and I am utterly clueless about the probability of that. In light of examples like these, it is more plausible to say that to hope for something, you need not think it improbable, but rather you must think that it is out of your control (see also McGeer 2004; Meirav 2009). This fits with the examples above: my hopes regarding the weather, Christmas presents, and your happiness are hopes because I don’t have control over these things. A common expression of a hope is the futile gesture of crossing one’s fingers: we do this in full knowledge that really, nothing we can do will help satisfy our hopes. In a note, I briefly reply to two objections to this view.11 11  One worry for this view is that there is some further more substantial kind of hope, with other features (Pettit 2004; cf. Martin 2013). I’m happy to permit that hope might come in other kinds too; I need only identify one important kind of hope. (On this, it’s helpful to think about fear—when I say, ‘I fear that we’ve missed the bus’, this is a legitimate use of ‘fear’ even if the state expressed is very unlike more substantial fears such as phobias.) A second worry for this view is that it fails to distinguish hope and despair: you might want p and believe it possible but out of your control, and yet despair about this rather than hope for p (e.g. Meirav  2009, 222–5). Here I think it best to say that despair is

Wanting, Wishing, Hoping  127 In short, we should think of wishes as characterized by our believing that their objects are impossible, and hopes as characterized by our believing that their objects are out of our control. Now return to our dispute between the guise of reasons and the guise of the good. It seems that we characterize our wants as wishes and hopes precisely in those cases where we could not have a reason to pursue the thing in question: when it is impossible or else out of our control. In this way, the guise of reasons, when combined with our acceptance of the reason-­implies-­can principle, makes perfect sense of why we distinguish wants and wishes and hopes: we wish or hope, rather than want, just because we cannot have reasons to bring things about if they are impossible or out of our control. In this way, the distinction between wants, wishes, and hopes vindicates the guise of reasons by showing how our desires are indeed constrained by our acceptance of reason-­implies-­can. And vice versa, the guise of reasons does a good job of explaining why we distinguish wants from wishes and hopes: our wants, unlike our wishes and hopes, are more tightly related to what we are actually able to do.12 In contrast, the guise of the good fares less well (see also Velleman 1992, 12–17). Defenders of that view will probably say that wishes and hopes are beliefs about goodness combined with the relevant beliefs about possibility. But given their view about desire, it seems somewhat arbitrary that we have names for these exact combinations of attitudes. I would not want to claim that this shows the guise of the good to be false, but it is clear that the guise of reasons explains more than the guise of the good: it makes better sense of why we draw these distinctions. Still, there is a natural worry lurking here. Though we should surely draw some distinctions between wanting, wishing, and hoping, don’t those three attitudes also have something in common, so that there are interesting connections between them? Yes. But defenders of the guise of reasons—and of desire-­as-­belief—can agree. The basic point is that it would be in the spirit of desire-­ as-­ belief to analyse wishes and hopes as some other kinds of normative belief. By doing that, we would capture the natural idea that there is some similarity between wanting, wishing, and hoping—all of these are

consistent with hope, and in this case both coexist. More generally, I say that unless you believe p is impossible, despairing about the chances of p entails that you hope that p. To whatever extent this claim sounds implausible, that can be accounted for by the previous admission that there is another richer notion of hope, and that might well conflict with despair. 12  See also §1.1 on the reduction of desiring [to __] to desiring [that __]. My claims here might make it more viable to say that all desires are desires [to __], though I think it better to instead treat my claims here just as placing restrictions on which propositions one can desire.

128  The Guise of the Normative normative beliefs of some kind. But we could nonetheless also capture the fact that these are different attitudes—they are all different normative beliefs. If we identify desires with reasons beliefs, what normative beliefs could we identify with wishes and hopes? There are a variety of options here, and I’ll neither give an exhaustive list of possibilities nor plump firmly for any specific view. My claims in this book are largely about desire, and desire-­as-­belief could be combined with a variety of nearby views about wishing and hoping. But let me mention one option. We might say that whereas desires are beliefs about the reasons you actually have, wishes and hopes are beliefs about the reasons you would have under other, more ideal, circumstances.13 That is, we might say that whereas desiring that p is believing that you do have a reason to bring p about, wishing that p is believing that you would have a reason to bring p about if it were possible, and hoping that p is believing that you would have a reason to bring p about if it were under your control. On this view, desires are distinct from wishes and hopes, but the attitudes are nonetheless extremely similar: whereas desires are beliefs about actual reasons, wishes and hopes are beliefs about counterfactual reasons. But again, other views are possible and I need not commit too firmly. So long as desire-­as-­belief is a plausible claim about desire, and so long as we agree that if desire-­as-­belief obtains, it would be plausible to hold some similar view about wishing and hoping, that would be enough to overcome any potential problem here. My claims elsewhere don’t hinge on what we might say about wishing and hoping, and for that reason I see no reason to investigate the issue any further here. In short, the guise of reasons—and in turn, desire-­as-­belief—promises to make good sense of the distinction between wanting, wishing, and hoping. We make this distinction because it tracks an important fact about reasons: that we have reason to do something only if we can do it.

6.6 Summary In this chapter I first argued that desire-­as-­belief is more plausible than the besire theory because it is more parsimonious and less revisionary to common sense. However, unlike the besire theory, desire-­as-­belief commits us to the guise of the normative. So second, I explored the guise of the normative. I 13  Or perhaps we should say, more carefully, that they are beliefs about the reasons you are disposed to have, thereby avoiding worries about the conditional fallacy.

Summary  129 contrasted the guise of reasons with the guise of the good. I claimed that there are numerous problems for the guise of the good, all of which are avoided by the guise of reasons. Since the guise of reasons is sustainable, desire-­as-­belief remains superior to the besire theory. And desire-­ as-­ belief should be dissociated with the guise of the good—it avoids the best objections to that view. Our next chapter turns to some mushy stuff. Is desire-­as-­belief a theory for the English alone, or is it also plausible for those who have feelings?

7 Desire and Feeling Desire-­as-­belief might seem objectionable because desires clearly interact with our feelings in numerous ways, whereas beliefs are intellectual states which might seem to have no special relationship to our feelings at all. So in this chapter, I explore a series of objections to desire-­as-­belief that hinge on the relationship between our desires and our feelings. In each case, I show how desire-­as-­belief explains that relationship. §7.1 addresses the relationship between emotion and desire. §7.2 addresses appetites such as hunger. §7.3 discusses the more general question of how prospective pleasures relate to our desires, and §7.4 uses those claims to explain how our desires can fluctuate over time. §7.5 addresses the special case of ascetics, who shun pleasure. §7.6 addresses the effects of desire on pleasure, as when my heart sinks or soars depending on whether journals accept my papers.

7.1  Emotions Causing Desire Shortly after we moved into our present house, I woke in the middle of the night to hear my partner say, ‘Alex, there’s something moving in here’. As I began to suggest that she had been dreaming, I noticed it too. Something was moving around in our bedroom. Then that something brushed against my head. By this point, drowsy curiosity was giving way to fear. What intruder was prowling around our room? After fumbling for the light, we discovered that the monster was our neighbours’ cat, Tibbles, who had climbed in through the window. In this situation, my fear influenced my actions—for example, it made me reach for the light. It’s easy to find other examples of our emotions influencing our actions. For example, you might mess up an interview because you are nervous, or be rude to a stranger because you are angry. Insofar as our emotions feed into our actions, they presumably do so by influencing our desires (§2.4.3). But if desire-­as-­belief is true, this amounts to the claim that our emotions influence our beliefs. But aren’t beliefs controlled by evidence

Desire as Belief: A Study of Desire, Motivation, and Rationality. Alex Gregory, Oxford University Press. © Alex Gregory 2021. DOI: 10.1093/oso/9780198848172.003.0008

Emotions Causing Desire  131 and reasoning, not by emotion? Desire-­as-­belief may seem to commit us to implausible views about the relationship between emotion and desire. In some respects, this objection is not too threatening. Plausibly, mental states might cause others in a variety of ways, and there’s no obvious principled reason to think that emotions couldn’t directly cause beliefs. Moreover, emotions might causally influence other mental states or processes in ways that indirectly affect our beliefs, as when your overwhelming fear of the fire causes you to reason carelessly about how to put it out. But on the other hand, these seem like somewhat coincidental effects of emotion, disconnected from their characteristic role in our psychology. We might think that emotions are supposed to be reflected in our desires, and perhaps rationally give rise to such desires. But how could emotions make beliefs rational? Your internal emotional state seems independent of what you should believe. To see how desire-­as-­belief can avoid this objection, I’ll first say a little about the nature of emotion. Giving a complete and unified account of the emotions is a very tall task, given that different emotions seem so very different to one another. Still, a helpful place to start is with the claim that emotions primarily serve to make us especially responsive to certain features of our environment (see also §2.4.3).1 This is an appropriately broad claim, since emotions seem to fulfil this general function as kludges: individual emotions need not achieve this goal in any systematic manner, and different emotions might achieve this goal in different ways. Plausibly, evolution furnishes us with a variety of emotions which get a variety of jobs done in a variety of ways. But despite emotions being something of a hodgepodge, we can still helpfully distinguish three more specific roles of the emotions that enable them to make us responsive to our environment. First, emotions make us responsive to certain features of our environment by preparing our bodies for relevant actions and/or by forcing certain physical reactions on us. For example, anger might move blood from your digestive system to your upper body muscles, making you more able to cope with 1  I doubt this is a complete account of the nature of the emotions. Most crucially, emotions also seem to play an important role in our lives by making themselves manifest in ways that provide information to others. For example, many emotions are hardwired to produce certain recognizable facial expressions, and we are hardwired to recognize those expressions (Ekman and Friesen 1971). Insofar as emotions make themselves manifest in ways that are out of our control, that makes them effective for conveying information to others in a way that they can take as reliable, and this can be beneficial in a variety of cooperative contexts (Frank 1988). For example, by making itself manifest, anger can make clear that our threats are genuine and not cunning verbal bluffs, and it thereby makes such threats more effective. And embarrassment, shame, and love can make us attractive cooperative partners because they can make us unable to effectively lie. But however important this signalling role of the emotions might be, it has no direct relevance for desire-­as-­belief, and so I set it aside.

132  Desire and Feeling physical threats, fear might paralyse you so that you avoid drawing attention to yourself, and surprise might make you flinch away from a potential threat or blink to protect your eyes. This bodily role of emotion is important, but has no direct relevance for our desires or desire-­as-­belief, so I set it aside. Second, emotions enable us to be responsive to our environment in part because they themselves are responsive to apparent reasons. We tend to be afraid in response to things that seem fearful, tend to be angry at apparent injustices, and tend to be happy at apparent goods. Obviously, our emotions are not always a perfectly rational response to our situation, but they are at least normally somewhat sensitive to the apparent state of the world. This doesn’t explain how emotions rationally give rise to desire, but does explain one kind of correlation between the two: often, they have a common cause. It might well be true that people who are angry also desire to seek retribution, but this might be explained, to at least some extent, by the fact that both anger and retributive desires tend to result from perceived injustices. Things are similar with other emotions: fear, and the desire to run away, are correlated in part because both arise in response to apparent danger; disgust, and the desire to avoid ingestion, are correlated in part because both arise in response to apparent threats to health; and happiness, and the desire to continue, are correlated in part because both arise in response to apparent goods. These correlations between emotion and desire are obviously consistent with desire-­ as-­belief, and allow it to explain why emotions are so strongly associated with various desires: because the two so often travel in tandem. But this is not the end of the story. Again, it seems that our emotions also rationally influence our desires. That is, it is not as though emotions and desires travel on independent but parallel lines. Rather, it seems as though our emotions cause us to have certain desires. But it also seems that such causation is somewhat rational. This is where the challenge for desire-­as-­ belief is strongest: how could it be rational to form certain beliefs in response to your own emotions? We can answer this question by appealing to the third role of the emotions. Emotions play an important epistemic role in our lives: they present us with new information about our environment (see also Brady 2009). This epistemic role of the emotions allows desire-­as-­belief to explain how our emotions rationally influence our desires. We should distinguish two ways in which our emotions can give us evidence that guides our desires.2 2  Actually, what matters is that emotions direct our attention in ways that make us seem to have new evidence, rather than in ways that in fact give us such evidence—our beliefs are responsive to

Emotions Causing Desire  133 Emotion (e.g. nervousness)

Beliefs (e.g. ‘I’m not well prepared!’)

Figure 7.1  A simple model of the influence of emotion.

First, your emotions might sometimes be evidence about what you have reason to do. For example, you might take your nervousness itself as evidence that you are not well prepared for your upcoming interview. See the pleasingly simple Figure 7.1. Some people have defended the theory that emotions are akin to perceptual appearances of the good (see e.g. Döring 2003; Tappolet 2016; for criticism, Brady  2013).3 On that theory, your emotions might sometimes generate desires because perceptual appearances are sometimes evidence for beliefs. For what it’s worth, I doubt such theories are correct. For example, it is not clear whether they capture the ways in which emotional states themselves arise in response to reasons, and can be rationally assessed (cf. §4.4, and see Brady 2013, 112–16; Helm 2015). Still, I need not take a firm stand on this. If emotions are perceptual states, that gives a clear explanation of how they rationally feed into our desires, given desire-­as-­belief. And if emotions are not perceptual states, then we can still agree that our emotions might serve as evidence about what we have reason to do (cf. Brady  2013, 129–33). For example, you might take some of your emotions to be subconsciously sensitive to certain normative facts, and thereby take such emotions as defeasible evidence for those facts. For example, you might think your nervousness about the interview results from some subconscious recognition that you are not well prepared, and as a result you might take your nervousness itself as evidence that you should prepare further. In short, our emotions might rationally feed into our desires because our emotions are evidence about what we have reason to do. But I think that our emotions more often rationally feed into our desires in a different way. Our emotions often feed into our desires indirectly: they direct our attention, in ways that reveal evidence which then feeds into our desires. For example,

apparent evidence, not evidence per se. But for ease, I’ll just talk about changes in evidence, taking this point as read. 3  Others have defended a view more parallel to desire-­as-­belief, claiming that emotions are, or involve, normative beliefs (Nussbaum 2004; Solomon 1976). That is not my view: we should distinguish desires and emotions. For example, desires are under a greater degree of rational control than emotions, and emotions have a starker bodily phenomenology than desires. Considerations like these explain why desire-­as-­belief is superior to desire-­as-­perception, and why emotion-­as-­perception is comparatively more plausible than emotion-­as-­belief (cf. §4.4).

134  Desire and Feeling Emotion (e.g. nervousness)

Redirection of attention

Beliefs (e.g. ‘I’m not well prepared!’)

Figure 7.2  A more complex model of the influence of emotion.

your nervousness about your interview might cause you to mentally focus on possible interview questions, and that might change your evidence about whether you are well prepared. See Figure 7.2. Michael Brady has defended a related view about the emotions (2009, 2013): that although emotions do present an initial guess at the relevant normative facts, their more important role is to direct our attention towards those facts and investigate them further. This is a tempting picture when we think about the effects of many emotions. For example, fear disposes you to check for possible dangers, as when your gaze darts about in the dark in order to check for monsters; anger disposes you to be alert to possible wrongs, as when your attention remains fixed on the decisions of a certain politician; and disgust disposes you to check for things you shouldn’t touch, as when your eyes scan around for the source of the ghastly smell in your garden. On this view, desire-­as-­belief permits a different explanation of how emotions feed into desires: because emotions redirect our attention in ways that affect the evidence we have about reasons. In summary, our emotions rationally feed into our desires, but this is not a problem for desire-­as-­belief. One possible view is that our emotions rationally feed into our desires directly, because emotions are themselves evidence about what we have reason to do. Another possibility is that emotions rationally feed into our desires indirectly, because they play a role in directing attention, and so our emotions make a difference to our desires by affecting the evidence we possess. Either way, desire-­as-­belief can explain how our emotions rationally influence our desires. For example, we might say that angry people often desire to harm others because their anger focuses their attention on (apparent) injustices and thereby inclines them to form beliefs about reasons for retribution. Or for another example, we might say that people in love often desire to spend time with their loved ones because their love focuses their attention on (apparent) good features of those ­people, and thereby inclines them to form beliefs about reasons for enjoying those goods. The rest of this section turns to a slight aside: is this epistemic influence of emotion a positive thing? Should we see our emotions as irrational forces that throw our desires into turmoil, or is the relationship between emotion and

Emotions Causing Desire  135 desire more positive?4 Inevitably, the answer is that it varies: the influence of emotion sometimes improves our desires, and sometimes worsens them. Again, emotions function as rough kludges that serve a variety of evolutionary purposes, and those purposes are sometimes good, and sometimes bad. On the positive side, our emotions sometimes enable us to stay in touch with facts of real significance. For example, when you are nervous about an interview, your nervousness might serve to improve your desires by making you aware of your unpreparedness. Equally, in some cases your anger might alert you to various injustices that ought to be rectified. Vice versa, emotional deficits might result in normative blind-­spots, as when a lack of empathy prevents you from noticing that your co-­worker needs help. (Perhaps at the extreme, emotional deficits can contribute to widespread moral blindness: perhaps this is part of the problem with psychopaths and those with certain kinds of brain damage (Damasio 2006).) But in other cases our emotions are more negative in their effects. Your nervousness might remain even after you’ve fully prepared for the interview, causing you to mentally attend to unlikely disasters that you should instead ignore. Similarly, your anger might make you obsess over minor slights, and blow them out of proportion. I said that our emotions direct our attention onto certain features of our environment: sometimes this causes us to notice facts of real significance, but at other times this focus is something that we can recognize as myopic once we are in a clearer state of mind. For example, you might wake in the morning, clearer headed, and recognize that yesterday’s sadness caused you to dwell on the most insignificant inconveniences as though they were life-­shattering disasters. Moreover, emotions often direct our attention onto facts that are themselves further reasons for those very emotions, as when your fear makes you notice further threats, and those threats themselves make you yet more afraid. In this way, being in the grip of an emotion is sometimes akin to being in the grip of a conspiracy theory, where you constantly cherry pick to find further evidence for your distorted views. In summary, emotions play a dual epistemic role in our lives, sometimes worsening our desires, and sometimes improving them. We can improve the epistemic effects of our emotions by revising our emotional dispositions over time, so that our emotions normally arise only when they are likely to be epistemically beneficial, and so that when they do arise, they cause us to focus on the important facts rather than obsess about unimportant ones (again, cf. 4  For relevant discussion, see e.g. Goldie (2004).

136  Desire and Feeling Damasio 2006). For example, you might train yourself to focus on the scenery when stuck in traffic, rather than descend into rage at the (apparent) idiots causing the jam. Or you might ensure that your nervousness focuses your mind on effective steps you can take to prepare adequately, rather than focusing your mind on embarrassing disasters that are unlikely to occur. But training our emotions in these ways is surely easier said than done.5 Let me summarize this section. Our emotions clearly have a close relationship with our desires. In part this may be because the two often have a common cause: emotions often arise in response to reason-­providing facts such as dangers and injustices. But mostly this is because our emotions feed into our desires. They might do so directly by providing evidence about what we have reason to do. But more often they do so indirectly, by channelling our attention in ways that generate more reasons beliefs. In these ways, desire-­as-­ belief is consistent with the close connection between emotion and desire.

7.2  Appetites Causing Desire We can now turn to another common objection to desire-­as-­belief. The worry is that even if some desires are somehow intellectual as suggested by desire-­as-­ belief, other desires involve bodily feelings and are clearly not beliefs. The most obvious candidates for such desires are hunger, thirst, and lust. These states seem to be desires, and yet it may seem implausible to claim that they are beliefs, as desire-­as-­belief seems to suggest. This thought is compelling even in this simple form, but it can be articulated more carefully by pointing out that bodily appetites like hunger are unlike beliefs in that they are not sensitive to reasoning (see e.g. Sinhababu  2017, 38–40). Do such bodily appetites refute desire-­as-­belief? The short answer to this question is that they do not, because bodily appetites like hunger, thirst, and lust are not desires (see also Green 1884, sec. 121). In what follows I focus on hunger, but the relevant points clearly generalize. 5  The above remarks all bear solely on the epistemic role of emotions. But I said that emotions also have bodily effects. Emotions might play a helpful epistemic role whilst having unhelpful bodily effects (and perhaps vice versa). For example, your anger at an email might be epistemically helpful because it focuses your attention on the problem at hand, but might nonetheless be unhelpful because it draws blood away from your digestive system and into your upper body muscles. Such bodily reactions might have their place in fighting tigers, but not when responding to emails. Again, the bottom line is that our emotions are a mixed bag, which do a variety of things, sometimes helpfully, sometimes unhelpfully.

Likings Causing Desire  137 We should distinguish two things that normally happen when someone is hungry. First, they have a collection of bodily feelings, primarily in their abdomen. Second, they have the desire for food. Just the first of these things is hunger: hunger is primarily a feeling, and one that you have in a specific part of your body. Distinguishing the feeling of hunger from the desire for food is both natural and necessary: it allows us to make room for obvious possibilities such as that you might desire food even when not hungry (e.g. if you are trying to take on calories in preparation for a long race) and that you might be hungry and yet only weakly desire food (if you are successfully implementing a diet). But although we should distinguish hunger from the desire for food, we should nonetheless allow that the former often leads to the latter. Desire-­ as-­ belief makes perfectly good sense of this. Hunger is uncomfortable, indicates that we would find food pleasurable, and also indicates that eating would be good for us. These all seem like good and transparent reasons for eating, and so desire-­as-­belief can explain how hunger gives rise to desire. Why is it so tempting to think that hunger is a desire? The answer is that people most often desire food in response to being hungry, and so if you say that you are hungry, it is natural for us to infer that you desire food. As a result, assertions about hunger often conversationally imply claims about desire. Things are similar in other contexts: being tired is a feeling, and not a desire, but nonetheless, if you say that you are tired, it would be extremely natural to infer that you want to sleep. Since claims about hunger often conversationally imply claims about desire, it is tempting to conclude that claims about hunger entail, or are, claims about desire. But this is mistaken. They are two distinct states, and a theory of desire need not be a theory of hunger. In summary: we should distinguish bodily appetites such as hunger from our desires. But we can nonetheless make good sense of the close relationship between such appetites and desire: desire-­as-­belief explains why appetites like hunger usually give rise to desires, and given this reliable connection, assertions about such appetites conversationally imply claims about our desires.

7.3  Likings Causing Desire Our appetites are hardly the only feelings that give rise to desire. Many other feelings generate desires, as when your sadness makes you desire to give your

138  Desire and Feeling relationship another try, or when the thrill of the rollercoaster makes you desire to get on. Desire-­as-­belief can explain these cases too: normal people believe that pleasures and pains give them reasons to do things. As a result, desire-­as-­belief predicts that normal people will desire things that they foresee will give them pleasure, and desire to avoid things that they foresee will bring them pain. (What about abnormal people, who deny that such feelings give us reasons? See §7.5.) We can present these claims in a different way, by appeal to likings. What is it to like something? On one view, likings are a subset of our desires (e.g. Sobel 2011). That view fits poorly with desire-­as-­belief. But on a better view, likings should be contrasted with our desires. We should say that to like something is for it to presently be giving you pleasure (see also Parfit 2011a, 52–6). This way of understanding likings captures various intuitive thoughts. For example, likings have a more limited range of objects than desires: it seems as though you can like only present experiences. That is why you cannot like being asleep, or like being dead. And this way of understanding likings captures the thought that you cannot now like past experiences, or now like future experiences: at best, we can say that you did like those things, or would like those things.6 And this way of understanding likings also captures the fact that we don’t like everything we want, not even everything we intrinsically want: I might want to be polite, without liking the experience of being polite (perhaps I find it unpleasant; more likely I find it neutral). And finally, this theory of likings attractively captures the thought that likings explain desires: normally, if you would like something, you will come to want it. On this view, this amounts to the claim that the expectation of pleasure normally gives rise to desire. This claim fits neatly with those above, and in turn with desire-­as-­belief. In short, we should deny that likings are desires, and instead claim that likings are independent states that tend to give rise to desires. The resulting view is a good fit for desire-­as-­belief, and captures plausible thoughts about the nature of likings and their relation to pleasure and desire. This view also helps to defuse objections that stem from conflating likings and desires: defenders of desire-­as-­belief can agree that likings are not beliefs of any kind. Before we move on, we should consider two objections.

6  This proposal fits awkwardly with some ordinary claims about likings: I might ordinarily say that I like chocolate, even when I’m not eating any. But I suggest that we treat such claims as compressed claims about what I normally like: when we say that I like chocolate, what we mean is that I normally like eating it.

Likings Causing Desire  139 First, some argue that pleasures and pains themselves must be analysed in terms of desire: roughly, pleasures are sensations we desire to have, and pains are sensations we desire to avoid (Heathwood  2007). Such views about pleasure combine awkwardly with desire-­ as-­ belief: whereas I said that pleasures give rise to desire because they transparently give us reasons, these views fit better with the idea that pleasures give reasons only because they involve desire. In that sense, such views are a better fit for the subjectivist tradition than the objectivist tradition to which desire-­as-­belief belongs (see also Sobel  2005,  2011). At any rate, desire-­based theories of pleasure seem objectionable. For example, they seem to put the cart before the horse: it seems that we want pleasurable sensations because they are pleasant; not that such sensations are pleasant because we desire them. As a result, I suggest that we reject such theories of pleasure, though I leave open which alternative theory of pleasure we should endorse.7 A second objection is that desire-­as-­belief overintellectualizes the relationship between pleasure and desire. Desire-­as-­belief explains the influence of likings on desire as resulting from our believing that pleasures and pains give us reasons to do things. But an objector might claim that pleasures and pains motivate us without any need for mediating reasons beliefs: this picture might seem to attribute sophisticated reasoning to people who instead form the relevant desires unreflectively. But this objection is difficult to sustain. We are generally happy to ascribe beliefs to people even when they have not consciously thought about the relevant issue—for example, you surely believe that the sun rose this morning despite having not explicitly thought about that exact question. We should say the same about the beliefs we form about reasons to pursue or avoid (un) pleasant feelings (Enoch 2011, 226–8). Remember that our beliefs are simply our map of the world around us, and it is tempting to think that a standard part of your map of the world includes information about the feelings you expect to have and the significance of those feelings for your choices. Perhaps the objection can be made more persuasive. If I foolishly try to remove a pan from the oven using my bare hands, I might drop it instinctively. And in some more harrowing case, I might writhe around furiously in a bid to prevent further pain. In these cases, we seem to be motivated in a manner that bypasses any intellectual apprehension whatsoever: by pure instinct. Similarly, Kate Manne writes of the ‘make it stop’ feeling generated by extreme pain, ‘the “let me go” inward lunge of a frightened animal’, and ‘the feeling of 7  For a helpful window onto the literature, see Pallies (2021).

140  Desire and Feeling struggling to draw breath’ (Manne  2017, 8). It’s true that in these cases we might be moved in a way that bypasses our reasons beliefs entirely. But we should think of these as cases where our behaviour is not driven by desire at all, but is instead unmotivated: we simply react.8 For example, when I drop a hot pan, this is hardly an action of mine, motivated by a desire, but instead a mere reflex. As a result, though there are cases where our behaviour should not be overintellectualized, those are precisely cases where our desires are just as irrelevant as our beliefs to the explanation of that behaviour (see also §2.4.3).9

7.4  Variation in Desire I have said that we form desires in response to expected pleasures and pains. Such desires play an important role in our lives. For example, if you tend to take cold showers rather than hot, tend to buy chocolate ice-­cream rather than vanilla, or tend to eat yogurt with a plastic spoon rather than a metal one (just me?), these choices are probably motivated by your desires for pleasure. Given that desires often focus on personal pleasure, we can understand one simple way in which desires can reasonably vary from person to person: you and I might have reasons to do different things because we gain pleasure from different things. To this extent, desire-­as-­belief permits there’s some truth in the slogan that you shouldn’t yuk someone else’s yum. Equally, though desire-­ as-­ belief implies that your desires can be assessed for correctness, it nonetheless permits that some of your desires might be comparatively arbitrary in the sense that they are based solely on arbitrary psychological facts about you, such as what you’ll find pleasant: such desires can be wrong only insofar as you are misguided about your own psychology. With many choices like those above, pleasure is the only relevant consideration, and so the desire for pleasure plays the dominant role in 8  Of course, after a few seconds pass we might manage to intellectually grasp what we are doing and form relevant beliefs about reasons (desires). Once formed, these might in some cases motivate the very same avoidance behaviour we were previously driven to do. 9  Some of Manne’s other cases seem to be somewhat different in kind. For example, she also writes of ‘the desperate need to sleep; the yen for silence when there are loud, jarring noises which you can’t control’, and ‘the frantic urge to protect those you love when they are in danger’ (Manne 2017, 8). In these cases it seems much more plausible that your actions are motivated by your reasons beliefs: such cases involve a more sophisticated recognition of your circumstances, such as an understanding of what the present danger might mean for one’s loved ones. In this respect they are quite different from the other cases where your behaviour requires no thought at all: a newborn baby can squeal out of pain, but cannot yearn for sleep or understand when others are in peril.

Variation in Desire  141 determining what we do. Other choices we make are clearly more complex: when you make a decision about which job to take, or whether to visit your family this weekend, you don’t deliberate as a simple hedonist, aiming to maximize the amount of pleasure you yourself can secure: you are also moved by a variety of other concerns, such as which job will give you the best opportunities to develop your skills, or whether your family have need of you. But still, we can reasonably take pleasure into account as one relevant factor in our decision-­making in these cases. And to the extent that we do, that might also justify some differences in desire between different people. I might prefer a job as a plumber, and you as a philosopher, not because we disagree in any deep sense about the respective merits of these professions, but simply because I predict that I will be more content as a plumber, and you predict that you will be more content as a philosopher. In this way, desire-­as-­belief can make sense of one way in which different people have different desires: because we are rightly sensitive to what we ourselves will enjoy. Given that desires respond to the prospect of pleasure, it’s also clear how a single person might find their desires changing over time. You might want cornflakes in the morning but not the evening, and want lasagne in the evening but not the morning. You might presently want a cup of tea, but half an hour later, no longer want a cup of tea. Or you might get a sudden urge—for the first time in years—for a particular kind of biscuit. Desire-­as-­belief is consistent with such fluctuations in desire: they involve our responding to the predicted amount of pleasure we might get from these things. When I fancy a cup of tea, I do so because I anticipate that drinking it will be pleasant at that time. This way of thinking about such cases fits the way they feel to us: I fancy a cup of tea just when I imagine it will be pleasurable because of the taste and heat. Insofar as we can predict which things will give us pleasure in future, we might thereby believe that we now have reasons to bring those things about. So, for example, if you predict that you will later get pleasure from a massage, a biscuit, or a cup of tea, you might now want to ensure that you’ll later get those things. But often, these passing desires are completely unpredictable: until the time arrives, it’s very hard to predict what food will give you pleasure at that time. I can tell you that I would now enjoy a cup of tea and would not enjoy lasagne, but it is much harder to tell you whether at this time tomorrow I would enjoy tea or lasagne. As a result, I might have no desires at all now about that time tomorrow, but acquire them nearer the time, as the relevant facts about pleasure become clear to me. In this way, my desires might fluctuate quickly over time, but only in ways predicted by desire-­as-­belief:

142  Desire and Feeling they fluctuate in response to unpredictable changes in independent reason-­ providing feelings.10 In summary, though it is true that desires vary between people, and can change fairly chaotically over time, desire-­as-­belief permits that some such variation can be rational: such variation is often variation in response to the reasons we have to secure, or avoid, various internal sensations.

7.5 Ascetics In the sections above I explained how pleasures and pains influence our desires. We believe that such feelings give us reasons and they thereby give rise to desire. But this explanation relied on the assumption that we believe that pleasures and pains give us reasons to do things. I assume that as a matter of psychological fact, the vast majority of us do believe this. Such reasons are amongst the most obvious and uncontroversial that we have. But if some unusual person denies that pleasures and pains are reason-­providing, desire-­ as-­belief implies that their desires will be insulated from such feelings. An ‘ascetic’ (as I will use the word) is someone who denies that pleasure and pain provide reasons for action, or who at least believes that such reasons are extremely weak. My claims above applied only to non-­ascetics—they suggest that ascetics will not desire pleasure, or will do so only very weakly. In many respects, this may seem extremely plausible. If you and I are equally hungry, but I am on a diet and you are not, it’s plausible that my preferences will be different to yours: unlike you, I might prefer to order a diet coke. Similar reasoning suggests that ascetics have different desires to the rest of us. Certainly, ascetics do behave differently from other people, and plausibly that is precisely because they have different desires. Reasoning like this can make it seem extremely plausible that our appetites and likings influence our desires indirectly, only via beliefs about reasons to seek pleasure. This is precisely why desire-­as-­belief promises to make good sense of these phenomena. But on the other hand, it is not clear that we have perfect control over our desires in these cases (see e.g. Setiya  2010, 36–7; Sinhababu  2017, 39–40;

10  Some people believe that reasons are time-­relative, so that you might now have more reason to avoid some future pain than you did a year ago. On one version of this view, we discount the weights of reasons by their distance in time: reasons are weightiest when they concern the present, and decline in weight as they extend into the future (for related discussion, see Parfit 1984, 149–86). Anyone who accepts such a view would also find their desires increasing in intensity as the relevant prospects came closer to the present.

Ascetics  143 Watson 1982, 101). If you are persuaded that pleasure is the work of the devil, is it really plausible that you will instantly cease to desire anything pleasurable? Can a decision to adopt a diet totally isolate your desires from feelings of hunger? Above I suggested that we have some influence over our desires, even those formed in light of prospective pleasures and pains. But it isn’t clear that we should jump to the extreme view that we have total control over our desires. Isn’t that what desire-­as-­belief implies? We have already seen how desire-­ as-­ belief can overcome objections involving motivational weakness (Chapter 5), and I won’t rehearse all of those points again. But one possibility I mentioned was that our beliefs can be irrational (§5.3), and this may be especially relevant in the case of ascetics. Asceticism conflicts with common sense. The reasons provided by the prospect of pleasure and pain seem to be the most obvious reasons that we have, and are often highly significant. Ascetics nonetheless resist the appearances here, by appeal to whatever wider commitments drive their asceticism: perhaps they believe that pleasure and pain are both convincing ploys by the devil to distract us from the awesomeness of God, or that the value of pleasure and pain is real but infinitesimal compared to the value of communion with God. Perhaps ascetics are irrational insofar as they lack sufficiently good grounds for denying the manifest significance of pleasure and pain (it sure seems that way to me). But let’s set that aside. Actual ascetics might be irrational in a further way. After all, given that they do think that pleasure and pain are not important, they should believe of every individual pleasure and pain that it is not important. I suggest that this is just what ascetics often fail to do. It is easier to convince yourself in the abstract that pleasure is worthless than it is to convince yourself that this particular pleasure you are now feeling is utterly worthless. Insofar as ascetics manage to form a general sceptical belief but fail to apply it to specific cases, they are irrational, and this might make sense of the way in which some ascetics’ desires seem out of line with their reasons beliefs. These two are not out of line in a manner that threatens desire-­as-­belief, but rather in a manner that threatens ascetics’ rationality: they struggle to bring themselves to believe things that obviously follow from their wider commitments. In this respect, ascetics may be somewhat like some external world sceptics, who profess in the classroom that the external world does not exist, but who nonetheless cannot escape the grip of that idea in their everyday lives (cf. Hume T1.4.7.8–9). So, too, ascetics may profess in the classroom that no pleasure has value, but they might well struggle to escape the grip of that idea in their everyday lives, and their desires might sometimes reflect the latter recalcitrant beliefs, not the former theoretical ones.

144  Desire and Feeling

7.6  Desires Causing Pleasure So far in this chapter I have discussed how feelings generate desires: our emotions, appetites, and likings all commonly cause us to form desires. But we should also consider the reverse possibility, where our desires generate feelings.11 For example, if you find out that your house purchase is going through, that might make you feel happy. And if you find out that your romantic relationship will end, that might make you feel sad. In general, it seems that we get pleasure when we come to believe that we have what we want, and feel displeasure if we come to believe that our desires are frustrated. (In a note, I mention a slight complication regarding the relevance of our prior expectations.12) How can desire-­ as-­ belief explain this correlation between desires and pleasure? It might seem that there are no obvious grounds for expecting beliefs alone to result in pleasure. In what follows I focus on the capacity of desires to produce pleasure, but I would obviously endorse parallel claims about the capacity of desires to produce displeasure. It seems to me that this objection is less intuitive than the other objections I have addressed in this chapter. It is at least initially unclear why reasons beliefs would be sensitive to feelings such as emotions, since these are unlike the ordinary considerations that bear on belief. But it is harder to motivate the puzzle here: would it be so surprising if some normative beliefs had hedonic consequences? It’s perfectly plausible that our ethical beliefs can impinge on our feelings. For example, imagine you believe that world peace would be good, and then world peace comes about. Under such circumstances, you might well feel happy. There is nothing surprising about this possibility. In fact, we might pursue this picture in two slightly different ways. First, perhaps only beliefs about goodness give rise to pleasure, so that we get

11  For related discussion, see Arpaly and Schroeder (2014, 116–25); Hobbes (L I.VI.11); Hume (T2.3.9.5); Strawson (1994, chap.  9); Schroeder (2004, chap.  3); Sinhababu (2017, 28–33). See also Long and Sedley (1987, sec. 65A4). 12  If you want Briggs to win the next election, you might get more pleasure from the unexpected and extremely promising polls before the election than you do from her actual win on the day. More generally, if you are already confident that you will get what you want, actually getting it might not bring much further pleasure, and if you are doubtful that you will get what you want, merely increasing the apparent likelihood that you will get it will give you pleasure. That is, what seems to give pleasure is not flat out believing that you will get what you want, but instead increasing your confidence that you will (Schroeder 2004, 94; Sinhababu 2017, 28–9). For ease of exposition, I set this point aside: though it might well be important for numerous theoretical and practical purposes, it isn’t important for my claims in this section.

Desires Causing Pleasure  145 Belief that I have a reason to promote p (i.e. desire that p)

Belief that p is good

Belief that p

Feeling of pleasure

Figure 7.3  How getting what you want might correlate with pleasure.

pleasure just when we think that good things have occurred. We can represent the resulting picture as in Figure 7.3. On this picture, pleasure results from beliefs about goodness, not desire. But desire and pleasure still correlate because of a common cause: desires, and pleasure, are both sensitive to our beliefs about goodness. On this view, it’s not strictly true that we get pleasure in virtue of our desires: rather, we get pleasure in virtue of believing that good things have happened, and we often desire things we believe are good. Is this view right? In Chapter 6 I argued that our desires should be identified with beliefs about reasons, rather than beliefs about goodness. That picture was certainly intended to be consistent with a rough correlation between these two kinds of belief: we often believe that we have reasons to bring things about precisely because we believe they are good. So what I said in Chapter 6 is consistent with a strong correlation between our beliefs about goodness and our reasons beliefs, and in turn consistent with the above explanation of the correlation between desire and pleasure. But I also said that in some cases people might believe they have reasons to bring about things that they don’t believe are good. At the extreme, perhaps Satan is like this, and more realistically, some of those who endorse deontological views in ethics might think like this. The view above was that people get pleasure when they believe that good things have occurred. It tells us that if our desires (i.e. our reasons beliefs) are somewhat independent of our beliefs about goodness, we will not get pleasure from getting what we want. Is that plausible? No. Plausibly, people who believe in the relevant de­onto­logic­al views might take pleasure in doing their duty, even if they believe that it secures nothing good. And so I think we are better off with a

146  Desire and Feeling Belief that I have a reason to promote p (i.e. desire that p)

Belief that p is good

Belief that p

Feeling of pleasure

Figure 7.4  How getting what you want might also cause pleasure.

more permissive view on which we get pleasure not only from seeing good things occur, but also from doing what we have reason to do. This view is represented in the more complex Figure 7.4. On this view, desires correlate with pleasure for two complementary reasons. The first is that we get pleasure when we believe good things have happened, and our beliefs about the good feed into our desires. But on this view we also get some (less) pleasure merely from doing what we have reason to do. That permits our view to explain why people whose beliefs about the good diverge from their reasons beliefs still get some pleasure out of doing what they want to do.

7.7 Summary In this chapter I have explained how desire-­as-­belief accommodates the relationship between our desires and our feelings. According to desire-­as-­ belief, our emotions feed into our desires by providing evidence, or by channelling our attention in ways that generate evidence, about what we have reason to do. Desire-­as-­belief also explains how appetites, and prospects of pleasure—likings—feed into our desires, via beliefs about reasons to pursue pleasure. Desire-­as-­belief can also explain some (supposedly non-­rational) variation in desires over time and between people by appeal to variation in what will give us pleasure. Finally, I discussed the role of desires in producing feelings. We get pleasure from outcomes we believe to be good, and since such beliefs correlate with our desires, we often get pleasure when we get what we

Summary  147 want. But we also get some pleasure directly from the thought that we have conformed to a reason. The next chapter turns to two further objections to desire-­as-­belief. First, must desire-­as-­belief say that degrees of desires are degrees of belief? That seems implausible. Second, does desire-­as-­belief have implausible implications about our capacity to revise our desires through reasoning?

8 Uncertainty and Reasoning This chapter addresses a pair of objections to desire-­as-­belief: first, about how degrees of belief relate to degrees of desire, and second, about our capacity to reason with our desires. The objections are distinct, but I lump them together because my solution to the first is useful for understanding my claims about the second. §8.1 addresses the first issue, about degrees of belief and desire. §8.1.1 sets up the basic objection, §8.1.2 rejects some possible replies, and §8.1.3 presents my preferred reply. §8.1.4 briefly ties up some loose ends about how desires affect motivation. §8.2 then turns to the second issue, about reasoning with our desires. §8.2.1 briefly clarifies the topic, and §8.2.2 sets up the objection. §8.2.3 sets aside some possible replies, and §8.2.4 presents my preferred reply. In two classic papers, David Lewis objected to desire-­as-­belief on the grounds that it made poor sense of degrees of desire, and of rational change in desire (Lewis  1988,  1996). My claims in this chapter undermine Lewis’ objection: I explain such matters in Appendix A, which should be read only after §8.1. If you aren’t independently curious about Lewis’ papers, that Appendix could be skipped.

8.1  Degrees of Desire 8.1.1  The Certainty Problem Desires come in degrees. One fairly common claim is that degrees of desire are degrees of motivational influence. In its strongest form, some commit to the view that your desires can be straightforwardly inferred from your actions.1 But as I argued in §2.3, this is not right (see also §5.5, §5.6). Our desires can motivate us to do things, but their strength is not the same as the

1 This is how many economists standardly understand preferences—see e.g. Samuelson (1947, 97–8). See Hausman (2011) for useful discussion.

Desire as Belief: A Study of Desire, Motivation, and Rationality. Alex Gregory, Oxford University Press. © Alex Gregory 2021. DOI: 10.1093/oso/9780198848172.003.0009

Degrees of Desire  149 amount of motivational force they exert.2 Some desires might exert no motivational force, as when you neglect extremely weak desires and they play no active role in your present motivational economy. Other desires might exert motivational force that is weak given how strong those desires are, as when you are weak-­willed and though you have the goal of losing weight, you struggle to summon much motivation to do so. Still other desires might exert motivational force that is relatively strong given how weak those desires are— perhaps your moderate desire to please your colleagues manages to override your strong desire to see your family and you find yourself working overlong hours. Since the simplest identification of degrees of desire with degrees of motivational influence fails, we need a more sophisticated view about degrees of desire. Given desire-­as-­belief, questions about degrees of desire are somewhat complex. I’ll first set up a puzzle for desire-­as-­belief about degrees of desire, and then solve it in §8.1.2 and §8.1.3. Then, in §8.1.4, I’ll briefly return to the relationship between degrees of desire and degrees of motivational influence. To get our puzzle into view, first note that according to desire-­as-­belief, our desires can be correct or incorrect, depending on whether we really have the reason in question. This, it seems to me, fits with common sense, on which desires can be more or less appropriate, depending on what they are for. But a related thought is that we ourselves might be unsure whether our desires are appropriate or not. After all, we can be unsure of our beliefs, and so if desires are beliefs, desire-­as-­belief must say that we can be unsure of our desires. But can we make sense of the idea that we can be uncertain with respect to our desires? Let’s get this puzzle into clearer focus. Desires seem to come in degrees of one kind: strength. For example, your desire to eat lunch might start as a mild desire, and grow in strength over time. In contrast, reasons beliefs vary along two distinct dimensions. First, such beliefs can vary along a dimension of certainty: a credence, a number from 0 to 1 representing subjective probability.3 You might be more or less confident of this belief, perhaps 2  Of course, one could simply stipulate that by ‘strength’ of desire you mean how much motivational force they exert at a particular moment. But no theory can be refuted by a choice of terminology: I could instead say that as well as strength, desires come in degrees of a different kind, and that that other thing is the one that matters in most contexts, such as in explaining general patterns of behaviour, or in assessing rational choice. Nothing hinges on our choice of terminology other than clarity, and I think it clearer to begin with the stipulation that ‘strength’ is whatever plays the important roles in these contexts, and then make the inference in the main text that the ‘strength’ of a desire is not simply the degree of motivational force it exerts at that particular moment. 3  Here and below, I treat certainty as a property of the attitude of belief. An alternative possibility is to treat certainty as part of the content of the relevant belief, so you might (flat out) believe [it’s 80 per

150  Uncertainty and Reasoning thinking of the issue as settled, or perhaps instead thinking of the issue as unclear and this as merely your best guess. For example, imagine you believe you have reason to strive for social equality, and then find out about the levelling-­down objection. If you are worried but not wholly convinced by this objection, you might continue to judge that you should strive for equality, but be less certain of this fact. Second, reasons beliefs vary along a dimension of importance. You might think that you have a very strong reason to bring this thing about, or else you might think that you have such a reason, but only a weak one that is easily outweighed. For example, perhaps you think that you have a very strong reason to eat organic food, but on further investigation you might decide that the benefits, whilst present, are only small, and therefore think that the reason to eat organic food is more easily outweighed by the extra cost or hassle. Certainty and importance are two different things. A reason belief might have high certainty, but low importance, and vice versa. For example, I am very sure that I have a weak reason to prevent minor pains, and very tentatively believe that I have a very weighty reason to donate only to the most effective charities. But again, desires come in degrees of only one kind: strength. We are left with a puzzle: desires come in degrees of one kind, whereas reasons beliefs come in degrees of two kinds. It doesn’t look like there’s any way to pair these up. As a result, desire-­as-­belief seems problematic.4 Before we canvass possible solutions to this problem, we should set aside a complication. This objection to desire-­as-­belief is easiest to navigate if we focus on non-­instrumental states of mind. Desire-­as-­belief identifies non-­ instrumental desires with beliefs about non-­ instrumental reasons. For example, it identifies believing you have a reason to secure happiness for its own sake with desiring happiness for its own sake. Such beliefs can be held with a greater or lesser amount of certainty, and the relevant reasons can be held to be more or less important, but the corresponding desires seem to vary cent likely that p]. I believe that the things I say below would equally apply to uncertainty understood in that way, though the vexed relationship between the two conceptions of certainty makes the issue complex. Another issue that arises for certainty, so understood, is whether you really count as ‘believing’ something if your credence is only very low. A similar question would arise regarding desire, given desire-­as-­belief. Again, I set this aside: the relevant issues would take me too far afield, and I can’t see any obvious difficulties that are unique to desire-­as-­belief. 4  This puzzle is best known as an objection to non-­cognitivism rather than as an objection to desire-­as-­belief (Smith  2002; Bykvist and Olson  2009a, 2011, 2017): non-­cognitivists also need to explain any apparent variation between the properties of normative beliefs and desires. In what follows, I focus on the problem as it arises for desire-­as-­belief, and the solution I pursue is unavailable to non-­cognitivists.

Degrees of Desire  151 only in strength. That is the objection to desire-­as-­belief, and the objection is especially easy to see when we focus on non-­instrumental states of mind. The problem does also arise for instrumental states of mind, but such cases introduce additional complexities that only obscure the underlying issues. If you instrumentally desire to do something, then in one loose sense this comes with a degree of certainty. For example, imagine that you instrumentally desire to visit Adventure Wonderland, because you think it will bring you happiness. In a loose sense this desire has a degree of certainty: you might be more or less confident about whether the visit will bring you happiness. But this is of little help to desire-­as-­belief: it amounts to only a kind of non-­ normative uncertainty about the means to the end of happiness. In contrast, desire-­as-­belief requires there to be a feature of desire that we can identify with normative uncertainty: uncertainty about ends, not means. That is why the problem is clearest when we focus on non-­instrumental states of mind, which focus our attention on uncertainty about ends, not merely non-­ normative uncertainty about means. In short, it is simpler to focus on non-­instrumental desires and beliefs about non-­instrumental reasons—this allows us to focus on the kind of uncertainty that causes problems for desire-­as-­belief. If desire-­as-­belief can make good sense of uncertainty in non-­instrumental desire, it can surely make sense of it in instrumental desire as well. In what follows, for ease, I’ll sometimes use the label ‘ultimate’ in place of ‘non-­instrumental’: in these terms, our focus is on ultimate desires, and beliefs about ultimate reasons, not instrumental desires or beliefs about instrumental reasons.

8.1.2  Some Bad Solutions With the problem in view, let’s examine how defenders of desire-­as-­belief might try to solve it. One simple option is to understand desire-­as-­belief so that it identifies desires only with fully certain beliefs, and to claim that uncertain beliefs never qualify as desires. But since even uncertain normative beliefs can motivate— as when I vote in an election despite my uncertainty about the right choice— we should maintain that even uncertain normative beliefs can qualify as desires. A second approach would begin by identifying the strength of desire with the certainty of reasons beliefs, so that strength of desire is the confidence

152  Uncertainty and Reasoning Table 8.1  Strength as certainty Belief that I have (ultimate) reason to bring about p

 

(Ultimate) Desire that p

Importance: How strong the reason is believed to be Certainty: How sure I am that I have that reason

Is the same as

???

Is the same as

Strength: How strongly I want p

Table 8.2  Strength as a compound Belief that I have (ultimate) reason to bring about p

 

(Ultimate) Desire that p

Importance: How strong the reason is believed to be Certainty: How sure I am that I have that reason

Together constitute

Strength: How strongly I want p  

with which you hold the relevant reason belief (cf. Lewis 1988, 327; 1996, 307; and Appendix A). If we said this, we’d then need to find some other property of desire to identify with the importance of the corresponding reasons belief. We can present views of this kind as in Table 8.1. We should avoid all views of this kind. Regardless of what we might try and insert at the top right of the table, this view is false because strength of desire and certainty in belief are clearly different things (Bradley and Stefánsson 2016, 707–11; Bykvist and Olson 2009b, 205; 2011, 2–3). To see this, note that certainty has a maximum whereas desire strength does not. That is, whereas you can be 100 per cent certain of some reasons belief, and no more certain than that, it is not clear what it would be to desire something 100 per cent, and for it to be impossible to desire anything more strongly. A third option is to understand strength in desire as a kind of compound, constituted by both certainty and importance. We can present this view as in Table 8.2. In particular, we might say that desiring p with strength S is identical to believing that you have a reason to bring about p with importance I and certainty C, where S=I*C. That is, we might treat strength in desire as a matter of the certainty-­discounted importance of the relevant reason. (The view may be easiest to understand if we consider a variant of desire-­as-­belief which treated

Degrees of Desire  153 desires as beliefs about the good: the present view would then say that the strength of desire is the same as expected value (cf. Broome 1991).) This view is more promising than the one above (I shall endorse a related claim in §8.1.4). But it is still dubious. It says that the strengths of your desires depend only the certainty and importance of your reasons beliefs. But that could be true only if everyone has the very same attitude to risk, so that everyone discounts the importance of their reasons by their uncertainty in the exact same way. But that is implausible. Plausibly, proportioning your desires so that they take risk into account systematically is a feat of rationality, and not a trivial task. For example, some people might be irrational and overweight small risks in their reasoning, and thereby end up with stronger desires than their more rational counterparts (Kahneman 2011, 310–33). Moreover, some people might positively embrace the overweighting of small risks as rational, and subsequently form desires on that basis (Buchak 2013). The proposed view rules out such possibilities, and should therefore be rejected. In short, this view says that differences between your attitude to risk and mine make no difference to our desires, but that is surely false. The above three options are implausible. So I shall pursue a different view.

8.1.3  The Best Solution: Being Unsure About What You Want I suggest that we identify strength in desire with the importance of the relevant reasons belief. To strongly desire something is to believe that you have a strong reason to bring it about, and vice versa. This claim is intuitive: in both cases, it is true to say that your state of mind represents the degree of attraction of the thing in question. But in order to maintain this claim, we need to find some other feature of desire with which we can identify certainty.5 We can represent this embryonic view as in Table 8.3. Our challenge now is to fill in the bottom-­right corner. To understand my proposal, it’s crucial to be crystal clear about the initial puzzle. Remember that desire-­as-­belief states an identity claim: it says that we have two different 5  There is another way to implement this strategy beyond the one I defend in the main text: One might treat certainty in desire as a matter of the strength of some second-­order desire. But this is not promising (Smith  2002, 317–18; Bykvist and Olson  2009a, 205–6; 2011). For example, it fails to address the worry that certainty, unlike strength, has a maximum. A further problem is that we might form such second-­order desires for instrumental reasons—such as the promise of payment for having the first-­order desire—and doing this is not a way of changing your degree of certainty in any reasons belief (cf. §5.7).

154  Uncertainty and Reasoning Table 8.3  Strength as importance Belief that I have (ultimate) reason to bring about p

 

(Ultimate) Desire that p

Importance: How strong the reason is believed to be Certainty: How sure I am that I have that reason

Is the same as

Strength: How much I want p ???

Is the same as

ways of picking out one and the same thing in the world. So the challenge is not to find some additional property of desire that is not certainty but which behaves the same as certainty. If desire-­as-­belief is true, the property of desire that corresponds to certainty is certainty: desire-­as-­belief says that desires literally are beliefs, and it follows that any property of a reasons belief is equally a property of that same state when it is described as a desire. So we are not seeking a property of desire that is distinct from certainty but which somehow corresponds to it. Rather, we are seeking a part of ordinary thought and talk about desire that corresponds to thought and talk about certainty. Desire-­as-­belief says that we have two different ways of picking out one and the same state of mind, and the theory is most plausible if ordinary thought and talk has two corresponding ways of picking out the properties of that state. With this in mind, remember how I explained desire-­as-­belief way back in §1.3: desire-­as-­belief says that ‘she desires’ is shorthand for ‘she believes has reason to bring about’. By choosing to describe someone’s mental states in terms of their ‘desires’, we opt to describe their mental states in such a manner that we present features of the contents of their beliefs as though they were some distinct attitude (again, remember the word ‘disbelief ’, which presents the negation in the content of a belief as instead part of an attitude). Now turn back to the problem at hand. Certainty is certainty about the content of a belief. But by describing reasons beliefs as ‘desires’, we present part of their content as though it were part of the attitude. So by the lights of desire-­as-­ belief, uncertainty about the content of a reasons belief should come out as uncertainty about the attitude of desire. See Figure  8.1 for a simplistic illustration. If this is right, then we should expect ordinary thought and talk about desire to pick out uncertainty by referring to our uncertainty about the attitude of desire. This gives us the solution we need. We often say things such as, ‘I don’t know what I want’ or ‘I think I want . . .’, and desire-­as-­belief permits us to make good sense of such assertions: they are indicating uncertainty

Degrees of Desire  155 Uncertainty Belief

I have reason to bring about ‘Desire’

p p

Figure 8.1  Uncertainty in desire.

about our reasons. To this extent, there is a part of ordinary thought and talk about desire that corresponds to ordinary thought and talk about uncertainty in our reasons beliefs. Let’s illustrate this with some examples. Perhaps you are unsure whether you have a reason to give a large proportion of your money away to charity: then you might say that you don’t know whether you want to give your money away. Or perhaps you are unsure whether your reasons to become a philosopher outweigh your reasons to become a nurse: then you might say that you are unsure whether you prefer to be a philosopher or a nurse. Vice versa, perhaps you say that you are unsure whether you want to go to the cinema tonight, or don’t know what you want to do on the weekend. Under these circumstances, it sounds as though you are unsure about what you have (most) reason to do. These claims all seem sensible, and that corroborates my proposal. There are two tempting objections to this proposal. First, an objector might complain that when you say (something like), ‘I don’t know what I want’, you are expressing a lack of self-­knowledge, and that is not the same as expressing uncertainty about what you have reason to do. In response, I deny that sentences like ‘I don’t know what I want’ express a lack of self-­ knowledge. To see this, think about how we resolve such uncertainty. Do you do so by seeing a psychiatrist? Do you ask your friends not for advice on what to do, but instead to guess what your state of mind is? No: the most obvious way to resolve uncertainty about what you want is to turn your eyes outwards onto the world, and to gain more evidence about the merits of the options open to you. In this way, deciding what you want is not a matter of discovering new facts about your own psychology, but instead a matter of resolving uncertainty about which course of action is best, just as my view predicts. For example, imagine you face a decision about what to do this weekend, and you say that you ‘don’t know what you want to do’. If you try to resolve this uncertainty via deliberation, you would consider facts like ‘the new Avengers film is supposed to be good’, or ‘the loft needs sorting’. In contrast,

156  Uncertainty and Reasoning you would surely not consider facts like ‘most weekends I have wanted to see a film’ or ‘my psychiatrist tells me I want praise from my parents’, as though the matter to be resolved was akin to predicting what a third party wanted. Deliberation about what you want is deliberation about what to want, and is a kind of worldly investigation, not a voyage of self-­discovery. To ensure the point is clear, it may be helpful to repeat the crucial clarification above: we are not seeking a property of desire that corresponds to certainty in our reasons beliefs. Rather, we are seeking a part of ordinary thought and talk about desire that expresses uncertainty in our reasons beliefs. And I claim that sentences such as ‘I don’t know what I want’ play this role. Those sentences are, in a way, very misleading: I agree that on their face they appear to express failures of self-­knowledge. But I claim that really, such sentences normally express uncertainty about the world—about what we have reason to do. My objector wrongly takes such assertions at face value, but that is unwarranted. A helpful comparison is with sentences such as ‘I don’t know what I think about that’. This sentence also looks like it expresses a failure of self-­ knowledge, as though you might utter it when you have some view on the matter at hand, but you are struggling to recall what it was (perhaps you have serious amnesia). But most often, someone using this sentence would use it as a roundabout way of expressing their uncertainty about the facts, not only their own state of mind. My proposal is that ‘I don’t know what I want’, and similar phrases, have the same function. Perhaps under strange circumstances, such sentences could be used to express failures of self-­knowledge. But more often, we use these sentences to express uncertainty about the facts—to express uncertainty about what we have reason to do. In summary, though my solution to the certainty problem refers to sentences such as ‘I don’t know what I want’, I claim that such sentences don’t express a lack of self-­ knowledge.6 And that is why they are plausibly understood as sentences that express a lack of certainty about what we have 6  With respect to this argument, I should briefly say something about transparency theories of self-­ knowledge. Those theories say that uncertainty about your own state of mind is resolved by attending to the facts. Similarly, such views might say that ‘I don’t know what I want’ expresses both uncertainty about our own state of mind and uncertainty about the facts. My claims in the main text can be read in ways that are consistent with this view: so long as such claims express uncertainty about the facts, that is enough for me, whether or not they might also serve to express failures of self-­knowledge. For discussion of transparency theories of self-­knowledge as applied to desire, see e.g. Byrne (2011, sec. 4 to end); Fernández (2007). Defenders of such views sometimes claim that the transparency method only delivers knowledge about some of our desires. They worry that phenomena such as akrasia, and the existence of appetites, show that some of our desires come apart from our assessments of our reasons (Byrne 2011, sec. 4 to end; Fernández 2007; see also Ashwell 2013; Moran 2001, 114–16). But insofar as my arguments in Chapters 5 and 7 are correct, we should reject such claims. For an argument from transparency theories to something like desire-­as-­belief, see Suikkanen (2018).

Degrees of Desire  157 reason to do. In this way, desire-­as-­belief is plausible insofar as there is a part of ordinary thought and talk about desire that corresponds to expressions of uncertainty in our reasons beliefs. I now turn to a second objection to my proposal. An objector might concede that ‘I don’t know what I want’ expresses uncertainty about what I have reason to do, but nonetheless insist that this uncertainty is always purely non-­normative uncertainty. When setting up the problem, I said that we should focus on ultimate desires and beliefs about ultimate reasons, because these are the cases where the problem for desire-­as-­belief is clearest. But an objector might complain that when you say, ‘I don’t know what I want’, this really amounts to your not knowing what to instrumentally want: you know full well what you ultimately want, but just don’t know how to achieve it. If that were right, then my proposal fails: uncertainty about what you want is not the counterpart to certainty in reasons beliefs, because the former is always uncertainty about means, not ends. This objection fails. When I faced the decision of whether to pursue a career in philosophy, or whether to have children, those decisions were difficult not only because I didn’t know the facts, but also because I didn’t know what I ultimately wanted. In this spirit, E.  J.  Bond (1983, vii) memorably writes: I remember being puzzled, as an undergraduate, when my professor and my fellow students all seemed to accept without question that . . . one simply had desires for certain things, and if one could . . . then one just went ahead and set out to do or get or keep them. . . .  Here, then were a couple of dozen or so people equipped with a set of ready-­made wants, which it was the business of their lives to set about ­satisfying . . . I was certainly the odd-­man-­out, for I did not have any such set of wants . . . and did not know what to do with my life. . . . My fundamental practical questions were not . . . ‘How can I best accomplish what I want the most with the least frustration of my desires along the way?’ but ‘What ends would be worth my while?’ or ‘What, of the things open to me, would be most profitable or rewarding?’ and ‘How can I realize the most worth or value in my life?’ And this was something I would have to discover. The wants which everyone else presumably had, and which made these questions needless for them, I simply did not have at all! I thought I must be very strange. I have since decided that I was not really so very strange.

158  Uncertainty and Reasoning As Bond implies, the strange view is really the one held by some philosophers according to which practical deliberation is always a matter of deliberating about how to satisfy pre-­existing and well-­defined desires. This is surely false: much practical deliberation involves thinking about what to want, and a decision is made by forming or sharpening up the relevant ultimate desires. If you don’t know what you want to do this weekend, or what career to pursue, or what to do with your life, your uncertainty is often partially instrumental, but it is also often partially ultimate: many of us are at least somewhat unclear about our fundamental goals in life. In short, it is implausible to insist that uncertainty about what we want is always a matter of instrumental uncertainty about the best means to predefined ends. Often, we are uncertain about what we want because we are uncertain about our ultimate ends. In this way, uncertainty about what we want is a plausible counterpart to certainty in our reasons beliefs. Let’s summarize our discussion in this chapter so far. Desire-­as-­belief may seem problematic because whereas reasons beliefs can vary in both importance and certainty, desires seem to vary only in strength. To solve this problem, we should adopt the view in Table 8.4. First, we should identify the strength of our desires with the importance of our reasons beliefs, so that you desire things in proportion to the weights of the reasons you think you have. Second, we should identify certainty in our reasons beliefs with uncertainty about what you want: to the extent that you are unsure about your reasons, you are unsure about your desires. One natural objection to this second claim is that this kind of uncertainty is simply a lack of self-­knowledge. A second objection is that this kind of uncertainty is always uncertainty about merely instrumental desires. Neither objection succeeds, since uncertainty about what we want is uncertainty about the world, and is sometimes uncertainty about ends, not means. I conclude that desire-­as-­belief makes good sense of the fact that desires and reasons beliefs come in degrees, including degrees of uncertainty. Table 8.4  Strength as importance, and certainty as certainty about desire Belief that I have (ultimate) reason to bring about p

 

(Ultimate) Desire that p

Importance: How strong the reason is believed to be Certainty: How sure I am that I have that reason

Is the same as

Strength: How much I want p How sure I am that I want p

Is the same as

Reasoning  159

8.1.4  Desire Strength and Motivation Before we move on, I can now briefly say a little more about the connection between degrees of desire and degrees of motivation. In §2.3, §5.5, and again in §8.1.1, I claimed that degrees of desire don’t perfectly correspond to degrees of motivation. A slightly better view says that degrees of desire correspond to the amount of motivation you are disposed to have, so that the strength of a desire is the strength of the relevant disposition. But given the claims above, even this is not quite right. For it amounts to the view that reasons beliefs are disposed to motivate us in proportion to the weights we believe the relevant reasons have. But a more plausible view is that the capacity of reasons beliefs to motivate us is affected not only by the weight the reason is believed to have, but also by the amount of certainty we have in that belief. For example, if you unconfidently believe that you have a strong reason to do something, you’ll presumably not be so strongly disposed to do it as you would if you confidently believed you had that reason. So the most plausible view is that a desire is disposed to motivate you in a way that is sensitive both to the strength of that desire and to how certain you are that you have it, just as it is plausible that a reasons belief is disposed to motivate you in a way that is sensitive both to the weight the reason is believed to have and to how certain we are we have that reason. I would guess that rational agents are those who combine these two factors in the obvious systematic way so that their dispositions to be motivated are proportionate to the product of importance and certainty. But no doubt many of us fall short of this ideal. If you wanted to read the Appendix on Lewis, now is as good a time as any.

8.2  Reasoning Having resolved the above puzzle about degrees of desire, I now move to a nearby topic. We often change our beliefs by reasoning, in order to ensure that they correctly track the state of the world. Can we modify our desires in the same way? Or are they just ‘internal weather’ that we can only passively observe?7 For ease, I’ll say that a state is under rational control if it is a state

7  I borrow the nice phrase ‘internal weather’ from Tamar Schapiro (2009, 232).

160  Uncertainty and Reasoning that we can change via reasoning. Since beliefs are under rational control, desire-­as-­belief suggests that desires are too. Is that plausible?

8.2.1  Preliminaries Some earlier sections bear on this issue. For example, I argued in §7.2 that we should distinguish desires from appetites, and so although our appetites are not under rational control, that is consistent with desire-­as-­belief. Equally, I argued in §7.3 that we should distinguish desires from likings, and so although many likings are not under rational control, that is also consistent with desire-­ as-­ belief. Finally, in §5.3, I pointed out that desire-­ as-­ belief permits that we might struggle to rationally control our desires, so long as we struggle to rationally control our beliefs to that same degree (see also §7.5). In what follows, I take these points as all read. It’s also helpful to distinguish the question of whether desires can be changed via rational processes like reasoning from the question of whether we can change our desires at will. It is one question whether desires are under rational control, and another whether they are under voluntary control. We can’t change our desires at will (see also §5.7; Millgram  1997, chap.  2; Kavka 1983). Imagine that I offer you a large sum of money to desire to drink some poison—I offer you the money not for drinking it, but rather just for wanting to drink it. This incentive won’t help you to actually form this desire, no matter how hard you try. Equally, you might recognize that some desire of yours is disadvantageous, but that alone isn’t enough to help you get rid of it. In this respect, desires are just like beliefs: we cannot form beliefs just because we choose to do so, and we cannot abandon beliefs just because we’d be better off without them.8 So though desires are insensitive to the will, so too are our beliefs, and in this way desire-­as-­belief is perfectly acceptable. The more difficult question is whether our desires are sensitive to reasoning. That is our focus.

8.2.2  Instrumentalism In short, the important question is whether desires are under our control in the same way that beliefs are: via reasoning. In fact, since desire-­as-­belief 8  For helpful discussion of some related issues, see McHugh (2012, 2014, 2017).

Reasoning  161 identifies desires specifically with reasons beliefs, the question is really whether desires can be controlled via reasoning in the same way that reasons beliefs are. Unfortunately, it’s contentious how we form and revise normative beliefs such as reasons beliefs, and that makes it difficult to evaluate whether our desires are formed and revised in the same way. We’ll plough on as best we can. We’ll proceed by focusing on the standard model for how desires are modified by reasoning (see e.g. Sinhababu  2017, 3–5, 38–41). This model originates with Hume, and I mentioned it in §3.5: Instrumentalism: All desires are either ultimate, or instrumental. Ultimate desires cannot be changed by reasoning. Instrumental desires can be changed by reasoning only insofar as we can change them by inference from other beliefs and desires we have.9

According to instrumentalism, our instrumental desires are formed and revised via reasoning, given our other beliefs and desires. But according to instrumentalism, you cannot modify your desires by reasoning in any other way, and cannot ever modify your ultimate desires by reasoning. (Instrumentalism permits that even our ultimate desires might be modified in other ways: by chemical changes, or a brick to the head, say. But it rightly says that such processes don’t count as reasoning.) Instrumentalism might seem to undermine desire-­as-­belief. Why would reasons beliefs be modifiable only in this very specific way? In what follows, I shall reject instrumentalism, and thereby provide grounds for optimism that we change our desires and our reasons beliefs in just the very same ways, just as desire-­as-­belief says. This isn’t quite a positive argument that we do change these states in the very same ways, but by knocking down instrumentalism, we remove the main barrier along the path to that conclusion. As I see it, instrumentalism faces numerous kinds of counterexample. But though I think this, I allow that some of these examples will fail to persuade my opponents, or else raise too many complications to be conclusive. In §8.2.3 I will describe three possible kinds of counterexample to instrumentalism, but I will set each aside since it raises too many complications. §8.2.4 presents a

9 Instrumentalism can be made more precise by specifying how the strength of instrumental desires depends on the prior desires and beliefs, and also by specifying how the strengths (or existence) of instrumental desires might depend on other related desires you have. For ease, in what follows I set these details aside and focus on the core instrumentalist claim in the main text.

162  Uncertainty and Reasoning different kind of case, one that most transparently undermines instrumentalism. If you like, you can skip to the main action in §8.2.4.

8.2.3  An Aside: Testimony, Intuition, Goodness So first, let me mention three different counterexamples to instrumentalism that I shall simply set aside since they raise too many complications. Perhaps we might rationally acquire ultimate desires via testimony.10 Perhaps you might form an ultimate desire to see a certain film because someone told you to see it. But this kind of objection is inconclusive: instrumentalists might reply that such testimony could change your desires only if it hooked up to some other desire you already had, such as a more general desire to see good films. Moreover, a defender of desire-­as-­belief might agree with this instrumentalist response. Pessimists about normative testimony believe that acquiring normative beliefs via testimony is problematic, unless you are really just using the testimony to gain non-­ normative information that you can combine with your prior normative beliefs.11 As such, these pessimists defend a view about the influence of testimony on normative beliefs that is instrumentalist in spirit. Someone who defended desire-­as-­belief but who was a pessimist about normative testimony would agree with the instrumentalist’s reply to the present objection. In short, to successfully undermine instrumentalism by appeal to desires formed by testimony, we would need to reject their reply and reject pessimism about normative testimony. Rather than attempt these tasks, I will just set this objection to instrumentalism aside. A second way in which we might rationally acquire ultimate desires is via rational intuition or some other kind of normative sense.12 Perhaps you might rationally come to desire something because you intuit how great it is (cf.  Darwall  1983, 39–40; Skorupski  2007b, 86). For example, if I present you with a choice you have never considered before, you might spontaneously come to desire one option just because it strikes you as correct—perhaps I present you 10  True, acquiring desires via testimony—and acquiring them via rational intuition, below—might not count as reasoning, exactly (I don’t know). Still, it is a rational process of desire formation of potential relevance to the prospects of instrumentalism, and in turn of desire-­as-­belief. 11 Pessimists include Hills (2009); Hopkins (2007); McGrath (2011). For optimism, see Sliwa (2012). 12  Some talk about a moral sense rather than a normative one, but since we are interested in all reasons beliefs, not just beliefs about moral reasons, the broader category is the interesting one for our purposes. For more on normative intuition, see e.g. Stratton-­Lake (2016).

Reasoning  163 with the option of entering a pleasant computer simulation for the rest of your life, and you spontaneously form the desire to stay out.13 But again, this argument is inconclusive: instrumentalists might reply that in such cases you really only form a new instrumental desire based on a prior desire you have, such as a desire for a good life (Sinhababu 2017, 52–5). Moreover, a defender of desire-­ as-­ belief might agree with this instrumentalist response: some defenders of desire-­as-­belief might be sceptical that we have a magical faculty of rational intuition (preferring, say, some more coherentist theory of belief formation). So again, rather than attempting to undermine the instrumentalist reply to this objection, or addressing the difficult question of whether we have a faculty of rational intuition, I shall simply set this objection to instrumentalism aside. A third possible objection to instrumentalism is that we might form ultimate desires by reasoning from normative beliefs, such as beliefs about goodness. For example, perhaps you might believe world peace would be good, and so come to desire it. But again, this argument is inconclusive. Instrumentalists might maintain that such reasoning could succeed only if it appealed to some prior more fundamental desire, such as a desire for good things. And yet again, some defenders of desire-­as-­belief might agree with this instrumentalist response: some defenders of desire-­as-­belief might think that beliefs about goodness can generate reasons beliefs only via the general belief that you have reasons to pursue good things.14 So again, rather than attempting to resolve these difficult wider issues, I shall simply set this objection to instrumentalism aside.

13  The example is of course from Nozick (1974, 42–5). 14  The suggestion in the main text is that you might be able to infer a reasons belief from a belief about goodness only via some prior reasons belief. Hume’s claim that you can’t get an ought from an is (T3.1.1.27) is often stated more precisely as Hume’s guillotine—the claim that no normative claim follows from non-­normative claims alone (for discussion, see Brown  2014; Prior  1960; Searle  1964; Wolf 2015). An inference from goodness to reasons would pass that test, so the proposed restriction on inference in the main text goes beyond Hume’s guillotine. As a result, it may be less plausible. But again, I won’t put any weight on that thought. Another possibility: remember that by ‘reasons belief ’ I just mean atomic first-­personal beliefs about reasons (§1.2, §3.4.1), and desire-­as-­belief, as stated, identifies only those with desires. So another possible counterexample to instrumentalism might involve inferring a reasons belief—in this narrow sense—from other beliefs about reasons. For example, you might infer a desire from the beliefs [that p] and [if p then I have reason to bring about q], and given everything I have said so far, desire-­as-­ belief treats neither of those states as desires. But cases of this kind also generate their own complications. For example, a natural extension of desire-­as-­belief would say that conditionally desiring [q, given p], is believing [if p, then I have reason to bring about q]. If we extend desire-­as-­belief in this way, then the above inference in fact does have a desire—a conditional desire—amongst its premises, and as such may be consistent with instrumentalism. Similar remarks apply to many other inferences to reasons beliefs from other beliefs about reasons. So yet again, I shall set such cases aside.

164  Uncertainty and Reasoning In short, I shall not press any of the above objections to instrumentalism, since they raise too many complications. In part this is because the instrumentalist has a reasonable reply to each: that we cannot modify our desires in the proposed manner unless we tacitly make that inference from some prior more fundamental desire. And in part this is because it isn’t clear whether defenders of desire-­as-­belief need agree with these objections: the relevant claims about reasons beliefs are themselves disputed. It will be better to assess instrumentalism in a way that raises fewer complications.

8.2.4  Against Instrumentalism In this section, I present my preferred main objection to instrumentalism. In what follows, I will sometimes talk about our uncertainty about our desires: for that notion, see §8.1.15 To understand why instrumentalism is false, we need to reconsider the ultimate/instrumental distinction itself. This is a useful distinction: for example, all normative reasons must be either instrumental or else ultimate. But though reasons must fall into one of these two categories, it doesn’t follow that our mental states must follow a similar pattern. Though I can believe [that I have an instrumental reason to v], and can believe [that I have an ultimate reason to v], I can also believe [that I have a reason to v], leaving open what sort of reason this is. That is, though reasons themselves must be determinately either ultimate or instrumental, our reasons beliefs might be indeterminate in this respect. With that claim in mind, we might say something similar about our desires. Instrumentalism says that our desires can be exhaustively divided into instrumental and ultimate. And I agree that many desires should be categorized in those ways. But on reflection, it is not clear why every desire should neatly fall into one of these two categories: perhaps some desires are somewhat indeterminate. That is, you might want something and acknowledge that you want it for its own sake, and you might want something and acknowledge that you want it only as a means to an end. But you might simply want something without having given much thought to whether you want it for its own sake or instead for the sake of something else. For example, though I am sure that I want 15  Perhaps the reasoning I describe below would in fact ultimately work via one of the methods above (say, via normative intuition). That would be fine by me: for my argumentative purposes the important point is just that the reasoning I describe below is possible, however you might ultimately understand it.

Reasoning  165 friendship, I am not at all sure whether I want this for its own sake or instead because of what it leads to (e.g. mutual happiness). Or, for another example, take the desire for health. Most people will simply recognize that health is worthwhile, and their wanting it moves little beyond this thought: the question of whether they want it for its own sake, or as a means to an end, will simply have not arisen. In turn, there will be no determinate answer as to whether their desire for health is instrumental or ultimate. In many cases like these we are surer that we want something than we are of the way in which we want it. If this is right, it makes room for a kind of reasoning with our desires that goes beyond instrumentalism: we might try to settle whether some desire is instrumental or ultimate. How might we do that? One way to do so is to systematize our desires in a manner that is non-­arbitrary and relatively simple (Smith 1994, 159–61; cf. Schroeder 2007a, chap. 10). For example, you might conclude that your desire for health is instrumental because you definitely instrumentally desire other things that lead to happiness, and so the simplest option is to desire health in this same instrumental way. You might corroborate this conclusion by noting that if your future promises only unhappiness, it seems that you should no longer desire health, and so again, the simplest option is to desire health only as a means to the end of happiness. Such reasoning is obviously highly defeasible, depending on further beliefs and desires that you hold, but the basic point stands that we might rationally reform our desires into ultimate or instrumental desires as a result of trying to find simple ways to systematize them. In short, we can modify our desires in ways that falsify instrumentalism. Even if instrumentalism were correct that we can only modify our instrumental desires in light of our other desires, and can never modify our ultimate desires at all, many of our desires are indeterminate as to whether they are instrumental or ultimate. We can modify those indeterminate desires by reasoning, reforming them to make them more determinately ultimate or instrumental. With this claim in hand, a bolder conclusion is available. Even with desires which are determinately instrumental or ultimate, we might be unsure about whether the desire really falls into that category.16 As a result, the reasoning we do with indeterminate desires will also extend to instrumental and ultimate desires. For example, imagine you think that you ultimately desire friendship,

16  Remember that despite appearances, this kind of claim is not one about self-­knowledge: the point is not that the desire is 100 per cent in one category but we don’t know which, but rather that although it is in one category, it is there only tentatively. Again, see §8.1 on uncertainty in our desires.

166  Uncertainty and Reasoning but are not completely sure of this. You might then note that you want friendship only when it leads to mutual happiness, which you are completely sure you ultimately desire. As a result, you might begin to desire friendship only instrumentally, thinking that it is attractive only as a means to the happiness of those involved. (Later, you might even modify that desire as you modify other instrumental desires.) In this way, instrumentalism is more broadly false: even in cases where you do think of a certain desire as ultimate, you might nonetheless revise that desire by first reconsidering its status as ultimate. In short, instrumentalism looks plausible only if we presuppose that all desires are either ultimate or instrumental. Insofar as some are indeterminate in this respect, we might modify them by reasoning. And once we notice this, it seems clear that we might modify other desires in the same way given that we can be unsure whether they are ultimate or instrumental. To ensure the argument is clear, here is a concrete counterexample to instrumentalism. I used to ultimately desire equality. I wanted a more equal society, and I saw this as something worth pursuing not only for its effects, but also for its own sake. If pressed, I might have allowed that I wasn’t 100 per cent sure that this is what I wanted, but I would nonetheless have maintained that I was very confident. But then I came across some of the philosophy written on the subject (Parfit 1997), and this changed how I thought about the matter: I came to see that some of the benefits of equality were clearly instrumental, and I became less clear what remained attractive about equality once those instrumental benefits were acknowledged. I came to want equality only as a means to an end, and not for its own sake. (Perhaps, as I age, further reasoning will lead me to abandon equality as even an instrumental goal.) This kind of case shows that instrumentalism is false: we can change our ultimate desires by reasoning, at least in part by coming to doubt whether they are ultimate desires. How might instrumentalists respond to this argument? They might insist that whereas our reasons beliefs can be silent about whether the reason is ultimate or instrumental, our desires must be either ultimate or instrumental. But it is hard to see why this must be the case. A good general principle is to assume that things are possible unless there is an explanation as to why they are not: the default view should be that we can desire things sans phrase, and not only instrumentally and ultimately. Our objector needs a positive argument for ruling this possibility out. One possibility is simply to insist that ‘ultimate’ and ‘instrumental’ are defined so that these two categories of desire are exhaustive. We might, for

Reasoning  167 example, use the label ‘ultimate desire’ to mean ‘any desire that is not instrumental’. But this terminological wrangling achieves nothing. For on this use of words, we now need to distinguish within the category of ‘ultimate’ desires those desires that are for things for their own sake, and those desires which are more indeterminate than this: those that are for things for something’s sake, we know not what. We could give these distinct kinds of desire different labels, and my claims could then be reformulated appropriately. You can’t turn three things into two merely by using words in a more coarse-­grained way. A better instrumentalist option is to treat the contrast between ultimate and instrumental desires as a contrast between the causes of these states: we might treat instrumental desires as those that were caused by other desires, and treat ultimate desires as those not caused by other desires in this way. This might secure the conclusion that every desire must fall into one of these two categories. But this suggestion is mistaken. Your desire to go running might start off as purely instrumental (you want to get fit) but develop into an ultimate desire over time as you come to recognize the intrinsic merits of running. The original cause of a desire is no guide to its present nature. An objector might move to a nearby view on which an instrumental desire is one that is presently sustained by another desire, and ultimate desires are those not sustained in this way. This is a better suggestion, and it might handle the above case, since over time your desire to go running ceases to be sustained by your desire to get fit, but instead takes on a life of its own. Still, this suggestion is also mistaken: instrumental desires, like children, can outlive their parents. Imagine that you desire to save up money, and discover that a distant petrol station is slightly cheaper, and so you form the instrumental desire to fill up there next week. But in the meantime, you win the lottery. Now you no longer desire to save up money. But you might fail to bring this information to bear on your desire to fill up at the distant petrol station. And so you might continue to instrumentally desire to fill up at the distant petrol station, despite having abandoned the relevant desire on which it was based. There are numerous such cases where we fail to rationally modify our instrumental desires in light of new information we gain. In this manner, when we are irrational, our instrumental desires can outlive their parents. The proposed view wrongly makes this kind of irrationality impossible, and should therefore be rejected. An objector might reply that if your desire genuinely persists and yet is not sustained by any other desire you have, then at that point it is an ultimate desire. But this is unconvincing. It might be true that over a matter of years,

168  Uncertainty and Reasoning we can form ultimate desires for things in virtue of having long instrumentally desired them. But it is doubtful that this process can take place over days, and yet it is clearly possible to irrationally hang onto an instrumental desire even when it was only recently formed, as in the above petrol case. Our objector has one last, and best, suggestion. They might say that instrumental desires are ones that we are disposed to change in light of certain changes in desire and belief, whereas ultimate desires are those that we are disposed not to change in that same way.17 This suggestion is highly plausible, and seems to handle the cases above (it is plausible in part because it mirrors a plausible view about the difference between beliefs about instrumental reasons and beliefs about ultimate reasons). But, unfortunately, it doesn’t give my opponent what they need. For if this is the distinction between ultimate and instrumental desires, then it supports, rather than undermines, the claim that some desires are neither ultimate nor instrumental. For the proposal is that an instrumental desire is one that we are disposed to change under certain conditions, and an ultimate desire is one that we are disposed not to change under those conditions. But this leaves open the possibility that we might have neither of these dispositions, or else that we might have both dispositions to a certain limited degree. Either way, there will be desires which do not fall neatly into either category, just as I said. I conclude that instrumentalism is false. If we assume that all desires are determinately either instrumental or ultimate, it might look attractive. But that assumption is false: many desires are indeterminate. Moreover, even those desires that do fall into one of these categories might be placed there only tentatively. As a result, we can revise our desires in ways that go beyond instrumentalism. Let me summarize our discussion. Desire-­as-­belief entails that reasoning modifies our desires and our reasons beliefs in the very same ways. This might seem problematic because of instrumentalism, the view that we can modify our desires only in very limited ways. I first set aside three possible objections to instrumentalism, since each raised too many complications. But then, in this section, I gave a neater counterexample to instrumentalism: some of our desires are neither instrumental nor ultimate, and we can revise those desires by reforming them into desires of one of those kinds. With this point 17  Let us here leave open whether we might be disposed to change our ultimate desires in some other way: that is what is under discussion when we debate instrumentalism. The crucial claim here is just that we are disposed to change our instrumental desires in a very particular way, and we are disposed to not change our ultimate desires in that same way.

Summary  169 established, parallel reasoning suggests that we might revise our desires even where they are presently ultimate desires. I conclude that instrumentalism is false. This does not positively demonstrate that we do change our desires and reasons beliefs in just the same ways. Showing that is beyond me, since it would require a comprehensive theory about how we form and revise normative beliefs, and I don’t have one of those. Still, I have at least removed the main obstacle to this conclusion, and so I conclude that we have sufficient grounds for optimism about desire-­as-­belief on this score.

8.3  Summary In this chapter, I discussed two related issues for desire-­as-­belief. First, I discussed how desires come in degrees. I suggested that the strength of a desire is the same as how weighty you believe the relevant reason to be. But beliefs about reasons also come in degrees of certainty. I suggested that desires have a corresponding feature: we can be uncertain about what we want. Second, I discussed our capacities to rationally change our desires through reasoning. I argued against the dominant instrumentalist model, in part by arguing that many desires are indeterminate as to whether they are instrumental or ultimate. In the next chapter, I turn to another objection to desire-­as-­belief: animals. They might seem to have desires, but lack beliefs about reasons. To reply to this objection, we’ll have to consider some questions about mental content.

9 Representing Reasons Imagine that your pet cat paws at the door. ‘She wants to get out’, you might say. But despite attributing this desire to her, you might doubt that she has any beliefs about normative reasons. This seems to count against desire-­as-­belief. More generally, it might seem as though various animals and small children have desires—for food, or sleep, say—but can have those desires without possessing the concept of a reason, and so without having any beliefs about reasons. In short, it might seem that desire-­as-­belief raises the bar too high for desiring, and wrongly entails that many creatures have no desires, when they clearly do (see e.g. Hawkins 2007).1 This objection is simple to state, but hard to assess. Deciding whether a given creature has some mental state is often extremely difficult, since it depends on difficult philosophical questions about what it takes to have that state of mind, and on difficult empirical questions about whether that creature has what it takes.2 The issues here touch on a wide range of questions, including many that are awesomely fundamental: what does it take to have a mind? How can a thought be ‘about’ the world? The difficulty of these issues is inherited by downstream questions such as whether robots can think, and whether lobsters can feel pain, and—our interest—whether animals have desires or beliefs about reasons. I’ll brave these deep waters as best I can. I don’t hope to settle all these questions comprehensively, but instead merely to give grounds for optimism about desire-­as-­belief on this score. To do so, I’ll sketch some views about mental content, and about animals. Objectors might complain that these commitments are a burden placed on us by desire-­as-­belief, and that it is already a mark against the theory that it forces us to take a stand on these 1  Those who have written about the normative beliefs of animals often focus on whether they have moral beliefs (see e.g. de Waal  1996; Hamlin  2013; for related discussion, see Machery and Mallon  2010). But as I have said elsewhere (e.g. §3.1.1), desire-­as-­belief analyses desires as reasons beliefs, and it is certainly possible to have reasons beliefs that are not about moral reasons, and perhaps possible to have moral beliefs that don’t correspond to any reasons beliefs (perhaps psychopaths are like this). As a result, I will simply set aside questions about whether children and animals have moral beliefs. 2 For helpful philosophical discussion and windows onto the wider empirical literature, see Andrews and Beck (2017); Beck (2012); Hurley and Nudds (2006).

Desire as Belief: A Study of Desire, Motivation, and Rationality. Alex Gregory, Oxford University Press. © Alex Gregory 2021. DOI: 10.1093/oso/9780198848172.003.0010

Conceptual Role Semantics  171 issues. But this is a burden only if desire-­as-­belief forces us to take the wrong stand on these issues: if it presupposes the right view, that hardly counts against it. So objectors must really complain that these wider claims of mine are incorrect. That complaint stands or falls with the details that follow. That is, objectors are in no position to complain that desire-­as-­belief forces us to take a stand on these questions. If they object to desire-­as-­belief by appeal to the mental states of animals, they too are taking a stand on the relevant issues. Just as rebutting this objection is hard work for me because it forces me to make wider claims about the nature of the mind, so, too, those pressing this objection need to press it by appeal to wider claims about the nature of the mind. Otherwise they are the ones in danger of presupposing falsehoods. We are all in deep water here, no matter which way we intend to swim. The chapter is structured as follows. In §9.1, I describe a general theory about the contents of beliefs. §9.2 and §9.3 apply this theory to reasons beliefs. Finally, in §9.4, I return to animals.

9.1  Conceptual Role Semantics What is it for a belief to have a certain content? What makes it true that a certain belief of mine is about France, rather than about Andorra? Or, more relevantly for our purposes, what determines whether a certain belief is one about reasons? There are many possible theories of mental content, but I will proceed by appeal to conceptual role semantics (CRS).3 Very roughly, CRS says that a belief has a specific content just in case it plays the right role in your psychology. We can describe this view schematically as follows: x is a belief [that p] = x has the role of being sensitive to certain input mental states and of producing certain output mental states.

3  One reason for this focus is that Terence Horgan and Mark Timmons’ ‘moral twin earth’ argument seems to rule out many rivals to CRS as theories about normative mental content (1991, 1992). See also footnote 6, below. CRS is sometimes labelled ‘inferentialism’, amongst other things. Some develop the view as a theory of linguistic content, rather than mental content, as I do here. Defenders include Block (1986); Harman (1982); Peacocke (1992). Defenders of CRS specifically about normative beliefs include Chrisman (2012,  2015); Enoch (2011, 177–84); Jackson and Pettit (1995); and Wedgwood (2001,  2007). CRS is in many respects also close to the ‘radical interpretation’ procedure found in Lewis (1974). For that comparison, see Peacocke (1992, 36–9); Block (1986, 645–6; 1987, 165–6).

172  Representing Reasons Output state 1

Input state 1

Input state 2

Belief (that p)

... Input state n

Output state 2

... Output state n

Figure 9.1  Conceptual role semantics.

Here, depending on what [p] is, there is some definitive list of which mental states are the inputs and outputs. We can also represent CRS as in Figure 9.1. CRS is perhaps most intuitive with respect to logical belief contents. For example, it’s plausible that the belief that [p and q] can be defined as the belief you are disposed to adopt if you believe [p] and believe [q], and which disposes you to believe [p] and to believe [q].4 I will pursue CRS as a view about reasons beliefs, so that a belief gets to be a belief about a reason in virtue of playing the right role in your psychology. But first, three clarifications are in order. First, when I say that the content of a belief depends on its psychological role, that role is not a matter of the actual causal inputs and outputs of that belief—for example, perhaps your belief was caused by some fallacious reasoning that is no part of its role. Some claim that the role of a belief is determined by what it ought to be caused by and ought to cause (Peacocke 1992; Wedgwood  2001,  2007). Alternatively, perhaps the role of a belief is determined by what it is caused by and causes under ‘normal’ circumstances, or else determined by the evolutionary function it has. But I won’t get bogged down in this: I will just talk about the roles of different states of mind—their ‘inputs’ and ‘outputs’, or what they are ‘sensitive to’ and ‘produce’—and leave it open exactly what constitutes such roles. Second, I don’t commit to CRS as the correct theory of all belief contents. Rather, I develop it as a natural option for reasons beliefs, and leave open whether other belief contents might be better analysed in some other way. Third, relatedly, as I pursue it, CRS says that the relevant inputs and outputs of a belief are always other mental states. Some defenders of CRS instead think 4  Here I focus on the role of one part of these belief contents: the concept . As the name suggests, CRS is normally developed primarily as a theory about concepts, from which we derive a theory of whole belief contents. But for presentational ease, I’ll ignore this issue and gloss over the difference between theories that focus on single concepts and those that focus on whole belief contents.

Inputs and Outputs of Reasons Beliefs  173 that the inputs and outputs of a belief can be actual features of the world, not just other mental states.5 But I will instead work with a ‘short-­arm’ kind of CRS on which inputs and outputs are always other mental states.6 Even this short-­arm kind of CRS explains how our beliefs represent the external world. We can say that a belief represents [p] in virtue of playing psychological roles that are somehow isomorphic to the worldly nature of [p] itself (Peacocke 1992, 133–43; Wedgwood 2001, 9–10; 2007, 87). For example, in a simple case, imagine you have a belief which you are inclined to abandon if you come to believe that grass is red, or blue, or yellow, or . . . Since this role is isomorphic to the fact that grass can only be green if it is not red, blue, or yellow, this role might make it true that your belief has the content that grass is green. Of course, whether this is really true would depend on the other roles of the belief in question, and other facts about grass being green. But this at least serves to illustrate the basic thought: the representational content of a belief is fixed by its roles, so that a belief represents whatever thing in the world best matches its roles.7 Or, more crudely, the view says that you represent a part of the world when the structure of your mind matches the structure of that part of the world. With this very broad sketch of CRS on the table, we can now apply it to reasons beliefs. CRS says that a mental state is a belief [that I have a reason to bring about p] just in case it has certain input mental states and certain output mental states. At this point, we need to do three things: (1) Identify the relevant inputs. (2) Identify the relevant outputs. (3) Show that those inputs and outputs are isomorphic to the worldly relations that reasons themselves stand in. I take these tasks in turn.

9.2  Inputs and Outputs of Reasons Beliefs What are the inputs to reasons beliefs? That is, if a state of mind qualifies as a belief [that I have a reason to bring about p], what must it be sensitive to? I suggest that the following two inputs are central: 5  See e.g. Harman (1982); Peacocke (1992, 113). For useful discussion, see Block (1986, 1987). 6  This assumption may be mandatory with respect to normative beliefs, given the argument in Horgan and Timmons (1991,  1992). See also Eklund (2017); Horgan and Timmons (2009); Sinclair (2018). 7  We might need to qualify this so that a belief doesn’t simply represent whatever worldly thing is the best match, but instead represents the best match if there is a sufficiently well-­matching contender. This might be necessary to make room for systematic misrepresentation. But I’ll set such possibilities aside here.

174  Representing Reasons Reasons beliefs are normally sensitive to instrumental reasoning. Reasons beliefs are normally sensitive to beliefs about goodness. The first of these inputs most obviously applies to beliefs about instrumental reasons: such beliefs are sensitive to changes in our beliefs about what we have ultimate reason to do, and changes in our beliefs about the available means to those ends. Reasons beliefs are beliefs that are sensitive to these other mental states in the obvious ways. The second of these inputs applies more broadly: reasons beliefs are sensitive to our beliefs about goodness in that we normally believe that we have reasons to bring about good things, and reasons to prevent bad things. I certainly don’t claim that these two inputs are the only inputs to reasons beliefs, but they are plausibly amongst the most important inputs to such beliefs, and they will be enough for me to make the points I wish to make. I now turn to our second task: what are the outputs of reasons beliefs? That is, if a state of mind qualifies as a belief [that I have a reason to bring about p], what must it tend to produce? I suggest that the following three outputs are central (though, again, perhaps not exhaustive): Reasons beliefs normally enter into instrumental reasoning. Reasons beliefs normally enter into reasoning about what you ought to do. Reasons beliefs normally produce motivation. These roles of reasons beliefs are easy to illustrate. The first is probably the most obvious: in order to discover what we have instrumental reason to do, we appeal to our prior beliefs about reasons—the latter beliefs have the former as an output. The second role is also intuitive: if you are trying to determine what you ought to do overall, you surely do so by weighing the reasons you have for or against various options. And the third role will be familiar from previous claims I have made (e.g. §3.1): reasons beliefs dispose us to be motivated. There are various objections to this broad proposal that I ought to address. One objection is that the view is circular. Can we really say that a belief is a reasons belief just in case it is sensitive to other reasons beliefs, and produces such beliefs? How do those other beliefs get their contents? We might worry that our account of these belief contents is going to be viciously circular. But this kind of circularity is fine (pace Wedgwood 2001, 5; see also Peacocke 1992, 10; Lewis 1970). We can say that no reasons belief gets its content determined

Inputs and Outputs of Reasons Beliefs  175 in isolation, but instead the system as a whole becomes contentful in virtue of the relations between the elements. Here is an analogy: I might describe a picture to you only by describing relations between the elements: dot A is five dots away from dot B, and dot C is three dots away from dot A and seven dots away from dot B, and . . . If my description is long enough, it might suffice to uniquely determine a particular picture, even though I have only ever told you how each dot relates to the others, not where any single dot is located non-­relationally. Similarly, according to CRS, we describe the roles of our beliefs in sufficient detail that the content of each is fixed by its relation to other states of mind. Even if we never break out of this circle, these relations can nonetheless determine the content of each belief. Another objection is that someone might have reasons beliefs that don’t play all of the roles above. As a stark illustration, take Milton’s Satan, who is disposed to believe that he has reason to bring about bad things, not good things (§6.3). This might suggest that the role of being sensitive to beliefs about goodness is not essential to reasons beliefs after all. More realistic cases are also available, such as philosophers who subscribe to confused views about reasons, and whose reasons beliefs thereby fail to play some of the appropriate roles, but who seem to have such reasons beliefs nonetheless. But this objection is mistaken, for three reasons. First, in §9.1 I said that the ‘role’ of a state of mind is not merely a matter of what it actually causes and is caused by. Rather, it is determined by something else, such as the evolutionary function of that state of mind. As a result, the beliefs of Satan and confused philosophers might still play the relevant roles even if they in fact fail to have the relevant causes and effects. Second, I assume that all reasons beliefs must play a significant number of the relevant roles, but I allow that not every reasons belief need play every single one of those roles (this marks a potential contrast with Peacocke 1992; Wedgwood 2001; 2007, 85). So Satan’s beliefs might be reasons beliefs, despite their insensitivity to his beliefs about goodness, because they are at least sensitive to his other reasons beliefs in the right ways, such as in instrumental reasoning. Similar remarks might apply to confused philosophers. A third—best?—reply to this objection is to abandon my earlier commitment to ‘short-­arm’ CRS and instead adopt a view which combines CRS with a concession to social externalism: perhaps others’ mental states can directly affect the contents of your own (Harman 1982, 247–8; Peacocke 1992, 27–33; see also Burge 1979; Williamson 2008, chap. 4). For example, perhaps Satan’s beliefs count as reasons beliefs because he expresses them using the word ‘reason’, and that makes his beliefs’ contents affected by how others use that word.

176  Representing Reasons In this way, Satan’s beliefs might count as reasons beliefs even though his personal psychology is abnormal. This manoeuvre requires us to move beyond short-­arm CRS, but is nonetheless consistent with the central claims I have made, since we might still think that the word ‘reason’ gets its meaning from the role it tends to play in our psychologies. Though I would be happy to extend CRS in this way, in what follows I largely set this possibility aside for ease of presentation. A further two objections to my overall view focus on the motivational role of reasons beliefs. First, we might worry that this is not part of the role of reasons beliefs at all. On this view, we should characterize reasons beliefs purely in terms of their roles in producing, and being sensitive to, other normative beliefs. Let’s imagine a whole community of people who have mental states that play these roles, but none of which have any bearing on their motivations or actions.8 Do these people have reasons beliefs? I am inclined to think not (see also §3.1). If these attitudes have no connection with motivation whatsoever, they seem too disengaged to count as normative beliefs. They sound more akin to beliefs about some unendorsed system of norms, such as our beliefs about the value system of the Incas, or about a code of etiquette that we understand but disregard. As a result, we should claim that the outputs of reasons beliefs include not just other normative beliefs, but also motivation. But another objection presses in the reverse direction. Impressed by the connection between reasons beliefs and motivation, we might try to define reasons beliefs solely by their role in producing motivation, and make no reference to their relations to other normative beliefs. But this seems wrong. Imagine creatures who are driven by certain attitudes to do certain things, but these attitudes bear no systematic relations to one another or to any other mental states of theirs such as their beliefs about goodness (see also the discussion of animals, below). It is far from clear that these creatures have reasons beliefs. If the states that drive them to move are impervious to even the most elementary kinds of reasoning—such as instrumental reasoning, or being weighed against one another—it is far from clear that we would even count them as motivational states at all. Plausibly, to even count as motivation, there has to be some rational processing going on such that the motivational states in question bear some systematic relations to one another. I conclude that the roles of reasons beliefs go beyond their connections to motivation, 8  I focus on a whole community of people to avoid any complications due to the possibility of social externalism, as discussed above. For related discussion, see Dreier (1990, 9–14); Tresan (2009).

How We Represent Reasons  177 and include their roles in relation to each other and to other normative beliefs, just as I said.

9.3  How We Represent Reasons We can picture all of the above roles of reasons beliefs as in Figure 9.2. The key idea behind CRS now is that a mental state represents something—p—in the world just in case its roles are isomorphic to the relations between p and other things in the world.9 Intuitively, the basic idea is that a mental state represents whatever thing in the world is the best match for its roles. So having described the roles of reasons beliefs, I’ll now just briefly explain why these roles are isomorphic to the relations that reasons stand in. I’ll address the roles of reasons beliefs in three stages: first, the roles relating reasons beliefs to other reasons beliefs; second, the roles relating reasons beliefs to other normative beliefs; and third, the roles relating reasons beliefs to motivation. In each case, I argue that those roles are isomorphic to relations that reasons themselves stand in. Since our reasons beliefs are a mirror for reasons themselves, the former represent the latter. First, I said that reasons beliefs play the roles of being sensitive to, and pro­ du­cing, other reasons beliefs. These roles of reasons beliefs are clearly iso­ morph­ic to the relations that reasons themselves stand in. For example, if you believe [that you have reason to bring about p], one role of that belief is that Beliefs about Ought Beliefs about Goodness Reasons Beliefs

Reasons Beliefs

Reasons Beliefs Motivation

Figure 9.2  The roles of reasons beliefs. 9  Ralph Wedgwood does things slightly differently: he says that a mental state represents whatever property in the world guarantees that the inferential roles in question are correctness-­preserving (Wedgwood  2001, 10–11; 2007, 86–7; compare McHugh and Way  2017). I worry that requiring a guarantee of correctness in output is plausible only to the extent that we restrict CRS to logical concepts only. But I won’t develop that thought here, and the claims I make below don’t depend on it.

178  Representing Reasons when combined with the belief [that q is a necessary means to p], you’ll come to believe [that you have instrumental reason to bring about q]. Reasons themselves stand in these relations to each other: if [you have reason to bring about p], and [q is a necessary means to p], then [you have instrumental reason to bring about q].10 Reasons beliefs are also sensitive to and produce other normative beliefs. For example, if you believe [that p is good], you will normally infer [that you have reason to bring about p]. This mirrors a fact about actual reasons: that if [p is good], then normally [you have reason to bring about p]. Or for another ­example, if you believe [that you have strong reason to v], and believe [that you have only weak reasons to do alternatives], you’ll normally infer [that you ought to v]. Again, this mirrors a fact about actual reasons: that if [you have strong reason to v], and [you have only weak reasons to do alternatives], then [you ought to v]. In short, the roles of reasons beliefs with respect to other reasons beliefs and other normative beliefs are plausibly isomorphic to relations that reasons themselves stand in. A slightly different case is the role of reasons beliefs in producing motivation. The roles above were roles in relation to other beliefs. The role of reasons beliefs in producing motivation is different in this respect. But a structurally parallel story applies: it’s in the intrinsic nature of reasons that they favour certain actions; this is just what they do. This is mirrored by the fact that reasons beliefs have the role of inclining us to perform those actions. That is, just as there is a relationship between the proposition [that you have reason to v] and the proposition [that you v] (the former means the latter is favoured), there is a parallel relationship between reasons beliefs and your motivations: the belief [that you have reason to v] tends to produce a motivation [to v]. So again, this role of reasons beliefs mirrors a relation that reasons themselves stand in. All of the points above could be described in more detail if we wanted a comprehensive theory of normative mental content. For example, perhaps this view has implications about whether normative properties are natural or non-­ natural, and/or for first-­ order normative theorizing (see, perhaps, Jackson 1998). But I won’t even begin to address such questions here. I suggest we return to the central point. I said that a belief is about whatever thing in the world best matches the roles of that belief. Reasons beliefs play numerous roles, such as their roles in producing and being sensitive to other reasons beliefs, their roles in producing and being sensitive to other normative beliefs, 10  For related discussion, see e.g. Kolodny (2018).

How We Represent Reasons  179 Other Normative Beliefs

Other Normative Beliefs

Reasons Beliefs

Reasons Beliefs

Reason Beliefs

Motivations to Act

Figure 9.3  The roles of reasons beliefs again.

Other Normative Truths

Other Normative Truths

Truths about Reasons

Truths about Reasons

Truths about Reasons

Favouring of Acts

Figure 9.4  The relations reasons stand in.

and their role in producing motivation. In each case, the role in question is isomorphic to a relation that reasons themselves stand in: truths about ­reasons stand in the relevant relations to one another, to other normative truths, and to truths about action. It follows that reasons beliefs, given their roles, represent reasons. That conclusion is no surprise(!), but the crucial thing is that we now have a good explanation of why it is true.11 We can represent our theory as in Figures 9.3 and 9.4. Since the diagrams are isomorphic, CRS implies that the reasons beliefs at the ­centre of Figure 9.3 represent the reasons at the centre of Figure 9.4. Before we return to animals, a slight aside: with all this out in the open, it’s interesting to apply the above claims to desire. Desires play roles with respect to other desires, such as the roles they play in producing instrumental desires. Desires also play roles with respect to other normative beliefs, such as being sensitive to our beliefs about goodness. Finally, desires play the role of pro­du­cing motivation. These roles of desire seem isomorphic to the relations that reasons 11  Or, to put things another way, I shouldn’t have proceeded by labelling certain beliefs as ‘reasons beliefs’ and then asked what roles they played. I should have started by neutrally identifying a mental state that plays certain roles, and then inferred that given its roles, that mental state must be a belief about a reason.

180  Representing Reasons themselves stand in, and so desires represent reasons, just as desire-­as-­belief says. To the extent that the above explanation of how reasons beliefs get their content applies equally to desires, it would favour desire-­as-­belief. This point is starkest with respect to instrumental reasoning, so let’s go through that case once more. It’s widely accepted that one role of desires is that they produce other instrumental desires through reasoning (§8.2). For example, if you want beer, and believe there’s beer only at the pub, you’ll normally form a desire to go to the pub. This role of desire perfectly mirrors certain facts about reasons: if it’s true that you have reason to get beer, and true that there’s beer only at the pub, then it’s true that you have reason to get to the pub. We might see this isomorphism between the role of desires, and entailments between reasons, as in some way coincidental. But according to CRS, a state of mind represents a part of the world just in case the role it plays in your psychological economy is a mirror for that part of the world. Since that is true in this case, it is plausible that desires represent reasons. I suggest that a similar story could be told about the other roles of desire. In this way, CRS promises another argument for desire-­as-­belief—so it seems to me. But I won’t press that point in any more detail here. Let me summarize our discussion. I sketched a general theory of how belief contents get determined: CRS. It says that beliefs have certain inputs and outputs, and a belief represents whatever thing in the world has an isomorphic structure. Reasons beliefs have reasons beliefs, other normative beliefs, and motivation, as inputs and outputs. Those inputs and outputs are isomorphic to the nature of reasons themselves. In some respects, this discussion has inevitably been somewhat sketchy: filling out a complete theory on all these issues would be a gargantuan task. But I hope to have said enough to help us to assess the objection from animals with which we began.

9.4 Animals I above suggested that to have a reasons belief is to have a mental state that plays certain roles. Since many of these roles connect reasons beliefs to other normative beliefs, in order to have reasons beliefs you need a network of normative beliefs that (mostly) stand in the right relations to each other. Plausibly, animals do not have a fully fledged network of such attitudes. But, equally plausibly, they do have states that approximate some aspects of such a network. In turn, I shall argue that animals count as borderline cases of

Animals  181 creatures with reasons beliefs, and so count as borderline cases of creatures with desires (henceforth: are borderline desirers). Let’s take my optimistic claims first. Animals clearly have mental states that affect their behaviour. Let’s stipulatively label those states ‘inclinations’. Animal inclinations are clearly sensitive to various kinds of instrumental reasoning: I presume this is the best explanation of how animals can develop (simple) tools for use, and can determine what to do by reasoning by elimination.12 Moreover, animals plausibly have some beliefs about the positive elements of various states of affairs (‘it’s nice when I have food’). Those beliefs might count as borderline cases of beliefs about goodness, and they plausibly interact with their inclinations in appropriate ways. Finally, animals plausibly sometimes balance their inclinations against one another in deciding what to do overall: they weigh alternatives in appropriate ways, at least some of the time. Such reasoning parallels inferences that we make, from our reasons beliefs to beliefs about what we ought to do. In short, inclinations play the roles of reasons beliefs, to some extent. But in other respects I am a pessimist: inclinations are not perfectly like reasons beliefs. For example, animals are clearly worse at instrumental reasoning than we are: we are capable of tracing out far more remote implications of what we have reason to do (this is a partial explanation of why our tools are vastly superior to animals’). I also said that their inclinations are sensitive to beliefs about the positive elements of states of affairs, but it is not clear if those beliefs themselves count as more than borderline cases of beliefs about goodness. For example, their beliefs about ‘goodness’ are also less sensitive to instrumental reasoning than ours are. And although animals can balance their inclinations against one another, they are plausibly less able to do this than we are, especially in cases where those attitudes can be weighed only if we know how to weigh risks against one another: competence at this requires greater numerical ability than many animals have. In short, inclinations play some of the roles of reasons beliefs, to some extent. But inclinations nonetheless don’t play all of those roles, and certainly not to the same degree. By comparison with our reasons beliefs, animals perform far fewer inferences with their inclinations, and do so within much more restricted bounds. I conclude that animals have attitudes which are borderline reasons beliefs. In turn, I claim that animals count as borderline 12  For some relevant discussion, see Camp and Shupe (2017); also Call (2006); Bermudez (2006); Erdohohegyi et al. (2007).

182  Representing Reasons desirers. Our mental processes are more sophisticated than animals’, and this fact brings with it the capacity to have full-­blown desires, rather than mere inclinations. Some might find this surprising. Perhaps it is plausible, in light of the above, that some animals count as borderline cases of creatures with reasons beliefs (see also Korsgaard 2009, 110–11). But surely some of those animals are clear-­cut cases of creatures with desires, and so not mere borderline cases, as I claim? A small initial point to make is that I above talked about ‘animals’ as though they are all the same. More realistically, some ‘lower’ animals such as slugs might not even qualify as having borderline desires since their psychology is so radically different from ours, and vice versa some ‘higher’ animals such as apes might qualify as having full-­blown reasons beliefs. Those animals can of course be clear-­cut desirers (I leave it open). Still, this doesn’t wholly disarm the worry: perhaps animals like apes might count as having full-­blown reasons beliefs. But what about animals like cats? Do I really say that they aren’t sophisticated enough to have clear-­cut desires? Yes. To sugar this pill, we can note two debunking explanations of why we might be misled into thinking that animals such as cats are clear-­cut desirers. First, I argued in §7.3 that we should distinguish likings and desires. Desire-­as-­belief permits that animals might like certain things even if they do not have the concept of a reason, and do not have desires. So when it seems tempting to claim that animals have desires, we should not be misled by the thought that they like certain things. Your cat can like hunting mice, or eating tinned herring, but that is not the same as her desiring these things: it merely means that she finds these things pleasurable. Second, ‘want’ is occasionally used in an older sense to mean needs. For example, when you point to the brickwork on my house and say, ‘it wants repointing’, you mean only that it needs repointing, not that it is in a certain state of mind. Equally, you might say that the grass wants watering, or that the car wants oil: again, these claims do not refer to mental states at all. The temptation to think that animals are clear-­cut desirers may come from our tacitly thinking of assertions that use ‘want’ in this sense. For example, when we say that the cat wants milk, or to be let out, those claims might be clear-­cut truths, but that is consistent with my claims if these are intended as claims about what these animals need, not about their desires, in our sense. So in one sense of ‘want’, animals are clear-­cut cases of creatures that ‘want’ things. But this sense is not the sense that is relevant for our discussion.

Animals  183 Let me summarize my claims in this section. Animals count as borderline desirers, since they have mental states that play at least some of the roles that reasons beliefs do. But they are not clear-­cut desirers, because those mental states do not stand in the rich array of relations that our reasons beliefs do. The claim that animals are clear-­cut desirers may look attractive because we conflate wants and likings, or else because we get misled by another use of ‘wants’ where it means ‘needs’. Before we move on, I will note two further routes to the conclusion that animals are not clear-­cut desirers. Both lines of reasoning raise issues beyond the scope of this book, but if you are sympathetic to either, that is further grounds for sympathy with desire-­as-­belief on this score. First, we might think that our behaviour differs from animal behaviour in many important ways, and we might explain those differences in behaviour by appealing to differences in the genesis of that behaviour. For example, we tend to ascribe human beings free will, but are at least more unsure about whether animals are free in that same sense. Equally, we tend to assess human decisions in a variety of ways—as rational, or moral, say—but we do not assess animal decisions in those ways (cf. Chapter 4). Finally, we hold people responsible for the things that they do, such as by blaming or praising them. But again, this is not something we do with animals. A natural way to explain these differences is to think that the mechanisms that result in our behaviour are very different from the mechanisms that result in animal behaviour (cf. Fischer and Ravizza 2000). Since our behaviour is explained by our desires, in line with ODM (Only Desires Motivate) (Chapter  2), we might infer that animal behaviour is not explained by their desires in the same way. Such reasoning would show, independently of desire-­as-­belief, that animals are at best borderline desirers, rather than clear-­cut desirers. Second, right at the start of the book (§1.1) I assumed that all desires have propositional contents: that every desire is a desire [that p]. I made this assumption because it seems plausible, but also in order to simplify the presentation of my claims. If we committed to this claim more strongly, it might provide another route to denying that animals have clear-­cut desires. We might think that animals only have attitudes directed at concrete objects, not abstract propositions, and so conclude that their ‘desires’ are quite unlike ours. More specifically, we might happily analyse our infinitival desires— desires to eat, to sleep, and so on—as propositional because they are really desires [that I eat], [that I sleep], and so on. But we might resist the same move for animals on the grounds that they lack the concept . If we

184  Representing Reasons accepted this argument, we would have an independent reason for thinking that animals do not have desires in the same sense that we do, since only ours are propositional attitudes.13 * So far in this section I have focused on animals, where that meant non-­human animals. But what about humans? Another objection to desire-­as-­belief is that many of our desires are long-­established parts of our mind, shared with even our pre-­human ancestors. But this continuity in desire might seem to contrast with a discontinuity in normative belief: we might think that normative concepts are uniquely human, or at least fairly recent additions, evolutionarily speaking. In reply, I will tell a somewhat speculative history.14 Our earlier ancestors were simple creatures. They responded to simple stimuli in their environment such as nearby food, offspring, or potential mates, and they performed some extremely simple mental operations, allowing them to do simplistic reasoning about how to get closer to food, or how to attract a mate’s attention. But over successive generations various biological and cultural factors enabled our ancestors to get better at such reasoning. Our more recent ancestors were those who were abnormally capable at representing the world and responding appropriately. The reasoning they could undertake was superior to their rivals, allowing them to perform logical inferences, probabilistic inferences, and other feats of reasoning that rendered their maps of the world, and their resulting behaviour, more effective and appropriate. We are the result of their success. This ability to make inferences is surely a large part of the success story that is the human race. Plausibly, the development of this ability is the explanation of how we became creatures with beliefs: states which are not merely indicators of how the world is, akin to the opening of the petals on a flower during daylight, but instead rational indicators which manifest a unique degree of re­li­ abil­ity and sensitivity to change in the world. These very same developments explain how we became creatures with desires: states which don’t merely kick 13  The argument in this passage is inspired by Paul Thagard (2006), who, admittedly, runs the argument in the reverse direction and concludes that our desires can’t possibly be propositional attitudes. 14  A different question is about human children. My previous claims about animals might well apply to young children: either they don’t have desires (babies, perhaps), or else they do have the concept of a reason. One thing to note in relation to the latter is that children might well possess the concept of a reason with comparatively greater ease than animals, since they might come to have beliefs about reasons in virtue of belonging to a certain linguistic community (see the possible concession to social externalism in §9.2). For example, perhaps children can come to acquire the concept of a reason merely in virtue of learning the word ‘reason’. In fact, given desire-­as-­belief, the word ‘want’ implicitly involves the concept of a reason, and so perhaps children might even learn the concept of a reason just by learning the word ‘want’.

Summary  185 us around, akin to the processes that cause plants to lean towards the sunlight, but instead rational inclinations that manifest an unusual degree of reliability and sensitivity to change in the world. The psychological architecture that produces our desires used to have a different function: it used to produce states which merely caused our ancestors to move around in certain ways, akin to the states that cause plants and many animals to move around. But as we began to develop rational capacities and used them to regulate these drives, we changed the nature of those states of mind: their inputs changed sufficiently that these states at last qualified as beliefs. This process is perfectly parallel to other processes which changed other psychological systems into ones that produce beliefs: all of our beliefs are produced by systems which at one point produced states that were not beliefs, and this process presumably involves increasing deployment of rational capacities onto the states in question. I intend for this broad sketch of our history to be sufficiently loose that it fits whatever more detailed story eventually emerges. But it nonetheless allows us to see how desire-­ as-­ belief is consistent with a naturalistic explanation of our desires and can permit the continuity between us and our ancestors. It stands in contrast to views on which our desires are long-­ standing states, had in common even with our most distant ancestors, and it is superior to such views because it rightly recognizes that we are much better than our distant ancestors at reasoning, and that we surely deploy those reasoning capacities on our desires. Those two facts together support the thought that our desires are very unlike the drives that influenced our distant ancestors: the functional roles of those states are very different, because they are regulated in very different ways. In this way, I agree with the Stoics who apparently regarded ‘reason as . . . a new disposition which totally alters all the pre-­rational endowment’ (Long and Sedley  1987, 322), so that we are not creatures with a simple brain plus some rational capacities, but instead creatures whose rational capacities infuse much of our mental lives with new significance.

9.5 Summary In this chapter I have explained the implications of desire-­ as-­ belief for animals. Desire-­as-­belief tells us that animals have clear-­cut desires only if they have clear-­cut reasons beliefs. This might seem objectionable. Assessing this objection requires us to address wider questions about the nature of the

186  Representing Reasons mind. I sketched a theory—CRS—that helps us to understand what it takes to have reasons beliefs. I then argued that animals—at least many animals—have borderline reasons beliefs and so have only borderline desires. The optimistic claim that they have borderline reasons beliefs is supported by the fact that they perform many of the inferences that are characteristic of reasons beliefs. The pessimistic claim that they have only borderline desires is supported by two debunking explanations of the temptation to say anything stronger.

10 Desires and Reasons Back in §3.5 I distinguished five different ‘Humean’ claims, and explained how desire-­as-­belief interacts with them. The fifth was postponed for this chapter: Subjectivism:  What you have reason to do always depends on your desires. We can contrast subjectivism with the obvious rival: Objectivism:  What you have reason to do is usually independent of your desires.1 As I said in the Introduction, and again in §1.2, desire-­as-­belief fits best with objectivism, since it says that our desires respond to pre-­existing reasons, as opposed to the subjectivist view on which our desires create reasons. Indeed, we can see my defence of desire-­as-­belief as a partial defence of objectivism, since it helps objectivists to acknowledge the importance of our desires without making concessions to subjectivism. This chapter says just a little more to situate desire-­as-­belief in this debate. I rehearse two common objections to subjectivism (§10.1), and then in §10.2, I give some reasons to think that objectivism—especially objectivism combined with desire-­as-­belief—can explain away the appeal of subjectivism. The most central of these is that desire-­as-­belief, combined with the Best Enkratic Requirement from §4.3, makes our desires highly important for rationality, 1  Subjectivism, or a nearby view, is endorsed by many, including Hobbes (L I.VI.7); Goldman (2009); Manne (2014); Markovits (2014); Schroeder (2007a); Sobel (2016); and Williams (1981, 1995). Subjectivism sometimes goes by the name ‘reasons internalism’ (Williams 1981, 1995), but I avoid that label since it is more easily confused with ‘motivation internalism’ from Chapter 3. Objectivism, or a nearby view, is endorsed by many, including McDowell (1998); Parfit (2011a); Raz (2000); Scanlon (1998,  2014); Shafer-­Landau (2003); and plausibly Plato (Euthyphro 9e–11b) and Aristotle (Metaphysics 1072a29). Objectivism says ‘usually independent’ because objectivists can allow that in some cases our desires are relevant to our reasons in some unusual way. For example, objectivists might allow that your desire to kill your mother generates a reason for you to see a psychiatrist. I won’t worry about such cases in this chapter, or more generally say anything further to pin down the dispute with more precision: the basic contrast is clear enough.

Desire as Belief: A Study of Desire, Motivation, and Rationality. Alex Gregory, Oxford University Press. © Alex Gregory 2021. DOI: 10.1093/oso/9780198848172.003.0011

188  Desires and Reasons and so subjectivism may seem attractive only because we confuse it with a parallel claim about rationality. Finally, in §10.3, I briefly discuss desire-­based theories of wellbeing.

10.1  Two Problems for Subjectivism Desire-­as-­belief fits poorly with subjectivism. The literature on subjectivism is too vast for me to comprehensively vindicate that commitment: again, this book is best thought of as just one part of a larger objectivist project. So instead of saying anything comprehensive, in this section I’ll merely remind you of two classic worries for subjectivism. If either is right, desire-­as-­belief is right to commit us against subjectivism. First, subjectivism conflicts with many of our ordinary views, especially our moral views.2 For example, we might ordinarily think that no-­one has any reasons to torture cats, but according to subjectivism you might have such reasons if you have the right (wrong) desires, and indeed, perhaps you outright ought to spend more time torturing cats if your desires are sufficiently malevolent. That is a surprising conclusion, to say the least. Objectivists will more plausibly say that no-­one has reason to torture cats—they will plausibly say that our desires are beside the point in this matter, and that malevolence in desire can’t justify cruelty in action. Many further examples highlight the implausible consequences of sub­ject­iv­ ism. For example, according to subjectivism, whether you have sufficient ­reason to commit war crimes depends on your desires, so that some people genuinely ought to commit them. Or, to give a nonmoral example, imagine that you are in some kind of self-­destructive mood where you want to harm yourself. According to subjectivism, you thereby have reason to do so. Examples like these are easy to come by: subjectivism says that no choice is right or wrong, wise or stupid, except insofar as it fulfils or frustrates your desires. In many cases this looks deeply implausible. Subjectivists have suggested very many different replies to this objection, and it would be impractical to try and address them all. Instead, I’ll just briefly address the most common reply. This is to qualify subjectivism. On the simplest interpretation, it says that you have reason to v when, only when, you want to v, so that what you have reason to do is always a perfect mirror of 2  The literature on this issue is vast. For a sample, see Foot (1972); Joyce (2007); Mackie (1990); Olson (2014).

Two Problems for Subjectivism  189 your desires. But subjectivists often adopt a more sophisticated version of their view, so that, say, you have a reason to v when, and only when, you would want to v after perfect deliberation (e.g. Williams 1981, 1995). Questions abound about what perfect deliberation consists in, but the basic idea is clear enough: some desires are based on confusion or ignorance, and subjectivists should disallow those desires from generating reasons. Vice versa, you might lack some desires because of confusion or ignorance, and subjectivists can say that you sometimes have reasons to do things because you would want to do them, if you thought things through more carefully. With this more sophisticated view in hand, subjectivists can insist that in problematic cases like those above, subjectivism gets more plausible verdicts since few people would have the relevant desires if they were to deliberate in less confused ways and with more information (for helpful discussion, see Street 2009). Of course, to count as subjectivist at all, they are likely to allow that in some cases the relevant conclusions might hold.3 But if they can show that such cases would be very unusual, that at least serves to limit the damage to their view. It seems to me that objectivists ought to agree that this manoeuvre does make it far harder to find concrete counterexamples to subjectivism, since for any apparent counterexample it’s very hard to conclusively rule out the possibility that further deliberation wouldn’t extinguish or generate some desires. At the same time, I think subjectivists ought to agree that this ­manoeuvre is a damage-­limitation strategy rather than a total vindication of their view. The reply permits that in principle, whether you have sufficient reason to commit war crimes depends on your desires, and that sounds like a surprising conclusion even if subjectivists can reassure us that few people would have the relevant desires if they deliberated more carefully. So although the reply has some force, the objection nonetheless continues to put pressure on subjectivism. A second classic objection to subjectivism—in both the simple and ­sophisticated forms—concerns the focus of deliberation. Imagine that you face a difficult moral decision, and you are deliberating about your options. Let’s more specifically imagine that it’s Christmas, and a relative of yours has said something that is ambiguously racist: it may have been racist, but the phrasing was sufficiently ambiguous that this isn’t clear. You might well be slightly torn between saying something and keeping quiet: on the one hand, it’s  ­enormously important to eradicate racism. On the other hand, it’s not 3  Though there is also the somewhat Kantian project of showing that such deliberation is so tightly constrained that it always delivers morally sound desires—for related discussion, see Markovits (2014); Smith (1994); Street (2009). I can’t address such views in the detail they deserve, but I join others in being sceptical that they can succeed.

190  Desires and Reasons clear that it was racist at all, and it’s a family member who you are perhaps obliged to give the benefit of the doubt. As you sit there torn about what to say, you might well deliberate and focus on numerous details of the case: exactly what was said, and how, by whom, in what context, and so on. But one thing it would be strange to do would be to start thinking about your own desires. Lots of details are relevant to what you ought to do in this case, but your psychology seems hardly relevant at all. It is very hard to square this fact with subjectivism, according to which your psychology is highly relevant to what you ought to do. In short, subjectivism has the consequence that your deliberations ought to often focus on questions about which desires you have, but this seems inappropriate in many cases where your eyes ought to be pointed outward, not inward. Mark Schroeder defends subjectivism from this objection by claiming that your desires are not themselves reasons, but instead background conditions on reasons: things that make other things into reasons (Schroeder  2007a, chap. 2; cf. Pettit and Smith 1990). The idea is then that the subjectivist can agree that deliberation won’t refer to your own desires, because deliberation focuses on what reasons you have. But the last claim is highly dubious: practical deliberation can reasonably appeal to anything relevant to what you have reason to do, whether those things are themselves reasons or instead merely somehow relevant to what you have reason to do (Gregory 2009; cf. Dancy  2004). So again, even Schroeder’s view faces the problem that it recommends inward-­looking deliberation. In summary, subjectivism faces two classic objections: it has surprising consequences about which reasons people have, and it makes surprising predictions about what good deliberation looks like. So we should probably reject it, and instead adopt some objectivist view—good news for desire-­as-­ belief. In fact, desire-­as-­belief, combined with objectivism, makes good sense of the problems facing subjectivism. Some people might have unusual desires to torture cats, to commit war crimes, or to self-­harm. But these people don’t thereby have reasons to do these things. Instead, we should say, with objectivists, that the reasons bearing on such choices are independent of our desires. And we should say, with desire-­as-­belief, that this is precisely why people who want to do these things are going wrong: their desires—reasons beliefs—fail to reflect the facts. Similarly, we should say, with objectivism, that good deliberation doesn’t make reference to our desires, since they are not relevant to what we have reason to do. And we should say, with desire-­as-­ belief, that our desires aren’t the objects of practical deliberation any more than our beliefs are the objects of theoretical deliberation (cf. Pettit and

Debunking Subjectivism  191 Smith  1990). They are relevant to such deliberation, but as the things you think with, not as things you think about.

10.2  Debunking Subjectivism Desire-­as-­belief might do other work for objectivists in characterizing how subjectivism goes wrong. And desire-­as-­belief might also play other roles in blocking various parts of subjectivism or arguments for it. But here I want to focus on how desire-­as-­belief can allow objectivists to defuse some of the attraction of subjectivism (see also Parfit 2011a, 65–70). Why is subjectivism attractive? One simple answer is that our desires seem to be normatively relevant somehow, and it might seem that subjectivism is best placed to accommodate this simple thought. We can undercut this line of thought by showing how desires are important even on an objectivist view. I’ll briefly explain two simple ways in which objectivists can accommodate the apparent importance of desire, and then discuss a third possibility at greater length. First, out of respect for autonomy, we might think political systems ought to be responsive to their citizens’ preferences, and more generally that people ought to respect others’ preferences. For example, if many citizens want a new motorway, that might make it true that a new motorway ought to be built. Or if I hear that you want to vote for a certain political party, that might make it true that I ought to permit you to do so, and even make it true that I ought not ridicule your choice. To that extent, our desires may be normatively relevant insofar as they affect what others ought to do. But these claims are perfectly consistent with objectivism, since they pertain to how third parties ought to respond to our desires, not how we ourselves ought to do so. The fact that I ought to accommodate your desires out of respect for your autonomy is not a step towards the subjectivist view that your desires make your choices right. So this is one way in which objectivists can permit that our desires have some normative relevance: third parties sometimes ought to take heed of our desires in order to respect our autonomy. Second, what we will find pleasurable—what we would like­—is clearly often relevant to what we ought to do (see §7.3). For example, perhaps you ought to see a certain film, or eat a certain dessert, because you would find that pleasant. To that extent, some reasons do depend on your state of mind. But again, this is perfectly consistent with objectivism, which can allow that our likings give us reasons. It is only if we confuse likings and desires that we may be wrongly led from this thought to subjectivism. So this is another way

192  Desires and Reasons in which objectivists can permit that our desires might seem to have normative relevance: our likings have normative relevance, and can be confused with our desires (Parfit 2011a, 67). So far I have explained two simple ways in which objectivists can capture some of the appeal of subjectivism. But even once we acknowledge these points, subjectivism still retains some appeal: though some of its appeal might come from cases where either autonomy or pleasure are at stake, it also has appeal in other cases. For example, imagine that a doctor prescribes a certain drug to her patient. And let’s imagine that the drug will lessen the patient’s pain but hasten her death. We might reasonably suggest that this was a good choice if the doctor wanted to end the patient’s pain, but a poor choice if she wanted to prolong their life. So this might seem like another kind of case which favours the subjectivist idea: whether the doctor acted sensibly depends on what she wanted to achieve. The remainder of this section discusses this case in order to draw out a third way in which objectivism—at least, objectivism combined with desire-­as-­belief—can defuse the appeal of subjectivism. It’s instructive to contrast the above case—where you assess the doctor’s choice as a mere spectator—with a case where you are instead offering advice to the doctor. Imagine that she asks you whether she ought to prescribe the drug. What might you say? True, it seems possible that you might reply by asking her what she hopes to achieve: to relieve the patient’s pain, or prolong her life. But this kind of response makes most sense if we assume that the doctor has some kind of authority on which of those things are at stake, and you don’t (perhaps you are a pharmacist, responsible only for telling doctors which drugs do what, and it is their responsibility to determine what is in a patient’s best interest). Under those conditions, you aren’t really offering her practical advice about what to do at all, but instead only theoretical advice about what the drugs do. So let’s try to exclude that possibility, and imagine that the doctor lacks this kind of authority. Let’s imagine that the doctor is a trainee, under your watch, and the patient is especially well known to you. Under these circumstances is seems like it would be extremely odd to tailor your advice to your trainee’s desires, which seem plain irrelevant to the best choice of drug for the patient. When you give her advice, you’ll surely think that whether your trainee ought to prescribe the drug is completely independent of her desires. Let me summarize the two central verdicts I have endorsed: first, there is some sense in which we might evaluate the doctor’s decision according to her desires, especially if we are imagining ourselves as spectators commenting on how sensible her choices are. But second, there is another sense in which we

Debunking Subjectivism  193 would surely not evaluate the doctor’s decision with reference to her desires, as seen by the kind of advice we might offer her in trying to help her reach a sound decision. Subjectivists have no trouble in accounting for the former verdict. But the latter verdict presents a problem. The connection with the objections in §10.2 will be obvious. Indeed, subjectivists can repeat their defensive manoeuvre from §10.2 and claim that what you ought to do depends on the desires you would have if you deliberated perfectly: perhaps we, as advisors, should offer the doctor advice that pertains to what she might want if she deliberated in full view of the facts, say. But on reflection this is implausible. After all, the problem is that our advice should not take her desires into account at all. But any subjectivist theory worth the name will say that her present desires are at least relevant to her choice insofar as they are relevant to the counterfactual question of what she would desire if she were to deliberate perfectly. To this extent, even sophisticated subjectivism cannot explain the attraction of saying that the doctor’s desires are completely irrelevant to our advice. Another way to defend subjectivism might be to insist that when we offer the doctor advice, we are responding to what we have reason to do, and that’s why our advice can be insensitive to the doctor’s desires: after all, the act of advising is an act of ours, not hers. But again, this thought isn’t sufficient to defend subjectivism. The problem is that to qualify as advice (rather than, say, sounding off), your advice has to be a sincere effort to help someone recognize what they ought to do. So the subjectivist can’t merely say that advice is rightly insensitive to the desires of the advisee because it is an act of the advisor, not the advisee. Of course, advice can be evaluated as to whether it was right of the advisor to give it (perhaps it was rude, or generous, or whatever), but as advice it is evaluated as to whether it helps the advisee determine what they ought to do. The problem for subjectivism is then that all good advice should be grounded in facts about the advisee’s desires, and that is implausible in cases like the one under discussion. So far I have argued that subjectivism fits poorly with the fact that good advice makes no reference to the advisee’s desires. But I did begin by granting that it fits well with the kind of evaluation that spectators make: as spectators, we might try to evaluate the doctor’s choices by referring to her desires. It might be thought that objectivism will have the reverse pro and con: that it will be able to explain why good advice ignores the advisee’s desires—because they are not relevant to what the advisee ought to do—but will not be able to explain why spectators may evaluate choices with reference to the subject’s desires. But in fact I think objectivism can explain the latter verdict as well as the former.

194  Desires and Reasons In §4.1, I distinguished between what you rationally ought to do, and what you non-­rationally ought to do. What you rationally ought to do is importantly dependent on your beliefs, so that it would be irrational to bet on Leicester winning if, so far as you know, they are likely to lose. On the other hand, what you non-­rationally ought to do is independent of your beliefs, so that in that sense you ought to bet on Leicester if it will bring you good fortune, and this remains true even if you don’t realize that it will.4 In fact, what it is rational to do depends not just on your non-­normative beliefs (say, about Leicester’s players), but also on your reasons beliefs. That is just the claim expressed by the Best Enkratic Requirement from §4.3 according to which rationality favours actions that you believe you have reason to do. This claim is highly intuitive: if you fail to do something that you think you have strong reason to do, you are surely to that extent irrational. With these thoughts in mind, let’s return to the debate between objectivism and subjectivism. Objectivism is a theory about what we non-­rationally ought to do: it says that in the non-­rational sense our reasons are independent of our desires. But objectivists can also allow that we sometimes assess the rationality of what people did, and when we do that, we assess what they did against their mental states—such as against their normative beliefs and their desires. So objectivists can allow that there is a genuine grain of truth in subjectivism: it is a plausible claim about rationality that what it’s rational to do always depends on your desires. This idea is made all the more compelling by desire-­as-­belief, which allows objectivists to accept the rational relevance of desire in a straightforward way (§4.3). Objectivist defenders of desire-­as-­ belief are in an especially good place to acknowledge that our desires are relevant to assessments of rationality, since for them this claim follows from the Best Enkratic Requirement. These claims allow us to have the best of all worlds. We get the best of objectivism because we have a theory on which what we ought to do is independent of our desires, so that moral duties, say, are universally binding and insensitive to what we want, and so that appropriate deliberation does not make reference to our desires. But we get the attractive features of subjectivism because we allow that what it is rational to do depends on our desires, in just the same way that it depends on our other beliefs: rationality consists in a kind of internal consistency, and you have more of that insofar as your choices

4  Though see also Chapter 4, footnote 3: it might be that the non-­rational ought depends on what evidence there is rather than merely on the facts. That would still be enough to deliver the important contrast between the non-­rational and rational ought.

Debunking Subjectivism  195 reflect your beliefs. So subjectivism is true at least if understood as a theory of rationality, and our desires have clear normative relevance at least to questions of rationality. With all this in hand, let us return to our doctor case, and the contrast between evaluating her choices as a spectator, and offering her advice. I said that objectivism was attractive insofar as advice to a trainee doctor should not make reference to her desires, and that suggests that what she ought to do is insensitive to those desires. But I said that subjectivism might seem attractive insofar as we, as spectators, might evaluate what the doctor did with reference to her desires. We can now see what is going on: the former claim is attractive insofar as we focus on what the doctor non-­rationally ought to do: what she non-­rationally ought to do is independent of her desires. The non-­rational ought is most salient in contexts of advice, where we try to help other agents do what they really ought to do. (To most starkly see this, note that we give other agents more information with advice, and this serves not to make them more internally consistent, but instead to ensure that their choices conform with the facts.) But the latter claim about spectators is attractive insofar as we focus on what the doctor rationally ought to do: what she rationally ought to do is highly dependent on her desires. The rational ought is often salient in third-­personal contexts of evaluation, where we try to see whether others acted sensibly given their perspective at the time. Let me summarize the basic point. In some ways, objectivism seems attractive—especially when we think about what good advice looks like. And in some ways, subjectivism seems attractive—especially when we evaluate agents as mere spectators. But objectivists can accommodate the latter thought by agreeing that our desires are relevant to what it is rational to do. Rationality is a matter of internal consistency, and you do manage that to a greater degree if your actions are responsive to your desires (=reasons beliefs). To that extent, subjectivists are onto something. But only a claim about rationality, and it is not in tension with the objectivist claim that what we non-­rationally ought to do is largely independent of our desires.5

5  In §10.1, I mentioned a more sophisticated subjectivist view on which what you ought to do depends on what you would want to do if you deliberated perfectly. Do my remarks here about rationality also serve to undercut the attractions of this more sophisticated view? I think so. On one view, what you rationally ought to do depends not on the mental states you in fact have, but instead on the mental states it would be rational for you to have. (For example, we might think that your actions rationally ought to be sensitive not to your actual beliefs, but instead to the beliefs you would have if you were correctly responding to your evidence.) I disagree.* Still, if it were right, this view about rationality would support the view that what you rationally ought to do always depends on the desires you would have if you deliberated perfectly, since those are—presumably—the desires it would be

196  Desires and Reasons Let me summarize where we are. Desire-­as-­belief fits poorly with sub­ject­iv­ ism, and much better with objectivism. To that extent, reasons to reject sub­ ject­iv­ism are good news for desire-­as-­belief. In §10.2 I explained two classic objections to subjectivism. Then, in this section, I presented three objectivist-­ friendly claims that help to defuse subjectivism’s appeal—that is, I have offered three possible grains of truth in subjectivism. First, I suggested that respect for autonomy might sometimes favour letting others do what they want to do. Second, I suggested that what we ought to do is often sensitive to our likings, which might be confused with desires. And third, I suggested that what we rationally ought to do is sensitive to our desires. These claims are all true, and might make subjectivism seem attractive. But all are consistent with objectivism. To that extent, we should be wary of subjectivism: its apparent attractions might be defused by these attractive objectivist-­friendly claims.

10.3  Against Desire-­Based Theories of Wellbeing There is a final thought I should address. Some objectivists might endorse a desire-­based theory of wellbeing, according to which a life goes well for a person if (roughly) they get most of what they want (e.g. Railton  1986; Singer 2011; for overviews, see Fletcher 2016, chap. 2; Heathwood 2015). One can endorse this view without subscribing to subjectivism, but this view alone is potentially problematic for desire-­as-­belief. We plausibly have reasons to promote our own and others’ wellbeing, and so this theory of wellbeing is enough to generate the conclusion that some reasons are sensitive to our desires. Again, this claim sits poorly with desire-­as-­belief. But desire-­based theories of wellbeing are implausible. As with subjectivism, addressing this topic in full would take us too far afield, but I can at least rehearse two powerful classic objections to such views, and then also explain how desire-­as-­belief sheds light on those objections. The first classic problem for such views is that satisfying altruistic desires need not contribute to your wellbeing (e.g. Adams 2002, 87–9; Overvold 1980). Imagine that your partner leaves you, but after a period of sadness you finally come to accept the situation and out of your persisting love for her, you strongly desire that she finds happiness in her new relationship. It’s far from rational for you to have. To that extent this sophisticated subjectivist view may seem attractive insofar as we confuse it with this more sophisticated view about the requirements of rationality. *  It’s irrational to believe that tea will cure your cancer, but drinking tea to cure your cancer is more foolish if you lack that belief than if you have it.

Against Desire-­Based Theories of Wellbeing  197 obvious that it’s in your interests for her new relationship to flourish: your desire is grounded in your concern for her wellbeing, and if you get what you want, that need not bear on your own wellbeing, and certainly need not bear on your own wellbeing in proportion to the intensity of this desire.6 A second problem for desire-­based theories of wellbeing arises in cases where your desires are misguided. For example, imagine that you desire to call at the garage on the way home in order to buy some petrol, but unbeknownst to you, the garage is out of petrol. Under these circumstances you get something you want if you call at the garage, but it’s far from obvious that your life goes better (even a little bit) if you get to call at the garage. We can construct other cases along similar lines: even if your main goal in life is to be thin, or to rule the world with an iron fist, or to turn into a flower, it’s not obvious that these things are therefore best for you. In these cases, it seems that getting what you want might fail to contribute to your wellbeing because your desires are foolish. Like subjectivism, desire-­based theories of wellbeing can be modified to say that desires only make a difference to wellbeing if they would survive perfect deliberation (§10.2), but as with the modified version of subjectivism, it’s not clear that this fully undercuts the objection since it seems that bad desires might survive such deliberation. It would take us too far afield to consider in detail the various replies that defenders of desire-­ based theories of wellbeing have offered to these objections—I can only note that the proposals I have seen seem unconvincing for a variety of reasons. But we can enhance our confidence in these objections by noticing how naturally they flow from desire-­as-­belief. By the lights of desire-­as-­belief, desire-­based theories of wellbeing say that a life goes well to the extent that you do what you believe you have reason to do. What people believe they have reason to do is bound to correlate to some extent with what’s in their interests, and this might explain the superficial appeal of desire-­based theories of wellbeing. But by the lights of desire-­ as-­ belief, desire-­ based theories of wellbeing will be implausible for reasons that parallel the two objections above: first, in cases where you believe that you have a moral reason to do something, doing it may not benefit you. Second, even if we restrict ourselves to beliefs about prudential reasons (those given by self-­ interest), such beliefs fail to track what is actually in your interests, because 6  This objection connects with another: desire-­based theories of wellbeing combine poorly with the claim that rational actions maximize expected desire satisfaction (on the latter, see Chapter 4). If you adopt both of those views, it seems that you are committed to the view that rational actions are always in your own interests. If it’s ever rational to be moral, this view is plausibly false, and we should backtrack on any sympathy for desire-­based theories of wellbeing.

198  Desires and Reasons such beliefs can be mistaken, as when you falsely think that you have reason to call at the garage. In this way, desire-­as-­belief predicts the standard objections to desire-­based theories of wellbeing: that provides further evidence for desire-­as-­belief insofar as it vindicates intuitive objections to this theory, and vindicates those objections insofar as they are supported by this attractive theory of desire. Let me summarize. Desire-­as-­belief may fit poorly with objectivist views that incorporate a desire-­based theory of wellbeing. So in this section I rehearsed two classic objections to such views, and also showed how they flow naturally from desire-­as-­belief.

10.4  Summary In this chapter I have discussed subjectivism and objectivism. Subjectivism says that all reasons depend on our desires, and as such it fits poorly with desire-­as-­belief. In §10.1 I presented two classic objections to subjectivism. Then, in §10.2, I presented three debunking explanations of subjectivism’s appeal: to the extent that subjectivism looks attractive, we may be confusing it with plausible claims about respecting autonomy, or the importance of likings, or the relevance of desires for rationality. Finally, in §10.3, I briefly noted that desire-­as-­belief also combines poorly with desire-­based theories of wellbeing, presented two classic objections to such views, and explained how those objections mesh neatly with desire-­as-­belief.

11 Conclusion In this book I have argued for, and defended, desire-­as-­belief, and tried to show how it illuminates various questions about motivation, rationality, and what we ought to do. Desire-­as-­belief says that what we call ‘desires’ are just beliefs about reasons, by another name. The view has a variety of payoffs, though I presented two as central. First, it allows us to hold onto the thought that all motivation is explained by desire whilst also accepting that our normative beliefs can motivate us (Chapters 2 and 3). Second, it explains the connection between our desires and rationality: we rationally ought to live up to our desires just because we rationally ought to live up to our own normative beliefs (Chapter 4). Much of the rest of the book addressed various objections to desire-­as-­ belief. For example, I argued that it is consistent with sensible understandings of the direction-­of-­fit metaphor (§1.4), permits significant irrationality in desire, motivation, and action (Chapter 5), permits that goodness can leave us cold (Chapter  6), can explain the relationships between emotion, appetite, pleasure, and desire (Chapter  7), can make good sense of degrees of desire and reasoning with our desires (Chapter 8), and has acceptable implications regarding the desires of animals (Chapter 9). With some objections desire-­as-­ belief not only escaped unscathed, but in fact shed positive light on the relevant phenomena. For example, I suggested that it makes good sense of akrasia (§5.5), illuminates what it is to be uncertain about what we want (§8.1), and meshes neatly with an attractive theory of mental content (§9.3). And I explained how desire-­as-­belief sheds light on the debate between subjectivists and objectivists (Chapter 10). We’ve also seen how the view is superior to various rivals, such as non-­ cognitivist theories that reduce normative belief to desire (§3.3), presentationalist theories that treat desires as akin to perceptual states (§4.4), the besire theory (§6.1), and the guise of the good which treats desires as constrained by representations of goodness (§6.3, §6.4).

Desire as Belief: A Study of Desire, Motivation, and Rationality. Alex Gregory, Oxford University Press. © Alex Gregory 2021. DOI: 10.1093/oso/9780198848172.003.0012

200 Conclusion Beyond these specific claims, it might be useful for me to briefly zoom out a little and restate the central implications of desire-­as-­belief in a less technical manner. Some pursue views on which we are subject to two conflicting internal forces: a cold, systematic, and intellectual force of Reason, and a hot, chaotic, and crude force of Passion. Those views are needlessly complex and fail to fit the facts: their view of Reason is unduly optimistic, and their view of Passion is unduly pessimistic. A better view thinks of us not as divided creatures with a good half and a bad half, but instead as unified creatures with a single but highly fallible power to be good. That view gives a simpler picture of the pipeline between thought and action (Chapters 2 and 3), and permits a better theory of what it takes to live well (Chapters 4 and 10). Such a view is often rejected out of hand for ruling out various kinds of stupidity—otherwise seen as products of the war between Reason and Passion—but this is premature when the view can permit that this power to be good is fallible (Chapter  5). The view might otherwise be rejected for neglecting variation between individuals, on the grounds that it subsumes everything to Reason, and Reason is universal. But this also is premature when Reason has person-­ relative elements (Chapters 6 and 7). * How confidently do I believe the claims in this book? Inevitably, my confidence is roughly proportionate to the specificity of the relevant claims. I’m most confident of the broad claim that our desires are sensitive to what we see as reasons, somewhat less confident of treating this kind of sensitivity as a matter of belief, and somewhat less confident again of the details of desire-­as-­ belief as I defend it. But I think progress is most easily achieved when you have clear views to build on or to oppose, and that is why I’ve tried, by and large, to commit to specific claims rather than hedge. There are numerous ways that you might modify the view here in interesting ways, such as by treating desires as somewhat belief-­like (say, as aliefs (Gendler 2008) or as in-­ between-­beliefs (Schwitzgebel  2001)), or by saying that only sufficiently rational desires qualify as beliefs whereas others do not but instead only aspire to that status. I leave such avenues open to others to explore. What am I least confident about? Though I very much like the positive arguments for the view, a positive argument for a view should be equivalent to an argument against every rival to that view, and I worry that some views might escape the net cast by my arguments. One alternative way of pitching desire-­as-­belief would have been to start by defending it from objections to show that it is viable, and only then noting the attractive implications it has

Conclusion  201 for motivation and rationality, leaving open whether other views might share similar attractions. That way of proceeding would have been more boring to read, but perhaps more honest. When it comes to objections, I worry least about akrasia, despite that being a historically dominant objection to the view. Instead, I’m most worried by questions about degrees of desire (§8.1), since that topic raises many complications, and so despite my best efforts to stitch the various threads together into an attractive view, it’s very hard to be sure that no loose ends remain. There isn’t much consensus about degrees of belief, or about the strengths of reasons, and the issues for desire-­as-­belief combine these issues with many more. Somewhat similar remarks apply to questions about animals (Chapter  9), where the topic too quickly threatens to become too unwieldy to make any progress at all. And again, similar remarks apply to what I said about reason-­implies-­can and the distinction between wanting, wishing, and hoping (§6.5): the issues there seem difficult to resolve without a more detailed view about the reasons-­implies-­can constraint. I like to think that I’ve made progress on these topics even if I’ve not had the final word. * It might be fruitful to finish by briefly mentioning a few things I’ve not done in this book, and which I see as work to be done. First, I stayed fairly non-­committal on the nature of reasons. I said a little to orient us towards the thing I have in mind by that label (§1.2), but nothing approaching a full theory of what reasons are. One particularly notable omission is that I used the word ‘reason’ to refer only to normative reasons, but there are also motivating and explanatory reasons. Some believe it important to unify these different uses of ‘reason’. Desire-­as-­belief might permit such a unification (see e.g. Singh  2019; and cf. Gregory  2016 which would interact with desire-­as-­belief in a more complex way). Second, more generally, I have stayed fairly non-­committal on many central debates in metaethics (see e.g. §1.2, though see also §9.3). I see this book as presenting a partial defence of objectivism, since it shows how our desires are enormously interesting and important even though reasons are independent of those desires. But I have said relatively little about the other issues that objectivism raises, such as about the metaphysics of reasons, how we might come to know about them, or what we in fact have reason to do. Third, I have said little about developments in neuroscience and how they might bear on our subject matter. This is a relatively principled omission,

202 Conclusion since I take states like belief and desire to be functional states whose essential features are independent of our biology. But we should nonetheless be interested in the question of how these functional states are realized in human beings. Finally, desire-­as-­belief says, obviously, that desires are beliefs. Some think that belief is just the pale imitation of real gold: knowledge. Is there some special status accruing to desire when it is constituted by a belief that amounts to knowledge? A tempting answer is that this is virtue (Little  1997; McDowell 1998; Plato, Protagoras). But I have said nothing about that.

APPENDIX A

Lewis on Desire-­as-­Belief Note: The material in this appendix briefly relies on some claims I defend in §8.1, so I’d suggest reading that first. In two well-­known papers, David Lewis argues specifically against desire-­as-­belief (Lewis 1988, 1996; see also Costa, Collins, and Levi 1995). Those papers are complex, and have spawned a sizeable technical literature. But the basic idea is clear enough: Lewis presents comparatively orthodox theories about how our beliefs and desires rationally ought to change when we receive new information. But those orthodoxies lay down two different sets of rules: they imply that the norms governing belief revision are different from the norms governing desire revision. It follows that desires cannot be beliefs. In this appendix, I explain and reject this argument. One complication: I’ve said that desires are beliefs about reasons rather than beliefs about goodness (see especially Chapters 1 and 6). But while discussing Lewis, I’ll set this aside and assume, with Lewis, that desire-­ as-­ belief reduces desires to beliefs about goodness. This does not affect the underlying points being made, and enables me to discuss Lewis without having to constantly translate his claims into my own preferred terms. (This topic is complex enough as it is.)

A.1  The Argument Lewis’ argument relies on three claims. (In what follows, you could get away with ignoring the formalizations.) Lewis’ first claim is a characterization of desire-­as-­belief. As Lewis understands it, desire-­as-­belief says that degrees of desire are degrees of credence (certainty) in propositions about the good (Lewis 1988, 327; 1996, 307). Following Lewis, let’s call this DAB: DAB : D[p] = C[Gp] Lewis’ second claim is more complex. To understand it, it’s helpful to first note Lewis’ assumption that your instrumental desire [that p] should correspond to the degree to which you desire the various possible outcomes of [p] (which we’ll label o1, o2 . . . on), weighted by your credence that they will occur: D[p] = ∑ i ( C[oi | p] * D[oi ]) With this in mind, we can understand Lewis’ second claim: when you learn new information, your instrumental desire [that p] should be revised only insofar as the information bears on your credences about the likelihood of different outcomes, given [p] (Lewis 1988, 327; 1996, 308–9). This principle is sometimes called invariance, but we’ll simply call it Lewis’ desire revision principle:



Lewis’ desire revision principle : De[p] = ∑

i

( C[oi | p & e]* D[oi ])

204  Appendix A Changed by new information

Act 1

State A.

State B.

Expected Value

Credence:

Credence:

C1

C2

(i.e. degree of rational instrumental desire)

Wanted to degree: Wanted to degree: D1

Act 2

D2

Wanted to degree: Wanted to degree: D3

D4

(C1*D1)+(C2*D2)

Changed only by changes to C1 and C2

(C1*D3)+(C2*D4)

Figure A.1  An illustration of Lewis’ desire revision principle. Here is a way of thinking about this. Decision theorists often describe decisions using a decision matrix, with rows representing possible acts, and columns representing possible states. We then insert a number for each column to represent our credences that these states will occur, and insert numbers in the boxes to represent our desires for these various outcomes, and we use all these numbers to calculate the expected value of each act. Lewis’ claim is that new information impacts on expected value only by affecting the credences we attach to states, never the desires we attach to outcomes. See Figure A.1. We’ll return to Lewis’ desire revision principle shortly. Lewis’ third crucial claim is that DAB combines with his desire revision principle to generate implausible conclusions about rational change in belief (Lewis 1988, 327–8; 1996, 308–9; the argument is related to that in his 1976, 1986b). The basic idea is that DAB tells us that degrees of desire are degrees of credence, but the desire revision principle is implausible as a claim about how credences should change in light of new information. I’ll simply trust that Lewis is right about this last claim, and object only to his first two claims. That is, I agree that Lewis is right that DAB and his desire revision principle make unhappy partners, but deny that DAB is a fair characterization of desire-­as-­belief, and deny that his desire revision principle should be accepted by defenders of that view.

A.2 Reply An initial worry with Lewis’ argument is that it threatens to prove too much (Oddie 1994; see also Bradley and Stefánsson 2016). Even if desire-­as-­belief is false, it is tempting to say that rationality requires your normative beliefs and your desires to march in step, so that your desires are somehow proportionate to your normative beliefs (objectors to desire-­as-­ belief commonly say that the view is wrong just insofar as we can be irrational and fail to do this—see Chapter 5). But Lewis’ argument seems to threaten even this view: it says that no-­one can simultaneously (a) update their credences about goodness rationally in light of new information, (b) update their degrees of desire rationally in light of new information, and (c) keep their desires proportionate to their beliefs about goodness. But we might have thought that a rational person would do all three of these things at once. Lewis’ argument thereby threatens not only desire-­as-­belief, but also any view on which we are rationally required to proportion our desires to our normative beliefs. So even those who reject desire-­as-­belief might well be sceptical about Lewis’ argument.

Reply  205 Beyond this point about having companions in guilt, I have two central objections to Lewis’ argument. First, DAB is an inferior version of desire-­as-­belief. Second, no defender of desire-­as-­belief should accept Lewis’ desire revision principle. I take these points in turn. First, Lewis’ characterization of desire-­as-­belief is problematic. As I argued in §8.1, desire-­as-­belief should not identify the intensity of desire with the certainty of the relevant belief. Rather, it should identify the intensity of desire with the importance of that belief (here, the degree of goodness). And it should identify uncertainty about what we want with certainty in the relevant belief. To this extent, Lewis’ argument undermines desire-­as-­ belief only if it is developed in the wrong way. Lewis’ argument undermines the idea that credences should respond to new information in just the same way that desiderative intensity should respond to new information. But insofar as the defender of desire-­as-­ belief fails to make that claim, Lewis’ argument misses its target. So Lewis’ argument relies on characterizing desire-­as-­belief as DAB above, but defenders of desire-­as-­belief should reject that characterization of their view for the reasons given in §8.1. In a note, I address two possible Lewisian replies.1 Second, no defender of desire-­as-­belief should accept Lewis’ desire revision principle. That principle says that when you learn new information, your instrumental desire that [p] should be revised only insofar as the information bears on your credences about the consequences of [p]. That is, Lewis’ desire revision principle says that new information never changes instrumental desires by changing how much you desire the relevant outcomes, but instead only by changing how likely you think [p] is to lead to those outcomes. By implication, Lewis clearly thinks that there is no rational change in our ultimate desires whatsoever. But no defender of desire-­as-­belief should accept these claims (Bradley and List 2008; Bradley and Stefánsson 2016, 699–706; Hedden 2015, 156–62; Stefánsson 2014). Defenders of desire-­as-­belief should say that all our desires ought to be sensitive to evidence about the worth of their objects: the idea is precisely that desires are themselves subject to normative standards in the same way that beliefs are. For example, you might rationally revise your ultimate desires because you have new evidence about which things are ultimately good— perhaps you come to ultimately desire equality less after discovering arguments against it (see also §8.2). And you might rationally revise your instrumental desires not only because you revise your views about the likelihood of various outcomes of [p], but also because you revise your views about the desirability of those outcomes themselves. In this way, Lewis’ desire revision principle simply presupposes that our desires are insensitive to reasoning about the non-­instrumental worth of their objects, but since no defender of desire-­as-­belief should accept that claim, Lewis’ argument presupposes too much. Lewis does provide a response to this second objection (Lewis 1988, 332; see also 1996, 303, 310–12). The idea is that we should think of the objects of ultimate desires as so 1  First: Lewis does acknowledge that his argument does not undermine every version of desire-­as-­ belief. But the main alternative he considers is desire-­as-­conditional-­belief, which identifies the intensity of the desire that [p] with the degree of credence in the claim that [p is good, given p] ([Gp|p]) (Lewis 1996, 310–11; see also Price 1989). Clearly, this still identifies intensity in desire with credence in some belief, and this is exactly what we ought not do. Second: Towards the end of his first paper Lewis addresses a more sophisticated version of DAB according to which degrees of desire (intensity) are affected by believed degrees of value (importance) (Lewis 1988, 330–1). But the view takes the form of the final view I rejected in §8.1.2, according to which degrees of desire are degrees of expected value. That is, this modified view permits that importance is one factor that matters for intensity of desire, but on my preferred version of desire-­as-­belief importance is the only factor that matters for intensity of desire, and degrees of certainty manifest themselves in other ways (§8.1.3). So this view is still not the version of desire-­as-­belief that I endorse.

206  Appendix A fine-­grained that all facts about their objects are built into their description. On this view, no-­one simply desires [a sandwich], but must instead desire [a sandwich which is thus-­ and-­so, in a world which is like this . . . and where that sandwich is good to degree n]. This being so, you cannot possibly change your mind about the properties of what you desire: you cannot come to learn, say, that the object of your desire is better than you thought, since the very description of that object already builds in a description of how good it is. As a result, Lewis claims that the only acceptable version of desire-­as-­belief collapses into ‘desire-­by-­necessity’, according to which we desire some things necessarily: perhaps we desire good things necessarily, and other things not at all. Lewis claims that such a view is perfectly acceptable, but is not a version of desire-­as-­belief (Lewis 1996, 307). These claims are unhelpful for two reasons. First, Lewis’ stipulation that the objects of desire are this fine-­grained is false with respect to actual people. Actual people don’t have desires whose objects are things like ‘that there be world peace and world peace is good’ (Byrne and Hájek 1997, 424). Rather, actual people, with our limited intellects, will have desires with much coarser contents (e.g. just for world peace, period). For us, Lewis’ argument will not hold. The norms governing our desires do leave room for our learning further facts about the value of the objects of our desires, as when you realize that the chastity you wanted is not so good. In short, for all actual people, new information about goodness should impact on our ultimate desires, and so Lewis’ claims are unhelpful if we are trying to construct a theory for how we ought to update our desires in light of new information. Second, most importantly, we might simply accept that desire-­ as-­ belief can be formalized as a version of desire-­by-­necessity (see also Hedden 2015, 160). Lewis happily admits that he has no argument against such a view, and I can’t see why it matters if desire-­ as-­belief gets interpreted by one (slightly extreme) formalism as a version of desire-­by-­ necessity. Lewis promised to argue against a certain substantive theory, and it’s disappointing if he really shows only that the theory must be formalized in a certain way if we first make some relatively strong stipulative assumptions. In summary, Lewis’ objection to desire-­as-­belief fails. In fact, its failure is overdetermined: his argument relies on three assumptions, and two of those assumptions are dubious. First, DAB is an inferior version of desire-­as-­belief: desire-­as-­belief should not identify degrees of desire with degrees of credence about goodness (§8.1.2). Second, no defender of desire-­as-­belief should accept Lewis’ desire revision principle, which permits too little rational change in desire. It might be helpful to step back a bit and forget the details of Lewis’ objection. His basic point is that desire-­as-­belief is inconsistent with some common assumptions in decision theory. Is that true? It might be, in two ways. First, some decision theorists might assume that you cannot rationally change your ultimate desires (cf. §8.2), whereas desire-­as-­belief tells us that we can rationally change our ultimate desires. We might want to extend decision theory to make sense of such changes. Second, in §8.1 I argued that we should think that desires vary not only in intensity but also in how sure we are that we have the relevant desires. As I implied in §8.1.4, rationality presumably requires us to take this into account when we decide how to act. Here, too, we might want to extend decision theory to take this into account (cf. Cyert and Degroot 1979). Doing this will be extremely difficult: it is not at all clear how to take normative uncertainty into account in decision-­making (see, for example, Gustafsson and Torpman 2014; Sepielli 2009). But however we should do that, we should do the same with uncertainty about what we want.

Bibliography Historical Works Aquinas, ‘DV’ – [1256–­1259]. Quaestiones disputatae de Veritate (‘Disputed Questions on Truth’). Translated by Robert  W.  Schmidt. Available at: https://isidore.co/aquinas/ QDdeVer24.htm Aquinas. ‘ST’ – Summa Theologiae (‘Summary of Theology’). Aristotle, ‘DA’ – 1986 [c.350BC]. De Anima (‘On The Soul’). Translated by Hugh Lawson-­ Tancred. London: Penguin. Aristotle, ‘DM’ – De motu animalium (‘On the Motion of Animals’). Aristotle, ‘EE’ – Eudemian Ethics. Aristotle, ‘EN’ – Nicomachean Ethics. Aristotle, Metaphysics. Augustine, 1961 [397–­400]. Confessions. Translated by R.S. Pine-­Coffin. London: Penguin. Hobbes, ‘L’ – [1651]. Leviathan. Hume, ‘T’ – [1639–­40]. A Treatise of Human Nature. Kant, ‘G’ – [1785]. Groundwork of the Metaphysic of Morals. Leibniz, Gottfried. 1982 [1765]. Leibniz: New Essays on Human Understanding—Abridged Edition. Edited by Peter Remnant and Jonathan Bennett. Cambridge: Cambridge University Press. Leibniz, Gottfried. 1989a [?1707]. ‘Comments on Spinoza’s Philosophy’. In Philosophical Essays, translated by Roger Ariew and Daniel Garber, 272–81. Indianapolis, IN: Hackett. Leibniz, Gottfried. 1989b [1686]. ‘Discourse on Metaphysics’. In Philosophical Essays, translated by Roger Ariew and Daniel Garber, 35–68. Indianapolis, IN: Hackett. Locke, ‘Essay’ – [1689]. An Essay Concerning Human Understanding. Plato, The Republic. Plato, Euthyphro. Plato, Protagoras. Spinoza, Ethics. 1994 [1677]. Translated by Edwin Curley. Princeton, NJ: Princeton University Press. * Adams, Robert Merrihew. 2002. Finite and Infinite Goods: A Framework for Ethics. Oxford: Oxford University Press. Altham, J. 1986. ‘The Legacy of Emotivism’. In Fact, Science and Morality, edited by Crispin Wright and Graham MacDonald, 275–88. Oxford: Blackwell. Alvarez, Maria. 2010. Kinds of Reasons: An Essay in the Philosophy of Action. Oxford: Oxford University Press. Andrews, Kristin, and Jacob Beck. 2017. The Routledge Handbook of Philosophy of Animal Minds. London: Taylor & Francis. Annas, Julia. 1981. An Introduction to Plato’s Republic. Oxford: Clarendon Press. Anscombe, Gertrude Elizabeth Margaret. 1963. Intention. 2nd ed. Cambridge, MA: Harvard University Press.

208 Bibliography Archer, Avery. 2016. ‘Do Desires Provide Reasons? An Argument Against the Cognitivist Strategy’. Philosophical Studies 173 (8): 2011–27. Archer, Avery. 2020. ‘Are Desires Beliefs about Normative Reasons?’ Analytic Philosophy 61 (3): 236–51. Armstrong, David Malet. 1993. A Materialist Theory of the Mind. London: Routledge. Arpaly, Nomy, and Timothy Schroeder. 2014. In Praise of Desire. Oxford: Oxford University Press. Ashwell, Lauren. 2013. ‘Deep, Dark. . . or Transparent? Knowing Our Desires’. Philosophical Studies 165 (1): 245–56. Austin, John Langshaw. 1965. Sense and Sensibilia. Edited by Geoffrey James Warnock. Oxford: Clarendon Press. Baker, Derek. 2017. ‘The Verdictive Organization of Desire’. Canadian Journal of Philosophy 47 (5): 589–612. Beck, Jacob. 2012. ‘Do Animals Engage in Conceptual Thought?’ Philosophy Compass 7 (3): 218–29. Bermudez, Jose Luis. 2006. ‘Animal Reasoning and Proto-Logic’. In Rational Animals?, edited by Susan Hurley and Matthew Nudds, 127–38. Oxford: Oxford University Press. Björklund, Fredrik, Gunnar Björnsson, John Eriksson, Ragnar Francén Olinder, and Caj Strandberg. 2012. ‘Recent Work on Motivational Internalism’. Analysis 72 (1): 124–37. Blackburn, Simon. 1984. Spreading the Word: Groundings in the Philosophy of Language. Oxford, New York: Oxford University Press. Blackburn, Simon. 1993. ‘Supervenience Revisited’. In Essays in Quasi-Realism, 130–48. Oxford, New York: Oxford University Press. Blackburn, Simon. 1998. Ruling Passions: A Theory of Practical Reasoning. Oxford: Oxford University Press. Block, Ned. 1981. ‘Psychologism and Behaviorism’. The Philosophical Review 90 (1): 5–43. Block, Ned. 1986. ‘Advertisement for a Semantics for Psychology’. Midwest Studies in Philosophy 10: 615–78. Block, Ned. 1987. ‘Functional Role and Truth Conditions’. Proceedings of the Aristotelian Society, Supplementary Volumes 61(1): 157–81. Bond, Edward Jarvis. 1983. Reason and Value. Cambridge: Cambridge University Press. Boyd, Richard. 1988. ‘How to Be a Moral Realist’. In Essays on Moral Realism, edited by Geoffrey Sayre-McCord, 181–228. Ithaca, NY: Cornell University Press. Bradley, Richard, and Christian List. 2008. ‘Desire-as-Belief Revisited’. Analysis 69 (1): 31–7. Bradley, Richard, and H. Orri Stefánsson. 2016. ‘Desire, Expectation, and Invariance’. Mind 125 (499): 691–725. Brady, Michael S. 2009. ‘The Irrationality of Recalcitrant Emotions’. Philosophical Studies 145 (3): 413–30. Brady, Michael  S. 2013. Emotional Insight: The Epistemic Role of Emotional Experience. Oxford, New York: Oxford University Press. Bratman, Michael. 1987. Intention, Plans, and Practical Reason. Stanford, CA: CSLI Publications. Brewer, Talbot. 2006. ‘Three Dogmas of Desire’. In Values and Virtues: Aristotelianism in Contemporary Ethics, edited by Timothy Chappell, 253–85. Oxford: Clarendon Press. Brink, David Owen. 1989. Moral Realism and the Foundations of Ethics. Cambridge: Cambridge University Press. Broome, John. 1991. ‘Desire, Belief and Expectation’. Mind 100 (2): 265–7.

Bibliography  209 Broome, John. 1997. ‘Reasons and Motivation: John Broome’. Aristotelian Society Supplementary Volume 71 (1): 131–46. Broome, John. 2004. ‘Reasons’. In Reason and Value: Themes from the Moral Philosophy of Joseph Raz, edited by R. Jay Wallace, 2004–28. Oxford: Oxford University Press. Broome, John. 2006. ‘Reasoning with Preferences?’ Royal Institute of Philosophy Supplements 59: 183–208. Broome, John. 2013. Rationality through Reasoning. New York: Wiley Blackwell. Brown, Campbell. 2014. ‘Minding the Is-Ought Gap’. Journal of Philosophical Logic 43 (1): 53–69. Brunero, John. 2010. ‘The Scope of Rational Requirements’. The Philosophical Quarterly 60 (238): 28–49. Buchak, Lara. 2013. Risk and Rationality. Oxford: Oxford University Press. Burge, Tyler. 1979. ‘Individualism and the Mental’. Midwest Studies in Philosophy 4 (1): 73–121. Bykvist, Krister. 2009. ‘No Good Fit: Why the Fitting Attitude Analysis of Value Fails’. Mind 118 (469): 1–30. Bykvist, Krister, and Jonas Olson. 2009a. ‘Expressivism and Moral Certitude’. The Philosophical Quarterly 59 (235): 202–15. Bykvist, Krister, and Jonas Olson. 2009b. ‘Expressivism and Moral Certitude’. The Philosophical Quarterly 59 (235): 202–15. Bykvist, Krister, and Jonas Olson. 2011. ‘Against the Being for Account of Normative Certitude’. Journal of Ethics and Social Philosophy 6: 1. Bykvist, Krister, and Jonas Olson. 2017. ‘Non-Cognitivism and Fundamental Moral Certitude: Reply to Eriksson and Francén Olinder’. Australasian Journal of Philosophy 95 (4): 794–9. Byrne, Alex. 2011. ‘Knowing What I Want’. In Consciousness and the Self: New Essays, edited by JeeLoo Liu and John Perry, 165–83. Cambridge: Cambridge University Press. Byrne, Alex, and Alan Hájek. 1997. ‘David Hume, David Lewis, and Decision Theory’. Mind 106 (423): 411–728. Call, Josep. 2006. ‘Inferences by Exclusion in the Great Apes: The Effect of Age and Species’. Animal Cognition 9 (4): 393–403. Camp, Elisabeth, and Eli Shupe. 2017. ‘Instrumental Reasoning in Nonhuman Animals’. In The Routledge Handbook of Philosophy of Animal Minds, edited by Kristin Andrews and Jacob Beck, 100–8. London: Taylor & Francis. Campbell, Douglas  I. 2018. ‘Doxastic Desire and Attitudinal Monism’. Synthese 195 (3): 1139–61. Chan, David K. 2004. ‘Are There Extrinsic Desires?’ Noûs 38 (2): 326–50. Chrisman, Matthew. 2012. ‘On the Meaning of Ought’. In Oxford Studies in Metaethics, edited by Russ Shafer-Landau, 7: 304–32. Oxford: Oxford University Press. Chrisman, Matthew. 2015. The Meaning of ‘Ought’: Beyond Descriptivism and Expressivism in Metaethics. Oxford: Oxford University Press. Coleman, Mary Clayton. 2008. ‘Directions of Fit and the Humean Theory of Motivation’. Australasian Journal of Philosophy 86 (1): 127–39. Cooper, John Madison. 1999. ‘Posidonius on Emotions’. In Reason and Emotion: Essays on Ancient Moral Psychology and Ethical Theory, 449–84. Princeton, NJ: Princeton University Press. Copp, David, and David Sobel. 2002. ‘Desires, Motives, and Reasons: Scanlon’s Rationalistic Moral Psychology’. Social Theory and Practice 28 (2): 243–76.

210 Bibliography Costa, Horacio Arló, John Collins, and Isaac Levi. 1995. ‘Desire-as-Belief Implies Opinionation or Indifference’. Analysis 55 (1): 2–5. Cyert, Richard  M., and Morris Degroot. 1979. ‘Adaptive Utility’. In Expected Utility Hypotheses and the Allais Paradox: Contemporary Discussions of Decisions under Uncertainty with Allais’ Rejoinder, edited by Maurice Allais and Ole Hagen, 223–42. Dordrecht: D. Reidel Publishing. Damasio, Antonio R. 2006. Descartes’ Error. New York: Random House. Dancy, Jonathan. 1993. Moral Reasons. Oxford: Blackwell. Dancy, Jonathan. 2000. ‘Should We Pass the Buck?’ Royal Institute of Philosophy Supplements 47: 159–73. Dancy, Jonathan. 2002. Practical Reality. Oxford: Oxford University Press. Dancy, Jonathan. 2004. Ethics without Principles. Oxford: Oxford University Press. D’Arms, Justin, and Daniel Jacobson. 1994. ‘Expressivism, Morality, and the Emotions’. Ethics 104 (4): 739–63. Darwall, Stephen L. 1983. Impartial Reason. Ithaca, NY: Cornell University Press. Davidson, Donald. 2001a. ‘Actions, Reasons, and Causes’. In Essays on Actions and Events, 3–19. Oxford: Oxford University Press. Davidson, Donald. 2001b. ‘How Is Weakness of the Will Possible?’ In Essays on Actions and Events, 21–42. Oxford: Oxford University Press. Davis, Wayne. 1981. ‘A Theory of Happiness’. American Philosophical Quarterly 18 (2): 111–20. De Sousa, Ronald B. 1974. ‘The Good and the True’. Mind 83 (332): 534–51. Döring, Sabine  A. 2003. ‘Explaining Action by Emotion’. The Philosophical Quarterly 53 (211): 214–30. Dreier, James. 1990. ‘Internalism and Speaker Relativism’. Ethics 101 (1): 6–26. Dreier, James. 1996. ‘Accepting Agent Centred Norms: A Problem for Non-Cognitivists and a Suggestion for Solving It’. Australasian Journal of Philosophy 74 (3): 409–22. Dreier, James. 2004. ‘Meta-Ethics and the Problem of Creeping Minimalism’. Philosophical Perspectives 18 (1): 23–44. Eklund, Matti. 2009. ‘The Frege-Geach Problem and Kalderon’s Moral Fictionalism’. The Philosophical Quarterly 59 (237): 705–12. Eklund, Matti. 2017. Choosing Normative Concepts. Oxford: Oxford University Press. Ekman, Paul, and Wallace  V.  Friesen. 1971. ‘Constants across Cultures in the Face and Emotion’. Journal of Personality and Social Psychology 17 (2): 124. Enoch, David. 2006. ‘Agency, Shmagency: Why Normativity Won’t Come from What Is Constitutive of Action’. The Philosophical Review 115 (2): 169–98. Enoch, David. 2009. ‘How Is Moral Disagreement a Problem for Realism?’ The Journal of Ethics 13 (1): 15–50. Enoch, David. 2011. Taking Morality Seriously: A Defense of Robust Realism. Oxford: Oxford University Press on Demand. Erdohohegyi, Ágnes, József Topál, Zsófia Virányi, and Ádám Miklósi. 2007. ‘Dog-Logic: Inferential Reasoning in a Two-Way Choice Task and Its Restricted Use’. Animal Behaviour 74 (4): 725–37. Ewing, Alfred Cyril. 1948. The Definition of Good. New York: Hyperion Press. Fernández, Jordi. 2007. ‘Desire and Self-Knowledge’. Australasian Journal of Philosophy 85 (4): 517–36. Finlay, Stephen. 2014. Confusion of Tongues: A Theory of Normative Language. Oxford, New York: Oxford University Press. Firestone, Chaz, and Brian  J.  Scholl. 2016. ‘Cognition Does Not Affect Perception: Evaluating the Evidence for “Top-Down” Effects’. Behavioral and Brain Sciences 39: 1–77.

Bibliography  211 Fischer, John Martin, and Mark Ravizza. 2000. Responsibility and Control: A Theory of Moral Responsibility. Cambridge: Cambridge University Press. Fletcher, Guy. 2016. The Philosophy of Well-Being: An Introduction. London: Routledge. Fogal, Daniel. 2018. ‘Deflationary Pluralism about Motivating Reasons’. In The Factive Turn in Epistemology, edited by Veli Mitova, 193–218. Cambridge: Cambridge University Press. Fogal, Daniel. 2019. ‘Rational Requirements and the Primacy of Pressure’. Mind 129 (516): 1033–1070. Foot, Philippa. 1972. ‘Morality as a System of Hypothetical Imperatives’. The Philosophical Review 81 (3): 305–16. Frank, Robert  H. 1988. Passions within Reason: The Strategic Role of the Emotions. New York: WW Norton & Co. Frankfurt, Harry  G. 1988a. ‘Freedom of the Will and the Concept of a Person’. In The Importance of What We Care About: Philosophical Essays, 11–25. Cambridge, New York: Cambridge University Press. Frankfurt, Harry  G. 1988b. ‘Identification and Wholeheartedness’. In The Importance of What We Care About: Philosophical Essays, 159–76. Cambridge, New York: Cambridge University Press. Friedman, Jane. 2013. ‘Suspended Judgment’. Philosophical Studies 162 (2): 165–81. Garrard, Eve, and David McNaughton. 1998. ‘Mapping Moral Motivation’. Ethical Theory and Moral Practice 1 (1): 45–59. Geach, Peter T. 1956. ‘Good and Evil’. Analysis 17 (2): 33–42. Gendler, Tamar Szabó. 2008. ‘Alief and Belief ’. The Journal of Philosophy 105 (10): 634–63. Gertken, Jan, and Benjamin Kiesewetter. 2017. ‘The Right and the Wrong Kind of Reasons’. Philosophy Compass 12 (5): 1–14. Gibbard, Allan. 1990. Wise Choices, Apt Feelings: A Theory of Normative Judgment. Cambridge, MA: Harvard University Press. Gibbard, Allan. 2003. Thinking How to Live. Cambridge, MA: Harvard University Press. Gibbard, Allan. 2012. Meaning and Normativity. Oxford, New York: Oxford University Press. Ginet, Carl. 1990. On Action. Cambridge: Cambridge University Press. Goldie, Peter. 2004. ‘Emotion, Feeling, and Knowledge of the World’. In Thinking about Feeling: Contemporary Philosophers on Emotions, edited by Robert C. Solomon, 91–106. Oxford: Oxford University Press. Goldman, Alan  H. 2009. Reasons from within: Desires and Values. Oxford: Oxford University Press. Green, Thomas Hill. 1884. Prolegomena to Ethics. Oxford: Clarendon Press. Greene, Joshua. 2013. Moral Tribes: Emotion, Reason, and the Gap Between Us and Them. London: Atlantic Books. Gregory, Alex. 2009. ‘Slaves of the Passions? On Schroeder’s New Humeanism’. Ratio 22 (2): 250–7. Gregory, Alex. 2012. ‘Changing Direction on Direction of Fit’. Ethical Theory and Moral Practice 15 (5): 603–14. Gregory, Alex. 2014. ‘A Very Good Reason to Reject the Buck-Passing Account’. Australasian Journal of Philosophy 92 (2): 287–303. Gregory, Alex. 2016. ‘Normative Reasons as Good Bases’. Philosophical Studies 173 (9): 2291–310. Gregory, Alex. 2017. ‘Might Desires Be Beliefs about Normative Reasons?’ In The Nature of Desire, edited by Julien Deonna and Lauria Federico, 201–17. New York: Oxford University Press. Gregory, Alex. 2018. ‘Why Do Desires Rationalize Actions?’ Ergo 5 (40): 1061–81.

212 Bibliography Grice, Paul. 1989. ‘Logic and Conversation’. In Studies in the Way of Words, 22–40. Cambridge, MA: Harvard University Press. Grzankowski, Alex. 2015. ‘Not All Attitudes Are Propositional’. European Journal of Philosophy 23 (3): 374–91. Gustafsson, Johan E., and Olle Torpman. 2014. ‘In Defence of My Favourite Theory’. Pacific Philosophical Quarterly 95 (2): 159–74. Haidt, Jonathan. 2001. ‘The Emotional Dog and Its Rational Tail’. Psychological Review 108 (4): 814–34. Hamlin, J.  Kiley. 2013. ‘Moral Judgment and Action in Preverbal Infants and Toddlers: Evidence for an Innate Moral Core’. Current Directions in Psychological Science 22 (3): 186–93. Harcourt, Edward. 2004. ‘Instrumental Desires, Instrumental Rationality’. Aristotelian Society Supplementary Volume 78 (1): 111–29. Hare, Richard Mervyn. 1952. The Language of Morals. Oxford: Oxford Clarendon Press. Hare, Richard Mervyn. 1963. Freedom and Reason. Oxford: Oxford Paperbacks. Hare, Richard Mervyn. 1981. Moral Thinking: Its Levels, Method, and Point. Oxford: Oxford University Press. Harman, Elizabeth. 2009. ‘“I’ll Be Glad I Did It” Reasoning and the Significance of Future Desires’. Philosophical Perspectives 23 (1): 177–99. Harman, Gilbert. 1982. ‘Conceptual Role Semantics’. Notre Dame Journal of Formal Logic 23 (2): 242–56. Harman, Gilbert. 2000. ‘Desired Desires’. In Explaining Value and Other Essays in Moral Philosophy, 117–36. Oxford: Clarendon Press. Hausman, Daniel M. 2011. Preference, Value, Choice, and Welfare. Cambridge: Cambridge University Press. Hawkins, Jennifer. 2007. ‘Desiring the Bad under the Guise of the Good’. The Philosophical Quarterly 58 (231): 244–64. Hazlett, Allan. 2018. ‘The Guise of the Good and the Problem of Partiality’. Canadian Journal of Philosophy 49(6): 1–22. Heathwood, Chris. 2007. ‘The Reduction of Sensory Pleasure to Desire’. Philosophical Studies 133 (1): 23–44. Heathwood, Chris. 2015. ‘Desire-Fulfillment Theory’. In The Routledge Handbook of Philosophy of Well-Being, edited by Guy Fletcher, 135–47. London: Routledge. Hedden, Brian. 2015. Reasons without Persons: Rationality, Identity, and Time. Oxford: Oxford University Press. Helm, Bennett W. 2015. ‘Emotions and Recalcitrance: Reevaluating the Perceptual Model’. Dialectica 69 (3): 417–33. Hills, Alison. 2009. ‘Moral Testimony and Moral Epistemology’. Ethics 120 (1): 94–127. Holton, Richard. 1999. ‘Intention and Weakness of Will’. Journal of Philosophy 96 (5): 241. Holton, Richard. 2009. Willing, Wanting, Waiting. Oxford: Oxford University Press. Hopkins, Robert. 2007. ‘What Is Wrong with Moral Testimony?’ Philosophy and Phenomenological Research 74 (3): 611–34. Horgan, Terence, and Mark Timmons. 1991. ‘New Wave Moral Realism Meets Moral Twin Earth’. Journal of Philosophical Research 16: 447–65. Horgan, Terence, and Mark Timmons. 1992. ‘Troubles for New Wave Moral Semantics: The “Open Question Argument” Revived’. Philosophical Papers 21 (3): 153–75. Horgan, Terence, and Mark Timmons. 2009. ‘Analytical Moral Functionalism Meets Moral Twin Earth’. In Minds, Ethics, and Conditionals: Themes from the Philosophy of Frank Jackson, edited by Ian Ravenscroft, 221–36. Oxford: Oxford University Press.

Bibliography  213 Horn, Laurence. 2001. A Natural History of Negation. Stanford, CA: CSLI Publications. Huemer, Michael. 2007a. ‘Compassionate Phenomenal Conservatism’. Philosophy and Phenomenological Research 74 (1): 30–55. Huemer, Michael. 2007b. Ethical Intuitionism. New York: Springer. Humberstone, I. Lloyd. 1992. ‘Direction of Fit’. Mind 101 (401): 59–83. Hurley, Susan, and Matthew Nudds, eds. 2006. Rational Animals? Oxford: Oxford University Press. Hursthouse, Rosalind. 1991. ‘Arational Actions’. Journal of Philosophy 88 (2): 57–68. Jackson, Frank. 1998. From Metaphysics to Ethics: A Defence of Conceptual Analysis. Oxford: Oxford University Press. Jackson, Frank, and Philip Pettit. 1995. ‘Moral Functionalism and Moral Motivation’. The Philosophical Quarterly (1950–) 45 (178): 20–40. Jacobson, Daniel. 2011. ‘Fitting Attitude Theories of Value’. In The Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta, Spring 2011. Metaphysics Research Lab, Stanford University. https://plato.stanford.edu/archives/spr2011/entries/fitting-attitude-theories. Jeffrey, Richard  C. 1974. ‘Preference Among Preferences’. The Journal of Philosophy 71 (13): 377–91. Johnson King, Zoë A. 2020. ‘Praiseworthy Motivations’. Noûs 54 (2): 408–30. Johnston, Mark. 2001. ‘The Authority of Affect’. Philosophy and Phenomenological Research 63 (1): 181–214. Joyce, Richard. 2007. The Myth of Morality. Cambridge: Cambridge University Press. Kahneman, Daniel. 2011. Thinking, Fast and Slow. London: Macmillan. Kauppinen, Antti. Forthcoming. ‘Rationality as the Rule of Reason’. Noûs Kavka, Gregory S. 1983. ‘The Toxin Puzzle’. Analysis 43 (1): 33–6. Kiesewetter, Benjamin. 2017. The Normativity of Rationality. Oxford, New York: Oxford University Press. Kolodny, Niko. 2018. ‘Instrumental Reasons’. In The Oxford Handbook of Reasons and Normativity, edited by Daniel Star, 731–763. Oxford: Oxford University Press. Korsgaard, Christine  M. 1986. ‘Skepticism about Practical Reason’. Journal of Philosophy 83 (1): 5–25. Korsgaard, Christine  M. 1997. ‘The Normativity of Instrumental Reason’. In Ethics and Practical Reason, edited by Garrett Cullity and Berys Gaut, 215–54. Oxford: Clarendon Press. Korsgaard, Christine  M. 2009. Self-Constitution: Agency, Identity, and Integrity. Oxford: Oxford University Press. Lewis, David. 1970. ‘How to Define Theoretical Terms’. The Journal of Philosophy 67 (13): 427–46. Lewis, David. 1974. ‘Radical Interpretation’. Synthese 27 (3): 331–44. Lewis, David. 1976. ‘Probabilities of Conditionals and Conditional Probabilities’. The Philosophical Review 85 (3): 297–315. Lewis, David. 1986a. On the Plurality of Worlds. Oxford: Blackwell. Lewis, David. 1986b. ‘Probabilities of Conditionals and Conditional Probabilities II’. The Philosophical Review 95 (4): 581–9. Lewis, David. 1988. ‘Desire as Belief ’. Mind 97 (387): 323–32. Lewis, David. 1989. ‘Dispositional Theories of Value’. Proceedings of the Aristotelian Society, Supplementary Volumes 63: 113–38. Lewis, David. 1996. ‘Desire as Belief II’. Mind 105 (418): 303–13. Little, Margaret Olivia. 1997. ‘Virtue as Knowledge: Objections from the Philosophy of Mind’. Noûs 31 (1): 59–79.

214 Bibliography Long, Anthony Arthur, and David Neil Sedley. 1987. The Hellenistic Philosophers: Volume 1. Cambridge: Cambridge University Press. Lord, Errol. 2015. ‘Acting for the Right Reasons, Abilities, and Obligation’. In Oxford Studies in Metaethics, edited by Russ Shafer-Landau, 10: 26–52. Oxford: Oxford University Press. Lord, Errol. 2018. The Importance of Being Rational. Oxford: Oxford University Press. Machery, Edouard, and Ron Mallon. 2010. ‘Evolution of Morality’. In The Moral Psychology Handbook, edited by John  M.  Doris and Moral Psychology Research Group, 3–46. Oxford: Oxford University Press. Mackie, John. 1990. Ethics: Inventing Right and Wrong. London: Penguin UK. Manne, Kate. 2014. ‘Internalism about Reasons: Sad but True?’ Philosophical Studies 167 (1): 89–117. Manne, Kate. 2017. ‘Locating Morality: Moral Imperatives as Bodily Imperatives’. In Oxford Studies in Metaethics, edited by Russ Shafer-Landau, 12: 1–26. Oxford: Oxford University Press. Markovits, Julia. 2014. Moral Reason. Oxford: Oxford University Press. Marks, Joel. 1982. ‘A Theory of Emotion’. Philosophical Studies: An International Journal for Philosophy in the Analytic Tradition 42 (2): 227–42. Martin, Adrienne. 2013. How We Hope: A Moral Psychology. Princeton, NJ: Princeton University Press. May, Joshua. 2018. Regard for Reason in the Moral Mind. Oxford: Oxford University Press. McDowell, John. 1998. ‘Are Moral Requirements Hypothetical Imperatives?’ In Mind, Value, and Reality, 77–94. Cambridge, MA: Harvard University Press. McElwee, Brian. 2017. ‘Supererogation across Normative Domains’. Australasian Journal of Philosophy 95 (3): 505–16. McGeer, Victoria. 2004. ‘The Art of Good Hope’. Annals of the American Academy of Political and Social Science 1: 100–27. McGrath, Sarah. 2011. ‘Skepticism about Moral Expertise as a Puzzle for Moral Realism’. The Journal of Philosophy 108 (3): 111–37. McHugh, Conor. 2012. ‘Epistemic Deontology and Voluntariness’. Erkenntnis 77 (1): 65–94. McHugh, Conor. 2014. ‘Exercising Doxastic Freedom’. Philosophy and Phenomenological Research 88 (1): 1–37. McHugh, Conor. 2017. ‘Attitudinal Control’. Synthese 194 (8): 2745–62. McHugh, Conor. 2018. ‘The Normativity of Rationality, by Benjamin Kiesewetter’. Mind 127 (508): 1245–53. McHugh, Conor, and Jonathan Way. 2017. ‘What Is Reasoning?’ Mind 127 (505): 167–96. McNaughton, David. 1988. Moral Vision: An Introduction to Ethics. Oxford: Blackwell. Meirav, Ariel. 2009. ‘The Nature of Hope’. Ratio 22 (2): 216–33. Mele, Alfred  R. 1995. ‘Motivation: Essentially Motivation-Constituting Attitudes’. The Philosophical Review 104 (3): 387–423. Millgram, Elijah. 1997. Practical Induction. Cambridge, MA: Harvard University Press. Montague, Michelle. 2007. ‘Against Propositionalism’. Noûs 41 (3): 503–18. Moran, Richard. 2001. Authority and Estrangement: An Essay on Self-Knowledge. Princeton, NJ: Princeton University Press. Moss, Jessica Dawn. 2012. Aristotle on the Apparent Good: Perception, Phantasia, Thought, and Desire. Oxford: Oxford University Press. Murdoch, Iris. 2013. The Sovereignty of Good. London: Routledge. Nagel, Thomas. 1970. The Possibility of Altruism. Princeton, NJ: Princeton University Press.

Bibliography  215 Nichols, Shaun. 2002. ‘How Psychopaths Threaten Moral Rationalism’. The Monist 85 (2): 285–303. Nichols, Shaun. 2004. Sentimental Rules: On the Natural Foundations of Moral Judgment. Oxford: Oxford University Press. Nozick, Robert. 1974. Anarchy, State, and Utopia. New York: Basic Books. Nussbaum, Martha. 2004. ‘Emotions as Judgments of Value and Importance’. In Thinking about Feeling: Contemporary Philosophers on Emotions, edited by Robert  C.  Solomon, 183–99. Oxford: Oxford University Press. Oddie, Graham. 1994. ‘Harmony, Purity, Truth’. Mind 103 (412): 451–72. Oddie, Graham. 2005. Value, Reality, and Desire. Oxford: Oxford University Press. Olson, Jonas. 2009. ‘Fitting Attitude Analyses of Value and the Partiality Challenge’. Ethical Theory and Moral Practice 12 (4): 365–78. Olson, Jonas. 2014. Moral Error Theory: History, Critique, Defence. Oxford: Oxford University Press. Overvold, Mark Carl. 1980. ‘Self-Interest and the Concept of Self-Sacrifice’. Canadian Journal of Philosophy 10 (1): 105–18. Pallies, Daniel. 2021. ‘An Honest Look at Hybrid Theories of Pleasure’. Philosophical Studies 178: 887–907. Parfit, Derek. 1984. Reasons and Persons. Oxford: Oxford University Press. Parfit, Derek. 1997. ‘Equality and Priority’. Ratio 10 (3): 202–21. Parfit, Derek. 2011a. On What Matters: Volume I. Oxford: Oxford University Press. Parfit, Derek. 2011b. On What Matters: Volume II. Oxford: Oxford University Press. Peacocke, Christopher. 1992. A Study of Concepts. Cambridge, MA: MIT Press. Pettit, Philip. 2004. ‘Hope and Its Place in Mind’. The Annals of the American Academy of Political and Social Science 592 (1): 152–65. Pettit, Philip, and Michael Smith. 1990. ‘Backgrounding Desire’. The Philosophical Review 99 (4): 565–92. Platts, Mark de Bretton. 1997. Ways of Meaning: An Introduction to a Philosophy of Language. Cambridge, MA, London: MIT Press. Pollock, John  L. 2006. Thinking about Acting: Logical Foundations for Rational Decision Making. Oxford: Oxford University Press. Price, Henry Habberley. 1969. Belief. London: George Allen and Unwin. Price, Huw. 1989. ‘Defending Desire-as-Belief ’. Mind 98 (389): 119–27. Prior, Arthur N. 1960. ‘The Autonomy of Ethics’. Australasian Journal of Philosophy 38 (3): 199–206. Quinn, Warren. 1998. ‘Putting Rationality in Its Place’. In Virtues and Reasons: Philippa Foot and Moral Theory: Essays in Honour of Philippa Foot, edited by Rosalind Hursthouse, Gavin Lawrence, and Warren Quinn, 181–208. Oxford: Clarendon Press. Rabinowicz, Wlodek, and Toni Rønnow-Rasmussen. 2004. ‘The Strike of the Demon: On Fitting Pro-Attitudes and Value’. Ethics 114 (3): 391–423. Rabinowicz, Wlodek, and Toni Rønnow-Rasmussen. 2006. ‘Buck-Passing and the Right Kind of Reasons’. The Philosophical Quarterly 56 (222): 114–20. Railton, Peter. 1986. ‘Facts and Values’. Philosophical Topics 14 (2): 5–31. Raleigh, Thomas. Forthcoming. ‘Suspending Is Believing’. Synthese. Raz, Joseph. 2000. ‘Agency, Reason, and the Good’. In Engaging Reason: On the Theory of Value and Action, 22–45. Oxford: Oxford University Press. Raz, Joseph. 2010. ‘On the Guise of the Good’. In Desire, Practical Reason, and the Good, edited by Sergio Tenenbaum, 111–37. New York: Oxford University Press. Ridge, Michael. 1998. ‘Humean Intentions’. American Philosophical Quarterly 35 (2): 157–78.

216 Bibliography Ridge, Michael. 2014. Impassioned Belief. Oxford: Oxford University Press. Ridge, Michael. 2018. ‘Expressivism and Collectives’. Mind 127 (507): 833–61. Roskies, Adina. 2003. ‘Are Ethical Judgments Intrinsically Motivational? Lessons from Acquired Sociopathy’. Philosophical Psychology 16 (1): 51–66. Rowland, Richard. 2019. The Normative and the Evaluative: The Buck-Passing Account of Value. Oxford: Oxford University Press. Samuelson, Paul Anthony. 1947. Foundations of Economic Analysis. Cambridge, MA: Harvard University Press. Scanlon, T. M. 1998. What We Owe to Each Other. Cambridge, MA, London: Belknap. Scanlon, T.  M. 2007. ‘Structural Irrationality’. In Common Minds: Themes from the Philosophy of Philip Pettit, edited by Geoffrey Brennan, Michael Smith, Robert Goodin, and Frank Jackson, 84–103. Oxford: Oxford University Press. Scanlon, T. M. 2014. Being Realistic about Reasons. Oxford: Oxford University Press. Schafer, Karl. 2013. ‘Perception and the Rational Force of Desire’. Journal of Philosophy 110 (5): 258–81. Schapiro, Tamar. 2009. ‘The Nature of Inclination’. Ethics 119 (2): 229–56. Schroeder, Mark. 2007a. Slaves of the Passions. Oxford, New York: Oxford University Press. Schroeder, Mark. 2007b. ‘Teleology, Agent Relative Value, and “Good”’. Ethics 117 (2): 265–295. Schroeder, Mark. 2008. Being For: Evaluating the Semantic Program of Expressivism. Oxford: Oxford University Press. Schroeder, Mark. 2010. Noncognitivism in Ethics. London: Routledge. Schroeder, Mark. 2011. ‘Ought, Agents, and Actions’. Philosophical Review 120 (1): 1–41. Schroeder, Mark, and Michael Milona. 2019. ‘Desiring under the Proper Guise’. In Oxford Studies in Metaethics, edited by Russ Shafer-Landau, vol. 14. Oxford: Oxford University Press. Schroeder, Timothy. 2004. Three Faces of Desire. Oxford: Oxford University Press. Schueler, George Frederick. 1995. Desire: Its Role in Practical Reason and the Explanation of Action. Cambridge, MA, London: MIT Press. Schwitzgebel, Eric. 2001. ‘In-Between Believing’. The Philosophical Quarterly 51 (202): 76–82. Searle, John  R. 1964. ‘How to Derive “Ought” from “Is”’. The Philosophical Review 73 (1): 43–58. Searle, John  R. 1983. Intentionality: An Essay in the Philosophy of Mind. Cambridge: Cambridge University Press. Searle, John R. 2001. Rationality in Action. Cambridge, MA: MIT Press. Sepielli, Andrew. 2009. ‘What to Do When You Don’t Know What to Do’. In Oxford Studies in Metaethics, edited by Russ Shafer-Landau, 4: 5–28. Oxford: Oxford University Press. Setiya, Kieran. 2010. Reasons without Rationalism. Princeton, NJ: Princeton University Press. Shafer-Landau, Russ. 2000. ‘A Defense of Motivational Externalism’. Philosophical Studies 97 (3): 267–91. Shafer-Landau, Russ. 2003. Moral Realism: A Defence. Oxford: Clarendon Press. Shah, Nishi. 2003. ‘How Truth Governs Belief ’. Philosophical Review 112 (4): 447–82. Shah, Nishi. 2006. ‘A New Argument for Evidentialism’. The Philosophical Quarterly 56 (225): 481–98. Sinclair, Neil. 2018. ‘Conceptual Role Semantics and the Reference of Moral Concepts’. European Journal of Philosophy 26 (1): 95–121. Singer, Peter. 2011. Practical Ethics. Cambridge: Cambridge University Press.

Bibliography  217 Singh, Keshav. 2019. ‘Acting and Believing Under the Guise of Normative Reasons’. Philosophy and Phenomenological Research 99 (2): 409–30. Sinhababu, Neil. 2009. ‘The Humean Theory of Motivation Reformulated and Defended’. Philosophical Review 118 (4): 465–500. Sinhababu, Neil. 2011. ‘The Humean Theory of Practical Irrationality’. Journal of Ethics and Social Philosophy 6: 1. Sinhababu, Neil. 2015. ‘Advantages of Propositionalism’. Pacific Philosophical Quarterly 96 (2): 165–80. Sinhababu, Neil. 2017. Humean Nature: How Desire Explains Action, Thought, and Feeling. Oxford: Oxford University Press. Skorupski, John. 2007a. ‘Buck-Passing about Goodness’. In Hommage à Wlodek: Philosophical Papers Dedicated to Wlodek Rabinowicz, edited by D. Egonsson, T. Rønnow-Rasmussen, B. Petersson, and J. Josefsson. https://www.fil.lu.se/hommageawlodek/site. Skorupski, John. 2007b. ‘Internal Reasons and the Scope of Blame’. In Bernard Williams, edited by Alan Thomas, 73–103. Cambridge: Cambridge University Press. Sliwa, Paulina. 2012. ‘In Defense of Moral Testimony’. Philosophical Studies 158 (2): 175–95. Smith, Michael. 1987. ‘The Humean Theory of Motivation’. Mind 96 (381): 36–61. Smith, Michael. 1994. The Moral Problem. Oxford: Oxford University Press. Smith, Michael. 2002. ‘Evaluation, Uncertainty and Motivation’. Ethical Theory and Moral Practice 5 (3): 305–20. Smith, Michael. 2004. ‘Instrumental Desires, Instrumental Rationality’. Proceedings of the Aristotelian Society, Supplementary Volumes 78: 93–109. Smith, Michael. 2011. ‘Scanlon on Desire and the Explanation of Action’. In Reasons and Recognition: Essays on the Philosophy of T.M. Scanlon, edited by R. Jay Wallace, Rahul Kumar, and Samuel Freeman, 79–97. Oxford: Oxford University Press. Smith, Michael. 2012. ‘Four Objections to the Standard Story of Action (and Four Replies)’. Philosophical Issues 22 (1): 387–401. Smith, Michael, and Jeanette Kennett. 2004. ‘Philosophy and Commonsense: The Case of Weakness of Will’. In Ethics and the A Priori: Selected Essays on Moral Psychology and Meta-Ethics, edited by Michael Smith, 56–72. Cambridge: Cambridge University Press. Smithies, Declan, and Jeremy Weiss. 2019. ‘Affective Experience, Desire, and Reasons for Action’. Analytic Philosophy 60 (1): 27–54. Sobel, David. 2005. ‘Pain for Objectivists: The Case of Matters of Mere Taste’. Ethical Theory and Moral Practice 8 (4): 437–57. Sobel, David. 2011. ‘Parfit’s Case against Subjectivism’. In Oxford Studies in Metaethics, edited by Russ Shafer-Landau, 6: 52–78. Oxford: Oxford University Press. Sobel, David. 2016. From Valuing to Value: Towards a Defense of Subjectivism. Oxford: Oxford University Press. Solomon, Robert C. 1976. The Passions: Emotions and the Meaning of Life. Indianapolis, IN: Hackett Publishing. Stalnaker, Robert. 1987. Inquiry. Cambridge, MA: MIT Press. Stampe, Dennis W. 1987. ‘The Authority of Desire’. Philosophical Review 96 (3): 335–81. Stefánsson, H. Orri. 2014. ‘Desires, Beliefs and Conditional Desirability’. Synthese 191 (16): 4019–35. Stevenson, Charles Leslie. 1937. ‘The Emotive Meaning of Ethical Terms’. Mind 46 (181): 14–31. Stocker, Michael. 1979. ‘Desiring the Bad: An Essay in Moral Psychology’. Journal of Philosophy 76 (12): 738–53.

218 Bibliography Strandberg, Caj. 2012. ‘Expressivism and Dispositional Desires’. American Philosophical Quarterly 49 (1): 81–91. Stratton-Lake, Philip. 2016. ‘Intuitionism in Ethics’. In The Stanford Encyclopedia of Philosophy, edited by Edward  N.  Zalta, Winter 2016. Metaphysics Research Lab, Stanford University. https://plato.stanford.edu/archives/win2016/entries/intuitionismethics. Strawson, Galen. 1994. Mental Reality. Cambridge, MA: MIT Press. Street, Sharon. 2009. ‘In Defense of Future Tuesday Indifference: Ideally Coherent Eccentrics and the Contingency of What Matters’. Philosophical Issues 19: 273–98. Streumer, Bart. 2007. ‘Reasons and Impossibility’. Philosophical Studies 136 (3): 351–84. Streumer, Bart. 2017. Unbelievable Errors: An Error Theory about All Normative Judgements. Oxford: Oxford University Press. Stroud, Sarah. 2014. ‘Weakness of Will’. In The Stanford Encyclopedia of Philosophy, edited by Edward  N.  Zalta, Spring 2014. Metaphysics Research Lab, Stanford University. https://plato.stanford.edu/archives/spr2014/entries/weakness-will. Suikkanen, Jussi. 2018. ‘Judgment Internalism: An Argument from Self-Knowledge’. Ethical Theory and Moral Practice 21 (3): 489–503. Sussman, David. 2009. ‘For Badness’ Sake’. The Journal of Philosophy 106 (11): 613–28. Svavarsdóttir, Sigrún. 1999. ‘Moral Cognitivism and Motivation’. The Philosophical Review 108 (2): 161–219. Swartzer, Steven. 2015. ‘Humean Externalism and the Argument from Depression’. Journal of Ethics and Social Philosophy 9 (2): 1–16. Tappolet, Christine. 2016. Emotions, Value, and Agency. Oxford: Oxford University Press. Tenenbaum, Sergio. 2007. Appearances of the Good: An Essay on the Nature of Practical Reason. Cambridge: Cambridge University Press. Tenenbaum, Sergio. 2008. ‘Appearing Good: A Reply to Schroeder’. Social Theory and Practice 34 (1): 131–8. Thagard, Paul. 2006. ‘Desires Are Not Propositional Attitudes’. Dialogue 45 (1): 151–6. Thomson, Judith Jarvis. 2015. Normativity. Chicago, IL: Open Court. Toppinen, Teemu. 2017. ‘Enkrasia for Non-Cognitivists’. Ethical Theory and Moral Practice 20: 943–55. Tresan, Jon. 2009. ‘The Challenge of Communal Internalism’. The Journal of Value Inquiry 43 (2): 179–99. Tversky, Amos, and Daniel Kahneman. 1973. ‘Availability: A Heuristic for Judging Frequency and Probability’. Cognitive Psychology 5 (2): 207–32. Ullmann-Margalit, Edna, and Sidney Morgenbesser. 1977. ‘Picking and Choosing’. Social Research 44 (4): 757–85. Unwin, Nicholas. 1999. ‘Quasi-Realism, Negation and the Frege-Geach Problem’. The Philosophical Quarterly 49 (196): 337–52. Unwin, Nicholas. 2001. ‘Norms and Negation: A Problem for Gibbard’s Logic’. The Philosophical Quarterly 51 (202): 60–75. Velleman, J. David. 1992. ‘The Guise of the Good’. Noûs 26 (1): 3–26. Waal, Frans de. 1996. Good Natured. Cambridge, MA: Harvard University Press. Wall, David. 2009. ‘Are There Passive Desires?’ Dialectica 63 (2): 133–55. Wallace, R. Jay. 2013. The View from Here: On Affirmation, Attachment, and the Limits of Regret. Oxford: Oxford University Press. Watson, Gary. 1977. ‘Skepticism about Weakness of Will’. The Philosophical Review 86 (3): 316–39. Watson, Gary. 1982. ‘Free Agency’. In Free Will, edited by Gary Watson, 96–110. New York: Oxford University Press.

Bibliography  219 Way, Jonathan. 2012. ‘Transmission and the Wrong Kind of Reason’. Ethics 122 (3): 489–515. Way, Jonathan. 2018. ‘Reasons and Rationality’. In The Oxford Handbook of Reasons and Normativity, edited by Daniel Star, 485–503. Oxford: Oxford University Press. Weatherson, Brian. 2019. Normative Externalism. Oxford: Oxford University Press. Wedgwood, Ralph. 2001. ‘Conceptual Role Semantics for Moral Terms’. The Philosophical Review 110 (1): 1–30. Wedgwood, Ralph. 2007. The Nature of Normativity. Oxford: Oxford University Press. Whiting, Daniel. 2014. ‘Keep Things in Perspective: Reasons, Rationality, and the A Priori’. Journal of Ethics and Social Philosophy 8 (1): 1–22. Williams, Bernard. 1976. ‘Ethical Consistency’. In Problems of the Self: Philosophical Papers 1956–1972, 166–86. Cambridge: Cambridge University Press. Williams, Bernard. 1981. ‘Internal and External Reasons’. In Moral Luck: Philosophical Papers 1973–1980, 101–13. Cambridge: Cambridge University Press. Williams, Bernard. 1995. ‘Internal Reasons and the Obscurity of Blame’. In Making Sense of Humanity: And Other Philosophical Papers 1982–1993, 35–45. Cambridge: Cambridge University Press. Williamson, Timothy. 2008. The Philosophy of Philosophy. New York: John Wiley & Sons. Wolf, Aaron. 2015. ‘Giving up Hume’s Guillotine’. Australasian Journal of Philosophy 93 (1): 109–25. Worsnip, Alex. 2018. ‘The Conflict of Evidence and Coherence’. Philosophy and Phenomenological Research 96 (1): 3–44. Yao, Vida. 2019. ‘The Undesirable and the Adesirable’. Philosophy and Phenomenological Research 99 (1): 115–30. Youpa, Andrew. 2007. ‘Spinoza’s Theory of Motivation’. Pacific Philosophical Quarterly 88 (3): 375–90. Zakkou, Julia. 2018. ‘The Cancellability Test for Conversational Implicatures’. Philosophy Compass 13 (12): 1–17.

Index Note: Figures are indicated by an italic “f ”, following the page number. For the benefit of digital users, indexed terms that span two pages (e.g., 52–53) may, on occasion, appear on only one of those pages. addiction  105n.8, 110–11 advice 192–5 akrasia  51–2, 82–4, 94–112 amoralists 50–1; see also Satan animals  170–1, 180–5 appetite 136–7 ascetics 142–4 besire 113–15 Bond, E.J.  157–8 Brady, Michael  132–4 Bratman, Michael  45–6 certainty in belief  26–7, 144n.12, 148–59, 164–6, 200–1, 203–6 children  63–4, 66, 122–3, 184n.14 conativism 55–72 conceptual role semantics  171–3 content  7, 10–17, 170 credence; see certainty in belief criticism  61–2, 76–7, 89–91, 125, 192–5 CRS; see conceptual role semantics Dancy, Jonathan  37n.11, 42, 118–19 decision theory  1, 27, 78, 78n.5, 92, 204, 206 depression 104 desire ambiguity of  6, 36–9, 100, 102 and duty  41 and motivation  25, 103–9, 148–9, 159 and preference  6 and want  6–7 as a propositional attitude  7, 10–12, 127n.12

as appearance; see presentationalism as correct  3, 121–3, 140–2, 149, 190–1; see also desire, as rational; desire, reasoning with as rational  85–6, 89–91, 96–9, 132–6, 143, 203–6; see also desire, as correct; desire, reasoning with as true  14–17; see also desire, as correct change at will  124n10, 160; see also desire, second-order consistency in  21–2; see also desire, variation in correct variation in  121–5, 136–42 degrees of  6, 148–59, 203–6; see also desire and motivation instrumental  88–9, 91, 110n.10, 150–1, 157–8, 160–2, 164–9, 173–4, 177–80, 203–6 reasoning with  88–9, 97, 132–4, 140–4, 155–69, 179–80, 203–6; see also desire, as correct; desire, as rational second-order 109–12 ultimate; see desire, instrumental desire-as-belief, defined  10–13 direction of fit  14–17 disbelief 12–17 dispositional theory of desire  79–82 dispositional vs. occurrent desire  35–6 distinct existences  73 embedding problem; see Frege-Geach problem emotion  23–4, 44–5, 130–6 enkratic requirement  82–7, 194–6 evolution  131, 134–5, 172, 184–5

222 Index Frankfurt, Harry  109–12 Frege-Geach problem  69–72 Gibbard, Allan  56n.6, 61–6 good  8, 116–28, 163, 173–4, 181 goodness, agent-relative  122–5 goodness, fitting attitude account of 124n.9 Grice, Paul  39–40, 100–1; see also implicatures guise of reasons  116–28 guise of the good; see guise of the normative, guise of reasons guise of the normative  116–18 Hazlett, Allan  124n.10 Hoping 125–8 Hume  3, 23–4, 27, 29–30, 55, 72–4, 143, 161, 163n.14 Humean theory of motivation  27; see also ODM Hume's negative claim  29–30, 72–3 hunger; see appetite Hursthouse, Rosalind  44–5 implicatures  19–20, 39–41, 100–3, 137 impulse; see urge inclination 181–2; see also urge inferentialism; see conceptual role semantics instrumentalism  73, 160–9; see also desire, instrumental intelligible 78n.5 intention  45–7, 78, 83n.10, 117 intuition 162–3 judgement internalism; see motivation internalism Lewis, David  10, 18, 107–8, 110n.10, 151–2, 171n.3, 174–5, 203 likings  7n.1, 90, 137–42, 162–3, 182, 191–2 lust; see appetite Manne, Kate  139–40 map  26, 49 McDowell, John  41, 112–14, 202 moral  7, 50–1, 61–2, 119, 135, 162n.12, 170n.1, 188–91, 197–8, 197n.6, motivation argument, the  53–5, 66–7, 88

motivation internalism  49–53, 57–60, 174, 176–7 motivation, defined  30–1, 86–7, 176–7 Muller-Lyer illusion  92, 92f Nagel, Thomas  28n.2, 36–7, 41–3 negation problem; see FregeGeach problem non-cognitivism  55–72, 150n.4 normative  2, 7–9 objectivism and subjectivism  3, 9–10, 73, 122n.7, 139, 187–96, 201 occurrent vs. dispositional; see dispositional vs. occurrent Oddie, Graham  87–8, 124n.10, 204 ODM  25, 72–4 only desires motivate; see ODM other-regarding normative attitudes  57–72 ought, and reasons  10, 53, 67n.10, 83–4, 107, 174, 178 ought, rational vs nonrational  76–9, 194–5 ought-implies-can; see reason-implies-can overwanting, objection from  95–106 parsimony  17, 28, 114–15 passion  20–4, 200 Plato  21–2, 107, 112, 202 pleasure  80–2, 120n.5, 137–46, 191–2 pragmatic influences on speech; see implicatures presentationalism 87–93 psychopaths; see amoralists quasi-realism  57, 68n.11, 69; see also non-cognitivism, conativism Quinn, Warren  79–82, 87 Radiohead 80–1 rationalize 78 rationality; see ought, rational vs nonrational; desire, as rational reason-implies-can 125 Reason (the faculty of mind)  20–4, 200 reasons 8–10 agent-relative  58–9, 122–5 normative vs. motivating vs. explanatory  8–9, 45, 201 strength of desire; see desire, degrees of

Index  223 wrong vs right kind of  111n.11; see also change at will, desire reasons beliefs, defined  8–9, 69–70, 163n.14 reasons externalism; see objectivism reasons internalism; see subjectivism Satan  119–20, 175–6; see also amoralists Scanlon, Tim  20–1, 58–9, 76n.2, 77n.3, 80, 101–2, 124n.9 Schroeder, Mark  11n.6, 39n.13, 70, 122, 124–5, 190 seemings; see presentationalism self-knowledge  117n.3, 155–7 simplicity; see parsimony Sinhababu, Neil  7n.1, 27–8, 86–7, 91n.21, 136, 142–3, 144n.12, 161–3 Smith, Michael  14, 26–8, 30n.5, 37n.11, 50–1, 52n.4, 60, 79–80, 90–1, 96, 104, 107–8, 113–14, 150n.4, 153n.5, 165, 189n.3, 190–1 snail  25–6, 47 social externalism  175–6, 176n.8, 184n.14 Stocker, Michael  96, 104, 120

Stoics  20–1, 97n.3, 104n.6, 107, 185 subjectivism; see objectivism and subjectivism supervenience 64–5 testimony 162 thirst 21–2; see also appetite twin-earth  171n.3, 173n.6 uncertainty; see certainty underwanting, objection from  95–106 urge  42, 80, 97, 107–8, 139–42, 184–5; see also inclination value; see good Velleman, David  119–20, 127 Wall, David  32n.8 want; see desire and want weakness of will; see akrasia wellbeing, desire based theories of  196–8 Williams, Bernard  22, 187n.1, 188–9 wishful thinking  98–9 wishing  32n.8, 125–8