University of Southern California Dissertations and Theses

On being bound: law, authority, and the politics of obligation

(USC Thesis Other)

On being bound: law, authority, and the politics of obligation

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content ON BEING BOUND: LAW, AUTHORITY, AND THE POLITICS OF OBLIGATION by Kory DeClark A Dissertation Presented to the FACULTY OF THE USC GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (PHILOSOPHY) August 2012 Copyright 2012 Kory DeClark ii Acknowledgements Students inevitably hear many horror stories about advisory relationships when they begin graduate school. They involve aloof, disinterested, unfriendly, and sometimes even downright callous professors, who are more interested in research than they are in teaching. My experience could not have been more different. The support I received, not just from my advisors, but also from fellow graduate students, administrators, staff, and other members of the broader University community was consistent and unqualified, touching many parts of my professional and personal life. I am, of course, most indebted to my two primary advisors—Gary Watson and Andrei Marmor—who went far beyond their institutional obligations to help me grow as a philosopher. Gary was endlessly patient with me as I tried to develop my (often inchoate) ideas. Even where we disagreed, his principal concern was always to understand what it was that I was trying to say, to see what (if any) value there was in looking at things as I was, and to help me to say it better. His ability to cut through superficial complications to the heart of whatever philosophical issue I was struggling with set a standard of philosophical integrity that I can only hope someday to meet in my own work. iii Andrei was a constant advocate during my time at USC, helping me to grow my idea for a dissertation from an initial course paper to the work it is today. He was unnecessarily generous with his time, and read more drafts of chapters and essays than he likely cares to remember. Though our views conflict on various issues, it will be clear to anyone familiar with Andrei’s work how significantly it has shaped my understanding of topics at the intersection of law and philosophy. It is a rare privilege to work so closely with a leading figure in one’s field, and I am indebted to him for the seriousness with which he took my work and the lengths he went to help me improve it. I am also deeply grateful to Gideon Yaffe for his mentorship during the past six years. Gideon was often the first person I turned to for serious concerns, both personal and professional, and it is no exaggeration to say that without his guidance and encouragement my life’s path would look much different than it does today. I would also like to thank the many people that have improved my work through their friendship, counsel, and tutelage: Sharon Lloyd for her philosophical instruction and constant encouragement; Jake Ross for showing me how to be creative in both my writing and teaching (and for teaching me the value of the dead lift); and Mark Schroeder for his professional guidance and commitment to making USC’s graduate program one of the best in the country. I am also grateful to Scott Altman for his helpful comments during my qualifying exam, to Gregory Keating for his thoughtful and constructive criticisms during my dissertation defense, and to both George Klosko and Christopher Wellman, who offered helpful comments on drafts of chapters concerned with their work. Finally, and most importantly, I am beholden to Lindsey Cordova, my partner and best friend, for her kindness, wisdom, and unwavering belief in my potential. iv Table of Contents Acknowledgements ii Abstract vii Preface viii Chapter 1: What is the Problem of Political Obligation? 1 1.1 Clarifying the Question 1 1.2 Success Conditions 10 1.3 Conclusion 18 Part I 20 Chapter 2: Acceptance and Goods Arguments from Fair Play 21 2.1 The Principle of Fairness 24 2.2 The Acceptance Approach 28 2.2.1 Acceptance 28 2.2.2 Willingly and Knowingly 32 2.2.3 Conclusion 42 2.3. The Goods Approach 44 2.3.1 Klosko’s Principle of Fairness 45 2.3.2 Objective and Subjective Interpretations 47 2.3.3 Objections to the Subjective Interpretation 52 2.4 Conclusion 71 Chapter 3: Samaritanism 73 3.1 The Broad Strokes 74 3.1.1 Legitimacy 75 3.1.2 Obligation 79 v 3.2 Problems Facing Samaritanism 82 3.2.1 Problems with Legitimacy 83 3.2.2 Problems with Obligation 103 3.3 Summary and Conclusion 119 Part II 122 Chapter 4: The Structure of Obligations 123 4.1 The Simple View of Obligations 124 4.2 Exclusionary and Preemptive Reasons 129 4.2.1 Hart and Raz on Deliberation and Action 132 4.2.2 Identifying, Evaluating, and Entertaining Reasons 136 4.2.3 Exclusionary Reasons and Obligation 140 4.3 Preemptive Reasons and Obligation 141 4.3.1 Decisions 144 4.3.2 Mandatory Norms 149 4.3.3 Obligation 152 4.3.4 Summary 155 4.4 Moving Beyond the Simple View 156 Chapter 5: Foundations of the Authority Theory 159 5.1 The Authority Theory in Outline 161 5.2 The Authority Theory in Detail 164 5.2.1 How Practical Authorities are Justified 165 5.2.2 How an Authority Generates Obligations 169 5.2.3 Conclusion 181 5.3 Objections to the Authority Theory 182 5.3.1 The No-Authority Objection 185 5.3.2 Raz’s Solution to the No-Authority Objection 191 5.3.3 A Better Response to the No-Authority Objection 208 5.3.4 Conclusion 217 5.3.5 The No-Obligation Objection: The First Strand 219 5.4 Conclusion 234 Chapter 6: Obligation and Justification Redux 235 6.1 The No-Obligation Objection: The Second Strand 235 6.1.2 Getting Obligations Out 237 6.2 Requirements 241 6.2.1 Austin’s Insight 241 6.2.2 The Structure of Requirements 242 6.3 Wrongings 247 6.3.1 Three Ways to Get It Wrong 247 6.3.2 Indexing 251 6.3.3 The Nature of Wrongings 255 vi 6.4 Updating the View 262 6.5 Conclusion 265 Chapter 7: Three Paths to Political Obligation 267 7.1 The Standard Argument 271 7.1.1 The Validity Question 272 7.1.2 The Applicability Question 279 7.1.3 Objections 289 7.2 The Proceduralist Argument 295 7.2.1 Political Autonomy 296 7.2.2 The Argument 299 7.2.3 Answering the Proceduralist and Gappiness Objections 303 7.2.4 Objections 306 7.3 The Subjectivist Argument 316 7.3.1 Having: Reasons and Obligations 319 7.3.2 Validity 323 7.3.3 Applicability 325 7.4 Advantages of the Authority Theory 330 Conclusion 340 References 342 vii Abstract I have two principle aims in this dissertation. I attempt in Part I to establish the need for an adequate theory of political obligation by highlighting the challenges facing what I consider to be the two leading contemporary theories: George Klosko’s theory of fair play, and Christopher Wellman’s theory of samaritanism. I then attempt, in Part II, to meet this need with a novel theory of my own—what I call the authority theory—which attends to important normative considerations many theorists working on the problem have ignored. While I argue that the authority theory more plausibly satisfies the conditions on a successful theory of political obligation, the most philosophically gainful feature of the theory is, from my perspective, the use it makes of the concept of authority in addressing the problem. viii Preface It is somewhat misleading for any contemporary writer to claim to deal with the problem of political obligation. Since Socrates’ inceptive treatment of it in Plato’s Crito, different writers have understood the problem in different ways. The particular question in which I will be interested in this essay is, roughly, which (if any) considerations obligate an individual in a particular political community to obey its laws, from which it can be determined in which political communities these considerations actually exist and to whom they apply. This is what I shall refer to as the problem of political obligation. 1 The argumentative landscape has changed dramatically in the last half-century, accelerating a historical trend as old as liberalism itself. After Socrates’ death, pioneering work on political obligation did not resurface until the mid 17 th century, when Thomas Hobbes took the first intellectual steps away from the traditional teleological worldview, which placed absolute authority in the monarch by virtue of God’s commission. Hobbes suggested that individuals’ obligations to obey the dictates of their monarchical government 1 My interests therefore diverge from those of, e.g., Gilbert (2006), Ladd (1970), McPherson (1967), and Singer (1973). Gilbert is interested in more extensive obligations than I am here, while both Gilbert and McPherson do not prioritize moral obligations over other possible types. Ladd, Singer, and a number of other writers are arguably not interested in obligation at all, but rather in the question of whether citizens ought to obey the law. I return to these differences below. ix were rooted in their preference for the peace and stability that only such governments could provide. Each citizen, because he was most interested in this peace and stability, had conclusive reason to agree to transfer certain of his natural rights to the monarch, giving him the absolute authority he required to legitimately provide the stability citizens desired. This generated, according to Hobbes, an obligation for the monarch’s subjects to obey him. Although Hobbes espoused an absolute monarchy, and would therefore seem anything but liberal by contemporary standards, his work was radical in its time, and planted the seeds of liberalism that would bloom forty years later with John Locke’s Two Treatises of Government. Locke deviated from Hobbes in various ways that more clearly grounded legitimacy and obligation in voluntary actions taken by individuals. He set limitations on the state’s authority and took other measures to dull the sharper edges of Hobbes’ view. The next 250 years accelerated the ideological shift towards democracy, beginning with the enlightenment ideals of Rousseau’s The Social Contract (1762) and Kant’s Philosophy of Right (1797). But the question was more tempered then than it is today. Writers such as Locke, Hume, and later John Austin, were concerned not with the question of whether citizens were obligated to obey the law—they assumed they were—but with the conditions under which they could justifiably disobey it. 2 It wasn’t until the turn of the century, following Kant’s work, that a nascent skepticism of governmental legitimacy emerged in the Western canon. Libertarian socialism, or anarchism as it’s sometimes called, represented a complete rejection of authority as a justifiable coercive force, and left a social, political, and intellectual mark on both popular and academic political thought. But while the anti- 2 Wasserstrom (1999, 18). x authoritarianism developed variously by writers such as Peter Kropotkin, Mikhail Bakunin, and Benjamin Tucker ushered into the political discussion new and critical perspectives on law, government, and authority, these and other anarchist views were consistently overshadowed by the concurrent evolution of liberalism, which included more renowned thinkers like Wilhelm von Humboldt, Jeremy Bentham, and J.S. Mill. Strands of these opposing philosophies did not come to a head over the question of political obligation until the middle of the twentieth century, when mainstream philosophers such as Robert Paul Wolff, M.B.E. Smith, A.J. Simmons and Joseph Raz began to develop thoughtful philosophical defenses of radical, anti-authoritarian viewpoints similar to those anarchists had been trumpeting for over 150 years. These philosophers have since become known as “philosophical anarchists.” One theme of the political narrative stretching from Socrates to Raz is a growing suspicion, at least in academic circles, of the state’s capacity to issue morally binding legal directives. My purpose, when I initially set out to write this dissertation, was to deepen this suspicion—to argue that no considerations could be offered which establish the existence of a general obligation to obey the law. In coming to understand the issues, however, I began to question my own position. While I agreed that nobody had offered a plausible theory of political obligation, it struck me that nobody had paid adequate attention to the purportedly authoritative relationship states bear to their inhabitants, or to the nature of the practical reasons this relationship allows states to create for them. Perhaps there were ways in which authority, properly understood, could be established nonconsensually? And if it could, then perhaps it could be part of a new sort of argument for political obligation? In what follows, I will argue that both are possible—that authority can be grounded nonconsensually, and that, xi when it is, the state has the capacity to generate moral obligations for its citizens to obey the law. I have split my argument into two main parts. My aim in Part I is to demonstrate a prevailing need for an alternative theory of political obligation by detailing the challenges facing what I take to be the two most plausible contemporary views: George Klosko’s argument from fairness (a descendent of H.L.A. Hart’s theory of fair play), and Christopher Wellman’s theory of samaritanism (a descendent of John Rawls’ theory of natural duty). In Part II I offer a competing theory—what I call the authority theory—which I believe to be superior to these and other theories currently on offer. As I will explain in detail in later chapters, the authority theory is unique in that it takes seriously a set of more fine-grained moral questions that I believe is too often omitted from the literature on political obligation, regarding obligations, authority, wrongings, and other concepts that bear on our practical reasoning. What does one have when one has an obligation to obey the law? Are obligations simply weighty reasons, or are they something else? Are political obligations different from other obligations in important respects, or are they merely different applications of the same moral requirement? Does having a political obligation entail that one must obey immoral laws? What is the relationship between authority and obligation? And so on. Having hopefully established the need for a new theory of political obligation in Part I, I begin in Part II by introducing and defending a framework for understanding the concepts of obligation and authority. In Chapters 4 (obligation) and 5 (authority) I reconstruct much of the work of Joseph Raz, and explain, roughly, how concepts similar to those he introduced can be used to create a novel theory of political obligation. My aim in these chapters is to explicate notions which are both plausible and sufficiently textured to be xii useful in our later argument. (Because this is meant to be a work of political rather than moral philosophy, I will not be able to investigate all of the many subtleties of the explanations I endorse, but will nonetheless do my best to defend the conceptions I introduce.) I go on in Chapter 6 to discuss and respond to two important problems facing the Razian view as I’ve described it. My treatment of these problems provides the foundation for the more detailed arguments I give in Chapter 7. Throughout this dissertation I will refer to the theory I’m developing, as well as the less developed view I credit Raz with introducing, as the authority theory. An authority theory is a theory according to which the authoritative relation between authorities and subjects plays a necessary part in the explanation of those subjects’ obligations. My particular version of this theory—which begins to diverge from Raz’s in Chapter 5, and is, by the end of Chapter 7, significantly different than anything he proposed—is made up of three arguments for political obligation, which I refer to as the standard, proceduralist and subjectivist arguments. I discuss these in detail in Chapter 7, and explain in what ways I take them to be, together, explanatorily superior to the views we’ll consider in Part I. Before doing any of this, however, we must be clear about the question we mean to ask. This is my aim in Chapter 1, to which I now turn. 1 Chapter 1: What is the Problem of Political Obligation? My aim in this short chapter is to bring the problem of political obligation into sharper focus—to say what we can expect any theory of political obligation to explain, why we are interested in obligations rather than what we have conclusive reason to do, what our sense of ‘obligation’ amounts to, whether we intend to justify obedience or simply compliance with the law, and so on—as well as to establish the conditions that a successful theory of political obligation must meet. 1.1 Clarifying the Question The theories we will consider in Part I, as well as the theory I advance in Part II, are narrow theories, meant to explain our moral duty (which includes both our obligations, properly speaking, as well as our natural duties) to comply with (rather than obey) the laws of our state. While many readers will no doubt already have a clear understanding of the differences between these theories and others that might be offered under the rubric of political obligation, I will say a few things about the nature of the question they are meant to answer. 2 Wide and Narrow Theories Theories of political obligation differ not just in the form of the explanation they aim to provide, but also in what they aspire to explain. Our political institutions mediate many of our relationships, most of which have clear deontic implications. How are we to decide which of the resulting obligations our theory of political obligation should explain? The answer depends on our theoretical aims. I will be concerned only with narrow theories of political obligation—those which purport to explain the conditions under which an individual within or bearing some other relevant relationship to a political community has an obligation to obey its laws. A wide theory, while it might also address obedience to law, is concerned foremost with what further political participation such a relationship might require. Narrow theories, in other words, try to provide an answer to the question “Why am I obligated to obey the law?” while wide theories try to provide an answer to the question “To what lengths am I obligated to go to support or uphold the political institutions of my society?” If we say that a theory which posits the existence of political obligations is successful if it correctly describes the considerations generating, as well as the actions required by, such obligations, then the success or failure of a wide or narrow theory does not logically entail the success or failure of the other. A wide theory might establish a threshold for political participation while at the same time relaxing constraints on obedience to law; it might claim, that is, that none of us has a special obligation to obey the law so long as we engage sufficiently in political activities which support our community’s broader political and social 3 institutions. 3 In this case each of us would have political obligations in the wide sense, but no obligation to obey the law. As we’ll see, a significant challenge facing the views I discuss in Part I is that they would fail as narrow theories even if they were to succeed as wide ones. Conversely, it might be that we have no general obligation to uphold the political institutions of our political community while at the same time each of us has an obligation to obey its laws. In this case there would exist a political obligation only in the narrow sense—a possibility that is consistent with the authority theory I defend in Part II. A.J. Simmons warns that by restricting our inquiry to narrow theories we may overlook obligations which are “intimately associated with the notion of citizenship.” Activities such as defending one’s country are, he suggests, “generally taken to be an important aspect of one’s political bonds” even where the law does not require them. 4 While I take Simmons’ point, the fact that the narrow question applies in any context in which directives are issued by a purported authority gives it a broader relevance within normative theory. My answer to the narrow question involves states and their subjects, but the theory 3 It is an interesting exercise to imagine a political institution that promoted civil disobedience as a preferred (or at least permissible) form of political expression. (Some theorists conceive of the current legal order as akin to this already. See, e.g., Freeman (1966, 235).) While surely an implausible political model, such a society would introduce some intriguing possibilities. A statute, appended to current laws, might specify proportionality requirements governing sanctions on actions recognized by the state to be legitimate forms of civil disobedience. Periodic votes would allow citizens to reconsider those laws their fellows opposed vehemently enough to consistently (but peacefully) disrupt public order. Costs in time, effort, embarrassment, and so on, attached to civil disobedience, and proportionally balanced in accordance with standards determined by a vote, would allow citizens to quantify their opposition to a particular law through action. By these means democratic practices could embrace the weighted dimension of voting practices proposed by some contemporary theorists of democratic economics, expressing not only their ordinal but also their cardinal preferences for particular laws. See, e.g., Albert (2004). 4 Simmons (1981, 5). See also Gilbert (2006, 13-14), who formulates the problem she engages accordingly. 4 I’ll defend applies to authoritative relationships more generally—to workers and their bosses, professors and their deans, privates and their officers, and so on. My aim throughout is to understand what about authoritative relationships makes them relevant to the reasons we have to act. Oughts and Obligation Even within the category of narrow theories, questions that must be distinguished are sometimes run together. For instance, regarding a particular state S in which P lives, we might ask whether P has reason to obey, has an obligation to obey, or simply ought to obey S’s laws (where ‘ought’ means, as it will mean throughout this essay, ‘has most reason’). The tendency for earlier writers on the subject to conflate these questions has moved most contemporary theorists to distinguish them carefully. 5 I, too, would like to highlight a few important distinctions here to explain our explicit focus on obligation—a concept I will examine thoroughly in Chapters 4 and 6. Each of us has many reasons to obey the law. We might want to avoid punishment, to appear in a certain light to our peers, etc., none of which directly concerns the special moral relationship that may or may not exist between us as members of the same political community. Because we will be interested foremost in this relationship, the question of whether people have reason to obey the law is overly broad. A theory of political obligation which told us that people never have reason to obey the law would obviously be false, 5 For a partial list of past offenders see Simmons (1981, 202 fn.2) to which these more recent examples can be added: R.P. Wolff (1970, 9) and (2000, 183, 188); Kavka (1986, 385); McPherson (1967, throughout). 5 whereas one which told us that people usually have such reason would tell us something we already knew. One way to narrow the question would be to ask whether we have a conclusive or even an absolute reason to obey. An absolute reason to do something is one that cannot be overridden in any context. Because no reasonable person could suggest that we have an absolute reason to obey the law—one that could not be overridden by any imaginable considerations—I will ignore this possibility. A conclusive reason to do something in a context C is, much more reasonably, one that outweighs the reasons one has in C to do anything else. When someone has a conclusive reason to do something we say that, all things considered, she ought to do it. 6 It may seem, at least initially, that this is precisely the question we should be asking: Ought we to obey the law? The problem is that, even within sufficiently just representative democracies, whether an individual ought to obey any particular law will depend on a variety of factors for which no theory of political obligation could account in the absence of complete information about the community. What we are interested in, alternatively, is the moral “bond” between subjects and states (or between subjects within states) that persists even in circumstances in which it is outweighed, for any particular individual, by other considerations. As I will explain in more detail in Part II, I believe this bond is obligation. 6 ‘Ought’ clearly has qualified uses which recommend actions within artificial parameters. If I know that you are concerned only with receiving good marks on a paper, and nothing else, then I might say that you ought to plagiarize. ‘Right’ and ‘wrong’ may have similar uses, but they fit less comfortably in many contexts. It does not seem, for instance, that it would be right for you to plagiarize, or (and especially) that it would be wrong for you not to, even if you ought to plagiarize given you aims. I discuss different senses of ‘wrong’ in Chapter 6. 6 The difference between ought and obligation is brought out nicely by the fact that answers to “ought” questions amount to advice, while answers to obligation questions amount, as I will explain in detail in Chapter 6, to requirements of a certain sort. Telling someone that she ought to φ (or, more weakly, that she has reason to φ) is merely to recommend that she φ given her circumstances. Telling her that she has an obligation to φ seems, comparatively, to simultaneously constitute both a stronger and weaker claim. 7 It is stronger in that it requires rather than merely recommends an action. It is weaker in that it might not carry with it an all-things-considered recommendation: Sometimes one ought to violate an obligation in order to accomplish something more important or pressing. But if obligations do not entail oughts, then why should we care whether we have political obligations? Aren’t we ultimately concerned with the way people should act? Suppose it turns out that all members of a particular political community incur obligations to obey its laws. If these obligations are so weak as to be overridden by even the insubstantial prudential reasons its members have to do something else, then why would these political obligations matter to us? And if they do not, then why ask about obligations at all? Why not ask what we have conclusive reason to do? As I will discuss at length in Chapter 4, the type of influence obligations exhibit on our practical reasoning can be explained structurally, without reference to the weight they have as reasons that bear on our practical deliberation. It is true, then, that without supposing that political obligations, if they exist, carry significant weight as reasons for action, our investigation will be little more than an analytical exercise. To be sufficiently 7 Simmons (1981, 10 fn.b). 7 interesting, a successful theory must explain why our obligations constitute weighty reasons to act as they specify. 8 I will have more to say about this later. Obligation and Duty To this point I have mentioned only one type of moral requirement. I should be clear, however, that while I will continue to use the language of obligation throughout much of this essay, the political “obligations” we have to one another are, according to the view I defend in Part II, particular applications of our more general duties. ‘Obligation’ and ‘duty’ are used interchangeably by most speakers, and may be substituted salva veritate in many instances. But while this is consistent with how I will generally use these terms throughout this essay, I should note that it is now common to distinguish between them on the basis of “paradigmatic uses” by native speakers. 9 A term’s use is paradigmatic just in case (i) its context of use is one in which no other term would, in that context, better capture the idea the speaker is trying to express, and (ii) the contexts in which the term seems most appropriate have certain features in common, such that a person’s use of the term would itself imply that those features are part of the context of utterance. 10 Whether two terms 8 It is possible to avoid this additional explanatory burden by maintaining, as, e.g., Simmons (1981, 7) and Gilbert (2006, 33) do, that obligations per se are inherently weighty. However, this approach brings other explanatory challenges that I’ll explain in Chapter 4. 9 See Brandt (1964), from which this discussion is drawn. 10 Brandt expresses the first condition in empirical terms. A term is said to “best capture” the idea that a speaker is trying to express in a context just in case (i) given a variety of similar available terms in that context, a vast majority of speakers will select that term to express that idea, and (ii) in other contexts in which multiple terms express that idea equally well, speakers have no preference between them. 8 have separate paradigmatic uses indicates, ceteris paribus, whether they represent legitimate conceptual divisions or accidental idiomatic variations. ‘Obligation’ and ‘duty’ both seem to have paradigmatic uses in Brandt’s special sense. ‘Obligations’ are historically related to promises and agreements, as they are to reciprocation for benefactions. They are also usually owed to someone, signifying both that multiple parties are involved, and that, when they are owed, the obligee has a correlative right to the relevant performances. 11 Duties, by contrast, are, at least in their more technical sense, owed to all persons if they can be properly said to be owed to anyone at all. Thus, we claim to “do” our duty, but to “discharge” or “satisfy” our obligation. 12 Paradigm uses also indicate that an obligation can be undertaken only through some voluntary action. For instance, I might obligate myself to someone by accepting her help (a phenomenon I discuss in Chapter 2). It’s this peculiarity that explains our comfort with “I’ve undertaken an obligation,” but not with “I’ve undertaken a duty.” 13 Of course, none of this is meant to suggest that there are not broader uses of ‘obligation’ that are perfectly acceptable today (what Brandt calls “extended” uses), or that certain uses of ‘duty’ cannot overlap with uses of ‘obligation’ in some contexts. Indeed, as I 11 Hence the term’s form: ob plus the verb ligare, which means “to bind.” It should be pointed out, however, that one’s performance need not always be “for” the person to whom one owes the performance. This will be particularly important for my argument in Part II. 12 Simmons (1981, 14). Note that this may not be true for positional duties. See next footnote. 13 However, the latter sentence might make sense when a person voluntarily accepts a position that comes with certain duties. These have been called “positional duties,” and should be distinguished from the sort of “general” duties we are concerned with here. Positional duties are discussed at length in Simmons (1981, 16-23). 9 said, in what follows I will generally use ‘obligation’ and ‘duty’ interchangeably. Where the distinction between these is important to the argument, I will be clear about which I mean. Obedience and Compliance A final feature of the problem of political obligation that I’d like to discuss involves obedience. An interesting fact about theories of political obligation is that they generally claim, not just that people have an obligation to comply with the law, but that they have an obligation to obey it. Ought the content of this obligation to shape our inquiry? In fact, I believe that it should. A genuinely obedient action entails something about the actor’s attitude towards the law that reflects her attitude towards its source. Because, as I’ll explain in Chapter 6, I believe (i) that moral obligations are constituted in part by the fact that wrongings follow from their transgressions, and (ii) that such wrongings ought to be understood as a certain form of disrespect, which is dependent on one’s attitudes, I believe that the attitudes we express through our behavior are particularly important to the question at hand. I think it would be worthwhile, therefore, to ask, of those theories that purport to explain our political obligations, whether they explain our duty, not just to do as the law says, but to do it for the reason that it’s the law. While I believe that this would be a worthwhile pursuit, however, the use of the term ‘obey’ rather than ‘comply’ in historical and even contemporary writing on the problem appears to be a meaningless artifact of the way the question has come to be asked. Though they talk explicitly about obedience, most political theorists are in fact interested only in whether we have an obligation to comply with the law, and appear not to see the value of asking the more demanding question about obedience. To avoid disengaging with the 10 literature on the subject, I will therefore use the terms ‘obedience’ and ‘compliance’ more or less interchangeably, as others have, and will follow other theorists in structuring our inquiry around the question of compliance. 1.2 Success Conditions Having clarified the question we aim to answer, I would like to say something about the conditions any theory must meet to be successful—the “success conditions,” “standards of success,” “success criteria,” or “satisfaction conditions” as they’re called in the literature. The conditions most commonly considered are, following A.J. Simmons, particularity, singularity, universality, and generality. 14 Most contemporary writers (including Simmons) reject singularity and universality but accept particularly and generality. 15 While I will ultimately subscribe to these same two conditions, I would like to approach the topic more cautiously than other theorists have. In particular, I’m skeptical that these conditions are as stark as they are often made to seem, and would like to be clear about the assumptions we make by endorsing them. I’ll have more to say about this as I discuss each criterion individually. 14 Simmons (1981, Chapter 2). 15 More accurately, Simmons does not fully accept generality, but says he will “adopt this wider criterion of success as a general standard against which to measure suggested accounts” (1981, 56). 11 Particularity According to the particularity requirement, to be a genuine political obligation one’s obligation must be to the political community of which one is a part. 16 The idea is that any principle which bound a person to multiple political communities, or to a foreign one, could not pick out the bond many of us intuitively take each individual to have only to “her own.” Since this bond is the one to which we mean to refer when we talk about political obligation, a principle that does not meet this requirement cannot be a genuine principle of political obligation. As Simmons says: The problem becomes particularly apparent if we imagine that I am living under an unjust government in a country at war with another, justly governed, country. Could we seriously maintain that my “political obligation” consists in opposing the efforts of my own country, in favor of a country with which I may have no significant relations whatever. While we may believe that I have a duty to oppose my country, this is surely ill described as my “political obligation.” 17 But while most people would accept Simmons’ conclusion in this case, it is not clear why it is correct. In its strongest form, particularity claims that one can only have a political obligation to one’s own political community—that any principle which bound one to another political community would not be a principle of political obligation at all. In its weaker form, particularity requires that the principle bind one to one’s own political community merely 16 See Simmons (1981, 31); Klosko (2005, 12). 17 Simmons (1981, 32 fn.g). In fact, the reasoning in this passage begs the question against any theory of political obligation that has the described consequences. For it stipulates that one has “no significant relations whatever” to the foreign country when the question is precisely whether one could have such relations, and on what grounds. I ignore this feature of the example. 12 above all others. However, both conditions seem too strong, at least pretheoretically. 18 Why should any justificatory considerations be incapable of binding individuals to political communities other than their own? Those which have traditionally been thought to ground our political obligations—consent, gratitude, fair play and natural duty, for instance, as well as the considerations at the heart of the authority theory I defend in Part II—are theoretically neutral between political communities. They bind us or they don’t, whomever they bind us to. Suppose, for instance, that we determine that some principle of gratitude G obligates individuals to behave in particular ways towards those who have benefited them, and that this principle binds members of some political community C to one another. In such a case we would say that people within C had political obligations, and that G described them. Now suppose that, having satisfied ourselves that G does the justificatory work we take it to, we find that people within C actually benefit much more from individuals from some other political community C*. Perhaps C is experiencing a serious drought, and the members of C*, who have no previous relations to members of C, take it upon themselves to dig an aqueduct to transport water from their own community to C, saving the lives and livelihoods of C’s members. By parity of reasoning, it would seem, ceteris paribus, that members of C are bound to C* over and above C. According to both the strong and weak form of particularity, however, this is grounds to dismiss G as a plausible principle of political obligation, despite 18 Simmons (1981, 31-35) seems not to notice (or at least not to be concerned with) the distinction between the strong and the weak interpretation, and glides back and forth between them. His interest in tying obligation to citizenship (an approach that shares certain characteristics with membership theories) is probably responsible for his apparent sympathy for the stronger version, while the weaker version, which both appears sufficient for his purposes and seems more plausible, follows clearly from certain key passages, particularly (1981, 32). Gans (1992, 9), as a comparative example, rejects the stronger interpretation but appears to accept the weaker one. 13 the fact that we previously took it to accurately describe the obligations members of C have to one another. 19 Advocates of particularity thus seem to be more interested in whether states meet the conditions necessary for the application of a particular justifying principle, rather than in that principle’s validity. But if particularity is problematic in both its strong and weak form, then why do essentially all political theorists working on the problem of political obligation accept it? One possibility is that they are implicitly committed to some sort of membership principle which is separate from the principle of political obligation for which they argue. Simmons, for instance, argues that political obligations track citizenship. 20 The danger is that he seems to simply presuppose what he aims to show—namely, that the fact that some group of people shares a political system obligates them to one another—and that this presupposition is subject to the same objections facing associative theories of political obligation, which are grounded in the simple fact that we do share such a system. 21 The alternative approach I suggest is to treat particularity as an indicative rather than an evaluative criterion. An evaluative criterion is one we use to evaluate the veracity, accuracy or completeness of a principle or theory. An indicative criterion, by contrast, merely indicates what we hope to find and where we might begin looking to find it. Failing to satisfy an evaluative criterion implies that a principle is not normative, or that its normativity does 19 Interestingly, Simmons (1981, 35), who was the first to explicitly discuss the particularly requirement, and who supports it, criticizes John Ladd (1970) and others for using the faulty intuitions we’ve discussed here to defend a constraint on universality, which I discuss below. 20 Simmons (1981, 31). 21 Dworkin (1986) proposed the best known associative theory. I address a number of problems with his view in Chapter 7. 14 not apply in a particular instance. Failing to satisfy an indicative principle implies only that the principle, whatever its normative status, is not interesting for our purposes. Were the particularity requirement an evaluative criterion, then the fact that principle G ultimately bound members of C more completely to C* would falsify the principle. Understood as an indicative condition, however, particularity sets boundaries, not on what counts as a right answer, but on which sorts of question are worth asking. As we’ll see in Part I, theories of both fairness and samaritanism have trouble explaining particularity, even as an indicative criterion. There is, according to these theories, nothing inherent in political communities to bind its members to one another. As I have tried to suggest, this does not mean that either theory is incapable of explaining political obligations. It does mean, however, that they cannot be used to explain and justify the common belief that we are bound to our compatriots in a way that we cannot be bound to members of other states. By comparison, the authority theory I advance in Part II is able to explain particularity, but it too rejects the suggestion that our obligations are due to our shared membership in a particular political community. Rather it depends, I argue, on pragmatic facts about the state’s epistemic position relative to its subjects. I will argue that compatriots are bound to one another in ways they are not bound to foreigners, but only contingently, and not for the reason that membership theorists or associativists believe. Singularity As an evaluative criterion, singularity in ground requires that political obligations be grounded in a single consideration (fairness, consent, etc.) rather than in some combination of principles. While it may seem obvious to some that there is no principled reason to accept 15 singularity as an evaluative criterion, it was only recently that political philosophers began experimenting with multiple-principle (MP) theories of the sort we’ll discuss later in this essay. I would emphasize that there is a substantial difference between accepting singularity as an evaluative criterion and believing that political obligation is grounded in a single consideration. One may assume that individuals are in fact singularly bound while rejecting the view that no set of principles could bind them. Even if one advocated the stronger modal view, one may remain committed to accepting any consideration that later proved one’s assumption false, just as a prosecutor can believe that only one piece of evidence could weigh conclusively against the defendant without being committed to forswearing another piece of evidence if it were discovered. Universality Universality requires that a successful theory of political obligation bind each and every person within a political community to obey its laws—that is, that a principle of political obligation cannot be genuine unless it binds an entire political community. Many past thinkers (and even some contemporary ones 22 ) accepted universality, or at least operated as though they accepted it. Some claimed, additionally, that it must bind everyone equally, and to all the laws. 23 The first constraint is now commonly rejected, while the second is used, like particularity, as an indicative criterion. 24 22 See, e.g., Ladd (1970, 27). 23 J. Wolff (2000, 182) calls the degree to which an obligation binds each of the individuals within a group its uniformity (though he ultimately rejects this as an evaluative principle); Ladd (1970, 17) calls obligations (including political obligations) totalistic in that an obligated person is “eo ipso required to 16 The problem with using universality as an evaluative criterion is, again, that there is no reason to think that a principle’s failure to bind everyone, or everyone equally, or everyone equally to obey all of the laws, excludes it from consideration as a principle of political obligation. 25 If it did, then to know whether certain political considerations obligate an individual to behave in a particular way, we would first need to know whether and to what degree those same considerations obligated everyone else. This is of course an implausible commitment. Except in highly artificial circumstances, others’ obligations, political or not, would appear to be completely independent of one’s own. Generality Generality is merely a weaker form of universality. Instead of requiring that all individuals within a political community be bound according to a particular principle for that principle to be an acceptable principle of political obligation, it requires merely that most of them be. Klosko, for instance, says that …an acceptable theory must be able to establish the obligation of all or most members of society. Ideally, obligations of all would be established. But since this could well be impossible, this standard can be relaxed to a certain extent, although do the whole of it,” thus explaining why a political obligation would apply to all laws. As a related but slightly different notion, Klosko (2005, 11) has called a valid principle’s applicability to all the political obligations we take ourselves to have its range. 24 See, e.g., Gans (1992); Klosko (2005); McPherson (1967); Simmons (1981); J. Wolff (2005). 25 J. Wolff (2000, 187) points out that there may be “political advantages” to uniformity, and one could surely say the same about universality. A state in which only some people were obligated to obey the law would be much more difficult to manage if justified disobedience were permitted. But, as Wolff notes, this is a political justification rather than a philosophical one. 17 the question of exactly how many or what percentage of inhabitants must have obligations could be a matter of controversy. 26 Given that any reasons we have to reject universality would seem to apply equally to generality, it is difficult to see how one could accept the latter while rejecting the former. Klosko does not explain for what reason “standards can be relaxed,” or why we should want to relax them. Again, I believe the problem is resolved by specifying that the condition is indicative—that we impose it artificially in hopes of finding a principle that tracks our intuitions. Burden of Proof The last issue I’d like to discuss, before moving on to consider the theories in Part I, is the position we take pretheoretically in addressing the problem of political obligation. As Jonathan Wolff 27 has pointed out, most writers assume that it is incumbent on the advocate of political obligation to refute the “philosophical anarchist”—the theorist who claims that there are no political obligations—rather than the other way around. 28 Philosophical anarchists, who generally champion values like liberty, autonomy, and other individualist benefits of liberal societies, might respond by invoking these features as grounds for anarchism’s claim to our fallback position. They might assert that, if no principle has been 26 Klosko (2005, 10). 27 Wolff (2000, 183). 28 The term “philosophical anarchist” is unfortunate, for the idea that we have no obligation to obey the law is at best only tangential to a truly anarchist position, which is concerned primarily with political coercion rather than with obligation. A person might consistently believe, for instance, that we have no political obligations but that we have most reason to have a strong, centralized state. Such a person would not have even an ostensible claim to being an anarchist. 18 found, liberty speaks in favor of concluding that no such obligations exist. As Wolff points out, however, the obligations that concern us in the debate over political obligation are moral obligations, not legal ones. Political obligations will put liberty at risk only in the eyes of those for whom morality is itself a threat to liberty. Even if our moral obligations to obey the law are generated by political considerations, we can still act freely within the confines of those requirements. Liberty can coexist with obligation. As Wolff suggests, then, we should demand positive arguments from theorists on both sides of the debate—from both philosophical anarchists as well as theorists advocating political obligation—privileging neither side. The latter must supply considerations which ground political obligations, and evidence that the circumstances specified by those considerations exist in contemporary states. From the political anarchist we expect an explanation of our intuition that in some cases we have obligations to others in virtue of the circumstances of our political lives. The failure of either side is not in itself a victory for the other. 1.3 Conclusion My goal in this chapter was merely to expand on the question that will guide us through the rest of this essay, and to introduce the conditions we will expect any positive theory to meet. Based on our discussion of these criteria, and particularly in light of the indicative status of the success conditions we discussed, I think it best to recast our question in two parts. What we aim to discover is (i) what sorts of considerations are capable of obligating an individual 19 in a particular political community to obey its laws, and (ii) in which (if any) political communities these considerations actually exist, and to whom do they apply. 20 Part I Theories of Political Obligation I now turn to Part I, in which I argue that the two most plausible contemporary theories of political obligation—George Klosko’s theory of fairness, and Christopher Wellman’s theory of samaritanism—fail to solve the problem of political obligation we elaborated in Chapter 1. Later, in Part II, I offer a competing theory of political obligation which I argue does a better job of addressing the problem. 21 Chapter 2: Acceptance and Goods Arguments from Fair Play Nearly all of the handful of attractive contemporary solutions to the problem of political obligation have distinguished histories, in most cases tracing back to Socrates, with some, like consent, receiving disproportionate attention through the centuries. But one of the most interesting and plausible solutions was first put forward in earnest only half a century ago, as H.L.A. Hart’s now-familiar “mutual restrictions” principle. According to Hart, “when a number of persons conduct any joint enterprise according to rules and thus restrict their liberty, those who have submitted to these restrictions when required have a right to a similar submission from those who have benefited by their submission.” 29 Hart’s focus on the mere receipt of benefits ultimately generates too many obligations, as I’ll explain below, but his general insight influenced various political and legal theorists who have since used it to develop more sophisticated approaches to the fair play argument. The two most central of these I call the acceptance and the characteristics of the goods approach (or simply the goods approach), each of which restricts Hart’s principle in a different way. The acceptance approach, which was first proposed by John Rawls and later discussed in detail by A.J. Simmons in his Moral Principles and Political Obligations, involves interpreting or refining Hart’s 29 Hart (1955, 185). 22 principle so that it applies only to individuals who have accepted (rather than merely received) benefits from cooperative schemes. 30 The goods approach, which has been advocated most forcefully in recent years by George Klosko, restricts the principle by applying it only to benefits with particular characteristics. 31 In this chapter I argue that neither approach provides a successful theory of political obligation. I begin in Section 2.1 by describing the features of Hart’s principle that make it a distinct and attractive basis for such a theory, as well as the advantages gained by interpreting it in accordance with one of the two approaches above, which I consider individually in Sections 2.2 and 2.3. It is important to recognize that on either approach two questions must be answered. The first and more fundamental question is whether the principle plausibly describes a genuine source of obligation—that is, whether there are any circumstances in which obligations are generated for the reasons described by the principle. The second question, conditional on an affirmative answer to the first, is whether the principle extends to the circumstances of political societies. The first question is therefore concerned with moral obligation much more generally, while the second is concerned with political obligation in particular. 32 We can refer to these as the validity question and the applicability question, respectively. We will return to them in later chapters. Perhaps the most vocal skeptic of political obligation, Simmons has focused his critical efforts on the applicability question. While he agrees with Hart and Rawls that the 30 See Rawls (1964); Simmons (1981, Chapter 5). 31 See Klosko (1992) and (2005) in particular. 32 To be clear, however, a political obligation just is a moral obligation one incurs due in part to certain facts about one’s political community. However cf. McPherson (1967). 23 acceptance approach indeed describes a genuine source of obligation, and so provides a satisfactory answer to the validity question, he denies that most people have in fact accepted the public goods available in states, and so he denies that citizens incur political obligations grounded in fairness. While I find Simmons’ critical argument mostly persuasive, I argue in Section 2.2 that problems with the principle of fairness run deeper than he and others suggest. While I do not deny that considerations of fairness plausibly bind in some cases—I leave that question open—I do suggest that they do not do so in virtue of acceptance, and that therefore the acceptance approach does not positively answer either of the questions that concern us. In Section 2.3 I turn to Klosko’s goods approach, which purportedly bypasses acceptance and thereby the problems I discuss in Section 2.2. I argue, however, that Klosko’s argument faces a dilemma. Interpreted objectively, it indeed avoids reliance on acceptance, but in doing so fails to ground obligations, and thus provides an affirmative answer to neither the validity nor the applicability question. On my interpretation of his view, Klosko avoids this criticism by endorsing a subjective conception of his principle. But this interpretation is doubly problematic, for it (i) unavoidably relies on a counterfactual notion of acceptance of the sort I reject in Section 2.2, while (ii) collapsing into a principle of hypothetical commitment, which opens it to three additional objections that I detail below. Together, these concerns leave the subjective interpretation looking even less plausible than the objective one. I conclude that, if there is some version of the fairness argument which grounds political obligations, it relies on neither the acceptance nor the goods approach. 24 2.1 The Principle of Fairness I take it to be relatively uncontroversial that, under the right circumstances, considerations of fairness can obligate a person to contribute to the cooperative projects of others. Consider a seemingly clear-cut case: Daycare Like me, a number of my colleagues have young children, and struggle to find reliable daycare. So they decide to organize their own daycare center on our university campus, each agreeing to staff the center once every few weeks. I refuse to participate. However, once the daycare center is operational I begin leaving my children there anyway, which (let us assume) members of the cooperative do not discover and could not prevent anyway (say, because of university policy). While nearly everyone agrees that it is wrong of me to “free ride” on the organized sacrifices of my colleagues, that I have an obligation to contribute to their scheme if I am going to take benefits from it, Hart was the first to suggest the principle at work in this and similar cases. Leaving aside conditions of his principle that are arguably extraneous—it is not clear, for example, why it should matter whether the scheme is rule governed in any strict sense, or even why a good must be cooperatively produced for a noncontributor’s receipt of it to be morally relevant 33 —its fundamental conditions, including those implicit in Hart’s description, 33 This will surely be controversial, as the cooperative nature of productive schemes has generally been central to fairness accounts. Even Simmons, who rejects a number of superfluous conditions Hart and Rawls suggest, does not reject this condition. Still, I see no reason to think that an individual producer cannot be taken advantage of, or why taking advantage of him is less objectionable than taking advantage of a group. Indeed, occasional examples from philosophers who 25 run as follows. When (i) an individual or multiple individuals undertake a burden (ii) in order to produce a good which is mutually beneficial, (iii) a person who benefits from that good has an obligation to contribute her fair share to the productive venture. Two characteristics make this a unique and attractive basis for a theory of political obligation. First, the principle does not rely on a deliberate undertaking or commitment, such as consent, a promise, or a contract, whose proper execution depends on one’s appropriate expression, in suitable circumstances, of an intention to undertake an obligation or transfer a right. 34 Suitable circumstances were presumably present in Daycare, where volunteering to participate would no doubt have signaled my commitment. However, I explicitly declined. And so because at no other time could I reasonably be thought to have given an appropriate expression in suitable circumstances, any obligation I incurred through my use of my colleagues’ service must have been grounded in something other than a deliberate undertaking. 35 Fairness therefore provides an explanatory advantage. Because so few constituents of any state have committed to obey the law, commitment theorists struggle to satisfy the generality condition, which requires as an implication of any successful theory include this condition seem to betray their recognition of its irrelevance. See, for example, Klosko (1992, 30). 34 See Hart (1955, 184-85); Simmons (1981, Chapter 5). But cf. John Ladd (1970, 18-24) who contends that fairness theories offer a way to “consent through participation rather than through contract.” 35 Nor did I tacitly consent (promise, contract) to participate when I later utilized the service. Consent is tacit when it is silent or otherwise passive. I tacitly consent to the passage of a bill, for example, by sitting motionless when it’s clear that the only way to vote against it (or abstain) is to raise my hand. Whether tacit or express, I cannot voluntarily commit to something without giving a deliberate and appropriate sign that I mean to commit myself. And so clearly in the case of Daycare, in which I explicitly refuse to undertake an obligation, I tacitly commit myself no more than I expressly commit myself. See Simmons (1981, 79-81, 126-28). 26 of political obligation that a significant number of the state’s constituents be obligated. Since Hart’s principle generates obligations when consent cannot, it stands a better chance of satisfying the success conditions on such principles. Second, it follows from Hart’s account that my obligation is not an instance of a more general duty not to harm others, or to refrain from reducing the benefits they receive relative to the burdens they bear. 36 It would, for instance, neither erase nor compensate for my obligation in Daycare if my child (who is older and more responsible than the others) made the chaperon’s task easier than it would have been had she not been there at all (say, by entertaining the other children). The special sort of obligation Hart picks out does not arise from the consequences of my actions in any particular case, whether by harming or increasing the burdens on others, but from the fact that I take advantage of them by riding free on their collective efforts. 37 This is the second advantage for the fairness theorist, who can explain why citizens have an obligation to obey the law even in cases in which breaking it would have essentially no repercussions. 38 Hart’s reliance on received benefits has, however, engendered serious problems for his account. As Robert Nozick observed in Anarchy, State and Utopia, Hart’s principle seems to empower individuals to obligate others simply by foisting benefits on them, even when they would prefer not to have those benefits at the cost. 39 This makes the principle an 36 Hart (1955, 183). For a competing view see Smith (1999, 81). 37 Simmons (2001, 33). 38 Indeed, some theorists who reject the fairness account as a complete theory of political obligation nonetheless endorse fairness constraints to secure this explanatory advantage in their own theory. See, for example, Wellman (2001), whose theory I discuss in detail in Chapter 3. 39 Nozick (1974, 94-96). 27 unattractive basis for a theory of political obligation because the central goods available in states are public. Public goods, as I will use the term, have two key characteristics: they are nonexcludable—meaning that they are impossible or prohibitively expensive to provide selectively—and unavoidable—meaning that they are impossible or at least unreasonably difficult for recipients to avoid. 40 Consider a simple case of a public good: Freeway Some friends and I decide to plant trees along major freeways in Los Angeles because we don’t like baking in the sun while stopped in traffic. You don’t like baking in the sun either, so you are happy to see the project develop. One day a member of our project comes to your neighborhood looking for volunteers. You politely decline. However the trees are eventually planted anyway, and you enjoy the shade they provide every day on your way to work. While it would have been generous of you to help with our tree-planting project, it seems unreasonable to suppose that by declining you violate a moral obligation you have to contribute. Public goods like those in Freeway are “thrust upon” their recipients without their approval, and in some cases even without their awareness. For this reason it has seemed plausible to many people (myself included) that the receipt of public goods lacks the binding 40 In economics a public good is a nonexcludable good which is at the same time nonrival, meaning that my consumption of it does not affect its availability to you. My use is therefore nonstandard in two ways. First, some central goods available in a state are indeed rival, such as, for example, local protective services and public health resources. Second, I include unavoidability as a characteristic of public goods. Because the receipt of some nonexcludable goods might generate obligations only because their recipient sought them out, it seems to me that, contra Klosko (1992, 36), it is unavoidability rather than nonexcludability that is the most important feature of public goods (a fact that seems implicit in some of his own examples—for example, (1992, 61 fn.39)). 28 force of the receipt of goods (like those in Daycare) that one elects to receive. 41 The central challenge for fairness theorists has therefore been to offer an account which explains why (i) I incur an obligation in cases like Daycare, and why (ii) you incur no obligation in cases like Freeway, but why (iii) each of us is nonetheless strapped with an obligation to obey the law simply for having received public goods generated by others’ cooperative obedience. 2.2 The Acceptance Approach 2.2.1 Acceptance One general strategy has been to either refine Hart’s principle (as Rawls 42 does) or interpret it (as Simmons 43 does) to include a condition on acceptance. On this conception, one must be a 41 This concern, which Klosko (1992, 37) refers to as the limiting argument, has been discussed extensively in the literature. See, for example, Nozick (1974); Dworkin (1986); Sartorious (1981). 42 Rawls (1964). 43 Simmons (1981, 120-23). I am skeptical of Simmons’ claim that Hart intended to offer an acceptance theory. Simmons’ primary reason for interpreting Hart this way seems to be that the more straightforward receipt-based interpretation would lead to highly implausible results—e.g., that Canadian citizens are bound to the United States’ political community (and visa versa) because they too benefit from U.S. citizens’ cooperative effort to maintain the rule of law (122). However even if the principle of charity would otherwise permit such a seemingly revisionary interpretation, acceptance does not eliminate the implausible implications Simmons highlights. For there is no reason from Hart’s perspective to think that only U.S. citizens can accept the benefits of law and order in their country. Simmons argues, of course, that in fact U.S. citizens do not accept them, but this is beside the point. For if Simmons interprets Hart correctly, then Hart held that they do. But if Hart is right that U.S. citizens accept these goods, then there is no reason that Canadian citizens cannot, for acceptance is, as I’ll explain below, just a matter of willingly and knowingly receiving a good. So if U.S. citizens do this from Hart’s perspective (on Simmons’ interpretation), then there is no reason to think that Canadian citizens can or do not. Thus, Simmons’ reason for rejecting the receipt-based interpretation of Hart’s position—that it leads to implausible results—is just as much a reason for rejecting Simmons’ interpretation. 29 participant in a cooperative scheme to incur obligations by receiving the goods it produces. 44 Most of the means by which one establishes one’s participant status are based in consent: one pledges support for, consents to be governed by, or plays an “active role” in the scheme in question. But one may also become a participant by accepting the goods produced by the cooperative scheme, which one can do without deliberately undertaking an obligation. In Daycare, for example, my effort to secure that cooperatively produced good sufficed to show, at least on an intuitive level, that I accepted it in spite of my explicit (and we can imagine, ongoing) refusal to deliberately undertake the obligations associated with the scheme. But how does acceptance explain the other two sorts of cases we mentioned? Again, the difficulty is to identify the relevant difference between goods of the sort you receive in Freeway, which seem to lack obligating force, and those provided by the cooperative efforts of members of political communities, the obligating force of which Hart and Rawls aim to account for. While Simmons accepts that Hart’s principle, grounded in acceptance rather than mere receipt, generates obligations of fairness to contribute to cooperative schemes, he goes on to argue that this more demanding conception makes fairness an implausible principle of political obligation. To have accepted a benefit, on Simmons’ view, one must have either (i) tried to get (and succeeded in getting) it, or (ii) received it willingly and knowingly. 45 It is the second condition that undermines the principle’s political application, on Simmons’ interpretation. 44 Simmons (1981, 119-42). 45 Simmons (1981, 129). 30 Apart from his empirical suggestion that western states are not the sorts of cooperative schemes to which questions of fair play are relevant, 46 Simmons contends that, even if they were, most people do not have the right attitude toward and beliefs about the benefits they unavoidably receive to be described as having received those benefits willingly and knowingly. Many people have not thought enough about the public goods they receive to have standing beliefs about them at all. Of those who have, most think of themselves as having purchased them directly from the government (through taxes), and many surely believe that the benefits are not worth the costs (though they continue to pay their taxes and obey the law to avoid being punished). 47 These people may be wrong about the source, nature, and even the value of the goods they receive, Simmons maintains, but they are not culpable 48 for their mistake, any more than one would be culpable for innocently drinking from a private water fountain which looked no different than a public one. Because a person cannot plausibly accept a good in a morally relevant way if she reasonably believed it to be free for the taking, citizens of states cannot plausibly be thought to accept and thus to be bound by the public goods they receive as a result of their compatriots’ cooperative behavior. To the extent that this is true, my receipt of the public good of domestic law and order will not obligate me to obey the law. Simmons’ impression of citizens’ beliefs is not “overly skeptical” in my view. 49 Nevertheless, his argument appears to have two weaknesses. First, though less importantly, 46 Simmons (2001, 39). 47 Simmons (1981, 138-39). 48 Simmons (2001, 32-33). 49 Arneson (1982, 633). 31 he seems to leave open the possibility that, if the government would only inform people that the benefits they enjoy are cooperatively produced, many of them would incur obligations. But while I think Simmons is right that the absence of nonnegligent ignorance is a necessary condition for obligation in fairness cases, it seems to me that the most central public goods satisfy it on the whole. It seems negligent not to know that one benefits from others’ obedience to law. Second, Simmons’ argument is empirical, which means that data may eventually reveal that he is simply mistaken, as some philosophers have already suggested. 50 Such a possibility represents a serious threat to Simmons’ anarchist position. For if he is right that accepting public goods amounts to receiving them while in a particular psychological state, then whether citizens accept them, and thus whether Simmons’ objections to the principle of fairness stand or fall, turns on contingent psychological facts. 51 I deny, however, that Simmons’ critical argument is all that prevents the acceptance theorist from successfully answering the applicability question. As I argue in the next section, there are deeper theoretical problems with the acceptance theorist’s answer to the validity question, and thus, to the applicability question as well. Because these problems go beyond the beliefs and attitudes people happen to have towards the goods they receive, they represent a greater challenge to the acceptance approach than Simmons’ argument does. By 50 See Gans (1992, 61-63) and Greenawalt (1987, 142). Song (unpublished) supports this objection by appeal to empirical data. 51 Were the data to reveal that citizens do accept public goods, Simmons would presumably fall back on the secondary argument I glossed above—viz., that the behavior of citizens is not genuinely cooperative, and so does not even represent the sort of scheme to which one could be obligated to contribute, for reasons of fairness, whatever one’s beliefs (2001, 39). I do not deal with this argument here, but note that Simmons’ overall position will be significantly weaker if it turned out that citizens’ attitudes towards public goods amount to acceptance. 32 extending our critique of acceptance from the applicability question to the validity question, we strengthen the anarchist position. 2.2.2 Willingly and Knowingly What does it mean to take a public good willingly and knowingly? 52 Although he addresses the question, Simmons’ conception of these conditions faces more serious difficulties than he or others have acknowledged. 53 The Willing Condition Without a condition on willing receipt, the acceptance approach would objectionably imply that you take advantage of individuals who provide you with unavoidable goods which you would reject if you had the choice. Think back to Freeway, only suppose that you hate the shade, and avoid it whenever possible. Clearly my friends and I cannot reasonably demand that you help us plant trees on the grounds that it is unfair for you to receive this good on our backs. After all, you would prefer not to receive it at all. We might suggest, then, that you receive a good willingly if you prefer having it to not having it. However this is overly inclusive. What seems relevant to considerations of fairness is not one’s preference for a good other things equal, but her preference in light of the burden she must bear to have it. You might have a slight preference for shade on the freeway, for instance, but if your fair share of the project requires that you plant trees eight 52 The willing condition (and possibly the knowledge condition) on the receipt of public goods is plausibly different than that on nonpublic ones. I focus here on public goods. 53 Simmons (1981, 128-133). 33 hours each week, then you would tolerate the sun. To demand that you help produce a good that you would prefer to forgo when bundled with its cost is clearly objectionable. 54 Simmons makes a different but related point when he says that, to have accepted a good willingly, we cannot “think that the benefits are not worth the price we must pay for them.” 55 Replacing the notion of preference with that of worth seems like a promising strategy initially; but in what sense must you think a good is worth its price to receive it willingly? One possibility is that you must think it is worth its price objectively. What this sense of worth amounts to is somewhat obscure, but it should be familiar. It is the sense a person uses when he says “X is worth Y” without at the same time being willing to pay Y for X. The problem with the objective interpretation is that the knowing receipt of a good with this sort of worth cannot plausibly ground an obligation. Consider this: I believe that some fancy wristwatches are worth more than $100, at least in this objective sense. I would never be willing to pay for one, however, because I don’t like watches. Now suppose that a local organization develops a cooperative scheme that, among other things, leaves me with just such a watch. Must I contribute something worth less than $100? They have provided me with a good, and I do believe that this good is worth more than what they have asked me to contribute, objectively speaking. Still, I would not be willing to pay this amount for this good. We can assume, in fact, that I would not be willing to pay anything for it. And so because, however objectively valuable I take the good to be, I would prefer not to spend my 54 This is one of the primary lessons from Nozick’s well-known examples (1974, 94). 55 Simmons (1981, 132). 34 time and effort to secure it for the price, it seems objectionable to demand that I contribute simply because I received it. 56 Perhaps, then, to take a good willingly it must be worth the price to the agent subjectively, by which I mean that she would agree to pay the price demanded for it were this the only way to get it. I take this to be Simmons’ considered view. But while this is surely more promising than the objective condition, the unfairness in question seems to derive any deontic force it has from the agreement I would be willing to make to get the good in question. That is, on this interpretation it is the fact that I would have agreed to pay the amount actually demanded by the good’s producers, had they been capable of controlling its distribution, that explains the wrongness of not paying it. Understood in these terms, the principle of fairness seems to collapse into a principle of hypothetical commitment. I will have more to say about the problems with this sort of condition when we discuss Klosko’s view in Section 2.3. I can point out here, however, that Hart, Rawls and Simmons would all presumably like to avoid this collapse, not least because one of the central advantages of the theory of fairness is that it does not depend on a deliberate undertaking, hypothetical or not. 56 Gans endorses a version of the objective interpretation according to which fairness generates obligations for goods that “people consider necessary from a moral point of view” (1992, 62). I find Gans’ view difficult to understand. The most coherent way to think of a good I receive as morally necessary (where this goes beyond mere objective value, or what it is good for me to have) is, I presume, to think of it as something it would be wrong of others not to give me. But the fact that I think it would be wrong of others not to give me some good does not even entail that I want the good, much less that I would willingly pay the price they demand for it. The difficulty of reconciling the “moral point of view” Gans seems to have in mind with the obligation-generating mechanism at work in fairness theories—namely, the agent’s receipt or acceptance of benefits—suggests that he has shifted his focus to a samaritan duty each of us has to provide goods for others. So far as this is right, Gans’ position diverges from what are traditionally thought of as fairness theories and anticipates views like the one offered by Christopher Wellman (2001), who, as I’ve already mentioned, utilizes considerations of fairness to justify the distribution of burdens that are necessary to supply the goods we owe to others for reasons that have nothing to do with fairness. I discuss this further in the Chapter 3. 35 Given our rejection of both the objective and subjective interpretation, it is difficult to see how a good’s worth to an agent, and thus her willing receipt of it, could bind her to contribute to its production. If this result seems counterintuitive—and it surely does at first—I suggest it is because we have failed to see that we ought to model our treatment of public goods on the (implicit) willing condition on nonpublic ones. Recall that in Daycare we have no reason to think that, from my perspective, the service my colleagues provide is (objectively) worth the time they spend to keep it up, or that I would have been willing to contribute to the project were contributing the only way to receive the benefit. Yet I still have an obligation to contribute to their scheme. It seems reasonable to me that we should think about nonpublic goods in the same way. If what obligates me in Daycare is the fact that I leave my kids at the daycare center—that I seek out the avoidable benefit my colleagues provide, irrespective of my beliefs about the value of their service—then perhaps to take a public good willingly it must be true that one would seek it out were it avoidable. 57 Interpreting the willing condition in terms of what one would seek out rather than what one would agree to pay is the only way I can see to maintain the principle’s obligation- 57 This condition may at first look no different than the subjective “worth it” condition. But while in many cases the goods one would seek out are the same goods one would be willing to pay one’s fair share to acquire, the two conditions come apart (as they do in Daycare, in fact). This happens most obviously when we consider goods which are nonexcludable but nonetheless avoidable—those a person might seek out in spite of the fact that she is unwilling to contribute to their production. So far as I can tell, Simmons has the subjective condition in mind, which he appears to extend to nonpublic goods as well (though this seems so clearly problematic that I suspect I’ve misread him). Thus he says that considerations of fairness concern one’s taking advantage of others, and that this happens “when one freely takes the benefits of cooperation with the requisite beliefs and preference structure, not when one merely unavoidably receives those benefits while going about one’s normally permissible business” (2001, 31; my underscore). So far as Simmons is describing nonpublic goods— which he seems to be—I do not see why preference structure matters. Even if I do not want daycare for my children, the fact that I freely (if irrationally) leave them at the daycare center obligates me to contribute. But then why think that to be bound to contribute to the production of public goods we must think that they are worth the price we must pay for them? 36 generating force without collapsing it into a principle of hypothetical commitment. It also has the general explanatory advantage of unifying the conditions of acceptance for public and nonpublic goods: for a nonpublic good (like that in Daycare) we ask whether I in fact sought it out, while for a public good (like that in Freeway) we ask whether I would seek it out were it avoidable. On all counts, then, this final interpretation appears to comport best with the spirit and force of Simmons’ interpretation of Hart’s principle. Sadly, though, even setting aside the practical difficulties of determining what a person would seek out in other circumstances, this account fails too. For it is difficult to see how the fact that you would seek a good out if my friends and I did not provide you with it entails that you owe us payment. The fact that you would have tinted your windows had we not planted trees, for instance, does not show that you owe us some contribution that you would not have owed otherwise. That this would be true even if you knew that your fair share of the tree-planting project cost less than window tinting demonstrates that the objection extends even to goods (such as personal protection) that you believe can be more cheaply provided by cooperative schemes (such as obedience to law) than by other means. Having exhausted the most plausible interpretations of Simmons’ willing condition, I do not see how it can play the significant theoretical role he claims it does. The Knowledge Condition The knowledge condition prevents you from incurring obligations for enjoying goods you nonnegligently believe to be free for the taking, or which you fail to recognize you receive. 58 58 See Simmons (2001, 32-33). 37 But one might wonder, as Klosko seems to, whether such a condition isn’t beside the point for public goods. 59 For if the point is that we can only be held responsible for consuming a good if our decision to take it was informed (or negligently uninformed), then the fact that we cannot choose to take public goods at all, ignorantly or otherwise, makes the condition irrelevant. Once we identify the sort of fairness Simmons is concerned about, however, the importance of the knowledge condition emerges. “Unfair self-selection,” as Simmons calls it, occurs when I unilaterally select myself to enjoy cooperatively produced benefits without bearing any of the productive burden. 60 This sort of unfairness explains the knowledge condition’s relevance in cases of public goods because one’s morally relevant choices extend beyond that of which goods one receives. While to unfairly select myself with respect to a nonpublic good I must choose to receive it by seeking it out (as I did in Daycare, for example), I can unfairly select myself with respect to a public good—one I had no choice but to receive—by choosing not to pay for it after I’ve willingly received it. In either case the condition’s role is pivotal. I do not unfairly select myself to receive a nonpublic good if I nonnegligently believe that the good is free for the taking. As with contracts or agreements, in pursuing a good I must be fully informed (or negligently uninformed) for my behavior to carry moral weight. For public goods, the less obvious point is that I cannot unfairly select myself to be a noncontributor if, through no fault of my own, I fail to recognize that the good was produced by a cooperative scheme which I’ve chosen not to contribute to. This is, as we 59 Klosko (1992, 52). 60 Simmons (2001, 30). 38 saw, precisely how Simmons argues against the political application of Hart’s principle. Even if obedience to law represents a genuine cooperative scheme, citizens cannot select themselves to be noncontributors if they are ignorant of it. 61 Supposing then that the knowledge condition is an integral component of acceptance: What must one know? Simmons’ initial answer is that “taking the benefits ‘knowingly’ seems to involve an understanding of the status of those benefits relative to the party providing them.” 62 However this is just another way of putting our question. What we want to know is what the status of a good is such that one’s nonnegligent ignorance of it prevents one from unfairly self-selecting. This initial way of putting things suggests that a good’s status is, for Simmons, more than the simple fact that it stands to some group of people in the produced by relation. It turns out, in fact, that it’s not, and that this is precisely the relation Simmons had in mind. (This condition is too weak, and I reject it below.) First, however, it’s important to see why no stronger condition can do the necessary work. Let’s return to the question: What must a person know when she receives a cooperatively produced good for her receipt of it to bind her to contribute? One possibility is the good’s deontic status—the fact that receiving the good obligates you to contribute your fair share to the scheme that produced it. The fact that many of the obligations we incur satisfy this constraint—promissory or contractual obligations, for example—provides this 61 Klosko may fail to see the relevance of Simmons’ knowledge condition because his own conception of the sort of fairness at issue diverges from Simmons’. While Simmons is concerned with how we choose to treat one another, fairness obligations are, for Klosko, “grounded on moral principles of the just distribution of benefits and burdens,” or the principle that “similar persons should be treated similarly” (1992, 54). However these ostensibly competing views may ultimately amount to the same thing, for Klosko’s conception reduces to Simmons’ if a “just distribution” of goods involves, as it would seem it must, the sort of acceptance Simmons endorses. 62 Simmons (1981, 132). 39 suggestion with an air of plausibility. On reflection, however, it is doubly problematic. First, it undermines an important distinction between fairness and consent accounts, and thereby eliminates an important advantage fairness offers as a basis for a theory of political obligation. As Simmons observes, “one can accept benefits without being aware of the moral consequences of so doing (while being unaware of the moral consequences of consenting defeats the claim that consent was given).” 63 If you needed to know the deontic status of a good for your receipt of it to bind you, then a receipt based fairness account would begin to look a lot like a consent account; receiving the good would be tantamount to acknowledging the obligation you thereby incurred. Second, in order for one to know that by receiving a good one thereby incurs an obligation, it must be true that one does. But whether it is true depends circularly on whether one knows it. Imagine that I am an acceptance theorist who wants to honor my obligations. Each time I receive a good I ask myself whether I’ve accepted it, for whether I have determines whether I incur an obligation, and thus whether I should contribute. By hypothesis, I have accepted it only if I received it knowingly, which I did only if I knew when I received it that I thereby incurred an obligation. But whether I incurred an obligation, and therefore whether I could know that I did, depends on whether I accepted it. But that is the very thing I’m trying to figure out! One might therefore weaken the constraint to say that what one must know is that by receiving a cooperatively produced good without contributing one takes advantage of its providers. (This occasionally seems to be what Simmons has in mind. 64 ) However this is no 63 Simmons (1981, 117). 64 Simmons (2001, 33). 40 better than the previous suggestion. First, I cannot plausibly know that I take advantage of you by doing something without knowing that (other things equal) I wrong you by doing it. And so this condition also conflicts with Simmons’ claim that acceptance does not require an awareness of the moral consequences of so doing. Second, as before, to know that I take advantage of you by acting in some way it must be true that I do. And so because whether I take advantage of you by receiving a good you produced depends on whether I accepted it, the circularity remains. In case this second possible account of the knowledge condition still seems appealing, I’ll add that whether by receiving a good produced by some cooperative scheme I take advantage of its members, or incur and violate an obligation I have to them, seems to depend on my preexisting obligations. This is suggested by the fact that in these cases the victim has the power to waive the obligation in advance. For instance, I seem clearly to take advantage of the actors at the local theater by attending their performance without paying. However this doesn’t follow if they invited me as their guest. For by inviting me they waive the claim they already had on me not to attend without paying. This indicates that the two conditions we’ve so far considered reverse the order of explanation. It is not that I incur an obligation because I know that I take advantage of a good’s providers by receiving their product without compensating them. Rather, I take advantage of them because I owe it to them not to take the good they’ve produced without compensating them. But if I already have an obligation to either reject a good or contribute to the scheme that produced it, then it is hard to see what we need the acceptance approach to explain. 65 65 I do not mean to imply that I cannot take advantage of you unless you can waive the obligation I owe you not to perform the offending action. There are arguably some ways of taking advantage of 41 The circularity of the two previous suggestions indicates that what one must know is a proposition whose truth does not depend on what one knows or ought to know about her deontic circumstances. Presumably this is one reason why Simmons did not adopt them. A final suggestion, then, and the one I think aligns best with Simmons’ considered view, is that one must know merely that others have cooperated to produce the good, and that they expect payment for their service. 66 This interpretation avoids the previous objections because, as a purely descriptive claim, its truth depends on neither one’s acceptance of a good nor the existence of a preexisting normative relation. However this account plausibly throws the baby out with the bathwater. For absent a preexisting obligation, it is difficult to see how satisfying this descriptive condition could ground one. Returning to Freeway, let us stipulate that you welcome the shade in whatever way is necessary to satisfy the willing condition. Is what determines whether or not you have an obligation to contribute simply whether you know that my friends and I have worked together to plant the trees? This seems, to me anyway, highly implausible. There seem to be at least some benefits that cannot bind you, whatever your attitude towards them, simply in virtue of the fact that you unavoidably receive them. (I will say more about this in Section 2.3.3.) So far as this is true, the fact that we expect you to contribute is irrelevant. For what people expect or demand of others has normative force only insofar as those expectations are justified, and the fact that you that you do not have the authority to permit, slavery being the obvious example. But cases in which you lack the authority to permit some self-regarding action are invariably those in which the obligation others owe you to treat you in a particular way is so important that even you cannot permit it. So in these cases more than in the others it seems that the obligation must exist for the charge of advantage taking to even make sense. 66 The first condition Simmons more or less states at (1981, 132); for the second see (2001, 32-33). 42 some people have cooperated to produce a good that another person enjoys does not show that they are, even if the recipient understands the nature of their productive scheme. 2.2.3 Conclusion At least as far as public goods are concerned, acceptance appears to be a less promising approach to the principle of fairness than its proponents have thought. On our best conception, an individual A accepts a public good G in a morally relevant way just in case (i) A receives G, (ii) A would seek G out if G were avoidable (the willing condition), and (iii) A knows, or can be held responsible for not knowing, some (yet unspecified) fact about the cooperative scheme responsible for G’s production (the knowledge condition). As we saw, conditions (ii) and (iii) (or something like them) are necessary for the acceptance interpretation to be at all plausible, but there seems to be no conception of these conditions such that a person who satisfies them incurs an obligation for receiving a good. Though I provided what I take to be the most plausible conception of the willing condition, the fact that I would seek a good out were it not provided by a cooperative scheme does not show that I owe some contribution to the scheme’s contributors for actually receiving it. Even if the good is beneficial for me, and I want it, without the cooperative scheme I might provide it differently, or more cheaply, or I might change my behavior so that another good is of more use to me and produce that. But even if I would simply take it, that is not what I do in the actual world-state. If the good is unavoidable, then facts about what I would do counterfactually seem incapable of justifying others’ expectations that I contribute. But even if the willing condition has greater binding force than I recognize, or a more plausible formulation than the one I’ve given it, of even greater concern is the 43 knowledge condition, which seems to have no plausible formulation at all. Its purpose, again, is to prevent obligations in cases of non-negligent ignorance. What remains unclear, however, is what one could be negligently ignorant of. One the one hand, demanding that I know the moral consequences of my actions—that I incur an obligation or take advantage of others by receiving a good—eliminates the distinction between fairness and commitment, and the circularity of this condition suggests that any obligation which emerges in such a case must depend on a preexisting one. On the other hand, modifying the condition so that I must merely understand the nature of the productive scheme generates obligations where there are none, as it did in our modified version of Freeway. There seem to be at least some goods which, though I understand their origin and have certain attitudes toward them, others are not morally justified in demanding that I help provide. While some may take all of this to imply simply that the acceptance approach needs clarification, it seems to me that, as Simmons hints at one point, “the principle’s force might in fact be captured by more basic principles grounding the rights and duties concerned with (collective) property.” 67 The existence of such principles seems likely if, as I argued earlier, the explanatory power of acceptance extends only to those cases in which obligations generated by other principles are presupposed. In Section 2.3.3 I argue that, besides the conceptual problems I raise here, there are clear cases in which acceptance does not entail 67 Simmons (2001, 29). Klosko too says multiple times that a noncontributor’s seeking out a cooperatively produced benefit could be viewed as “theft” (1992, 51, 146). Indeed, it seems to me that Klosko’s application of his principle to excludable goods (1992, 36)—which entails, among other things, that producers “would be justified in refusing to provide benefits” to noncooperators—is maximally coherent only against the backdrop of (something like) Lockean property rights. For a consumer’s fairness obligation to contribute to a productive scheme does not always or obviously entail a producer’s liberty to prevent her from consuming its fruits, at least not as obviously as property rights do. 44 obligations. These too suggest that acceptance relies for its deontic force on more fundamental principles, perhaps including property relations. 2.3. The Goods Approach In light of the challenges facing the acceptance approach—the worries I’ve recounted concerning its central principle, as well as Simmons’ objections to its political application—it may be more promising to modify Hart’s receipt-based principle in a way that doesn’t involve acceptance at all. The main advocate of this strategy has been George Klosko, who offers what I call a characteristics of the goods (or simply a goods) approach. 68 Klosko’s argument has evolved in important ways over the years. Most notably, he now offers a “multiple principle” theory according to which considerations beyond fairness are necessary to explain our obligation to support the full range of state functions (a condition he calls comprehensiveness). However, Klosko argues that without the principle of fairness any theory of political obligation will fall short of generality—that is, it will not explain why nearly everyone in the state has an obligation to obey the law. It follows from this argument that if Klosko’s principle of fairness cannot ground obligations, then the multiple principle theory will fail as well. Below I try to show that this is precisely what happens. 68 For the most recent and comprehensive presentation of Klosko’s view, which includes the multiple principle argument I mention below, see Klosko (2005). Because Klosko’s discussion of the principle of fairness is more elaborate and, in my view, clearer in his 1992 book, I draw most often on that work, in addition to many of his papers, while reconstructing and responding to his view. 45 2.3.1 Klosko’s Principle of Fairness Klosko accepts that I can incur an obligation to a nonpublic scheme (like that in Daycare) by seeking out the good its members produce. 69 But the more difficult question again concerns public goods (like those in Freeway). According to Klosko, acceptance would only redundantly ground obligations in cases in which the public goods one receives have the characteristics listed in his central theoretical principle: Fairness A person S is obligated to contribute her fair share C to any scheme from which she receives benefit B if all of the following conditions are satisfied: EF B is worth C to S. AL B is indispensible to S’s living an acceptable or satisfactory life. RFS It would not be possible for S to supply B either on her own or by coordinating with another group. 69 Klosko (1992, 41). However he later indicates (1992, 48) that the reason I incur an obligation for seeking out nonpublic goods is that my behavior indicates that I take the good to be worth the cost of participation. Were Klosko right, then his theory would, like the acceptance theory with the conception of the willing condition I proposed above, unify the conditions of obligation for public and nonpublic goods. I believe, however, that Klosko unifies these conditions in the wrong direction. Whether I incur an obligation in Daycare, for instance, has nothing to do with my evaluation of the value of that service. What matters is that I use it. Therefore it seems better to adopt conditions on public goods that match those we adopt for nonpublic ones, our judgments about whose cases are less controversial, rather than the other way around. 46 BN The benefits and burdens associated with B are fairly distributed. 70 Although, as Klosko says, the idea of an acceptable or satisfactory life is vague, some goods are undoubtedly necessary to have one, such as food and water, shelter, relative security, and basic medical care. But his conception goes far beyond mere subsistence. I cite Klosko’s characterization in full, as the details will be important later: By ‘acceptable’ lives I mean lives in modern industrial societies, as we know them. These societies are relatively safe, have functioning economies, and allow travel and a wide range of occupations, activities, and modes of life. ‘Acceptable’ lives are led by people who are integrated into such societies and take advantage of the amenities they provide. 71 Taken together with Fairness, Klosko’s conception of a satisfactory life explains how you could have an obligation to obey the law without having an obligation to help me and my friends plant trees in Freeway. Even if we suppose that the shade they provide would be worth the sacrifice to you (EF), and that each volunteer’s workload would be fairly assigned according to the number of hours she would spend in the shade (BN), the benefit our scheme provides nonetheless fails to satisfy AL and RFS. Because Klosko believes that “in recent years, the main objection to political obligations under the principle of fairness has concerned [RFS],” 72 one of his central aims 70 I have divided Klosko’s principles differently than he does. AL and RFS are adapted from principles “EI” and “RSA” in Klosko (2004, 146). I rename them to indicate that their content has been updated to match Condition 2 from Klosko’s most recent and comprehensive presentation of his principle in Klosko (2005, 6, 19). 71 Klosko (2005, 19). 72 Klosko (2004, 147). 47 has been to show that indispensible state-provided benefits—national defense, domestic law and order, a legal system, environmental regulations, and major public health services— satisfy it. 73 For our purposes, however, I am happy to grant that Klosko’s argument for RFS is sound, as well as that contemporary states satisfy BN (though I believe this is highly implausible 74 ). I shall focus instead on AL and EF. 2.3.2 Objective and Subjective Interpretations Klosko presents his argument in more or less the same way in his various books and articles, and in each case it exhibits a tension between an objective and a subjective interpretation of Fairness—his central principle. Though he says explicitly at one point that “benefit is a subjective notion,” 75 it is unclear just what role a noncooperator’s beliefs are supposed to play. Simmons’ position is, as we saw, expressly subjective: for a good I receive to bind me I must believe that it is worth the price demanded for it (at least in some sense), and I must believe, or it must be unreasonable for me not to believe, that the good was produced in a particular way. In claiming that his view avoids the problems philosophers such as Simmons have raised for the political application of the acceptance approach, Klosko implies that his view diverges from Simmons’ subjectivism. As we shall see, however, his view is at odds with an objective interpretation as well. Below I argue that both interpretations fail, and that, 73 Klosko’s approach is to consider and reject the most plausible non-state strategies for providing these goods, such as Michael Taylor’s (1987) analysis of N-person prisoner’s dilemma supergames, as well as the libertarian arguments for robust non-state entities advanced by Nozick (1974) and Friedman (1973). 74 See Carter (2000). 75 Klosko (1992, 51). 48 therefore, whichever interpretation he endorses, Klosko cannot ground our political obligations in the goods we receive in political communities. The Objective Interpretation Gregory Kavka recommends an objective conception of the fairness principle in his 1983 review of Simmons’ Moral Principles. “It seems more reasonable,” he says, “…to say that a participant is bound if the benefits actually exceed the burdens (for him) whether or not he agrees with this accounting or realizes from whence the benefits flow.” 76 Klosko’s language echoes Kavka’s when he says, for example, that certain indispensible public goods “can be presumed to be beneficial without taking the recipients’ views into account.” 77 But in spite of these deviations, I do not think Klosko means to offer an objective interpretation. Still, we’ll consider the possibility of such a view. On the objective conception of Fairness the satisfaction conditions for EF and AL are determined independently of the recipient’s beliefs, desires, and preferences. On this account a good is indispensible for you if without it your life would be objectively unacceptable, and it is worth its cost if there is an accurate objective metric according to which it is. To see why receiving a good satisfying this conception cannot plausibly generate obligations, consider a functioning economy of the sort we have in modern industrial societies, which, as I noted earlier, Klosko highlights as a good which is indispensible to an 76 Kavka (1983, 228). 77 Klosko (1992, 51). Klosko’s discussion of paternalistic laws (1992, 60 fn.36) similarly suggests that on his view an agent’s beliefs do not determine a benefit’s worth relative to her. 49 acceptable life. Assume that this economy (call it E) satisfies the four conditions of Fairness, interpreted objectively, with respect to Fred, a farmer who is wholly self-sufficient and would prefer that E did not exist at all. One might think that, while Fred does not take advantage of the range of opportunities E offers, the volatile life he leads in the mountains would be substandard without at least access to E, that, like everyone else, Fred genuinely benefits from the safety net a functioning economy provides. The question, however, is whether by refusing to bear some share of the burden of supporting E—say, paying taxes that support regulative bodies and the like—Fred behaves unfairly. It seems plain to me that he does not. For how can a person be said to take advantage of those who provide him with a good that he would genuinely prefer not to receive? The fact that very few people share Fred’s perspective is beside the point, which is that a person’s attitudes towards goods appear integral to the obligations they incur by receiving them. However you may still disagree. Fred benefits, you say, and so he is obligated despite his preferences. He is simply a fool, and fools, like everyone else, must contribute to the cooperative schemes from which they benefit even if they don’t realize or agree that they benefit from them. There are two problems with this objection. First, Fred’s preferences need not depend on ignorance for the objection to apply. On the objective conception the agent’s preferences do not matter at all, even in determining what goods are valuable or what counts as a worthwhile life. And so to evaluate this conception we are free to suppose that Fred decided to forgo E under even ideal epistemic circumstances. 78 Second, if considerations of fairness bind Fred in this case, then there is no reason to think that 78 Views that incorporate Fred’s preferences in any form in determining the binding force of <good, cost> pairs fall under the subjective interpretation below. 50 someone could not similarly bind you to contribute some fee by giving you something of objective value—clean air, e.g.—even if you did not want or appreciate it. The person working in the adjacent cubical purchased an air purifier months ago, suppose, and has finally made it over to your cubical to collect your contribution. Must you pay him? Klosko would likely point out that the slight improvement in the air quality does not satisfy AL, and that it therefore falls outside the scope of his principle. But it is hard to see why, if the receipt of something objectively valuable binds in some cases, it should not bind in all of them. Klosko provides no reason for endorsing AL on the objective picture. In fact, one begins to suspect that its sole function is to exclude goods which, like the goods in Nozick’s examples, provide the clearest counterexamples to Klosko’s view. But rather than accept what appears to be an unprincipled application of a binary condition (you have an obligation or you don’t) to a scalar relation (both goods are objectively valuable, but one is more valuable than the other), we should, until an acceptable argument is made, suppose that the apparent correlation between the indispensability of the received good and its obligatory force is explained by another (plausibly subjective) principle. If receiving an objectively valuable (though not indispensible) good cannot ground a fairness obligation for you, then there is no principled reason to think that the objective value of the good Fred receives grounds an obligation for him. And if this is true then it is implausible to think that it would ground an obligation for anyone else, even if their preference is to have the good in question. 79 79 I should note that rejecting the objective interpretation of Fairness does not entail that public goods should not be provided. Nor does it delegitimize all paternalistic laws (see Klosko (1992, 60 fn.36)). It entails only that a good’s allegedly objective value cannot explain the obligation one incurs 51 The Subjective Interpretation Klosko seems at times to acknowledge the failure of the objective interpretation. (“If someone finds life in a Hobbesian state of nature acceptable,” he says, “she will not choose to enter civil society and so might not have political obligations.” 80 ) On the subjective interpretation of Fairness, EF and AL are determined by the recipient’s beliefs, desires, and preferences. A good is indispensible if without it you could not lead a life that you find satisfactory, and it is worth its cost if you believe it is. 81 The subjective interpretation of Klosko’s principle, as I understand it, subtly incorporates acceptance in order to capture the binding force absent in the objective receipt- based interpretation. His argument proceeds as follows. 82 The most important goods the state provides are indispensible to the sorts of lives most of us find satisfactory (AL), and so to the extent that to secure any particular good one does not have to give up another good that is more indispensible, it will be worth one’s effort to help provide it (EF). Because the costs the state imposes to provide these goods—obeying the law, paying one’s taxes—do not amount to a forfeiture of indispensible goods, and because without the state these goods by receiving it. To the extent noncontributors have obligations to cooperate, these must be justified another way. 80 Klosko (2004, 19). To be clear, she would avoid obligations because of her preferences, not because she did not enter civil society, which may not have been an option. 81 This subjectivity can be incorporated on different levels of the theory. For example, it might be necessary only that a good is indispensible for a life the agent finds acceptable, but it might also be necessary that she realize that it is. My arguments below apply equally to any version of the subjective interpretation. 82 This line of reasoning, in nearly identical form, can be found in Klosko (1992, 41-42), (2004, 147) and (2005, 6-7, 19). 52 would not be provided at all (RFS), one could not but and thus would (on pain of irrationality) pursue and thus accept any one of them at this price were doing so the only way to receive it. Therefore, so far as the benefits and burdens associated with these goods and their production are fairly distributed (BN) (which we can assume arguendo), those of us living in states where these benefits are provided have no justification for failing to contribute our fair share for their provision. We are therefore obligated to contribute. For to do otherwise would be to take unfair advantage of sacrifices others make to produce benefits without which each of our lives would be, even by our own standards, unsatisfactory. 2.3.3 Objections to the Subjective Interpretation If Klosko’s theory succeeds then there are at least two ways in which receipt of a benefit can generate an obligation. The first is if she accepted the benefit by pursuing it, as we saw in Daycare. The second is if the benefit she received has the characteristics specified by Fairness. The connection between these two obligation-mechanisms is integral to Klosko’s fair play argument, for we can understand the second only by reference to the first. With respect to public goods, such as national defense …it is not clear how [a man, Grey] could pursue them even if he wished to. Because the benefits of national defense are indispensable, we can presume that Grey would pursue them (and bear the associated costs) if this were necessary for their receipt. If we imagine an artificial choice situation analogous to a state of nature or Rawls’s original position, it seems clear that under almost all circumstances Grey would choose to receive the benefits at the prescribed cost, if he had the choice. 83 83 Klosko (2005, 7; original emphasis). Klosko makes the same point at (1992, 41-42) as well as (2004, 147). 53 Anticipating a likely objection, Klosko stresses that Grey’s obligation does not stem from hypothetical consent, but from the fact that he receives the benefits. But while I think he is right that obligations of fairness do not collapse into obligations of commitment—as I explained in Section 2.1, obligations of commitment depend on a deliberate undertaking (consent, a promise, an agreement), while considerations of fairness obligate those who unfairly free-ride on the labor of others—I am not convinced that Klosko’s view does not depend on hypothetical commitment. 84 For while it seems clear that we can incur obligations by pursuing and thereby accepting nonpublic goods without agreeing to contribute to their production, and in fact even while explicitly refusing to contribute, we saw in Section 2.2 that it is unclear just what acceptance amounts to for public ones. Things only become less clear when the resulting obligation is purportedly justified by the fact that one would accept them, though one has not actually. For what can it mean that one “would pursue them (and bear the associated costs) if this were necessary for their receipt” but that one would agree to undertake the cost in question to receive the goods? If it does not mean or imply this, then presumably we should be able to imagine a case in which an agent incurs an obligation by receiving a good which she would not accept for the price. However, because endorsing such an example amounts to endorsing the objective interpretation we rejected in the last section, it seems that the subjective interpretation must be more closely related to hypothetical commitment than Klosko concedes. Whatever this concern amounts to, however, I believe the objections below are sufficient to make one reasonably skeptical of Klosko’s theory, so I set this line of objection aside. 84 Also see Simmons (1981, 115-17). 54 It is important to see that whether a person “could not but” accept a good, or whether it “must” be provided, 85 as Klosko often says, can be determined only relative to an unacceptable state of affairs that would follow from rejecting it. The man at the ATM cannot but hand his wallet to the gunman if he wants to live. The student must take the final exam if she wants to pass the class. As we saw, the objective approach to establishing a baseline for these unacceptable states of affairs does not explain how the corresponding benefits generate obligations. This is, I think, precisely why Klosko construes AL, EF and RFS in the way I’ve suggested he does. The recipient of some good is immune to obligations, according to Klosko, whenever she can legitimately claim she would prefer not to have the good at the specified price. 86 But she can only legitimately make such a claim if the good in question is not indispensible to the sort of life she views as acceptable (¬AL), is not worth its cost to her (¬EF), or if she could provide the good by some alternative means (¬RFS). Thus the conditions of Fairness determine whether an agent could reject a particular benefit given the living standards she has set for herself, taking her alternatives into account. 87 If she could not then we may suppose that she would accept it, in which case Klosko contends that her obligation is based in receipt alone. 85 Klosko (1992, 43). 86 I take this to be an implication of Klosko’s (1992, 44) response to Nozick: “If the members of the scheme demand that she contribute, she can respond that she would prefer not receiving the benefits to having to make the specific contribution they demand. Because the benefits are relatively insignificant, it would be difficult for the scheme-members to demonstrate that generalizing the advantages of noncooperation would lead to consequences the noncooperator could not accept.” 87 Thus: “One rule of thumb for determining if fairness obligations are established in a particular case is to ask if [the recipient] would be willing to generalize her own noncontribution. If this would mean that particular indispensable goods were not supplied, then she could not allow general abstention from contributing” (2004, 8; my emphasis). Also see Klosko (1992, 42). 55 Klosko’s indirect reliance on acceptance shows that his theory more closely resembles Simmons’ than its structure first indicates. (A fact not lost on Simmons. 88 ) However the two theories are importantly different. Klosko’s emphasis on what one would rather than what one does accept allows him to avoid Simmons’ most forceful objections to the fairness principle’s political application. While it may be, as Simmons claims, that most people do not accept the public goods they enjoy (due to their nonnegligent ignorance, for example), Klosko is confident that, given their preferences, the vast majority of people could not but and thus would accept these goods at the price demanded for them were this the only way to secure them. In Section 2.2 I argued that the acceptance of public goods is less well understood than most philosophers acknowledge. While I believe that Simmons is right to think that a satisfactory theory of acceptance must include conditions on willing and knowing, our unsuccessful attempt to provide a plausible analysis of those conditions suggests that another principle likely underlies their ostensible deontic force. If Klosko also relies on acceptance, then these gaps are as much a problem for his answer to the validity question as they are for Simmons’. However, even if we suppose that we have a satisfactory conception of acceptance—one that corresponds with our general intuitions about what it means to willingly and knowingly receive something, based on the clearer example of pursuing nonpublic goods—Klosko’s theory faces at least three additional problems. 88 Simmons (1987, 273). 56 Actual Acceptance Does Not Guarantee Obligation Because obligations generated by Fairness ultimately depend on the fact that the recipient would accept goods satisfying its conditions, were that the only way for her to secure them, it is important to ask whether cases of acceptance always obligate. For if a benefit can actually be accepted without the recipient incurring an obligation then the fact that the recipient would accept it entails nothing. The clearest sort of case to consider involves a good that is nonexcludable but avoidable: 89 Orchestra My neighbor Brian hosts a group of jazz musicians on Friday nights. Every musician prefers listening to playing, but playing to no music at all, so each week the group rotates who plays and who listens. The wall separating our apartments is not thick enough to absorb the sound, so I hear the music in my apartment more or less as clearly as they do in Brian’s. Lucky for both of us, I really enjoy what they play. In fact, sometimes I skip other events because I would prefer to stay home to listen. Let’s assume that it is impossible or prohibitively expensive for Brian and his friends to control who hears their music. Still, their good (call it G) is not public because I can still coherently seek it out, which I do when I either stay home to listen to them play when I would otherwise go out, or come home when I would otherwise stay out. It follows from 89 The following case closely resembles Nozick’s well-known example of the public address system. I use a fresh case because people familiar with Nozick’s may, intentionally or not, build important features into it which Nozick neither intended nor explicitly excluded. 57 this, and from the fact that that I understand the scheme Brian and his friends have organized to produce G, that I satisfy our presumed conception of the willing and knowing conditions and thus accept the good they provide. Assuming, then, that I am a musician capable of contributing, and that Brian and the others would like to reduce the load on all of the participants by having more musicians participate in their scheme, does the fact that I accept G entail that I have an obligation to contribute? There are at least two reasons to think that it does not. The first concern resembles Nozick’s when he says that you “may not decide to give me something…and then grab money from me to pay for it, even if I have nothing better to spend the money on. […] One cannot, whatever one’s purposes, just act so as to give people benefits and then demand (or seize) payment.” 90 Of course, our case is slightly different. Because G is sometimes avoidable, the relevant circumstances for us are those in which I choose to be home to enjoy the good Brian and his friends produce. Unlike Nozick’s characters, then, they do not simply give me something and then demand payment. Thus one might argue that I am responsible in these cases and so owe them compensation because I unfairly select myself as a noncontributor. While I agree that Orchestra is different than Nozick’s case in this respect, I think the spirit of his concern nonetheless applies. Brian’s demand that I contribute to their scheme is unwarranted, not because G is unavoidable, but because Brian and his friends have unilaterally modified my environment—one I have every right to occupy—so as to make actions which were once permissible (coming home on Friday evenings, in this case) 90 Nozick (1974, 95). 58 impermissible absent payment. What is objectionable, in other words, is that I owe others something for doing what I used to be able to do for free, that by behaving as I used to they acquire the standing to demand payment from me. Of course, this is too general a way of putting things. You may permit me to pass across your land today while justifiably forbidding it tomorrow. But the point to notice is that cases like this invariably involve claims we already have on one another. My neighbors know, for instance, that when my gate is open they are free to pick apples from my tree, but that when it is closed I expect them to help rake my yard if they pick apples. By closing the gate, I unilaterally modify their environment so as to make their previously permissible behavior—picking apples in my yard without raking—impermissible. But in this case closing the gate does not explain why I can justifiably demand that they help me rake if they pick apples. It is rather a way of expressing an expectation that is justified independently. 91 Fairness, so far as it generates the obligations Klosko says it does, functions altogether differently. Because the claims Fairness generates for the providers of goods does not depend on any preexisting rights or obligations (such as property rights and their correlative obligations), anyone can unilaterally create circumstances in which others cannot perform actions they once could without owing others. In Orchestra, I cannot come home as I once could without owing it to Brian and his friends to contribute to their scheme. On Klosko’s theory my preference structure gives them the standing to demand my cooperation if I do. 91 Some cases are obviously more complicated, but the general principle is the same. For instance, I have no right to stop you from picking apples from trees on some parcel of abandoned land. But I can acquire that right by purchasing the land, thereby unilaterally modifying the environment to make your previously permissible behavior impermissible. However this case is merely a more complex version of the case above. For here the right I already have is the right to acquire the right to forbid you from picking those apples. 59 So while Brian and his friends do not “act so as to give [me] benefits and then demand (or seize) payment,” as Nozick says, a consequence of their scheme is that I am no longer free to come or stay home when, given my preference structure, I would otherwise stay or go out were they not playing, and as a result they are justified in demanding payment when I do. This view is objectionable because my coming or staying home seems justifiable whatever goods I receive as a result. Though in Orchestra I pursue and thus accept G, there are other considerations in light of which I nonetheless have no obligation to pay for it, namely, that I am justified in being in my home whenever I’d like to be. The important and overlooked general principle to which this case points is that in some circumstances no good qualifies as “reasonably avoidable,” where that condition is used, as I explained before, to determine whether someone pursues, and thus accepts, and is thus bound by a good he receives. Whether a good is reasonably avoidable determines, not whether it is pursued and thus accepted, but whether one’s acceptance of it generates obligations. What I am arguing, then, is that in cases like Orchestra—those in which one is justified in occupying a particular space or behaving in this or that way—goods one receives will categorically fail to bind, for it will never be reasonable to expect one to avoid them. In these cases their producers must decide whether they are worth their effort to produce given that they cannot justifiably demand payment from those whose enjoyment constitutes nothing more than a positive externality. Here the connection between Orchestra and the problem of political obligation emerges. Just as in Orchestra I have a right to be home when I would like to be, each of us has a right to exist somewhere. Together with the fact that it is unreasonable to expect one to emigrate to avoid his state’s requirements, this indicates that the unavoidable goods provided 60 in states cannot bind one to obey the law any more then Brian and his friends can bind me by playing pleasant music. To generate this sort of obligation, the state (or the aggregate of its members) would need to have a prior disjunctive claim on individuals within its territory to either emigrate or else to pay for the unavoidable public goods they receive. However this creates a dilemma. For if the state has this prior claim, then it must be based in something other than fairness (at least as it relates to benefits received), which depends on it. But if it lacks this claim, then its demands that citizens “love it (and pay) or leave it” are unjustified. 92 This leads to my second concern. Everyone agrees (though for different reasons) that I can only accept a nonpublic good if I pursue it. It is agreed, for instance, that I owe no contribution to Brian’s scheme if I come home to make dinner only to find that Brian and his friends are playing music that I happen to enjoy. For many philosophers, however— including both Simmons and Klosko—things change if I come home in order to listen to their music. For they argue that in this case I seek out and thus accept G, and therefore incur an obligation to contribute. On this view, then, my motivating reasons determine (i) whether I seek a good out, and thus (ii) whether I accept it, and thus (iii) whether I thereby incur an obligation, and thus (iv) whether ceteris paribus it would be impermissible not to contribute. What it is worth considering is whether my reasons for performing an action can affect its permissibility, as this view assumes. Some philosophers maintain that they cannot. According to T.M. Scanlon, for instance, what makes an action permissible or impermissible are considerations which count decisively in favor of or against it. 93 Are there considerations 92 Wellman (2001, 736) advances a similar dilemma against the consent theorist. 93 Scanlon (2008). However note that my purpose in highlighting this concern is slightly different than Scanlon’s. Very roughly, Scanlon’s primary aim is to show that doing the right thing for the 61 that count decisively in favor of or against my going home and listening to Brian and his friends play music if I do not intend to contribute? As I’ve argued already, the fact that it is my home must surely count in favor of that action’s permissibility, so much so, in fact, that it is unreasonable to expect me not to perform it in order to avoid any attendant benefits. What counts against it? If Scanlon is right, then anything that does must do so independently of whether I go home with the guiding intention of listening to Brian and his friends play music or doing something else. Since the only reason I seem to have not to go home if I’m not going to contribute depends on my intentions—the fact that by doing so in order to hear them play entails that I accept the benefit, and thus that I unfairly self-select—Scanlon’s thesis entails that I have no reasons not to go home. And so even if I am wrong about the strength of the considerations that weigh in favor of my action’s permissibility, if Scanlon is right, then none weigh against them, further strengthening the case that I can accept some goods without incurring an obligation to help provide them. But if this is on the right track, then one might wonder how we explain the normative tension implicit in the cases we have considered. For it seems that, even if I do not incur an obligation to pay for the goods I receive through others’ efforts, my deontic circumstances after having sought a good out must be different than if I had received nothing at all. Here I think Scanlon’s view provides the basis for an attractive answer. It may be that my reasons for performing an action that affects another person, while not affecting its permissibility, nonetheless determine its meaning—the “significance that person has reason to assign to it, given the reasons for which it was performed and the person’s relation” to wrong reason doesn’t make it the wrong thing, while I am suggesting more generally that one’s intentions in performing an action cannot determine whether that action is right or wrong. 62 me. 94 If, for instance, after promising to attend your wedding I renege because the stars of my favorite soap opera will be touring through my town, my behavior says something negative about our relationship even if it is permissible because my doctor recently informed me that I have cancer, and strongly advised against traveling. 95 What we are interested in, however, is how the meaning of one’s previous actions can affect the appropriateness of future actions. For instance, while my promissory obligation to attend your wedding was no doubt defeated by the fact attending would have aggravated my newly discovered cancer, and while, therefore, not attending was permissible, it may nonetheless be appropriate to admit and apologize for what little weight I gave to your feelings in my deliberation. Perhaps my behavior makes sending you an apology appropriate in a way that it would not have been had I skipped your wedding out of concern for my health. What is important to see is how this relates to Orchestra. The fact that on some occasions I sought G out may make certain future actions particularly appropriate, of which contributing something to Brian’s scheme may be one. Just as my reasons for not attending your wedding made certain conciliatory actions appropriate despite the fact that my behavior was perfectly permissible given the facts that counted in favor of it, contributing to Brian’s scheme in the future, or at least letting him know how much I appreciate the music he and his friends produce, may be appropriate in virtue of my pursuit of that good even if it is not obligatory. Just as the meaning of my action says something about me and my relationship with Brian, how I respond in light of the benefits he and his friends provide may also signify 94 Scanlon (2008, 54). 95 Scanlon (2008, 66). 63 something about these things. While Brian is not justified in demanding that I contribute to their project, offering to contribute would say something positive about me. So far as this is true, it explains why claiming that no obligation is generated in some cases of acceptance does not preclude the possibility that some normative relation is thereby established, and so it explains how one can argue against the acceptance approach, as I have, while maintaining that acceptance is normatively significant. Before leaving this line of argument, it is important to note that Klosko cannot reject Orchestra because G does not satisfy the AL condition. Recall again that the role of Fairness is to specify conditions in which merely receiving a benefit binds in the same way pursuing it would. On the subjective interpretation, this amounts to identifying goods one would accept. But in Orchestra we assumed that I actually pursue and thus accept G. So while it is true that the good in Orchestra does not meet AL, it does not need to, for it satisfies a stronger condition than the hypothetical condition AL exists to ensure. Indeed, Klosko would put himself in a precarious position were he to deny this. Because the normative force Fairness produces is grounded in what one would accept, which is parasitic on the normative force of actual acceptance, by denying that accepting a good can bind independently of Fairness Klosko would undermine the normative ground of his own principle. For if acceptance is insufficient to bind, then what does it matter whether a person would accept a good? Further, it is unclear why I would incur an obligation even if G did satisfy Fairness. If, say, the music emanating from Brian’s apartment is the only thing keeping me from spiraling into a fatal depression, one I could avoid by no other means, then, even if I am aware of this fact, it remains unclear why my special needs obligate me to contribute. If another neighbor enjoys 64 Brian’s music more than I do, but does not have my depressive condition, is it really plausible that only I have an obligation to contribute? That One Would Accept a Good Does Not Guarantee Obligation Even if we suppose that the successful pursuit and thus acceptance of a good guarantees obligation, the goods which are ultimately relevant to political obligation are, again, public, and so they cannot be accepted in the way goods like those in Orchestra or Daycare can. It is precisely this problem that Fairness is designed to overcome, by showing that a person could not reject certain public goods, and thus would accept them. Some philosophers have responded that one could in fact rationally reject the goods receipt of which Klosko claims obligates. As Alan Carter has pointed out, for example, some political anarchists believe (and with good reason, Carter thinks) that states justify themselves by preventing harms which states themselves are ultimately responsible for causing. 96 So far as this is true, a state is like a local gang that justifies its exploitation of the neighborhood by pointing out that it defends its members from nearby gangs. However nobody thinks it would be irrational for someone to reject the benefits the gang offers, even when those are indispensible in Klosko’s sense. Why, then, think that it would be irrational to reject state-provided benefits like national defense? While I’m sympathetic to Carter’s position, I think he misinterprets Klosko’s argument. Carter’s point is that, if the political anarchist can reject state-provided goods for 96 Carter (2001). 65 sound reasons, then anyone can. 97 However Klosko is not concerned with what just anyone could rationally reject. He admits that a small number of people like the genuine political anarchist Carter considers might very well reject state-provided goods. 98 He is concerned, instead, with what the majority of citizens could reject. Given that most people are different than the political anarchist both in the way they view the state and in their preference structures, they could not rationally reject, and so would accept, many of the goods the anarchist would not. These people incur obligations simply by receiving these goods, Klosko argues, even if the anarchist does not. However Carter is right to be skeptical of the principle at the heart of the subjective interpretation. That principle seems roughly to be this: Hypothetical For any good G, cost C, and person P, if P would pay C were this the only way to receive G, and C is a fair price for G, then P has an obligation of fairness to pay C for receiving G if G’s producer expects her to. The problem for Fairness is that Hypothetical is false. The fact that one would pay some amount or make some sacrifice if necessary to receive a benefit does not imply that one has an obligation to pay that amount, or to pay any amount, simply because he received the benefit. Consider the following case: 97 Carter (2001, 241). 98 Klosko (1992, 115-16). 66 Gamble You don’t bet for the enjoyment of it; you bet because each bet provides an opportunity to make money. Thus, you’re willing to bet on any team, for any game, and for any amount up to $100, if you are given better odds than those published before the game. There are no odds for the upcoming USC v UCLA football game—it’s expected to be an even matchup. Nonetheless, I’m willing to give you unprecedented 2-to-1 odds that USC takes the game. The problem is that I cannot reach you. You haven’t been in the office all week, and you aren’t answering your phone. However I go ahead and assume the bet is on anyway. I intend to give you $200 if UCLA wins, and expect you to pay me $100 if they lose. The following week, after USC emerges victorious, I come into your office asking for my $100. Given the plausible assumptions (i) that a smart gamble can constitute a benefit, and (ii) that you would have taken this gamble had I been able to reach you, it follows from Hypothetical that you have an obligation to pay me. 99 However this result is clearly absurd. That you would have accepted does not mean that you did accept. I should explain once more before leaving this section why it is no objection to the example that Gamble, like Orchestra, does not satisfy all of Fairness’s conditions—most centrally, AL. The reasons are much the same as they were when we discussed Orchestra. 99 The first assumption is highly plausible. Someone who sold you a lottery ticket for $.01 rather than $1 would provide you with a benefit in spite of the fact that what he sold you was merely a better gamble. And we can build the second assumption into the case by supposing that you simply admit that you would have taken the bet had I reached you. 67 Recall that Orchestra was relevant, in spite of the fact that it did not satisfy Fairness, because it included the relation between the good and its recipient (actual acceptance) from which we are to derive the relation Fairness exists to provide (hypothetical acceptance). Assuming that I actually accept the good therefore satisfies the aim of the principle even if it does not satisfy the conditions it happens to employ to reach that aim. Let me put the same point more formally. Klosko’s theory reduces roughly to a two- step strategy. He claims, with respect to good G and recipient S, that: (1) If G satisfies Fairness with respect to S → S would accept G; and (2) If S would accept G → S is obligated to contribute his fair share to G’s production. Orchestra and Gamble challenge (2), not (1). Orchestra bypasses (1) by assuming actual acceptance, which is stronger than hypothetical acceptance, while Gamble builds hypothetical acceptance directly into the case. Both cases, then, satisfy the aim of Fairness—they both grant the consequent of (1)—even if they do not satisfy the conditions the antecedent of (1) employs to ensure that aim is achieved. We see then that because Orchestra and Gamble challenge the normative significance of hypothetical acceptance—that is, because they challenge (2)—it is no objection that they do not satisfy the conditions of Fairness, or satisfy (1). Fairness Generates Too Few Obligations I’ve argued that Fairness’s obligating mechanism—what one would accept—is insufficient to generate obligations, both because acceptance itself is insufficient, and because the 68 hypothetical basis of Fairness is problematic. But Fairness also fails to capture important cases of obligation by relieving those who fail to satisfy any one of its conditions. The problem is that, on Klosko’s view, a good’s failing to satisfy one of the conditions of Fairness entails that its recipient does not have an obligation from fairness. Thus, because he argues that competing theories of political obligation (from natural duty or consent, for example) cannot capture the most important political obligations we have, the truth of his broader theory guarantees that any important obligations missed by Fairness cannot be accounted for. Let me explain by focusing on RFS. That this condition or something like it is necessary to avoid counterexamples is clear from this example from Simmons: [T]here is a severe drought in my rural neighborhood, where we are all dependent for water on our wells, wells that are now drying up. I am hard at work, successfully digging a new, much deeper well in my backyard to supply my family. But my neighbors, instead of doing the same, opt to dig a long trench along our neighborhood road and beyond, diverting water from a river several miles away, so that all will have access to running fresh water in front of their homes. If I decline to participate in my neighbors’ scheme, have I breached an obligation of fair play by benefiting as a free rider? 100 Klosko agrees that in this case Simmons would not be a free rider because “his plan for alternative supply is unobjectionable.” 101 By unobjectionable, Klosko means that it meets his “plausibility condition” for alternative supply—a three-part condition which specifies when RFS is false with respect to a particular good and a particular person. It includes that the alternative mechanism for supply be plausible in light of the “human condition,” and not 100 Simmons (2001, 34). 101 Klosko (2004, 67). 69 significantly violate others’ rights. 102 Because Simmons can dig his well and thereby provide his own water without violating any of these conditions, RFS is not satisfied, and so the good he receives does not satisfy Fairness and thus does not bind him to contribute to his neighbors’ trench digging project, just as we would expect. But claiming that a good’s failure to satisfy RFS relieves its recipient of an obligation she would otherwise have—which is precisely what Klosko must claim to avoid Simmons’ counterexample—opens Klosko’s principle to other objections. Consider Warren Buffet. Klosko would claim that the primary goods available in states—things like national defense, domestic law and order, environmental regulation, public health services, etc.—satisfy RFS with respect to Buffet because even Buffet could not afford to provide them on his own dime. But while Klosko is right that Buffet could not afford to provide these to everyone, he could arguably avoid the costs of not having them for himself. For instance, while it is likely true that in an anarcho-capitalist North America Buffet could not afford to prevent large-scale environmental degradation, or provide a private security service equivalent to our current national defense program, as Klosko argues 103 , it is not clear that he must if his aim is merely to enjoy privately benefits which are equivalent to these services. What matters for Fairness is not whether the environment is protected, or whether the nation is safe, but whether without a healthy environment or a safe nation Buffet’s life would be acceptable to him. It is significant then that many burdens, like those of environmental degradation, are largely proportional to wealth. A person who lives 102 Klosko (2004, 64-65). 103 Klosko (2004, Chapter 2). 70 on rich farmland along seaside bluffs has much less to fear from rampant urban pollution or rising sea levels than do the urban poor in Los Angeles or New Orleans. Similarly, national defense is important because without it one’s country might be invaded, putting one’s life or lifestyle at risk. But invasion surely matters much less to someone like Buffet, whose private jet is on call to whisk him and his family to another country—to which he would have no political obligations either, for the same reasons I’m detailing here. Myriad details would have to be filled in here before we could be satisfied that Buffet’s plan for alternative supply met Klosko’s plausibility condition. I doubt we could ever know with certainty whether it did. But for the broader point I’m trying to make it is enough to see that richer people have access to goods and services that others do not. And when a person is as wealthy as Warren Buffet, it is highly plausible that she could live her life with most of the amenities of modern societies, and more of some of them, without at least a large number of the primary services on which the rest of us depend. Now, it might be suggested that a person could not become as wealthy as Buffet without a structured legal system to protect private property. Supposing that this is right, Klosko may respond that state provided public goods come in packages, and that receiving one indispensible good like the protection of private property obligates one to support the protective institution as a whole, even if it provides other goods as well. 104 However this does not solve the problem. For suppose that Buffet made his money in another country, but that, living here now, he enjoys all of the state provided benefits you and I do, only that, while these are indispensible for us, they are dispensable for him. It might be that in this case 104 See Klosko (1992, Chapter 4) and (1987b). For a response to this sort of “bundling” argument, see Simmons (1987). 71 Buffet owes political obligations to the country where he made his fortune, but clearly he owes them here too, if anyone does, even if his wealth can be protected in another country and only spent here. Wealth seems, if anything, to increase and strengthen one’s obligations, not to reduce and weaken them. “There is, then,” as Rawls says, “another sense of noblesse oblige: namely, that those who are more privileged are likely to acquire obligations tying them even more strongly to a just scheme.” 105 This seems particularly true if that wealth leads one to enjoy more of the benefits state’s offer than others do. Klosko’s principle suggests, however, that the less one must have, the less one owes, and this seems highly counterintuitive. 2.4 Conclusion I have argued that obligations from fairness are less well understood than many philosophers have acknowledged. It is unclear how these considerations account, not just for political obligations, but for any obligations, particularly in cases regarding public goods. Though Simmons has argued forcefully that citizens’ beliefs and attitudes entail that they have not accepted the central public goods provided in states, we saw that the conditions those citizens must meet to incur obligations are entirely unclear (Section 2.2.2), and that, anyway, acceptance does not plausibly ground obligations in cases in which the supposed obligees have unilaterally modified the environment so as to restrict the supposed obligor’s permissible actions (Section 2.3.3). I also explained why shifting away from an acceptance approach would not improve the fairness theorist’s position. Whether interpreted as an 105 Rawls (2003, 100). 72 objective or a subjective conception (Section 2.3.2), a principle like Klosko’s cannot account for the obligations many people take citizens to have. Interpreted objectively, it is unclear how one’s receipt of a benefit can obligate her to contribute; interpreted subjectively, receipt based arguments face the same objections the acceptance approach faced, and more (Section 2.3.3). It is worth emphasizing once more that none of this entails that considerations of fairness do not bind in some cases, or even that none of those cases involve public goods. At best, the arguments I’ve offered show that acceptance and receipt based approaches to the principle of fairness are deeply problematic, and that, besides failing to explain political obligations, they likely fail to explain any obligations we incur in virtue of these considerations. 73 Chapter 3: Samaritanism In the last chapter we evaluated the two most plausible approaches to the argument from fair play—the acceptance and the goods approach—and found that neither succeeded. The initial appeal of the acceptance approach faded as we tried unsuccessfully to interpret its main components in a manner that could plausibly ground any obligation, not to mention political ones. The goods approach, on the other hand, attempted to attach obligations to benefits individuals may not want, or to invoke a type of acceptance more problematic than the sort we were forced to reject on the acceptance picture. In this chapter we consider the markedly different approach taken by Christopher Wellman. Though probably best classified as a natural duty theory, samaritanism draws liberally from theoretical traditions that emphasize both the enormous benefits states provide, and the equal responsibility each of us has to contribute our fair share to their success. Wellman has two distinct but related aims. One—and the one we are more interested in—is to account for the obligations members of states purportedly have to obey the law. To understand the samaritan case for these obligations, however, we must first understand Wellman’s other aim, which is to show that the state is justified in enacting and enforcing laws in the first place. Putting pressure on the purportedly separate arguments for these aims 74 causes them to converge in important ways that will be important later when we consider objections to Wellman’s view. I explore all of this below, but I point it out in advance so that its clear why I spend so much time grappling with the samaritan case for legitimacy. The objections I raise will apply equally to the samaritan case for political obligation, and the distinction between legitimacy and obligation provides an important contrast between Wellman’s theory and the sort of theory I develop in the second part of this dissertation, in which I defend political obligation without legitimacy. In what follows we will consider samaritanism’s two primary arguments separately: the first for state legitimacy (or “statism,” as Wellman sometimes calls it), the second for political obligation. My aim in Section 3.1 is to present these arguments in their simplest form, which I do by distilling the fundamental principle on which each argument is implicitly based. It is in virtue of the view’s structure that some of its most central elements become clear only in response to the objections I raise in Section 3.2, so the expository work carries over into that section as well. As I try to argue throughout, a closer look at these further details reveals serious challenges facing the samaritan approach, particularly its account of political obligation. 3.1 The Broad Strokes I’ll begin with the samaritan case for state legitimacy, which provides, on Wellman’s view, a foundation for its case for political obligation. 106 I should say at the outset that, while these 106 Wellman (2001, 742). 75 two arguments look very alike, and employ principles which look very alike, their subtle differences are important to Wellman’s case. Part of our aim in evaluating samaritanism in Section 3.2 will be to determine to what extent Wellman is right to endorse these differences. For now, however, we will be careful to keep his two lines of reasoning separate. 107 3.1.1 Legitimacy A state S is legitimate with respect to some person P and realm of action R if S has a liberty- right to coerce P to act in some particular way with respect to R. 108 Following Hohfeld, we will say that a person (or entity) has a liberty-right to do something if and only if she (or it) does not have a duty not to do it. I have a liberty-right to drink from a particular fountain, for instance, just in case I have no duty not to drink from it, and I lack this right if such a duty exists (say, because the fountain belongs to you). Thus, generally speaking, S is legitimate with respect to P if S does not have a duty not to act towards P in the ways S does. 109 I will call a principle which specifies those circumstances in which a state is legitimate a liberty limiting principle because it describes the conditions in which a state may coercively limit a person’s liberty. As an initial gloss, samaritanism holds that the state is legitimate—is 107 I believe that samaritanism would be simplified and probably strengthened if Wellman were to reduce its two fundamental principles (outlined below) to a single normative foundation, such as the duty of samaritanism. While simplifying the theory in this way would, I think, do nothing to help it overcome the objections I raise later in this chapter, it would preempt important interpretive concerns which I have for the most part set aside in what follows. 108 Hohfeld (1917). 109 “Generally speaking” because no state can reasonably be expected never to violate duties it has to its citizens, and because such occasional violations do not undermine the legitimacy of an otherwise exemplary state. 76 justified in coercing its population in the ways that it does—when such coercive state action is necessary to save everyone from the perils of the state of nature. Rather than explicate the precise liberty limiting principle on which his account depends, Wellman suggests its content through examples. The most central of these, and the one we’ll use to outline his theory, I quote directly from Wellman: Heart Suppose that Alice and Beth are walking down a street together when, all of the sudden, Beth suffers a heart attack. Suppose further that the facts are quite clear: unless Alice can get Beth to a hospital immediately, Beth will die on the sidewalk. As it turns out, there is only one car in the vicinity. (Neither Alice nor Beth recognizes the car nor knows its owner, Carolyn.) Serendipitously, however, Carolyn has left the doors unlocked and her key in the ignition. Wasting no time, Alice puts Beth in the car and speeds to the hospital where her life is saved. As soon as Alice has done all she can at the hospital, she returns Carolyn’s car and leaves an appreciative and apologetic note explaining why, and to what effect, she commandeered the car. 110 While not a perfect descriptive match with the complex social realities of a state, Wellman claims that the details in Heart faithfully capture the relevant moral relationship between a 110 Wellman (2001, 744-745). The case (which I’ve named ‘Heart’) appears with different characters in (2005, Chapter 1). 77 state and the individuals within it. 111 If this is true, then Heart furnishes an important insight into the sort of liberty limiting principle samaritan considerations provide. The relevant details seem to be these: (i) Beth is in peril; (ii) Alice can save Beth from peril by taking Carolyn’s car without her permission; and (iii) taking it does not impose an unreasonable cost on Alice, Carolyn, or anyone else. Thus, since taking Carolyn’s car and thereby infringing on her rights in a manner that would normally be impermissible is the only way for Alice to save Beth, she is permitted to take it. In Section 3.2 I raise some concerns about the comparison between Heart and the political context. To see how the argument unfolds, however, let us accept for now the moral relationship Wellman posits between them. Samaritanism’s liberty limiting principle runs as follows: JC A may justifiably coerce C in the process of φ-ing if (i) some person B is in or is heading towards peril, (ii) A’s φ-ing will help B get out of, or prevent B from ending up in, peril, (iii) φ-ing will not impose an unreasonable cost on C or anyone else, and (iv) φ-ing is necessary to save B from peril (the necessity condition). 112 111 In his original presentation, Wellman (2001) provides an additional example of a bus driver who has forcibly commandeered his passengers’ bus parts to fix the broken down bus out of peril. For a number of reasons this latter case is a better descriptive match with the political context. Because its morally relevant details are no different than those in Heart, however, and because Heart remains Wellman’s primary focus throughout his arguments in (2001) and (2005), I omit it. 112 I use the term ‘coercion,’ as I believe Wellman does, broadly to refer to any behavior which encroaches on the rightful dominion of another. 78 If JC is a sound principle, then to make his political argument Wellman need only show that the political circumstances in states satisfy its conditions. Some theorists, typically either libertarians or socialist-anarchists, will reject his claim that they do. 113 But while they (particularly libertarians) sometimes offer empirical arguments to show that societies could be saved from peril without states (usually basing these arguments on economic “laws” that describe coarse regularities in human behavior), we know far too little about human psychology to know whether or how social stability would emerge in an industrial or post- industrial anarchist society of any ideological stripe. 114 And so rather than getting tangled up in this empirical guessing game, we will simply grant arguendo that no non-state entity could prevent peril to the extent that states can. 115 Granting this empirical point, Wellman’s argument appears straightforward and plausible. States may coerce their inhabitants for the same reason Alice may take Carolyn’s car. Like Beth’s impending peril, a state of nature would (i) be perilous for all those within it, so since (ii) the state helps its citizens avoid this unacceptable fate, and (iii) the laws through which it does are not unreasonably burdensome for its inhabitants, the state may coerce 113 But note that theorists in these two camps disagree vehemently about the reasons for this. Both groups agree that a stateless society would be (or at least could be) harmonious. However while socialist anarchists (also called anarcho-communists) argue that this would require the absence of capitalism, which they believe would result in widespread oppression, libertarians believe that it is only through capitalism that society could thrive without the state. For different approaches from within the socialist-anarchist camp, see, e.g., Albert (2004) and Bookchin (1986). For an anarcho- primitivist account, see Zerzan (2002). Friedman (1973) and Nozick (1974) offer libertarian accounts. For Wellman’s brief defense against objections of these sorts, particularly from the libertarian camp, see Wellman (2005, Chapter 1). 114 Though, I particularly like the utopian vision offered in William Morris’s 1890 fiction News from Nowhere. 115 Later I will question whether certain ways in which the state prevents peril are necessary. What I am willing to assume here is simply that no non-state entity could prevent peril as well as the state. 79 them, as Alice coerced Carolyn, because (iv) this is the only way to prevent everyone from falling into peril. 3.1.2 Obligation One might think that, if Wellman’s samaritan argument can establish legitimacy, then it can establish obligation too. It is hard to imagine, for instance, that Alice could be justified in taking Carolyn’s car unless Carolyn, had she been present when Beth collapsed, would have had an obligation to surrender it. Wellman does not explicitly explore this question, but, barring an important complication which arises when we move from Alice’s liberty to Carolyn’s duty, he seems to agree. 116 We will discuss this additional complication in a moment. First, however, let’s extract from this modified version of Heart the samaritan principle governing obligation: DoS C has a duty of samaritanism to B to φ if and only if (i) B is in or is heading towards peril, (ii) C’s φ-ing will help B get out of, or prevent B from ending up in, peril, (iii) φ-ing will not impose an unreasonable cost on C or anyone else, and (iv) φ-ing is necessary to save B from peril (the necessity condition). 117 116 Wellman (2001, 748). 117 It is important to notice that there is no biconditional connecting the necessity condition in DoS and the necessity condition in JC. One might be satisfied when the other is not, as I explain presently. 80 Now, as I mentioned, at this point the analogy between Heart and the circumstances in states breaks down. Unlike states, which satisfy the necessity condition in JC, their citizens do not satisfy the necessity condition in DoS. 118 The significance of this failure is evident only if one remembers that the necessity condition is integral to both arguments. Had Alice had a car of her own, she would have had no right to take Carolyn’s, and Carolyn would have had no obligation to surrender it. But while we have already granted that states are necessary to save their constituents from peril, just as taking Carolyn’s car is necessary to save Beth, no citizen is quite like Carolyn after we modify the case so that she is on hand when Beth collapses. While Carolyn must act to save Beth, you don’t need to act to save the rest of us. Of course, you will not contribute much to the state’s peril-preventing project by breaking the law and cheating on your taxes. But the lives of your 300 million compatriots will not be solitary, poor, nasty, brutish, and short because of it. But if we do not need your voluntary cooperation to avoid the perils of the state of nature—if your contribution is more or less inconsequential—then samaritanism fails to explain why any one person has an obligation to obey the law, even if the state has the general liberty (and perhaps even the duty) to enforce it. It is like trying to explain why Carolyn must give up her car when Alice already has a car of her own: she doesn’t. 119 118 Of course, states satisfy DoS with respect to their citizens, and so one might wonder whether the state’s duty to its citizens might transmit to those citizens themselves, all of whom would then have a political obligation. Unfortunately, this explanation misses the distinction between the duties in question. The duty states have is to coerce citizens in whatever way is necessary to keep everyone from falling into peril, not a duty to obey the law. 119 One might wonder at this point why the state may legitimately force a citizen to cooperate if her cooperation is not needed to save others from peril. As we’ll see in Section 2.2, Wellman is ultimately forced to eliminate this asymmetry between his two arguments by invoking considerations of fairness. Later I argue that these considerations cause special problems for JC. 81 Wellman calls this the necessity problem. And he finds a solution, interestingly, in the components of a theory that he rejects as a stand-alone explanation of either legitimacy or political obligation: fairness. 120 Wellman rejects fairness as a defense of statism because he believes that it is paternalistic. On his interpretation, fairness theory holds that citizens can be coerced for their own benefit, which is precisely the sort of justification about which he believes we, as liberal theorists, should be skeptical. His objection to the fairness argument for political obligation is similar. In line with Simmons 121 , Wellman holds that the only way a good can bind an individual is if she accepts it. He also believes, however, that a person cannot accept anything she cannot reject. Thus, since on his view citizens cannot reject the public goods states provide, Wellman claims that they cannot accept them either, from which it follows that considerations of fairness cannot account for political obligations. 122 Despite his belief that considerations of fairness cannot alone account for our political obligations, however, Wellman believes that they play an integral role in the samaritan account, patching normative gaps left by our samaritan liberties and duties. The problem, again, is that while DoS can explain our collective responsibility to contribute to the state’s peril-preventing effort, it cannot explain why that responsibility falls on any one of us, 120 Wellman (2001, 737). 121 Simmons (1981, Chapter 5). 122 I’m skeptical of Wellman’s case against fairness. His first objection seems to be based on a misunderstanding of the fairness theories he criticizes, whose authors do not contend that citizens can be coerced for their own benefit. And the key premise of his second objection is implausible: it seems that a person can indeed accept a good she cannot reject. I return to this issue in Section 3.2. Before moving on, however, I must point out that the success of these objections would leave a significant gap in Wellman’s broader argument for samaritanism, which begins with the claim that previous theories (such as fairness) cannot explain all that needs explaining. If his criticisms of fairness are unsuccessful, then, at least from his theoretical perspective, there may be no need for a samaritan theory. 82 since no individual’s contribution is necessary to achieve the good that grounds the obligation. In response, Wellman points out that leaving the samaritan chore to others would be unfair, and that each of us therefore has an obligation to contribute our share—in this case, paying our taxes and obeying the law. By utilizing considerations of fairness for this limited distributive purpose, Wellman argues that samaritanism can overcome the necessity problem without falling prey to either of the objections facing freestanding theories of fair play. Because the state’s right to coerce me is grounded in the good it will do, not for me, but for others, samaritanism is not paternalistic. And because our collective responsibility to aid the state in its peril-preventing project is grounded in a natural duty of samaritanism—to the satisfaction of which each of us, for reasons of fairness, must contribute—we need not accept any goods, or do anything else for that matter, to incur these obligations. They are simply an extension of the more basic obligation we have to help others. By using considerations of fairness to distinguish our individual responsibilities while grounding our political obligations in the duties we have to prevent the sorts of consequences we see in Heart, samaritanism succeeds as a theory of both legitimacy and political obligation where theories of fair play failed. 3.2 Problems Facing Samaritanism The most attractive feature of Wellman’s theory is that it adopts and explains the insights of failed theories before it. Like the benefit theorist, Wellman starts with the moral significance of the public goods states provide. But rather than contend that we can be bound simply by receiving such goods, he argues, along with the natural duty theorist, that we are bound 83 because other people need them, and because we have a duty to help people get what they need when doing so is not unreasonably burdensome. Yet the natural duty theorist fails to account for the prescriptive vagueness of such an obligation. Must I help you if you do not need my help? Not obviously. And so it is difficult to see how natural duties would be relevant to the question of political obligation. Here Wellman takes his cue from fairness theory, the kernel of which is that it is unfair that some of us do all of the necessary work. But while the fairness theorist has this general point correct, she has, Wellman claims, been concerned primarily with the wrong sort of fairness. We should not focus on the goods you unfairly enjoy at my expense, he says, but rather on the disproportionate amount of work you do to ensure that the samaritan obligation you and I share is satisfied. The priest and the Levite who ignored the man victimized by thieves on his way from Jerusalem to Jericho did not act wrongly because they rode free on anyone else’s labor, as I do when, say, I walk on the streets you’ve paved. They acted wrongly because they failed to help a man who needed it. And in doing so they wronged the man who chose to help, that is, they wronged the Samaritan, for they left it to him to discharge an obligation which belonged to them all. Despite gleaning the insights of competing theories, however, samaritanism faces substantial challenges. Below I canvas the most serious of these, beginning with the samaritan case for legitimacy. As we’ll see, the two strains of the argument begin to converge, leaving them open to similar objections. 3.2.1 Problems with Legitimacy The intuitive force of the samaritan case for legitimacy depends largely on the analogy Wellman draws between Heart and the political context. But while the morally relevant 84 features in Heart surely warrant Alice’s appropriation of Carolyn’s car, it is not clear that states are sufficiently similar to support a parallel argument. In particular, it is questionable whether state coercion satisfies JC’s conditions (iii) and (iv). We will evaluate these conditions separately. What is Reasonable? How much does Heart really have in common with the political context? To extend the conclusions we drew about Alice to the behaviors of the state, the two cases must be sufficiently similar. Yet on closer inspection the analogy seems strained. One consideration is that Carolyn is not around for Alice to consult prior to taking her car. Though I’m sure it was not Wellman’s intention, this fact invites us to imagine that Carolyn would have consented had the opportunity arisen. Most people are reasonable, after all, and so Carolyn probably is too, and no reasonable person would jeopardize another person’s survival for something so trivial as a borrowed car. Yet Heart offers no support for the samaritan account if the permissibility of Alice’s behavior depends at all on what Carolyn might have agreed to. Wellman’s story would therefore have been more compelling if Alice had asked Carolyn for permission to use her car and Carolyn had refused. Could Alice have taken the car anyway? I think so. Still, there is an important moral difference between a person’s failure to give consent merely because she is not around to give it, and her failure to give consent because she has opted not to—a difference that is easy to overlook because its significance is revealed only when the severity of the coercion is amplified. You may permissibly administer a blood transfusion to save my life even if, because I am unconscious, I cannot consent. But the absence of my consent is significant if I have expressly refused to give it (as, for instance, 85 a devout Jehovah’s Witness would). All of this indicates that our conclusion in Heart is at least less clear than first impressions might indicate. It may be that Alice could take Carolyn’s car to save Beth even if Carolyn tried to prevent it. But when the stakes are raised, as they are in the political context, the fact that citizens have the opportunity to give their consent to their government, but do not, at least does not help the analogy. 123 This is, however, simply an instance of a more general problem, which is that Heart downplays the significant costs of state coercion. 124 For, even if Carolyn had explicitly denied Alice’s request to use her car, Alice’s taking it remains importantly different from the sort of coercion administered by states, which is at once more complex, invasive, consistent, and severe. A better example would perhaps be that of the violinist in Judith Thomson’s article, A Defense of Abortion. 125 Imagine that the Society of Music Lovers has, without your permission, connected the body of a dying violinist to yours through a piece of medical equipment. His blood is poisoned, and his kidneys are no match for it. In fact, few kidneys would be capable of eliminating such a powerful chemical. But yours are. And so here you find yourself, sitting in a hospital room with your body attached to the violinist’s through a plastic tube. Was it 123 Admittedly, it is not clear just how much we can conclude from the mere absence of consent. Simply because you have not consented to something to which you could have consented does not mean that you have done so intentionally (in a way that carries moral weight). Perhaps the thought of consenting to something that already regularly occurs never crossed your mind. It does seem, however, that if an agent’s failure to give consent is due to her ignorance of her option to give it, and the party whose liberty is in question (Alice, the state, etc.) is capable of bringing that option to her attention, then her lack of consent should, I think, at least in cases in which she is not the beneficiary, be treated more like an express refusal of consent until the choice is presented to her. Such is the case in states, in my opinion. 124 See Klosko (2003). 125 Thomson (1971). 86 permissible for Society to kidnap you? And must you stay in the hospital now that you’re here? The violinist is in peril. And your kidneys are uniquely able to detoxify the blood. Being taken to the hospital against your will has of course been burdensome, and lying in the hospital, as you now are, is not how you would prefer to spend the next few days of your life. But surely the cost to you is relatively minor if we consider that the violinist will otherwise die, and that this procedure is (let us imagine) common and safe. Yet as Thomson suggests, while it would be very nice of you to remain in the hospital hooked to the violinist, nobody can claim that you are duty bound to stay, let alone that the Society of Music Lovers was warranted in kidnapping you in the first place. I think, however, that this sort of coercion and sacrifice more closely resembles what we see in states than what we see in Heart. States restrict our behavior in invasive and personal ways. They have the authority to quarantine me if I am sick, to vaccinate my children before they attend (mandatory) school, to force me (if I’m a woman) to have a trans-vaginal ultrasound before getting an abortion 126 , and to lock me away in prison for as long as they see fit for whatever actions they deem unacceptable. Granted, not all of these powers seem as severe as those the Society (illegitimately) exercises in Thomson’s case, but hers is surely a better analogy than Wellman’s. Thomson’s case also highlights another discrepancy. Besides downplaying the invasiveness of state coercion, Heart exaggerates the samaritan benefits states provide. For one thing, the circumstances in Heart remove uncertainty. We simply stipulate what will happen to Beth if Alice fails to act: she will die. Yet the perils of anarchy are at once less 126 See, for instance, Erik Eckholm’s New York Times article “Ultrasound: A Pawn in the Abortion Wars” (February 25, 2012). 87 certain, less immediate, and, at least in many cases, less dire than in Heart. It is true that without a state some people would face an untimely and gruesome death, but many others would survive—as people presumably did before the first state ever formed—and some would even flourish. The fate of these people is nothing like Beth’s. For everyone else, things are at least less certain. This raises a question important enough to flag before moving on, though we will not pursue it further. When we make comparative claims about the benefits of a state, which circumstances should we use to determine whether those claims are true or false? Wellman’s example suggests that he takes the proper comparison class to be something akin to certain death, which, if true, makes his view more Hobbesian than Hobbes’ own. But are these the right circumstances to compare? There are in fact two questions here, which are easy to run together. The first is what a stateless society would actually look like, and how it should be evaluated. Hobbes’ conception is significantly more perilous than Rousseau’s, for instance. Which should we use? And even if we can agree on this, from whose perspective should it be evaluated? That people in the state of nature do not necessarily see things as we do is nicely illustrated by one anthropologist’s meeting with a Hadza bushman. When the anthropologist asked why the Hadza, who only hunt and gather, would not take up agriculture, which the anthropologist considered to be clearly superior, the bushman asked, “Why should we plant when there are so many mongomongo nuts in the world?” 127 Why, indeed. 127 Quoted in Sahlins (1972, 27). 88 The second question is whether we should consider such a society as if it had developed naturally—as primitive societies did for the majority of Homo sapiens existence—or whether we should consider what our society would look like if the state suddenly disappeared. Wellman appears to endorse the latter perspective: Imagine that a supremely powerful government official approached you with an extraordinary offer: If you are content to continue obeying the law and paying your taxes, then she will keep your current state in place as is; if you would rather be freed from your state’s imposition, on the other hand, then she will entirely disband your current state. How would you choose? 128 Is this really the relevant question? The disappearance of state institutions in a society habituated to depend on them does not approximate an anarchist society, at least as that term is normally understood. What Wellman describes would be more like a failed state. Even critics would not argue against anarchist forms of political organization by citing the peril in, for instance, Somalia, or post-invasion Iraq. That states save us from immanent death because such a fate would be likely were our government to completely dissolve poses a false dilemma. By the same reasoning we could justify airplane hijackings, since, with a dead pilot, most passengers would prefer that the hijacker fly the plane than that he disappear altogether. But then to what social arrangement should we look to determine the significance of the benefits states offer? Arguably, we should look to the best sort of stateless society we could plausibly have. Some will immediately object that we simply cannot know what this would look like. And they are right. But at the very least this fact does not license the nightmare scenario Wellman envisions. In any case, more must be said. 128 Wellman (2005, 18). 89 The Reasonability Concern Returning to our previous line of thought, I suggested that state coercion is more severe than what Carolyn experiences, and that the benefits it provides are less significant and less certain than those Beth enjoys. So it seems reasonable to recast the concern with which we began: Even if JC can justify coercion is some cases, such as Heart, can it account for state coercion? This is the reasonability concern. Klosko initially raised the reasonability concern in response to what later appeared to be a misleading statement in Wellman’s original presentation of samaritanism in 2001. “Compared to the benefits derived from political community,” Wellman claimed, “the costs of citizenship are also relatively trivial.” 129 However what Wellman seemed to have in mind were the benefits coercees enjoy. This would be equivalent to citing some benefit Carolyn receives to justify Alice’s behavior. As Klosko noted, however, this reduces a necessary feature of the samaritan argument to precisely the sort of fair play consideration Wellman previously rejected as paternalistic. In a later article, Wellman seemed to set things straight. He claimed that he was “emphatically not suggesting that states are justified in coercing their constituents because this coercion ultimately benefits each of them [the coercees].” 130 He had, it seemed, simply been misunderstood. What Wellman claims to have meant all along is that the burdens of citizenship I bear are trivial, not relative to the benefits I receive, but relative those you receive. Thus, because it is your benefit which explains why I may be coerced to such an 129 Wellman (2001, 746 fn.11). 130 Wellman (2005, 18). 90 extent, Wellman claims that samaritanism cannot be accused of the sort of paternalism to which fairness theories allegedly fall prey. Were this the final word, then the disagreement between Wellman and Klosko would likely end in gridlock: Klosko would claim that our samaritan duties do not warrant the significant burdens states impose; Wellman would claim they do. But Wellman, who seemed to have been dissatisfied with this result, went on to give a more elaborate (and, as we’ll see, more problematic) argument for the reasonability of the serious costs states impose. Although these costs are substantial, Wellman now claimed, “the net costs of citizenship are not unreasonably great because these net costs consist of these admittedly considerable initial costs minus the even greater benefits of political stability.” 131 In other words, while the state’s provision of self-regarding benefits—the benefits I receive while I am at the same time coerced by the state—cannot explain state legitimacy, it can explain why the costs states are justified in imposing for other reasons are reasonable for citizens to bear. “It is,” Wellman says, “as if the state forced each of its constituents to give up a hundred dollar bill but in return gave back ten twenties.” 132 There is an air of creative accounting in this argument, which seems merely to confirm that Klosko’s interpretation of Wellman’s 2001 argument was mostly correct. Let me explain what I mean. On Wellman’s view, state coercion is justified if it satisfies JC. However this includes condition (iii), which specifies that the state’s action not impose unreasonable costs on anyone. Wellman seems to accept that state-imposed costs would be 131 Wellman (2005, 17). 132 Wellman (2005, 17). 91 unreasonable but for the benefits they simultaneous provide for the very people they coerce. But this seems squarely inconsistent with Wellman’s rejection of fair play. On the fair play account, states may coerce citizens because those same citizens receive goods which benefit them. Wellman’s view diverges only in that he adds an additional condition: it must be sufficiently good for others too. But adding another necessary condition does not eliminate his view’s reliance on the same self-regarding benefits that fairness theorists do. It merely supplements it. And so in the end Wellman argues, like Klosko, that states are justified in coercing me because of the benefits I receive. Consider this example: Aqueduct You are one farmer among many living in a valley. Rain in the valley is somewhat unreliable, so the state decides to install an aqueduct, which it determines will make you all much better off. The project is expensive, however, and you’ll be expected to contribute your fair share. To simply somewhat, the fairness theorist will claim that you owe your contribution because you benefit from the aqueduct. That you’ll benefit is, in other words, necessary and sufficient to explain your obligation to contribute. However Wellman rejects this approach. It is paternalistic, he says, to expect you to contribute simply because you receive a good that will benefit you. If you owe a contribution, he says, it is because the other farmers need the aqueduct, and you have a duty to contribute to projects that help meet needs of this sort. But it is a part of the samaritan view that these burdens can’t be unreasonable. That it is unreasonable is, however, precisely what you claim. Why should you have to pay this amount for an aqueduct you didn’t ask to have? According to Wellman, it is because you are going to 92 benefit too. This is, again, like contending that Alice is not justified in taking Carolyn’s car unless Carolyn receives some benefit which makes her net sacrifice reasonable. But if we must cite this sort of self-regarding benefit to make the argument, then it would seem that Wellman’s objection to fairness theory applies to his own view as well. On both views, it is the fact that you will benefit which justifies coercing you to pay for a good you’ve received. 133 Is It Necessary? Recall that, to overcome the necessity problem, the samaritan argument for political obligation relies on considerations of fairness in a way that the argument for legitimacy does not. But one might wonder, given the similarity between the two arguments, why the necessity problem does not apply to the legitimacy argument as well. For if state coercion is justified because it saves society from peril, and, as Wellman concedes in his argument for political obligation, not every member of a society needs to obey the law to prevent our collective slide back into the perilous state of nature, why may the state coerce everyone in its territory? That would be like saying that Alice is justified in taking all of the cars on the street, even though Carolyn’s car alone is sufficient to get Beth to the hospital. While absent from his initial argument, Wellman responds to this objection in precisely the way he responds to the necessity problem for political obligation—by claiming that coercing only the necessary number of people would be unfair. “Because samaritan 133 In some passages Wellman seems to try to avoid this problem by claiming that net benefits do not matter to his argument. He writes, for instance, that a state is justified in coercing me even if I am not benefited because the state must coerce everyone within its borders to provide benefits to others (2005, 19). However this interpretation is problematic, for it implies that the state may impose significant burdens on me irrespective of self-regarding benefits, and thereby undermines Wellman’s claim (2005, 17) that the net benefit argument explains why state coercion is reasonable. If it is reasonable without the net-benefit argument, then why include it? 93 responsibilities fall equally upon all of us,” he writes, “it would be objectionable for a state to allow a fraction of [citizens] to shirk their fair portion of the chore.” 134 Yet while Wellman claims that fairness is a source of political obligation on the samaritan account, he denies it this status in his argument for political legitimacy. “With political obligation, fairness is necessary as a source of duty, but with political legitimacy, fairness only limits the manner in which a state may permissibly carry out what it has (for other reasons) a right to do. 135 It is important to see why Wellman wants to limit the role of fairness in this way. It may be reasonable to coerce you to save some other person from peril, but it is much less reasonable to coerce you simply to keep things fair. There is a lot of unfairness in the world, and surely not all or even most of it warrants coercive state interference, even when what is unfairly distributed is some form of necessary coercion. Consider, for instance, my neighborhood near downtown Los Angeles, which occasionally hosts high-profile charity races that require shutting down the streets for several blocks around my house. If I do not park elsewhere the night before, or leave the area before the race begins—usually before 6am—I get blocked in until it’s over. I believe, and imagine that the other people in my neighborhood would agree, that the funds these events raise for cancer research or domestic violence awareness clearly warrant the limited logistical burdens they create. Yet given that none of us agreed to host these races—the city presumably chose the most convenient (for them) place to disrupt traffic—it does seem unfair that we have to bear these costs, while 134 Wellman (2001, 756). 135 Wellman (2001, 756 fn.27). 94 people only three blocks north of us do not. Is it therefore acceptable for the city to block everyone in, just so that nobody is unfairly burdened? Of course not. One might object, of course, that the samaritan good provided by states is so important that we are warranted in fairly distributing the coercive burdens necessary to achieve it. However this depends on the erroneous assumption that the reasons one has to impose burdens in the first place affect the weight of the reasons one has to distribute those burdens fairly. Modifying our example seems to show that they do not. Imagine that one of two races will be run around my neighborhood. The first raises money for schoolbooks (an important but not perilous issue) while the other raises money to pay for a medical procedure without which a particular person will die (a perilous one). Does the race that’s run affect the roads the city may block? Presumably not. The reason for which either race is run appears to affect neither the sort of unfairness to which the people in my neighborhood are exposed, nor the weight of the reasons that unfairness gives us to resolve or avoid it. So while the second (perilous) cause may give the city more reason to have the race, it does not give them a more compelling justification for blocking the streets north of mine. What emerges from all of this is, I think, a discrepancy between samaritan and fairness considerations which interferes with the samaritan line. A samaritan consideration gives us a reason to do something (coerce an unspecified subset of people, e.g.), while a fairness consideration gives us a reason to do it in a particular way (distribute the coercion equally across the whole set). Yet the second sort of reason does not inherit the weight of the first. This implies that, absent some special argument, the fact that the state is justified in using some unspecified amount of coercive force does not entail that it may coerce me when 95 it does not need to in order to avoid social peril. Coercion is justified, it seems, but not any particular instance of it. You might be skeptical of this argument. If the state is permitted to coerce some subset of people, then you might think the justificatory work is already done, and that the only question remaining is that of how we are to distribute that coercion. According to this way of thinking about the problem, there are two stages. At the first stage we justify the coercion itself through the samaritan argument. At the second stage, then, we are no longer looking for a justification for coercion, only a justification for distributing that (now justified) coercion in a particular way. Thus, you might think, it is no objection to point out that fairness may not provide a justification for coercion, since that’s not its role in the argument. While I see the force of this response, I am not convinced that fairness or some lesser principle does not play a justificatory role, at least as the argument applies to states. Think back to my example about the races in my neighborhood. The city could not permissibly block everyone in simply because blocking everyone in would be fair, and it needed to block some of us in to have the race at all. But why not? Presumably the answer involves the fact that by blocking everyone in the city would increase the overall level of coercion beyond what it needed to be. Blocking everyone in would be a heavy-handed way to create fairness—a sort of “leveling down.” A better option would surely be to compensate the people in my neighborhood using taxes from the rest of the city, in which case all of the neighborhoods in Los Angeles would bear their share of the coercive burden, thereby limiting the burden on any one person. This way of distributing the coercion would be much more difficult to object to than blocking everyone in for the morning. 96 The lesson, then, is twofold. Considerations of fairness underdetermine how the necessary burden can be distributed, and the way to select amongst the options is to find the (fair) distributive mechanism with the least coercion. Of course, things don’t always happen as they should. My neighborhood does not get compensated when the city blocks us in. Why not? Presumably because the city bears other considerations in mind. In this case (as in most cases in the political context) the limiting consideration is efficiency. Keeping track of who owes whom for what would be a hassle, and presumably not worth doing from the city’s point of view. But now the criticism I raised a moment ago reemerges one level up: If one is permitted to coerce some subset of people, and there are multiple fair ways to distribute the coercion, as there almost always are, then any principle one uses to select anything other than the least coercive option is playing a justificatory role. This objection is relevant, not to the validity of the samaritan principle, but to its application. There are many fair ways to distribute the coercive burden in states which would at the same time limit the total coercion exercised. Perhaps political scientists could be trained to find the minimum number of people the state can coerce while preventing the populace from falling into peril—as Wellman himself suggests—and then the state could have a fair lottery to determine who those people are. This would surely be more difficult for them, but, again, so far as concerns of efficiency are brought into play at this stage in the argument, they appear to be playing an expressly justificatory role. Note, however, that efficiency in not an inherently moral consideration, in the way most people accept fairness is. When circumstances are unfair, someone is not getting her due. But this does not necessarily follow from something’s being inefficient, which implies only that something could be done better in one way or another. But if this is right then it 97 would seem that considerations of fairness are more likely than considerations of efficiency to justify coercion. Yet, as I argued above, even considerations of fairness do not seem up to the task. This would seem to imply that, at best, Wellman’s argument justifies coercing members of groups beyond the minimum number necessary only when that coercion is distributed according to the least coercive fair scheme. Presumably a state that embraced this principle would operate much differently than any we see today. Too Little The final concern I would like to raise about the legitimacy argument involves the scope of Wellman’s liberty limiting principle, which seems at once to justify too little and too much. Wellman has partially addressed the worry that samaritanism justifies too little. 136 Clearly states do a great many things that are not necessary to prevent our political community from falling into peril. They certify marriages, zone commercial and residential properties, and stipulate what will count as an organic egg. Faced with these functions, Wellman responds that samaritanism provides no obvious justification. He is content to say that, by itself, samaritanism may justify no more than a night-watchmen state. 137 While some might be disappointed that samaritanism justifies so little, justifying even a night-watchmen state would be a significant contribution to our understanding of state legitimacy. But does samaritanism justify even this much? Whether it does depends on which conditions make the state of nature perilous. Would it be perilous to live in a state that did 136 Wellman (2001, 758). 137 Though he goes on to suggest that other principles likely extend the samaritanism’s justificatory reach. 98 not, for example, enforce contracts, regulate fraud, or protect our individualistic conception of private property? It would likely be less efficient. 138 But samaritanism’s powerful justification for statism is grounded in concerns about peril, not efficiency. And so it’s unclear why samaritanism would support a state which did anything more than protect its members from others’ aggression. Yet, to justify this would be to justify significantly less than the classical liberal picture permits. Too Much More interesting, however, is the question of whether samaritanism justifies too much. JC seems to warrant a variety of peril-eliminating projects, many of which are more controversial than statism. Consider homelessness, for example. Each night in any major city, thousands of people, many of them children, live in a state of nature that exists in civil society’s shadow. With inadequate shelter, insufficient food, and little or no external support, these people are conspicuous targets for violent criminals, sexual predators, and drug dealers. At the same time, thousands of hotel rooms go vacant every night, sometimes only feet away from the sidewalks these people sleep on. Since samaritanism justifies coercion in cases in which “the peril of others generates weightier moral reasons than the presumption in favor of each individual’s dominion over her own affairs,” 139 JC would seem to justify 138 Though even this is not obvious. We can imagine, for instance, that if the state did not enforce contracts, a group of individuals would likely incorporate to facilitate guarantees between third parties for a fee. This corporation would then honor its contracts in order to maintain the reputation necessary to make a profit selling future trade policies. People might eventually refuse to cooperate with those who were not covered by a trade policy, giving everyone and incentive to join. Eventually such a company might regulate transactions as efficiently as the state does. 139 Wellman (2001, 746). 99 commandeering these rooms to save the homeless from the daily peril they face. It would be difficult to argue, after all, that sacrificing otherwise unoccupied rooms on any particular night is unreasonably burdensome to save someone from the perils of skid row. Wellman responds to too-much problems by pointing out that they could just as well be solved in other ways. Governments could build more shelters, and aid organizations could start work-live programs to get homeless people back on their feet. If these solutions are possible, then it is not necessary to commandeer hotel rooms in order to provide shelter for the homeless. JC’s necessity condition is therefore not satisfied. This is what separates the perils of homelessness from the perils of statelessness, on Wellman’s view: “it is not merely that political coercion is a possible solution to the harmful circumstances of the state of nature; it is the only viable solution….” 140 Rather than solving the problem, however, Wellman’s response seems to cast critical light on the necessity condition itself: What exactly does Wellman mean when he says that, to be justified, a particular coercive action must be necessary to avoid some perilous circumstance? Presumably he means that achieving the desired end would be impossible otherwise. 141 But impossible in what sense? He cannot mean logically impossible. There is, after all, nothing contradictory about a society living in harmony without a coercive state. Nor would such a society be nomologically impossible, or impossible given the laws of physics. I take it, then, that he means something like actually impossible, which is probably better 140 Wellman (2005, 38). 141 Weakening the necessity condition to mean, e.g., “most cost-efficient” opens JC to obvious counterexamples. For example, it would permit Alice to take Carolyn’s car even if Alice had her own car, so long as Carolyn’s car were more fuel-efficient and Carolyn wouldn’t need it. 100 interpreted as highly improbable. It’s highly improbable that I can be in New York by noon today, given that it is 9am and I am in Los Angeles. As things actually are, I have no way to get there by that time—though of course its possible that I’ll meet someone with a private jet. But notice that this more inclusive sense of “impossibility” cannot separate cases of homelessness from political cases, as Wellman contends. State coercion is effective because it provides a sufficient incentive for people to behave in ways they might not otherwise behave. But alternative, equally powerful incentives could be applied in these cases. Iceland, for example, has a population of roughly 300,000. Bill Gates is worth approximately $56 billion. Divided amongst Iceland’s inhabitants, Gate’s fortune would afford each Icelander roughly $200,000. Would it be possible to entice everyone in Iceland to follow rules which made a legal system unnecessary? Imagine for the moment that Icelandic citizens would obey the law to an extent sufficient to avoid peril if Gates offered them enough money. Since enticing them in this way represents an alternative to state coercion, it follows from JC that political coercion is unjustified in Iceland. But of course this is absurd. For while it is possible in each of the first two senses I mentioned above that Gates will sacrifice his entire net worth to entice everyone in Iceland, he almost certainly will not. But if this is the sort of “possibility” Wellman means to invoke in JC, then it again seems to apply to the perils of urban homelessness, not to mention a panoply of other cases. True, there are many ways to provide homes for homeless people. It could be, for instance, that everyone will wake up tomorrow and adopt one. But that’s about as unlikely as Gates deciding to use his fortune to entice Icelanders. So why is a human rights organization not as justified in commandeering hotel rooms as Alice is in taking Carolyn’s car, or my 101 government is in coercing me? In the absence of such coercion it is highly unlikely that we will achieve the good which justifies it on the samaritan account. It is worth emphasizing, before moving on, how far this principle actually extends. It is not just hotel owners who could contribute more towards eliminating peril, and not just human rights organizations who could force them to. So long as I do it fairly, there is no reason that I cannot go around coercing money from people on the street, so long as it is ultimately in the service of peril reduction, and the amount I take and the way I take it are not any less reasonable than the state’s methods. While I agree that a world without peril is something worth working for, I am skeptical that this is the best, or even a permissible, way to get there. There is one other way in which the samaritan argument seems to justify too much. For if coercion must be distributed fairly to be permissible, then Wellman must explain why the government is justified in coercively extracting more of a contribution from some people than it does from others. Consider for instance that the state forces the employed to both obey the law and pay taxes, while it forces the unemployed only to obey the law. So far as the state is justified in its treatment of the employed, then, it would seem that the unemployed are not doing their fair share to support society’s collective samaritan effort. You might think, of course, that the unemployed have nothing to give, and so that the state has nothing to collect. It is as though another woman, Denise, were present when Beth collapsed, but did not have a car. Obviously Alice cannot take Denise’s car if she does not have one. But that does not entail that Denise is not doing her share, you might think, for there is nothing more she can do. 102 But this analogy is flawed. Some amount of any person’s labor has the same peril- preventing value as the tax dollars I pay. So why, if I can be compelled to contribute my tax dollars, can the unemployed not be compelled to contribute their labor? Perhaps forced labor is a different and more significant form of coercion than taxation, in which case we might think that it is an unreasonable sacrifice to force people to make merely to ensure that they do their fair share to save others from peril. Yet this concern seems unsupported by the samaritan account. To see this, let’s divide Heart into two modified cases. In (HA), Beth collapses, but there is no car on hand, only a taxi. The taxi fair is $10, and the driver, who does not speak English, and does not seem to understand the gravity of the situation, will not drive to the hospital without fare. Alice does not have the money. However she sees a stranger, Carolyn, holding $10 on the street. JC permits Alice to take Carolyn’s $10 to get Beth to the hospital, even if Carolyn is unwilling to give it up voluntarily. In (HB) things are as they were in (HA), except that Carolyn is not standing on the sidewalk, but pulled over to the side of the road in her car, looking at a map. She does not have $10 for Alice to take for cab fare. The only way to get Beth to the hospital is therefore to force Carolyn to drive her there. Does JC permit this? I think Wellman would say that it does. (I would agree.) But this looks very much like the difference between taxes and forced labor. JC permits taking wages from people, but it also permits us to force them to perform certain actions. Unless there is a principled reason why the samaritan argument would permit the state to take my money without allowing it to take my labor, we must conclude that it would allow both, just as it would in (HB). To be clear, this is an objection to the use to which samaritanism puts considerations of fairness. Remember from our discussion above that, to avoid the charge that fairness is 103 doing justificatory work, Wellman had to claim that it functions as a two-level theory, such that the question of fair distribution comes up only after the permissibility of coercion is established. Thus, the unemployed person cannot claim that the government does not have the justification to force her to contribute; the most she can say is that forcing her in this way would be unfair. Whatever is wrong with forcing the unemployed to work, however, it does not seem to be that it’s unfair. More plausibly, it has something to do with restricting her freedom, which for whatever reason seems worse than taking money from wage earners. But then, the state is in precisely the business of restricting people’s freedom—indeed, that is how it saves all of us from peril. So the samaritan seems again to be in a dilemma. Either considerations of fairness do not permit restricting people’s freedom, in which case the state cannot coerce any particular individual to obey the law, or they do, in which case, since samaritan considerations permit the state to force the employed to contribute their fair share, they can distribute this burden fairly by forcing the unemployed to work. 142 3.2.2 Problems with Obligation We now shift our focus from legitimacy to obligation. It is more difficult to establish that A may justifiably coerce C to help B than it is that C has an obligation to help B. For this reason some of the objections I raised in the last section will not apply. For instance, my claim that it seems unprincipled to say that the state may force me to do my fair share of the 142 Robert Nozick (1971, 170) has made essentially the same point. “If it would be illegitimate for a tax system to seize some of a man’s leisure (forced labor) for the purpose of serving the needy, how can it be legitimate for a tax system to seize some of a man’s goods for that purpose?” An obvious response would be that the employed enjoy more of what society has to offer than do the unemployed and so that they owe more in return. However this line of reasoning is not open to Wellman, who holds that coercion which is justified by appeal to self-regarding benefits should be rejected. 104 samaritan chore when it may not permissibly force me to act fairly in other aspects of my life is irrelevant to the question of political obligation—to the question of whether I have an obligation to obey the law—because that question does not concern coercion. However, the structural similarities between JC and DoS allow us to apply most of our previous objections to the argument for obligation, most centrally, those focused on conditions (iii) and (iv). Reasonability and Necessity, Again Because we are now focused on obligation, we shift our focus from Alice (the coercer) to Carolyn (the coercee/obligor). It would be unreasonable for Carolyn to refuse to loan her car to Alice if doing so is necessary to save Beth’s life, and so it seems that she is obligated to loan it. As we saw above, however, Heart diverges from the political context in two important ways. First, Beth’s impending death is both more certain and more perilous than the fate of individuals in a stateless society. Second, the sacrifice demanded of Carolyn to prevent it is trivial. The question, then, is not whether we have samaritan duties to others—I am happy to grant that we do—but whether it is reasonable to demand that we do as the state prescribes in light of them. Earlier we substituted Thomson’s violinist case to try to better capture the costs imposed by a state. Thomson argues that, while remaining in the hospital would be laudable, you are not morally bound to bear that cost to save the violinist. Whether we agree or not, the relevant point is that the burdens the state places on us are so dramatically different from those Alice places on Carolyn that we must question the extent to which Heart usefully models the samaritan reasons we have to help others in our political communities. 105 Recall that to reach the opposite conclusion—that Heart faithfully represents the moral relationship between a state and its members—Wellman argued that each of our sacrifices are reasonable in light of the net-benefit states provide to those they simultaneously coerce. We conceded that most people certainly do enjoy net self-regarding benefits, but pointed out that this line of argument is unavailable to Wellman, who considers accounts which depend on them paternalistic. This problem, which we developed when we discussed the samaritan account of legitimacy, applies equally to political obligation. But let us set the reasonability concern aside and suppose that the perils of the state of nature resemble Beth’s peril in Heart closely enough to draw conclusions from it. While it remains impermissible, for the reasons provided in the previous section, for the state to force individuals to discharge their samaritan obligations, it begins to look much more plausible that these obligations would require us to bear costs as great as those the state imposes. Let’s suppose for the moment that they would in order to see how the samaritan case for political obligation explains the necessity problem. Recall that on the present account I have a samaritan duty to φ only when φ-ing is necessary to prevent some sort of peril. Yet my contributions to the samaritan effort are, not just unnecessary, but inconsequential. Here the case for political obligation purportedly diverges from the case for legitimacy. Just as it would be unfair for the state to force only some of its inhabitants to contribute to its samaritan aims, the samaritan account claims that it would be unfair of me to let others make the sacrifices necessary to save society from peril. But with respect to the argument for political obligation, fairness is not just a distributive 106 mechanism, but also a source of obligation. 143 While I think this is correct, it raises problems similar to the ones we discussed in the last section. For the fact that it is reasonable to φ if φ- ing is necessary to save someone from peril does not entail that it is reasonable to φ simply because not φ-ing would be unfair. Because which actions are reasonable depends on precisely what is at stake, shifting justificatory weight from samaritan considerations to concerns about fairness reduces the cost any person can be expected to bear. Return to our earlier case. Though it is unfair that my neighborhood is effectively quarantined each time there is a charitable race in downtown Los Angeles, the government cannot reasonably quarantine my northern neighbors simply to keep things fair. What fairness prescribes is, if anything, forcing other neighborhoods to compensate us for our trouble. The samaritan argument for political obligation is open to a parallel concern. Just as I might choose to keep a promise that the state cannot force me to keep, residents of nearby neighborhoods might choose to remain at home so as to minimize the unfairness residents of my neighborhood endure, or, if this is a poor choice based on some secondary principle, to mail some compensation if the state is not going to mandate that others contribute. If considerations of fairness are themselves a source of obligations on the samaritan account, as Wellman claims, then it follows that this is not just a nice thing to do, but a moral requirement. However it seems implausible that other neighborhoods have an obligation to limit their mobility or send me a check simply because we lose some mobility, or even that they should. This suggests that the normative force of our samaritan reasons do not transfer 143 Wellman (2001, 754). 107 to considerations of fairness, even when what is unfair is the distribution of the burdens necessary to discharge samaritan duties. Fairness and Transference Examples like these illustrate what I take to be one of the most serious difficulties facing the samaritan approach. To answer the necessity problem, samaritanism relies heavily on considerations of fairness. However these considerations will be weighty enough to do the work they must only if the force of samaritan obligations can be transferred from one claimant to another—something I do not find plausible. The transferable obligations with which we are most familiar are financial debts. The local bank from which I took my home mortgage, for example, eventually sold it to a commercial bank. But the fact that the local bank no longer has a claim on me to repay my loan does not mean that I no longer have the debt. The claim holder has changed, but the content of the debt remains the same. But not all obligations are transferable. If I promise you that I will attend a conference at which you’re speaking, you cannot transfer that obligation to someone with a greater interest in my attendance. By skipping out, I wrong you, or, if you’ve released me from the obligation, I wrong no one. But in no circumstance do I wrong anyone else. The same is true of obligations we owe to those in need. If I owe it to you get you a pharmaceutical you need to stay well, you cannot transfer your claim on me to a person willing to buy you cigarettes in exchange for pharmaceuticals. You might ultimately make that exchange after I’ve given you the drugs, of course, but I do not owe this service to anyone but you. 108 What reasons do we have to think samaritan obligations are like financial debts rather than other, nontransferable obligations? Before we can answer this question we need to look at a peculiar property of samaritan duties which I shall call elasticity. An elastic obligation is a collective obligation for which each member of the collective is fully liable. 144 If a bank loans $1m to an investment group on the condition that each member of the group be independently responsible for the total value of the loan, then the debt the group owes the bank is elastic in my sense. Samaritan obligations are elastic. Imagine that you and I are on a walk together when we see Andy and Bill drowning in a shallow pond. While the samaritan obligation to pull them out falls equally on both of us, each of us is nonetheless individually accountable for the “total value” of the obligation. This follows from the fact that, if I refuse to wade into the water to save either of them, then you must save them both. It would be no excuse for letting Bill drown that you already did your share of the job by saving Andy. Are elastic obligations transferable? The cases in which they are seem to be those in which there exists some sort of implicit understanding between the parties specifying that each of them owes a particular portion of their collective debt. For instance, four students might sign a lease agreement for the total rental cost of a house for which each of them understands that she will be accountable to the others for one quarter. If when the monthly rent comes due three of them cannot produce their share, the fourth will be liable to the landlord for the total amount. But if she pays it, then the others clearly now owe her their unpaid share. In essence the fourth roommate has purchased the debt of the other three in 144 A person strapped with an elastic legal obligation is said to be “jointly and severally” liable. 109 the same way that the commercial bank bought my mortgage. Each roommate’s debt is still owed, but now it is owed to another party. It seems to me that this is not a plausible model for samaritan obligations. On the samaritan account, we all share an elastic obligation to save others from peril. But if others are already doing the job, then why should I contribute? The samaritan claims that those who discharge our obligation have essentially bought my portion of it. My samaritan obligation to Andy and Bill has transferred to you, the samaritan says, because by saving them you did what I also had an obligation to do. True, the obligation to save both of them was yours all along (because the obligation is elastic), and so by doing nothing I cannot be accused of forcing you to do anything you were not already bound to do. Yet I am still obligated to repay you, the samaritan claims, for it would be unfair not to contribute, and this unfairness, as we already saw, is thought to carry the weight of the original samaritan obligation, just as one roommate’s rent debt transfers to the one who picked up her share of the monthly fee. The claimant changes, but the content of the obligation—the amount that’s due—stays the same. This is precisely what bad citizens owe good ones, who carry our collective samaritan burdens on their shoulders. I believe that this is an implausible model of samaritan obligation because samaritan obligations do not admit of implicit understandings between those on whom they happen to fall, and because these understandings seem necessary for elastic obligations to transfer. After saving Andy and Bill from the pond, you cannot go around collecting contributions from bystanders who could have done what you did. But the sense in which you “cannot” is not practical. It is certainly within your means to circle the area soliciting contributions. I mean, rather, that by discharging the samaritan obligation on your own you do not “buy” the 110 portion of it that fell on everyone nearby. While it is difficult to show definitively that this is true, thinking of samaritan duties this way runs counter to daily experience. I have never been asked, for the sake of fairness, to share in a cost born by someone acting for samaritan reasons. And never has it occurred to me to ask others for contributions, even when I would have welcomed them. Once, while travelling through Brussels, I was on a crowded subway platform when an elderly woman collapsed and fell (quite far down) onto the rails. Without much time for thought, three of us jumped into the pit and lifted her back onto the platform, climbing back out an uncomfortably few seconds before the train arrived. There were cheers all around. As I collected my things, however, I realized that I had damaged my laptop when I dropped my bags on the cement floor. It seems unfair that I had to bear most of the cost of the rescue. Yet it is strange to think that anyone standing on the platform owed me anything. (The same would be true if damaging my laptop was necessary to help.) Once we boarded the train, I sat with a group of onlookers. They found the action the three of us had taken commendable, and they were very sorry that my laptop had been damaged. Yet nobody offered to contribute to its repair. Why not? Everyone believed, I presume, that the circumstances had strapped us all, collectively, with an elastic obligation, one that most of us could have been held fully accountable for if everyone had failed to act. Everyone saw that a few of us had discharged that obligation alone, and that in doing so some of us had incurred costs. And everyone knew that these costs were strictly monetary, so anyone could easily have contributed any amount they desired simply by handing me money. Yet nobody offered. One possible explanation is that everyone thought the action was supererogatory. However, I doubt they did, just as I doubt it was. (I don’t know that I would have helped if I 111 did not feel like I had to help.) Perhaps, then, these were simply less than morally praiseworthy people; they knew what they ought to do, but chose not to. However this seems unlikely too. I talked to them, and they had the appearance of morally decent people who were genuinely sorry that I had suffered a loss. These seemed, not like people who wanted to avoid doing what they thought they were obligated to do, but like people who did not think they were obligated to do anything. It was unfortunate that my laptop suffered the fate that it did, they agreed, but it was not their responsibility to compensate anyone for it. Since all of the onlookers I spoke with appeared to know that a samaritan duty had been owed to the woman, and that it had been discharged by only a few of us, they must have believed, if they were the morally decent people I took them to be, that there was no question about fairly distributing the costs of our collective duty. This is, I think, what we see overwhelmingly, if not exclusively, in daily affairs. Samaritan duties that attach to the members of a group are often discharged by only a few of them, yet there is no question about fairly distributing the costs those people incur, even when everyone agrees about the facts of the case. We seem to think that in cases in which there is no prior agreement or understanding between members of the group, elastic obligations do not admit of such concerns. And so it may not be merely that the force of our samaritan duties does not transfer to considerations of fairness; it may be, more problematically, that in these cases fairness is not a relevant consideration at all. If this is the case, then considerations of fairness are not available for samaritanism to invoke to answer the necessity problem. If I am wrong about this, then at the very least we are owed an explanation for why samaritan duties must be distributed in this way, when similar duties need not be. 112 Particularity Wellman says at one point that satisfying the particularity requirement is the biggest challenge facing samaritanism. 145 I agree. For even if we concede that we have samaritan obligations to save others from peril, that it would be unfair for others to carry the samaritan burden alone, and that the seriousness of this unfairness gives us, not just a reason, but an obligation to contribute to the state’s peril-eliminating project, it is still not enough to show that we have a political obligation. A political obligation, at least in the sense that concerns us here, is an obligation to obey the law, and it is not clear why obeying that law is the only way I can acceptably discharge my samaritan duties. This is the first challenge of what I will call the particularity concern: Why must I discharge my samaritan duty in this particular way? 146 While relevant to all theories of political obligation, particularity has been especially problematic for theories which are, like Wellman’s, based in natural duty. John Rawls advanced what is perhaps the best-known natural duty account in A Theory of Justice, in which he proposed that each of us has a duty to “support and to comply with just institutions that exist and apply to us.” 147 Because compliance is built directly into the content of our duty on Rawls’ picture, and our state’s law applies to us, it follows that each of us has an obligation to obey its laws. It therefore avoids the particularity concern’s first challenge. But in doing so it opens itself to a second one. 145 Wellman (2004, 97). 146 The “particularity requirement,” first introduced by John Simmons (1981, 31), corresponds to the second challenge of what I am here calling the particularity concern. The similar nature of these two concerns should make it clear why I’ve grouped them together. 147 Rawls (1999, 99). 113 We conceive of political obligations as, in Simmons’ words, “special relationships” between states and citizens which in some important sense tie them together. 148 It is in virtue of this relationship that it would be strange were I to pay taxes in Canada, rather than in the United States, where I live and work. Presumably there is something special about the fact that the United States is my country that explains why I have special bonds to it. Yet it is difficult to see how Rawls’ theory accounts for this. 149 There are, after all, many just institutions that seem to apply to us: some are states, some are non-state political institutions, and some are non-state non-political institutions. Yet we do not think that we are obligated to comply with all of them. This is the second challenge of the particularity concern: Why must I comply with this particular state, rather than other institutions? In other words: What makes my state special? 150 To answer the particularity concern’s two challenges without abandoning the normative force of the Rawlsian approach, Wellman must explain (i) how the somewhat nebulous samaritan duty to save others from peril requires the specific behavior demanded 148 Simmons (1981, 34). 149 One might suggest that Rawls is not open to this objection simply because his theory is supposed to apply only to those people with whom I cooperate to produce goods. But even if this is the relevant metric, some Americans cooperate more with people in Mexico than they do with people in, say, Alaska. And even if we were to artificially limit Rawls argument to domestic institutions, this still would not explain why domestic non-state organizations could not bind me by telling me what to do. 150 Simmons (1979, Chapter 6). Jeremy Waldron (1993) responded to Simmons’ critique by offering a more accommodating account of Rawls’ application condition. He argued that we should be concerned, not with merely internally just institutions, but with those institutions that promote justice. Wellman (2004) puts pressure on Waldron’s account, claiming that, to draw the distinctions he must, Waldron depends, not on the justice promoted by such institutions, but instead on the peril those institutions usually help individuals avoid (104-05). Absent these samaritan considerations, Wellman claims that Rawls’/Waldron’s natural duty account cannot answer the particularity concern. According to the dilemma at which he gestures, the justice account either fails to explain what it aims to explain, or collapses into a samaritan account. 114 by political obligations, without (ii) tying us to foreign governments or non-state institutions in the same stroke. 151 He addresses both challenges the same way. Imagine you have a neighbor, Sam, who, though he has an income that is significantly less than yours, chooses to sacrifice significantly more than you do in taxes. Only rather than send a check to the government each April, he sends it to Haiti, where he believes it will do more to eliminate peril. Is Sam doing everything that can be expected of him on the samaritan account? According to the argument we have seen to this point, it would seem that he is. There are two basic obligations to consider: one’s fundamental samaritan responsibility to help others in peril (explicated in DoS), and considerations of fairness in distributing the burdens imposed by that responsibility. Yet Sam seems to satisfy both. He satisfies his samaritan responsibility by so devotedly addressing the peril in Haiti, and he does his fair share, at least with respect to you and me, because he sacrifices significantly more than we do. How could anyone say that he is not doing his part? Wellman claims that, though Sam does more than the rest of us, there remains an important respect in which he also takes advantage of us, and that his obligation not to act in this way explains why he has an obligation to contribute to the state. Wellman’s contention is that, by selecting for himself how he will discharge his samaritan obligations, Sam unfairly helps himself to a benefit that would not exist without our hard work: discretion. Like all goods, we value the discretion we have to make choices in our lives, choices about the careers we pursue, the people we marry, and the religions we practice. Sam enjoys the discretion he exercises when he chooses to send money to Haiti rather than to the U.S. 151 The particularity concern’s two challenges are divided into three separate but related questions in Wellman (2004, 99-100). 115 government. But it is available to him “only because the rest of us obey the law.” 152 That is, it is only because we take care of the peril at home that Sam has the luxury of choosing to eliminate peril abroad. By enjoying his discretion without contributing to our local project, however, he takes advantage of us; for in making the particular sort of contribution we do, we cannot enjoy the discretion we produce for people like Sam. This makes Sam a free-rider in spite of the fact that, when all is said and done, he eliminates more peril than the rest of us do. To permissibly enjoy the discretion he takes, Sam must first contribute to our collective political project at home. The response to the first challenge of the particularity concern is essentially the same. When deciding how to support one’s own state, there are many options. But since we “could reserve this discretion for ourselves only against a backdrop of general compliance in which others are denied this discretion, clearly it would be wrong to unfairly make an exception for ourselves.” 153 Thus, we must obey the law of our state for the same reason that we must support it before we support other political and non-political peril-eliminating institutions: to use our discretion would be to ride free on the non-discretionary sacrifices of others. There are two problems with this argument, the first of which ties in nicely with our discussion in Chapter 2. Up to this point in the argument, Wellman has cited distributive fairness to extend samaritanism’s normative reach. This is important, for he rejects theories of fair play that, like Klosko’s, depend on the unfairness that results when individuals fail to pay for the benefits they unavoidably receive. But to avoid the particularity concern, 152 Wellman (2005, 42). 153 Wellman (2005, 45). 116 Wellman makes a subtle change to his argumentative approach. He contends, not that one should do one’s fair share of the peril-eliminating project, but that one should not unfairly take a good others have worked to produce. In doing so he offers, not a samaritan justification, but precisely the sort of reciprocity concern on which Klosko builds his theory of fair play. Recall that at the heart of Klosko’s account was the claim that it is unfair, and so impermissible, for me to consume goods others cooperatively produce without contributing to their production myself. I must do my part to contribute to the infrastructure, medical services and military defense that others provide for me. Wellman makes the same argument here. Only he is concerned with discretion rather than these other public goods. It is unfair, Wellman says, that Sam has the discretion to decide where to send financial support when that benefit “is available only because the rest of us obey the law.” We produce Sam’s discretion, so he should contribute to our collective project. Wellman rejects the fair play account of political obligation because he believes that people can only be bound by receiving or enjoying goods they have accepted in a morally relevant way. However, on his view one cannot accept a good in this way unless she is free to reject it, so he infers that, since public goods in states cannot be rejected, they cannot be accepted either. 154 He concludes therefore that fairness theory cannot explain the obligations we have to contribute to our political community’s collective project. Two things are worth noting here. First, Wellman’s objection to fairness theory is suspicious. Why must one be able to reject a good to accept it a morally relevant way? 154 Wellman (2001, 738). 117 Imagine that I have legally immigrated to the U.S. with my children, and do not realize that in the U.S. children are legally mandated to attend school. I see them off to school on the first day of term anyway, taking advantage of this publicly available good. According to fairness theory, I have an obligation to contribute to this public good now that I benefit from it. But this is a benefit I could not have rejected. Had I tried to keep my children from going to school, the police would have taken them there anyway. Wellman is therefore committed to saying that fairness theory gets things wrong in this case because schooling is not a benefit I can accept in a morally relevant way. But it is Wellman, and not the fairness theorist, who seems to get things wrong in this case. For the fact that I could not have rejected public school for my children does not seem to show that I did not accept it. 155 But even if Wellman’s critique were sound, why would it not apply equally to his own theory? Since there is no question that he relies on same sort of unfairness that theorists like Klosko do—the sort of unfairness one displays by enjoying a publicly produced good without contributing to it herself—Wellman must think that discretion is importantly different from the public goods theorists like Klosko discuss. In particular, he must think that discretion can be accepted in a morally relevant way. But his theory then implies that people like Sam can reject discretion. What would this mean if it were true? It seems clear that Wellman is not talking about practical discretion, or the general freedom to choose how one will act. If anything, it is Sam who loses his practical discretion, for practical discretion is what one loses when one is coerced into doing something, and the government applies coercive pressure on Sam to pay 155 This example is merely an adaptation of Harry Frankfurt’s well-known argument against the principle of alternate possibilities, an important principle in the free will debate. See Frankfurt (1969). 118 his taxes. What I think Wellman is interested in, however, is the moral freedom we have to discharge our obligations in one of a number of permissible ways—what we might call moral discretion. A person with only one avenue by which to discharge an obligation lacks moral discretion in the same way the bank teller lacks practical discretion during the holdup. The amount of discretion a person has is a function of the morally permissible options she has to choose from. This conception of discretion makes sense of Wellman’s claim that we provide Sam with discretion by fulfilling our moral obligations at home, for by doing so we provide Sam more ways by which to discharge his own samaritan obligations. His behavior is impermissible because he takes an unfair share of this sort of discretion. The second point, then, is that even if Wellman were right that goods cannot be accepted unless they can be rejected, he cannot treat moral discretion as a good without presupposing the very thing he has set out to prove, namely, that our duty of samaritanism directs us to obey the law in our own political state before we support other peril-eliminating institutions. Look again at Wellman’s argument. The central claim is that Sam unfairly exercises his discretion when he sends money to Haiti rather than the U.S. government. 156 It is, he says, “only because the rest of us obey the law” that Sam able to address samaritan concerns abroad, for we create the context in which Sam is able to behave this way. And that’s not fair, Wellman says. But to suppose that our work at home allows Sam to send his money abroad implies that, were we to shirk our samaritan chore, Sam would have to redirect his efforts domestically. For if our samaritan deeds make his choice permissible— and thus “possible” in the relevant moral sense—then their absence should make it 156 The argument here is the same whether we focus on the first or second challenge of the particularity concern. I focus only on the second challenge for simplicity. 119 impermissible. But why would Sam’s morally permissible options change if we acted differently? For our actions to matter, one must assume that there is already a natural lexical ordering to the perilous circumstances individuals are morally bound to address. One must assume, in other words, that Sam’s duty, like each of ours, demands that he eliminate the peril in his own political community before addressing peril abroad. But this is just the proposition that Wellman’s argument was supposed to prove. 157 It seems, therefore, that Wellman’s argument fails to answer both parts of the particularity concern. 3.3 Summary and Conclusion I began this chapter by outlining the samaritan case for both state legitimacy and political obligation. There arguments turned out to be parallel in important ways, and we saw that by developing problems facing the legitimacy argument we were able to discover important problems with the argument for political obligation as well. The most important concerns we raised involved the reasonability and necessity conditions of both JC and DoS. With respect to the legitimacy argument, it seemed far from clear that the state could reasonably coerce its inhabitants. And while coercion would seem more permissible were we to take into account the benefits states provide for the people they coerce, this net-benefit argument was shown to draw on precisely the sort of considerations Wellman rejects as paternalistic. 157 Another suspicious feature of this argument is that it implies that I act unfairly when I discharge my obligations in ways it would not have been possible to discharge them unless you had performed some other action to which I did not contribute. Yet this seems absurd. Suppose again that you and I find Andy and Bill drowning in a pond. Andy doesn’t know how to swim, and so I intend to try to save him before worrying about Bill, who does. Before I can do anything, however, you dive in after Andy. I now have a choice—I can help you with Andy, or I can head over to Bill. Wellman’s argument suggests that I somehow take advantage of you by trying to save Bill and leaving you to save Andy on your own (something you’re fully capable of doing). However this seems clearly false. 120 But even were we to concede that coercion is reasonable in the political context, it is unclear how the samaritan considerations that support this conclusion apply when no particular individual’s participation is necessary to achieve the conditions that justify it. This concern about samaritanism’s necessity condition extends to its arguments for both legitimacy and obligation. In each case, samaritanism invokes fairness to explain why the burdens it justifies can be distributed across the population. In the legitimacy argument, we saw that choosing anything other than the least coercive fair distribution would require a morally relevant principle with heft enough to carry justificatory weight, and that, without an explanation, it is not clear that a principle of efficiency has what it takes. The argument for political obligation, in which Wellman allows that fairness carries justificatory weight, presupposes that considerations of distributive fairness govern our samaritan obligations, which are elastic. I argued that elastic obligations seem to be governed in this way only when they fall on groups whose members have a preexisting understanding about the distribution of those obligations between them. Since no such understanding exists in the political context, it seems plausible, not just that the force of samaritan concerns do not transfer to considerations of fairness, but that the concerns of fairness seem detached from samaritan worries altogether, at least in typical cases. If this is right, then it leaves samaritanism with no way to explain why the state may coerce everyone within its borders, or why its members have an obligation to contribute to the state’s peril reducing project when they need not. Finally, we saw that samaritanism’s case for political obligation faces an important dilemma with respect to the particularity concern. If Wellman relies on the notion that discretion is a good, then he not only utilizes an argument he claims should be rejected, but presupposes the very fact he sets out to prove. On the other hand, giving up the argument 121 from discretion would leave him with no way to explain why we must support our state over other institutions, or why we must obey the law rather than support our state in another way. He leaves us with nothing to say to Sam, who, besides doing his samaritan duty, sacrifices more than the rest of us, and so cannot be said to do less than his fair share. 158 This leaves Wellman’s theory looking no more capable of answering the particularity concern than Rawls’ theory was—another victim of the problem that brought down the natural duty theories preceding his own. 158 I should note that at points it seems as if Wellman is attempting to preempt this sort of objection. He says, for instance, that “the perils of the state of nature are always the most pressing,” indicating that other perils, such as those in Haiti, might sometimes take priority. Yet it is hard to see how this is consistent with his other claims. “Of course,” he writes, “virtually all of us would prefer to determine the content of our sacrifice, but the political solution to the perils of the state of nature is predicated on denying individuals this sort of dominion, so an integral part of the sacrifice is to have little to no say in what our sacrifice must be. If one reserves for oneself this individual dominion (by ignoring the law so that one might contribute more to some apolitical peril, by contributing to political stability elsewhere, or even by contributing to one’s own state in an extralegal manner), then one is helping oneself to an unfair portion of discretion, free riding on others who have forgone this discretion” (2005, 45-6). As I said, it makes sense to say that someone like Sam exercises his discretion unfairly only if we suppose that Sam is able to support Haiti only because others are doing what Sam ought to or would otherwise have to do. But this is precisely the question. Why would Sam be morally obligated to put his own state above states like Haiti if others did not? To assume that he would is to assume rather than show that his samaritan obligation is a political obligation. 122 Part II The Authority Theory of Political Obligation My goal in Part I was to demonstrate the serious problems facing current theories of political obligation. I would now like to begin developing my own positive theory, which I believe to be explanatorily superior to the two theories we’ve so far considered, as well as others currently on offer in the literature. I begin, in the following two chapters, by reconstructing Joseph Raz’s work on both obligation and authority, which I use as the foundation for the authority theory I defend in subsequent chapters. 123 Chapter 4: The Structure of Obligations Obligations are the normative currency of communities—the relations we use to organize our lives together. They provide us with reasons to perform the actions they recommend, or constrain the force or applicability of the reasons we have to do something else, while our capacity to unilaterally impose them on one another partially constitutes the status and significance of our relationships. The aim of this chapter is to explore the way in which the formal structure of obligations differs from that of reasons and requirements of other sorts. I hope to show, first, that understanding what it is to have an obligation is just to understand the special way in which an obligated person is “bound” by practical reasons, and, second, that to be bound in this sense is just to have practical reasons which are structured in a particular way. Determining what obligations are is therefore a much different sort of enterprise than that of determining the weight they carry as reasons for action, or what a person with an obligation ought to do, both of which depend on various contextual factors. The structure of obligations is, or so I will begin to argue in this chapter, unified across cases. We will begin by looking at the intuitive thesis that obligations are simply reasons of abnormally great weight. After presenting three problems with this view that I suggest any 124 theory of obligation must answer, I outline a theory, presented by Joseph Raz, which seems to answer them. We are interested in Raz’s theory largely because of the novel theory of political obligation it can be used to construct—what I call the authority theory of political obligation—which is the focus of later chapters. I believe that Raz’s views on obligation and authority have been misunderstood, and that understanding them properly provides valuable insights into the role obligation, and thus political obligation, plays in our practical reasoning. I highlight these below. Later in the essay I consider and try to answer some objections to certain features of Raz’s view. 4.1 The Simple View of Obligations To tell someone that she has an obligation to do something is to tell her something importantly different than that she ought to do it. You ought to do something if you have most reason to do it. But obligations seem to do more than recommend some course of action. They are, as A.J. Simmons has put it, “limitations on our freedom” and “impositions on our will.” 159 One way to flesh out this difference is to say that to have an obligation to φ is to be required to φ. Understanding obligations as requirements captures, Simmons suggests, the special way in which obligations exert “a serious pressure to perform.” 160 Because a requirement is, as he says, a “special” sort of reason for action, it follows that to have an obligation to φ is to have this special sort of reason to φ, one that places serious pressure on 159 Simmons (1981, 7). 160 Simmons (1981, 7). I discuss the relationship between obligations and requirements in detail in Chapter 6. 125 one to act in a particular way. 161 W.D. Ross nicely illustrates this suggestion in a passage from The Right and the Good: When a plain man fulfils a promise because he thinks he ought to do so, it seems clear that he does so with no thought of its total consequences, still less with any opinion that these are likely to be the best possible. He thinks in fact much more of the past than of the future. What makes him think it right to act in a certain way is the fact that he has promised to do so—that and, usually, nothing more. 162 That the plain man thinks more about the past—about the obligation he is strapped with— than about the future evidences the sort of pressure Simmons highlights. Our aim here is to find the right way to understand it. A popular way of understanding the unique force of obligation has been to explain it in terms of another force that we presumably understand better. Some commentators have suggested, for instance, that obligatoriness is akin to a demand on our behavior. 163 But while it may seem that obligations stand to reasons as demands stand to recommendations, I am skeptical of the explanatory value this likeness offers. A demand must have an issuer whose status relative to her addressee determines and explains its unique force. It is hard to see, then, how this conception of obligation could add clarity without a freestanding theory of demands which explained in what sense they are binding on addressees independently of 161 The reasons we are concerned about here are normative reasons, or reasons that count in favor of acting in a particular way. I will consider reasons, following Raz, as facts rather than statements or beliefs. (See Raz (2002, Chapter 1).) Note also that I will, for simplicity, sometimes speak as though actions and events are reasons for particular actions. By this it should be understood that the fact of some action or event’s occurrence provides a reason for action. 162 Ross (1930, 17). 163 Anscombe (1958) takes a harder line, suggesting that, indeed, ‘obligation’ has “no reasonable sense” beyond a law conception of ethics. Pink (2004) rejects Anscombe’s attempt to identify obligations with demands, but holds that the force of obligation can be best understood by reference to demandingness. Pink’s more moderate view is what I have in mind here. 126 those addressing them, or in what sense an obligation should be understood to have an implicit addresser in each instance. Another suggestion, which I shall call the simple view, has it that an obligation to φ is “special” in that it is simply a very weighty reason, and proposes that an actor, to decide which action she ought to take in any given context, is simply to weigh her obligations to φ, along with any non-obligation reasons she has to φ, against any reasons she has not to φ. The action supported by reasons of the greatest aggregate weight is the one she ought to perform. 164 Here the supposed weightiness of an obligation-reason can be used to explain why an obligation to φ, even where φ-ing is itself a relatively trivial action, might condemn failing to φ even when the availability of a more attractive but incompatible alternative exists. For instance, suppose I promise to help A move tomorrow in a context in which I thereby generate for myself an obligation to help A move. Should a conflicting opportunity arise when I am scheduled to help A—say, B invites me to a concert I would love to see—then it seems that, even if my reasons for going to the concert would far outweigh those I would have had to help A sans my promise, the fact is that I did promise, and the obligation generated by my promise is a weighty reason that overrides the reasons I have to go with B. In such a case the balance of reasons, which includes my weighty reason, favors helping A, which I ought therefore to do. The same considerations would also seem to explain the serious pressure which inclined Ross’ “plain man” to look into the past when deciding what to do. Because the great weight of the reason provided by his obligation to keep his promise in all likelihood 164 H.J. McCloskey (1969, 243) seems to advance a view of this sort about prima facie obligations. 127 outweighs any reasons that the consequences of his performance might generate against it, he doesn’t need to probe into the future to have a good sense of what he ought to do. Whether or not the simple view provides an intuitive explanation in these cases, three problems highlight its inadequacy as a theory of obligation. First, the simple view does not explain why obligations, even those created by the same obligation-generating action type, do not always or even usually have similar weights. An obligation created by one promise, for example, can be much weightier than one created by another. If the fact that A’s reason to φ is an obligation-reason is supposed to explain why it should have an abnormally great weight in A’s deliberation, then it would seem that other obligation-reasons should have a similar weight simply in virtue of their type. This is most clearly true in circumstances in which A has two promissory obligations, O 1 and O 2 , which require her to perform two equally trivial acts, φ and ψ. Supposing that A would have reasons of roughly equal weight to φ and ψ sans her promise, the simple view would seem to predict that by promising to perform these actions A would generate reasons for each action which, though weightier than those she would have had otherwise, are presumably still roughly equal relative to one another. But this is obviously not the case in many circumstances. My promise to you to meet you for lunch and my promise to my mother to go for a walk do not necessarily generate obligations of similar weight. Whether they do will depend on other facts about the case. But then what features do these obligations share as obligations? It could be argued that they share an unusual weight as reasons, one which distinguishes them as a class despite the fact that weights within that class can be quite different. One would then need to explain how the way in which one undertakes certain obligations causes them 128 to have greater weight than others, even than others created by the same type of obligation- generating act. The problem with this threshold account, however—and this is the second problem with the simple view—is that some obligations simply do not provide reasons of great weight. In fact, some obligations have little enough weight that they can be defeated even by prudential reasons, or by reasons that rest on inclination. 165 My promise to join the large crowd of your friends cheering you on at your swim meet may be outweighed by the fact that I’ve been offered an important job interview at that same time. 166 The third difficultly runs in the opposite direction: identifying obligations merely with reasons of a particular weight forces us (until we have some explanation otherwise) to identify reasons that are particularly weighty as obligations. But this is clearly wrong also. I would have weighty reasons to take a well paying philosophy job in a major city in California where I could surf. But I would not have an obligation to take it. Just as one can have an obligation that amounts to a relatively weak reason to φ, one can have very strong reasons to φ without having an obligation to φ. An entailment between obligations and weighty reasons runs in neither direction. 165 Many philosophers reject the idea that reasons related to self-interest are capable of outweighing obligations. Simmons (1981, 7): “Obligations…must be discharged regardless of our inclinations”; Gilbert (2006, 40): “If I owe you this action, how can I appropriately argue that my own inclinations or considerations of self-interest, as such, permit me not to give it to you?”; Gans (1992, 21): “[A duty] has absolute precedence over certain kinds of reasons, such as those of personal convenience or financial interests…” However, in rejecting the claim that moral reasons (which I take obligations to be) cannot be outweighed by prudential considerations, I am also in good company: see, e.g., Foot (1978a). 166 A defense of this claim would require a lengthy treatment of the contextual factors affecting obligations. Below we will discuss some such factors which bear on the structure of obligation- reasons. A more complete account of the contextual determinants would take us too far afield. 129 What the simple view fails to do, then, is provide an explanation of the “special” sort of reasons obligations are, those which place “serious pressure” on individuals that have them, in a way that takes seriously the variable and sometimes-minimal weight reasons of this sort can have. In the next section we explore one way in which obligations can be understood to provide these special reasons without appeal to their weight. These considerations illuminate phenomena that any acceptable theory of obligation must explain, and are an essential component of the authority theory of political obligation that we will consider later. 4.2 Exclusionary and Preemptive Reasons A device by which to gain traction over some of the challenges in the last section is suggested by the following vignette of a woman, Ann, who is looking to invest her money. 167 Ann is presented with a complicated investment opportunity to which she must commit tonight if she is to invest at all. The drawback is that she knows she is too tired tonight to intelligently weigh the reasons for or against investing. Recognizing this, she decides against hearing the reasons on either side. When told that deciding against even hearing the reasons is tantamount to deciding against the investment, Ann concedes the point. She also concedes that the fact of her tiredness is not a reason not to invest, for that fact does not bear positively or negatively on the merits of the investment. Ann maintains, however, that though the reasons may ultimately weigh in favor of investing, and though she may even be 167 This example is taken from Raz (1999a, 37-38). 130 brought to believe that they do, it is still reasonable, given her tired state, to refuse to consider the case for the investment at all, since she is too exhausted to trust her ability to deliberate well. Ann’s claim cannot be easily explained on the deliberative model provided by the simple view (other than to say that it’s false). For according to the simple view, if the reasons for investing outweigh the reasons against it, then Ann ought to invest. The insight Raz provides here is drawn from Ann’s experience. She claims that, though her tiredness is not a reason to invest or not, it is a consideration that excludes considering the reasons on either side. Raz calls this an exclusionary reason. Exclusionary reasons are considerations that defeat certain reasons for action in virtue of their categorical standing relative to those reasons, rather than by outweighing them. To support this observation Raz contends that reasons are of different orders, exclusionary reasons being, on this picture, one breed of what he more generally refers to as second-order reasons. 168 Second-order reasons are reasons that either count positively in favor of acting for certain reasons, or negatively against acting for certain of them. The reasons they count either in favor of or against are first-order reasons—reasons of the sort we cite to justify our actions on a day-to-day basis, which depend directly on our interests, needs, desires, beliefs, values, etc. That I am tired is a first-order reason to take a nap, and that you are lonely is a first-order reason for you to visit a friend. 168 Raz, (1999a, 39-41). 131 An exclusionary reason is a second-order reason of the negative sort. 169 To illustrate, say that r is a first-order reason for A to φ while n is a first-order reason for A not to φ. To decide what she ought to do, A simply weighs r and n against each other. The introduction of an exclusionary reason like Ann’s complicates this picture. If A has an exclusionary reason e for refraining from acting for n, then A must now, because of e, abstain from including n in her practical deliberation (in a manner to be explained below), leaving her with only r left to consider when deliberating about whether or not to φ. Applying this model, Ann acts rationally when she elects not to consider the reasons for or against investing (which is tantamount to deciding against investing) because her tiredness is a reason to refrain from considering reasons both for and against the investment. To this point we have equivocated about the precise role exclusionary reasons play in our practical reasoning. Do they exclude or simply weigh against those first-order reasons to which they apply? And must one refrain only from acting on excluded reasons, or must one also refrain from considering them in one’s practical deliberations? These are open questions for which Raz’s occasionally loose language is partly responsible. Regarding the first, Raz says that an exclusionary reason is “a second-order reason to refrain from acting for a reason.” 170 This is misleading because a reason e need not be truly “exclusionary” to be a reason to refrain from φ-ing for reason r—e might just weigh against r. For example, you 169 Second-order reasons of the positive sort are less common, but should still be familiar. That my daughter is sick is a reason to stay home. But I also have a reason to stay home for this reason. (Raz calls acting for a reason complying with it. Though I shall not adopt his usage.) The idea is that something is amiss if the reason I stay home is, say, that I would like to watch a soap opera, and not that my daughter is sick. In cases such as this, I not only have a first-order reason to perform a particular action, Raz claims, but I also have a second-order reason to perform that action for a particular reason. See Raz (1999a, 178). 170 Raz (1999a, 39). 132 may be greatly upset with me if I lie to you, that is, if I tell you a falsehood for the reason that I will thereby come to deceive you. So, say that φ represents some particular act of telling you a falsehood, and that r represents the reason that I will thereby come to deceive you. That you will be upset with me, not if I φ, but if I φ for r is a second-order reason to refrain from φ-ing for r. It counts not against telling you a falsehood—which I might do accidentally—but against doing so in order to deceive you. A charitable reading of Raz avoids this confusion by adopting something closer to the description I provided above, namely, that exclusionary reasons defeat first-order reasons in virtue of their categorical standing. 171 This means that exclusionary reasons do not weigh against but rather exclude those first-order reasons to which they apply. 172 4.2.1 Hart and Raz on Deliberation and Action Determining whether exclusionary reasons apply to both action and deliberation is more challenging. 171 It is not clear just what Raz means by ‘categorical.’ One possibility is that exclusionary reasons that truly exclude are “categorical” in virtue of their content-independence. Content-independent reasons are typically understood as reasons with no direct connection to the actions for which they are reasons. (On this see Hart (1982, 254); Raz (1988, 35). I modify this definition in Chapter 5 for reasons I explain there.) Promises and authoritative commands are examples. It is unclear, however, why a reason should be content-independent to exclude. Raz’s own example of Ann seems to include an exclusionary reason which is not at the same time content-independent. Whatever Raz intended, the notion of obligation I go on to develop below and in Chapter 6 presents a clear sense in which obligations are categorical relative to first-order reasons. 172 This reading is supported by Raz’s claim (1999a, 40) that conflicts between first-order and exclusionary reasons “are resolved not by the strength of the competing reasons but by a general principle of practical reasoning which determines that exclusionary reasons always prevail, when in conflict with first-order reasons.” Note, however, that this can be misread too. Not all reasons are excluded by exclusionary reasons. They exclude only those which are within their exclusionary scope. 133 Taking what appears to be a stronger, deliberative position, H.L.A. Hart claims that the subject of a command (a paradigmatic sort of exclusionary reason) is meant to “take the commander’s will instead of his own as a guide to action and so to take it in place of any deliberation or reasoning of his own.” The command is supposed to “cut off any independent deliberation by the hearer of the merits…of doing the act.” 173 While there is some confusion surrounding Raz’s view 174 , he suggests on numerous occasions that exclusionary reasons exclude only acting for certain reasons. 175 “Surely what counts, from the point of view of the person in authority,” Raz says, “is not what the subject thinks but how he acts.” 176 Since it is Raz’s view that authoritative directives limit the subject by providing exclusionary reasons, his response seems to indicate that exclusionary reasons (at least those that are constituted by authoritative commands) only restrict the reasons on which one can legitimately act. It seems to me, however, that the deliberative interpretation is compatible, on the appropriate reading, with everything Raz says. However this can be seen only once we disambiguate the modes by which we deliberate or think about our reasons. Combined with 173 See Hart (1982, 253-55). 174 For further notes on this confusion, see David Owens (2008). Owens and I appear to favor a similar resolution. However his approach to the question is framed by a debate, introduced explicitly by Raz (1999a), involving conformity versus compliance with reasons. Since these terms are misleading in their own right, I cannot be sure that we agree outside of this muddied context. 175 For example, see Raz’s wording in (1999a, 39, 40, 41, 183, 184). Moreover, the definition he provides in the 1999 postscript to the book in which he introduces the notion of an exclusionary reason says explicitly that exclusionary reasons are “reasons not to act for certain reasons” (1999a, 183). The same line of reasoning appears elsewhere as a response to H.L.A. Hart’s discussion of authority. 176 Raz (1988, 39). In a footnote, however, he concedes to earlier holding the view he there criticizes. Also see Raz (1974). 134 the interpretive principle that, unless there is overwhelming evidence to the contrary, one should interpret a text in a manner that supports the strongest view compatible with it, these two considerations lead me to read Raz as supporting the stronger act/deliberation view. 177 One reason for rejecting the action-only interpretation is that it fails to explain the role of exclusionary reasons in our practical deliberation. For reasons that will become clearer as we move along, the exclusionary reasons we are interested in are justified so far as they save time or labor, minimize error, or are constituted by directives issued by an authority either (i) to secure coordination for certain activities or (ii) in a manner that causes one to act in accordance with her own reasons better than she could have were she to try to act on the balance of reasons in those cases in which the authority’s directive applies. When one’s circumstances include an exclusionary reason satisfying these conditions, weighing the reasons it excludes undermines the value promoted by the fact that constitutes the exclusionary reason. A timesaving rule is an example. A rule that is justified because it saves a decision maker time is useless to a person who insists on weighing all of the reasons extant in those circumstances in which the rule is meant to apply. Consider a person who makes a rule always to follow the directions provided by Google when on a long trip. She has occasionally found Google’s directions to be sub-optimal, but she has also determined that, notwithstanding these anomalies, faithfully following her Google directions in all cases will end up costing her less time, over her many trips, than that which she spends cross- referencing maps to ensure that her directions are optimal in those cases in which she has 177 See Lloyd (1992, 15-26). The principle I cite combines Lloyd’s plausibility and theoretical power criteria of interpretation, which she attributes to Ronald Dworkin. 135 doubts. If her aim is to minimize the time she spends traveling overall, then her rule does nothing for her if she treats it only as a reason not to act for certain reasons. Even if she eventually acts only on non-excluded reasons, and so does everything right according to the action-only interpretation, this does nothing to restore the time she loses deliberating about excluded ones. And so were we to understand exclusionary reasons of this sort as reasons merely not to act for certain reasons then their utility in practical reasoning would dissolve in these cases. 178 Since Raz contends that these sorts of exclusionary reasons are useful in practical deliberation, we should, if we can, interpret his view so that it supports his contention. And there is, as I said, textual evidence that Raz supports something like the deliberation view Hart defends. But if this is right, then why the competing interpretations? The confusion can be traced back to the fact that “deliberate about,” “attend to” and “think about” are coarse terms which are used interchangeably to describe importantly distinct mental processes we perform with respect to reasons. I propose that a more finely grained specification of these functions may align Hart and Raz’s views. If this is right, then the same concepts can be used to determine in which ways we can go criticizably wrong when we are bound by obligations, as we shall in Chapter 6. 178 Similar arguments could be made for exclusionary reasons justified in any of the other ways mentioned above. 136 4.2.2 Identifying, Evaluating, and Entertaining Reasons There are at least three different mental operations we perform with respect to reasons. 179 The first and most fundamental is to identify a reason, which is to take initial stock of it as a reason, and so far as it is necessary to do this, to assign it a preliminary weight. I identify the fact that my library book is due today as a reason to go to Doheny Library to turn it in, not when I recognize that the book is due today, but when I recognize that its being due today counts in favor of going to Doheny, or to put it more formally, when I see that its being due today is a premise in a practical inference the conclusion of which is that I ought to go to Doheny. Once I have identified a reason, I can evaluate it, which involves analyzing it in any of a variety of ways, including (but not limited to) assigning it a weight, weighing it against other reasons, asking why it has the weight that it does, and considering whether others would assign it the weight I have were they in my shoes. It may also include reviewing how I would weigh it against other reasons in hypothetical circumstances, such as those I consider when I plan for the future. In such cases I attend to the reasons I’ve identified seriously, however—and this is what separates evaluation from the final mental operation we will discuss—when I merely evaluate reasons I do not open myself to being motivated by them, either because I am not yet in a position to be motivated by them (say, because they do not yet exist 180 ), because they are not reasons for me to do anything, or for other reasons. 181 179 Taken together, these resemble William Solomon’s (1975) “two-tiered” (or, more accurately, three-tiered) model for practical reasoning. 180 Thus I can reason practically about what I would do if such and such occurred. 137 When a person opens herself volitionally in this way to a reason she evaluates, I will say that she entertains it. To entertain a reason is to weigh it against other reasons as part of one’s own deliberative process; it is to open oneself from a first-personal standpoint to being motivated by the reason in a way that one need not and sometimes cannot while merely evaluating it. 182 Because identifying, evaluating and entertaining reasons are activities to which we are all accustomed, I will not try to explain their phenomenological attributes. It is illustrative, however, to observe that we occasionally position ourselves so as to increase or decrease the likelihood that we will entertain reasons to do things we take to be important or foolish. For instance, the reasons I have to pick up the novel I’m currently reading are particularly motivating—I’m particularly open to being moved by them—after I’ve been reading dense philosophy articles for a few hours, even though the weight of the relevant reasons stays the same. So I make it a point never to bring novels to campus, knowing that I might be less productive in the afternoon if I have the option of reading them. This case illustrates the 181 Philosophers call attention to the fact that we can and do perform these operations on reasons that are not our own for a variety of argumentative purposes. As Raz points out (1978, 5), for instance, in a more general point about the nature of practical reasoning, this capacity entails that the conclusions of practical inferences are in fact deontic statements (e.g., that “one ought to φ”) rather than actions or intentions formed by the deliberating agent. Alternatively, Nagel (1986, 150) cites our capacity to evaluate reasons—a capacity which “enables us to say what others have reasons to do, and also what we would ourselves have reasons to do if our desires were different”—as evidence of the objective standpoint which is at the heart of his defense of ethical realism. 182 Two points. First, ‘entertain’ is sometimes used in ordinary language interchangeably with ‘evaluate’ and even ‘identify.’ It would be perfectly reasonable to say, for instance, that we can entertain whether Socrates should have fled Athens in spite of the court’s order that he drink hemlock. This is of course perfectly fine in casual speech, but this informal use should be kept separate from the technical use I’m employing here. Second, there is a sense in which I can open myself to your reasons, but this happens only when the action open to you is also open to me and the facts that count as reasons with respect to you are also reasons with respect to me. In such cases I evaluate the reasons with respect to the action open to you and entertain the reasons with respect to the numerically distinct action open to me (that is, my φ-ing rather than your φ-ing). 138 more general point that entertaining reasons sometimes leads to unpredictable results: We don’t always know, going into a genuine deliberative process, what we will be moved to do by the reasons we entertain. 183 The distinction we have drawn between identifying, evaluating and entertaining reasons provides evidence that Hart and Raz may not disagree after all. Recall Hart’s claim: …the commander characteristically intends his hearer to take the commander’s will instead of his own as a guide to action and so to take it in place of any deliberation or reasoning of his own: the expression of a commander’s will that an act be done is intended to preclude or cut off any independent deliberation by the hearer of the merits pro and con of doing the act. 184 That the commander intends for his men to refrain from deliberating or reasoning as a guide to action indicates that Hart is using ‘deliberate’ to mean entertain, which concerns an evaluative process in which one opens oneself to acting for the reasons he considers. (Evaluative processes alone cannot directly motivate action, for once a person opens herself to being moved by the reasons she is evaluating, she is no longer merely evaluating them.) The commander’s men should not, in Hart’s view, be opening themselves to reasons that might move them, even if they ultimately fail to. Raz’s comments appear consistent with this interpretation. He says, for instance, while ostensibly arguing against Hart’s analysis: 183 Antonio Damasio, a neuroscientist at USC, studies individuals who have lost the functioning of their “emotional brain” as the result of a tumor or other damage to their prefrontal cortex. His subjects perform as well on IQ tests as they did before sustaining cortical damage, but they no longer experience certain forms of emotion. An interesting effect is that such individuals will meticulously scrutinize each and every potential reason they have for and against each of some range of actions available to them, weighing them against one another soundly, but endlessly. A number of things could be happening here phenomenologically, but on one interpretation Damasio’s work strikingly illustrates the divisions I have drawn by presenting individuals who can evaluate but cannot entertain reasons. See Damasio (1994). 184 Hart (1982, 253; my emphases). 139 Reflection on the merits of actions required by authority is not automatically prohibited by any authoritative directive… Hart would agree, so long, anyway, as Raz is using ‘reflection’ to mean evaluation, which the context of his comment seems to imply. So far as this is right, Hart and Raz both concern themselves foremost with reasons’ effect on action. And so it seems wrong to attribute to Hart a deliberative analysis according to which a person under orders cannot reasonably identify or evaluate competing reasons for action—which is, anyway, likely impossible for all but the most well trained ascetic. Hart does not want to say that a commander expects his soldiers to fail to see reasons to act other than he directs them. He simply doesn’t want them to think about these reasons in such a way that they might be motivated by them. He doesn’t want them to entertain excluded reasons. It is just as problematic, I think, to attribute to Raz an analysis according to which only action is excluded. One problem is that, even if Raz were concerned merely about action, a person with an exclusionary reason cannot predict the extent to which it might motivate her to act if she entertains it. Indeed, Raz seems to recognize this fact. And so while, in an attempt to separate himself from what he takes to be Hart’s position, Raz says that an exclusionary reason “can [but should not] be confused with a reason to avoid thinking, considering, or attending to certain matters,” he qualifies his claim shortly after: “so long as one knows that one’s reflections will not affect one’s action.” 185 In other words, a person with an exclusionary reason is free to engage in those mental processes which will not influence action—that is, to identify and evaluate reasons—but is not free to entertain them. Distinguishing in which cases Hart and Raz mean to be talking about indentifying, 185 Raz (1999a, 184), my emphasis. 140 evaluating, and entertaining reasons by their terms ‘think about,’ ‘deliberate,’ and so on, brings their views much closer together than their exchange would indicate. 4.2.3 Exclusionary Reasons and Obligation Exclusionary reasons are relevant to obligation primarily in cases in which the fact that constitutes the exclusionary reason (to refrain from acting on first-order reasons not to φ, or to ψ, where φ-ing and ψ-ing are incompatible) is also a first-order reason to φ. (When two separate facts occupy these roles only together, there is no obligation.) For instance, the fact that I promised you to φ is a fact which constitutes a first-order reason for me to φ and an exclusionary reason to refrain from acting on certain first-order reasons not to φ. My promise therefore constitutes reasons of both orders, and is called a preemptive reason. 186 In the next section we explore the relation between preemptive reasons and obligations. 187 186 I should note that Raz’s terminology is not always consistent. Though his preferred usage seems to be preemptive reason ((1988, 42) and (1990, 121)), he refers to these reasons sometimes as peremptory reasons (1977, 223), and other times as protected reasons ((1979, 18) and (1999a, 191, postscript)). ‘Preemptive,’ as Hart noted, means “cutting off deliberation, debate, or argument,” and derives ultimately from Roman law, “where it was used to denote certain procedural steps which if taken precluded or ousted further argument” (1982, 253-54). Because, as we will see, Raz believes that obligations can be defeated by other considerations, ‘preemptive’ is preferable to ‘peremptory,’ which means that something is final, and not open to challenge. In any case I will use ‘preemptive’ instead of the less descriptively accurate ‘protected’ to refer to these reasons. 187 In the postscript to the 1999 edition of Practical Reason and Norms, Raz says the idea that rule-like considerations involve “recognising the validity of exclusionary reasons” is “the only novel thesis in this book” (196). It is in this context William Edmundson (1993, 329) says that the exclusionary reason is “Raz’s principal contribution to practical philosophy,” and that Practical Reason and Norms is “perhaps the most significant work in jurisprudence since H.L.A. Hart’s The Concept of Law.” 141 4.3 Preemptive Reasons and Obligation We begin with Raz’s notion of obligation rather than that of another theorist, as I said, not because it is necessarily the right or only one, but because it introduces and takes seriously a number of observations that any theory of obligation must eventually explain, many of which turn on the idea of a preemptive reason. It will be helpful to state Raz’s position in advance of his argument so that we see upfront the view he is building towards. A concise summary is this: An action is obligatory if it is required by a categorical rule, i.e. a mandatory rule which applies to its subjects not merely because adherence to it facilitates achievement of their goals. 188 Before unpacking this statement, a few preliminary remarks are in order. Most important is that Raz uses ‘obligation’ and ‘duty’ (interchangeably) to express what he calls the “narrow concept” of obligation, i.e. one which is meant only to specify the “formal features” of obligation that explain the distinctive role it plays in our practical reasoning. 189 Two points should be made explicit here. First, the narrow concept of obligation clearly diverges from a more common application of the term as an expression of moral requirement. (We will return to this more robust conception of obligation later in later chapters.) Second, Raz imports his narrow concept into the broader political theory he develops in subsequent work 188 Raz (1977, 223); also see Raz (2010, 291). Note that this notion of ‘categorical’ is different than that used to describe the status exclusionary reasons have relative the reasons they exclude. 189 Raz (1977, 233). Raz defends his view of practical reasoning in his earlier work, Practical Reason and Norms (originally published 1975). The view I present below is a reconstruction from material found in his 1975 (republished 1999), 1977, and 1986 (republished 1988) works. 142 without making this fact sufficiently explicit. 190 No doubt these two factors are responsible for some of the confusion surrounding Raz’s authority theory of political obligation. Indeed, out of context they can cause his larger political theory to look inadequate or even incoherent. Obligations involve what Raz calls mandatory norms, which are more-or-less general binding rules specifying how one is to act in particular circumstances. 191 “Other things equal, one ought to do what one has promised to do” expresses a mandatory norm, as do “drivers are to pull over to the right of the road when they hear sirens” and “an invitation in the third person should be answered in the third person.” Mandatory norms are important to understanding obligation because they provide independent, preemptive (and thus exclusionary) reasons for action. 192 To understand the manner in which mandatory norms exhibit these characteristics it will help to gain a better understanding of the constitutive notion of “bindingness,” and then work our way down. To my knowledge Raz never provides an explicit definition of binding, so it must be extrapolated from the various things he says about it. The definition most conducive to his project states that a rule or some other consideration is binding just in case in those circumstances falling under its description one opens oneself to the possibility of certain forms of criticism by acting, trying to act, or in some cases entertaining acting merely on the 190 Here I have in mind particularly Raz (1988). For evidence that Raz does in fact import his narrow conception of obligation see pp. 60, 186, and 195. I take this up in greater detail in later sections. 191 This is not exactly right. Raz uses ‘mandatory norm’ to refer to a broad grouping of rule-like entities, some of which are particular norms that are not properly described as rules in some cases. See Raz (1999a, 49, 77-78). This makes no difference to the exegesis. 192 Raz (1999a, 58). 143 balance of reasons. 193 The appropriate form of criticism will depend on the type of norm or rule one violates. For instance, we would appropriately criticize someone who violated the rules of rationality by judging them irrational, someone who violated the rules of etiquette rude, and someone who violated moral principles blameworthy, cruel, or even monstrous, depending on the circumstances. The distinctions these forms of criticism draw are, or so I shall argue in Chapter 6, important to the nature of the requirements associated with them. 194 It may initially seem strange that one could ever be open to criticism for opening herself to acting on the balance of reasons, but in fact one of Raz’s most insightful observations is that obligations play precisely this preemptive role in our practical reasoning. To understand why we’ll consider a condensed version of his argument, which takes us through a number of different types of preemptive considerations leading up to obligation. I will first explain the role of decisions in our practical reasoning, as they offer probably the clearest example. I will then explain the connection between decisions and mandatory 193 “The balance of reasons” refers only to first-order reasons. To be consistent with Raz’s own terminology I will also refer to one’s acting on the balance of reasons in a particular case as her acting “on its merits.” By contrast, “all things considered” includes, for Raz, all those considerations that bear on one’s practical reasoning in a particular context, including, beyond first-order considerations, exclusionary reasons, mandatory norms, obligations, etc. Note that this is a non-standard use of “all things considered,” which others commonly use interchangeably with “on the balance of reasons.” 194 This explanation of bindingness is, I believe, superior to explanations which do not provide the reason for which, and thus do not explain the sense in which, one can be “bound” to act in a particular way. Compare, for instance, Leslie Green’s explanation of what I take to be the same concept of bindingness: “B regards the fact that p as a binding reason to φ only if B regards p as providing reason to φ and a reason not to act on some of the reasons for not φ-ing” (1988, 40). Green’s notion of bindingness says, then, that B takes a consideration to be binding only if B also takes that consideration to be a preemptive reason. It will become apparent later why Green’s definition is problematic. I can say here, however, that even were it correct, I take it that the reason B would regard p as binding is the fact that, were B to act other than p recommends, she would, given p’s preemptive status, open herself to some sort of criticism. I discuss this in greater detail in Chapter 6. 144 norms, and then, finally the characteristics which separate mandatory norms from obligations. 195 4.3.1 Decisions Decisions are mental actions by which we form intentions to perform a later action after some length of deliberation. 196 For A to have made a decision it is a condition that A must have both come to some conclusion about what to do (or believe) and formed the belief that the time has come to stop (or at least suspend) deliberation. A person that comes to a conclusion about what to do in some instance but plans to deliberate further about whether to do it, or continues actively to seek reasons why she should or should not do it, has not actually come to a conclusion, and therefore has not decided what to do. This is not to say that a person who has decided what to do cannot change her mind later, but it does mean that she would not generally do so barring a significant change in the circumstances upon which she has made her decision conditional. This is supported by the observation that a person who frequently “decides” to φ, only to later “decide” to ψ, only to “decide” again to φ, all without significant changes in the information available to her, would soon be thought by her friends, when she claimed to have decided on some course of action, to be still deliberating about it. That to make a decision one must come to the conclusion that it is time to stop deliberating entails that decisions are exclusionary, and therefore binding. This is why, 195 For the complete argument see Raz (1999a, Chapter 2). 196 See Raz (1999a, 65-73), from which this exposition is elaborated. 145 barring significant new information, we are reluctant to argue with a person who has already “made up her mind,” i.e. decided on a course of action. Our reluctance follows from the fact that such a person has left the deliberative phase, and has thereby closed herself off to a certain class of reasons that might compete with those she now considers sufficient for her intentions. Unless we can offer new reasons why she should φ that fall outside this excluded class, she will not entertain them. 197 This condition on deliberative cessation, which distinguishes decisions from mere intentions—which, though we often do not reconsider them, do not constrain reconsideration—is responsible for their exclusionary quality and explains their bindingness. If A’s decision to φ constrains her search for and ability to (rationally) entertain or act for reasons that would compete with those she already determined to be sufficient to φ, then A’s decision is exclusionary, for it excludes from practical deliberation certain reasons that would otherwise weigh against her first-order reasons to φ. 198 To the extent that A, having made a decision, is open to entertaining any new information, however trivial, or information she already entertained while deciding, to that extent A has begun to go back on her decision. 199 197 In this way decisions involve a reasoning-centered constraint that parallels the constraints imposed by intentions. My intention to go to Doheny Library today constrains the set of intentions I can rationally form to those consistent with going to Doheny today. Similarly, if I have truly decided to go to Doheny today then that decision will constrain, among other things, the information I can rationally entertain regarding whether I will go to Doheny. On reasoning-centered constraints as they concern intentions, see Bratman (1987, 17). 198 Importantly, I am not asserting that decision makers are always justified in treating their exclusionary reasons as such. The claim is merely that we do in fact consider them to be exclusionary in the sense specified. It is not always the case that a decision ought to have been made, but once it has been made, a decider will treat it as an exclusionary reason. See Raz (1999a, 68-69). 199 Note that it is compatible with this analysis that A can make a different decision by adding a caveat to the conclusion she reaches—say, to reconsider her decision if certain new information comes to 146 She opens herself to criticism because she continues to deliberate when, having made a decision, she has formed the belief that it is time to stop deliberating. While the exclusionary nature of decisions explain their bindingness, we said earlier that exclusionary reasons will be most interesting to us when they are preemptive—that is, when the fact that constitutes the exclusionary reason is itself a first-order reason for the action protected by the exclusionary one. While it is surely a more contentious point, Raz argues that a decision is also a first-order reason and therefore meets the conditions for a preemptive reason. 200 I believe that it may be possible to show this for decisions generally, but because our interest is in decisions as an illustration of a binding preemptive reason, we need only show how one could take decisions to be first-order reasons in at least some cases in which they are also exclusionary. Tie-breaking cases are probably the most plausible example. 201 Suppose that I take myself to have equally good reasons to φ and to ψ without light. However, a situation in which A has “decided” to φ with the caveat that she continues looking for any additional reasons for or against her decision is a case in which she has not decided at all. An intermediate case would be one in which she has made a decision with limited caveats. 200 Gilbert (2006, 127-34) argues in favor of this view. Bratman (1987, Chapter 2) objects that if intentions or decisions were reasons for action then one could unacceptably provide oneself with reasons that would in some instances tip the balance in favor of performing an action that is irrational in virtue of one’s other reasons. 201 Some may be skeptical that decisions could be reasons in particular circumstances but not in others. If a decision is a mental action that plays a particular role in practical reasoning, how could it fail to play that role in some cases? One answer, I think, is that decisions can be made for different reasons, only some of which provide significant justification for their status as first-order reasons. In tie-breaking cases in which one must make some decision (say, between competing but time-sensitive job offers) to avoid an unwanted loss (losing both offers and being unemployed), a decision more plausibly counts in favor of acting as one has decided to act in a way that it does not if one has an infinite amount of time to decide. Mortality makes all decisions time sensitive to some degree, and is approximated as closely as possible when a person faces no deadline to decide. In these cases a decision to φ may provide an imperceptibly small reason to act, leaving one with what seem like the same reasons one had before deciding. But small reasons are still reasons. (Though cf. Broom (2001).) Even if this is right, however, it is not a conclusive argument against the existence of decisions that are purely exclusionary. One might think, for instance, that sometimes one’s decision is 147 believing that I have conclusive reason to do either. Yet I may, because I believe that the circumstances provide a conclusive reason to make some decision, decide to φ, and thereafter φ. In such a case it is clear that I at least treat my decision as a reason to φ, for in the circumstances no new facts other than that of my decision weigh into my deliberation. 202 When asked later why I φed, I might coherently cite my decision, which, due to the circumstances of the case, would not appear to be a means of citing the reasons I considered to weigh in favor of φ-ing prior to making my decision, reasons which were, by hypothesis, too weak to weigh conclusively in favor of φ-ing. Whether I made the right decision or not, it was only after I decided to φ that I took myself to have conclusive reason to φ, and therefore it seems that I must have also taken my decision as a reason to φ. 203 If decisions are indeed both first-order and exclusionary reasons to do as one decides, then they are preemptive reasons. This makes sense when we consider that simply an acknowledgment of the reasons one already has to perform the action one chose, adding nothing to either side of the justificatory scale. Alternatively, it may be, as some have suggested, that the purportedly reason-giving force of a time-sensitive decision comes solely from the costs of reconsideration, and therefore cannot be attributed to the decision itself. Because these issues do not affect the present argument, I do not pursue them further. 202 Gilbert (2006, 28) highlights this distinction terminologically by claiming that whether or not I have a reason to act as I have decided in such a case, which may imply that the action or its consequences has some “good making” feature, I have reason to so act. 203 One might find this explanation paradoxical. If one must have formed the belief that it is time to stop deliberating in order to have made a decision, how can one include one’s decision as a fact that weighs in favor of anything? That is, if by deciding to φ a person completes the process of weighing reasons, how can her decision subsequently tip the balance? There are a few available responses, however I will give only one. Because, as I mentioned already, decisions are not boundlessly exclusionary, all decisions must contain what are essentially caveats, which create normative space to weigh reasons that will matter to what we should do. In other words, the “scale” on which we weigh reasons does not disappear simply because we have made a decision. Rather, the decision constrains the reasons we place on it. A decision, at least one with the right sort of justification, may not restrict its own access to the scale, which we then cite later to justify our action. On one interpretation of this response, decisions qua reasons only help us evaluate the wisdom of particular actions. On another, they also play an action-guiding role by providing agents with reasons to act. 148 decisions are deliberative tools that allow us to settle matters for ourselves prior to taking action, or determine a course of action prior to being confronted with a confusing, misleading, or otherwise undesirable influx of reasons which would be too time consuming or difficult to consider in the circumstances. Decisions, like norms, intentions, and other tools of practical reasoning, make our volitional lives more manageable. This is an important point to see to understand why preemptive reasons “bind.” Earlier I said that it follows from the fact that a rule is binding that in those circumstances falling under its description one opens oneself to criticism by acting, trying to act, or in some cases entertaining acting merely on the balance of reasons. So it follows from the fact that one is bound to act in some way that one ought not necessarily act (etc.) on the balance of reasons, which is true when a preemptive consideration is in play. Because decisions preempt (by excluding and replacing) some set of first order reasons, to neglect the preemptive reason one has at t 2 by acting on the balance of reasons is, ceteris paribus, to undermine the purpose the decision one made at t 1 was meant to serve. Such a person forms the belief that the time has come to stop deliberating, or that she would most benefit from coming to a conclusion about what to do in the circumstances, and then continues to deliberate nonetheless. Because this sort of behavior opens one to criticism, decisions are binding in Raz’s sense. This is, again, not to say that circumstances cannot arise that properly lead one to disregard one’s decision, by cancelling it or providing reasons to reopen deliberation; it is, rather, that having made a decision one can be criticized, however insignificantly in some trivial cases, for considering reasons one’s decision is meant to preempt. 149 4.3.2 Mandatory Norms While decisions and rules share some characteristics, decisions are not rules. Whereas decisions apply only to the decision maker, and apply only because she made the decision, rules can apply to multiple people and do so whether or not they countenance the rule. 204 However there is an important respect in which rules are like formal and enduring decisions for those who recognize and follow them. Just as entertaining the reasons preempted by a decision can subvert its practical purpose and thereby eradicate it, so too does it subvert the purpose of certain rules to entertain the balance of reasons in those circumstances to which the rule is meant to apply. Such rules are, like decisions, meant to take the place of some set of reasons that bear on the agent in the circumstances. Mandatory norms are rules and principles of this sort. 205 But whether a purportedly mandatory norm is a binding preemptive reason depends in part on its justification. Raz argues that justifications for preemptive norms include that they save time or labor, minimize error, or are directives issued by an authority either (i) to secure coordination for certain activities or (ii) such that the directive, if followed, will cause one to act in accordance with her own reasons better than she would were she to try to act on the balance of reasons 204 Raz (1999a, 73). 205 See Raz (1999a, 49, 58). Note however that, while they are not the ones we are most interested in, some norms clearly don’t provide a first-order reason for action. For instance, a rule specifying that one is to ignore all rules of etiquette is purely exclusionary (1999a, 77). Returning to an earlier point, it follows from the fact that merely exclusionary reasons can bind that Green’s explanation of binding, which provides a two-part necessary condition, one of which includes a first-order reason, would appear to be the wrong one. See Green (1988, 40). (This is true, of course, only if Green and Raz in fact mean to refer to the same notion of bindingness.) 150 in those cases to which the directive applies. 206 A rule that is justified in one of these ways and that an agent considers to be a mandatory norm is like a decision in that it provides a first-order and exclusionary reason for her to act in a particular way, and is therefore binding in that she will open herself to criticism by acting for reasons it excludes. Because (by hypothesis) she takes the rule to serve a value that justifies it, by acting other than it prescribes she would fail to serve that value, which she takes there to be justification for her to serve in her very circumstances. An example will hopefully illustrate the point. Parties in court have the power to move to strike claims made by third parties when those claims are inadmissible to the case and an abuse of court proceedings. When the motion is sustained, the jury is directed to disregard them. The judge’s directive in these cases is a mandatory norm which the jurors are meant to follow in their later deliberations. Suppose, however, that while the piece of evidence stricken is inadmissible, it seems clear to everyone on the jury that it verifies the defendant’s guilt, and also that without this piece of testimony there would remain reasonable doubt that the defendant committed the crime. How should the jury treat the stricken evidence? If they treat the judge’s directive as a mandatory norm, then they will disregard it. But how would this make sense? If the evidence seems clearly to condemn the defendant, would it not be a mistake, at least in this particular case, to ignore it? Suppose that the justification for the judge’s directive, and thus the rule, is that it protects the impartial procedure of the court in the service of justice by preventing jurors from being moved by unreliable evidence, to which this sort of attestation overwhelmingly 206 Raz (1999a, 74). The last two justifications are central to the authority theory of political obligation I propose in Chapters 6 and 7. 151 amounts. But if the evidence is thought to be both credible and decisive by the jury then it would seem to them that the rule is without justification in this particular case, for it appears that justice will be served by taking the evidence into account and finding the defendant guilty. In such a case wouldn’t they do better by disregarding the rule? The answer is no. Not, at least, if the justification for it takes into consideration circumstances like these for violating it. For if the rule is valid then this is because juries complying with it will serve the values that justify the rule (justice, in this case) better than they would were they to attempt to serve them directly. The rule is valuable in virtue of the fact that juries are poorly positioned to know whether certain sorts of evidence are credible, which is, for all the jury members know, perhaps illustrated by the fact that they find this bit of stricken evidence compelling. It is, in other words, the general unreliability of juries’ beliefs about whether this sort of evidence is reliable that provides justification for the rule. Thus, so far as the jury treats the judge’s directive as a mandatory norm they will take it to exclude the stricken testimony from consideration, even in cases in which they believe they could better serve the value served by the rule were they to try to do so directly. 207 In such a case the jury will 207 Of course, circumstances do arise that cancel the rule. For instance, were the judge found to be drunk or to have stricken evidence that was clearly admissible by the rules of law, then one may have reason to ignore the rule, since in these circumstances it is obvious that the rule will not serve the value purportedly justifying it. However, one has to be vigilant here. Some rules are designed to prevent errors even in cases in which, and sometimes especially in cases in which, one takes oneself to have good reasons to act other than the rule suggests. I may make it a rule never to use my credit card to purchase drinks, so as to make it more difficult to get ahead of myself with my friends at the bar and wind up with a tab I later regret having paid. After having a drink, however, and running low on cash, it might seem to me that the reasons I have to use my card to buy drinks in this case are not accounted for by the rule. This would be a mistake in most cases, for the rule is supposed to prevent me from acting on those very reasons I now take to be outside its exclusionary scope. This is reminiscent of a point Wittgenstein once allegedly made to Elizabeth Anscombe. Anscombe told Wittgenstein that it was understandable that people once believed the sun went round the earth. When Wittgenstein asked why this was, she told him that that is how it looks. “What would it look 152 consider themselves bound to act as they do. And so far as the justification is valid, they will be. 208 4.3.3 Obligation We are now in a position to understand Raz’s notion of obligation. According to the narrow conception, P has an obligation to φ in circumstances C only if a mandatory norm specifies that P φ in C. Recall that the mandatory norms we are talking about are like decisions in that they constitute binding preemptive reasons, and thus that it is entailed by the fact that P is obligated to φ that P has a preemptive reason to φ. This is to say that on Raz’s view it is entailed by P’s having an obligation to φ that P is bound to φ by a fact that constitutes a first- order reason to φ and an exclusionary reason that excludes certain first-order reasons to refrain from φ-ing or to ψ (where φ-ing and ψ-ing are incompatible). Were mandatory norms sufficient for obligations, however, then, because mandatory norms are preemptive reasons, from the fact that there is a preemptive reason to φ it would follow that one has an obligation to φ. But this is inaccurate. Suppose that I’m at USC in Los Angeles and want to get to Hollywood as soon as possible. I know that the 10 bus goes to Hollywood from USC, but don’t know along which route it runs. Because I’m unsure whether the 10 is the quickest route, I inquire of a person standing beside me, who appears like,” Wittgenstein replied, “if the sun went round the earth?” The line between what is obvious and what is “obvious” is very thin in some cases. 208 Raz observes that our actions reveal important facts about our understanding and acceptance of exclusionary reasons: “Often we feel that the rule should have been followed even though it was known in advance and proved after the event that deviating from it would have been beneficial, and even if it is established that it would not have undermined the chances that the rule would be followed on other occasions. This reaction is conclusive proof that the rule is regarded as a rule, that is, as an exclusionary reason” (1999a, 75). Also see also pp. 63-64. 153 also to be waiting for the bus. She tells me to take the 110, a route she takes daily, because it is an express bus to Hollywood and so has fewer stops. Assuming that I take this person to be an expert on the matter, her directive (“Take the 110”) qualifies as a binding mandatory norm. That is, so far as my goal is to get to Hollywood as soon as possible, then, ceteris paribus, if I take her to be an expert on the matter, I will be open to criticism for taking the 10. But surely in this situation I don’t have an obligation to do what she says. Cases like this are frequently offered as objections to Raz’s view. However, they are based on a misunderstanding. Raz agrees that mandatory norms are not obligations, and he avoids this result by adopting a proviso specifying that the value justifying the exclusionary force of the norm must be “independent of the person’s own goals.” 209 This proviso is worth examining. Mandatory norms to which we are subject in virtue of our desires are structurally similar to obligations. According to the proviso, however, obligations are unique in that an individual bound by one cannot avoid falling subject to it by modifying her preferences. That is, some mandatory norms but no obligations are justified merely in virtue of the fact that by acting on or satisfying them one will achieve (or will at least have the greatest chance of achieving) something one values. In this way non-obligatory mandatory norms are like Kantian hypothetical imperatives, which are contingent on one’s willing a particular end. But the general division goes back at least as far as Hobbes, who in 209 Raz (1977, 224). Thus: “Not all mandatory rules…impose obligations. Many of them apply only to persons who pursue certain goals and are binding on them because they help promote these goals.” Hart marks this distinction terminologically, using obliged rather than obligated to indicate that the rule’s bindingness is partially determined by self-interest. See Hart (1961, 80). 154 distinguishing commands from council observed that our reasons for acting on obligations could not be “drawn from our own benefit.” 210 Consider an example. A small business owner might discover that she has a bias for certain personality types. For reasons she doesn’t completely understand, when she conducts face-to-face interviews she personally enjoys and is disposed to hire people who turn out to be less competent than those she hires based solely on written applications. So far as she desires optimally competent employees, she may adopt a rule specifying that she forgo interviews for all new hires and consider only written applications. This rule binds her because she would be open to (in this case, rational) criticism were she to violate it by holding interviews during a later hiring cycle. However she is bound just so long as she aims to hire only maximally competent people. Were she to become indifferent to performance, and, say, to come to value most highly some particular personality trait that her face-to-face hiring practice tracked, she would no longer be open to criticism, for she would no longer have reason to eliminate interviews and would, in this case, even have some reasons to have them. She would therefore no longer be bound. By contrast, one cannot escape one’s obligations by changing one’s preferences or desires. I remain obligated to feed my daughter even if I’ve grown somewhat indifferent to feeding her, or have developed a desire never to feed anyone (say, because I’ve come to believe that feeding others is emasculating). We remain open to criticism for violating obligation-norms irrespective of our interests. 210 Pufendorf held a similar view. For discussion see Korsgaard (1996, Section 1.3). 155 4.3.4 Summary Pulling these ideas together, we can finally return to the statement of Raz’s narrow conception of obligation with which we started, but which we are now in a position to fully appreciate: An action is obligatory if it is required by a categorical rule, i.e. a mandatory rule which applies to its subjects not merely because adherence to it facilitates achievement of their goals. 211 On this conception, then, to have an obligation to φ is to have a structured reason to φ that excludes some class of first-order reasons to act other than the obligation directs, and that applies to us independently of our aims. It is worth emphasizing, as some commentators have, that this conception diverges from our more common notion of moral obligation. Raz appears to intend this, however, for he claims that moral obligations are merely one type of obligation. 212 Thus his analysis does not conflict with, for example, Philippa Foot’s attempt to dissolve the hypothetical/categorical division between moral and non-moral normative institutions. 213 As I see it, Raz is not committed to an equivalence between what have traditionally been thought of as categorical and hypothetical imperatives, but only to the 211 Raz (1977, 223). It is worth noting that the somewhat enigmatic term ‘categorical’ occurs only twice in Practical Reason and Norms, despite the fact that in the introduction Raz distinguishes categorical rules as one of the “most important kinds” of rules (1999a, 9). 212 “Nothing about my account of promises or obligations suggests that they are exclusively moral notions” (1977, 225). 213 Foot (1972). 156 possibility of there being non-moral values capable of justifying rules which bind individuals irrespectively of their goals or interests, such as, for example, rationality. 214 4.4 Moving Beyond the Simple View Understanding the exclusionary considerations surrounding obligation allows us to see how the Razian notion of obligation overcomes the objections to the simple view that we considered earlier. These were based on three observations. First, it is sometimes the case that two obligations—O 1 to φ and O 2 to ψ—exert distinct degrees of pressure, which can be true even when O 1 and O 2 are generated by the same act-type (say, promising), and even when φ and ψ are acts of the same sort (say, meeting someone for lunch). To account for this fact, the simple view must appeal to weight in two different ways: first to separate obligations from mere reasons, and then again to distinguish one obligation from another. I do not claim that this cannot be done, only that these distinctions are much easier to draw, and seem to better represent the nature of our requirements, when there are two separate dimensions in play, each performing a different function. This is precisely what the narrow concept does. Whether a person has an obligation is determined by the structure of her reasons, whereas the weight of that obligation is determined independently. This structural 214 Rules of etiquette and rules of games are other possible examples. If I am playing soccer (and am not the goalie) then there is a sense in which I am bound, irrespective of my goals, not to touch the ball with my hands. A game in which players use their hands cannot be soccer, though it can look a lot like it. Thus, so long as I am playing soccer there is a categorical rule against using my hands. Nevertheless, this does not seem to be a moral rule. Along with Foot’s similar treatment of rules of etiquette, this suggests that categorical imperatives may be divisible into moral and non-moral categories, the distinction between which may turn on the range of reasons those rules exclude. The difficulty here, as everywhere, is to justify the application of those rules for the sake of values the person to whom they apply does not care about. 157 explanation explains what two promissory obligations that are nonetheless of dramatically different weights have in common. The second and third observations together attack the implicit biconditional the simple view posits between weighty reasons and obligations. Some obligations are of very little weight, while some very weighty reasons are not obligations. Again, the narrow concept provides ample explanatory resources. A view of obligation that recognizes their exclusionary qualities has an easier time accounting for an obligation’s special force in a case in which it appears to be outweighed by a non-obligation reason. The explanation is that this obligation, like all obligations, excludes a certain class of reasons from practical deliberation, but that this does not entail that it cannot be outweighed by reasons that it does not exclude. To use our earlier example, my promise to join the large crowd of your friends cheering you on at your swim meet may be outweighed by the fact that I have been offered a job interview that conflicts with it. The interview defeats the obligation even though I’m not obligated to take it. Nonetheless, my obligation to you still has a special force, accounted for by its exclusionary nature. Determining, therefore, that an obligation is outweighed by only a non-obligation reason does not entail that what one had was not an obligation after all, or that, though it was an obligation, it did not exert the special sort of pressure an obligation characteristically does. These same considerations explain, finally, why reasons are not obligations merely in virtue of their significant weight. The financial security and professional prospects I would enjoy by taking a particular job offer may be very strong reasons to take it, but in no sense am I obligated to do so. Weak or strong, obligations are special in virtue of the binding 158 preemptive force they exert on our practical reasoning. Weighty reasons are merely weighty reasons. 215 We will return in Chapter 6 to the nature of obligations, adding to Raz’s narrow conception to account for the various species of requirements individuals can have. We will then use the final version of the conception we arrive at there to explain political obligation in Chapter 7. Before doing that, however, we move to a preliminary discussion of authority in Chapter 5. 215 How can one tell the difference? It depends on the case. In cases of promises it seems relatively clear that the reasons they provide are exclusionary. For one reason, a promiser can wrong her promisee if she opens herself to being moved by certain reasons to do something other than she promised to do. For instance, you would be justified in being upset with me if I told you that, although I promised to help you move into a new apartment later, I am thinking of going to see a film instead. (I discuss this phenomenon in depth in Chapters 6 and 7.) A plausible explanation in these cases is that the reason I considered was excluded, and that my obligation included a deliberative restriction that corresponded with the practical one. In these circumstances entertaining excluded reasons is a way (though presumably a less flagrant way) of violating the obligation. In other cases too there is good evidence that certain reasons are exclusionary. Our discussion of decisions and mandatory norms was supposed to show, among other things, that such considerations are best understood in terms of their exclusionary role in practical deliberation. For more on this point see Raz (1999a, 41). 159 Chapter 5: Foundations of the Authority Theory The goal of this chapter is to understand another foundational component of the theory of political obligation I offer in Chapter 7: authority. In the context of political obligation, an authority theory is, as I explained in the introduction, one according to which the authoritative relation between authorities and subjects forms a necessary part of the explanation of those subjects’ obligations. 216 Joseph Raz, whose narrow concept of obligation we considered in Chapter 4, offers one such theory in The Morality of Freedom, and this is the version we shall start with here. I should be clear from the start that my goal is not to defend Raz’s theory, but to learn from it. As we proceed through this chapter, I’ll offer objections and explain the answers Raz has either given in the literature or seems to be committed to giving based on his comments in various places. I then go on to give alternative answers to these same objections. By the end of the chapter our approach will have begun to diverge from Raz’s in important ways that I will not adequately develop until Chapters 6 and 7. Despite the fact that the theory is itself a work in progress, I will continue to refer to it throughout as “the authority theory.” 216 It should be obvious from this definition that in calling the theory I defend the authority theory, I do not mean to suggest that it is the only one of its type. 160 One might think I am confused to attribute a theory of political obligation to a theorist who explicitly denies 217 that there is a general obligation to obey the law, and who is, indeed, usually counted as a so-called “philosophical anarchist.” 218 But in fact there is no confusion. Raz’s claim is not that people cannot be, or even that they are not, strapped with an obligation to obey the law. It is that, given the ways in which one can be strapped, not enough people actually are to describe the extant state of things as one in which there are, properly speaking, political obligations. What Raz denies, in other words, is that the generality condition can plausibly be satisfied, even by his own theory. On this point I agree—at least so far as we consider only the theory as Raz developed it. Our aim, however, is to answer this and other objections by developing the view atop the foundations Raz provided. I begin this project in earnest in Chapter 6, and offer the most explicit versions of the related arguments under the authority banner in Chapter 7. Our aim presently, however, is to understand the authority theory as Raz developed it, so that we may usefully build on it in the next two chapters. Prior to providing a full exposition of the authority theory, it will be helpful to first outline the view and explain, roughly, why it counts as a theory of political obligation. Following this I will present two parts of a more detailed analysis: the first concerning the normal justification of authority, the second the nature of normative capacities. After wrapping up these expository remarks I will turn to a number of objections that have recently been offered against Raz’s view, and will consider two lines of response. The first is 217 Raz (1979, 233-34). 218 Recall that “philosophical anarchism” is a position the central tenant of which is that there can be no general obligation to obey the law. 161 Raz’s own, which, for reasons I will explain, I find unconvincing. The second is an alternative solution that I believe is more defensible. While I argue that this second solution is superior to Raz’s own, we will see in Chapters 6 and 7 that two important objections remain, and that the answers to these objections shed valuable light on the nature of requirements, obligations, authority, and ultimately, political obligation. Explaining these implications is left for those chapters, when we consider the authority theory in its most interesting and plausible form. 5.1 The Authority Theory in Outline As I said in Chapter 1, the aim of a theory of political obligation is to provide considerations which would obligate an individual in a particular political community to obey its laws, and to determine in which political communities these considerations actually exist and to whom they apply. Like consent theories, the authority theory locates individuals’ political obligations in the “special relationships” they bear to authorities. 219 They diverge, however, with respect to the conditions under which such obligations emerge. Consent theory is transactional, and focuses on the relations individuals create for themselves though voluntarily undertakings. As Hume 220 and many others have observed, however, it is difficult to find this sort of consent anywhere in the political world. This has led theorists to argue 219 Raz (1988, 104). Of course this is not to prejudice the question of whether authoritative relations between a sovereign and her subject do not ultimately depend on or trace back to relations the subject bears to other members of the political community. 220 See “Of the original contract” (1748). 162 that the sort of consent which ultimately matters for legitimate authority is hypothetical—a suggestion that already begins to pull uncomfortably against our ordinary conception of voluntary. The authority theory avoids these problems by focusing on non-voluntary authoritative relations which are, as we will see presently, transaction-independent. To put the authority theory in its simplest possible terms, it is a theory about the normative power legitimate authorities have over those we will call, for lack of a better term, their “subjects.” 221 It involves explaining (i) the conditions a person must satisfy to be a legitimate authority, and (ii) the special way such an authority can affect what her subjects are to do. These two parts of the theory are interdependent. The authority theory claims, roughly, that person A is a legitimate authority for person B if B would, on the whole, act in accordance with his independent 222 reasons for action better by following A’s directives than he would by trying to act in accordance with those reasons on his own. When A satisfies this condition with respect to B, she is in a position to affect what B is to do by way of her directives. Specifically, it is claimed that A’s directive that B φ will in this case constitute for B (i) a content-independent reason for B to φ—that is, a reason that does not entirely depend on 221 The term ‘subject’ should not lead one to suppose that the realms of action with respect to which the authority is an authority must be vast. A child, whose mother is an authority for him within almost all realms of action, is his mother’s subject no more than a professor is her dean’s subject or an art student is his teacher’s subject. All subjects are subjects equally on the present conception, and differ only with respect to the realms of action within which they are subjects. 222 Independent reasons are those which apply to a person independently of the authoritative directive. Raz calls these ‘dependent reasons’ because he argues that an authority’s directives ought to depend on them (1988, 47). However he appears to use the term, quite confusingly, both to refer to the reasons on which directives are to be based as well as to refer to the sort of preemptive reasons the directive constitutes (1988, 41). For clarity, I use the term ‘independent’ to refer to reasons of the former sort. 163 the independent reasons B has to φ, 223 and (ii) an exclusionary reason for B to φ—a reason which categorically excludes certain reasons not to φ (Chapter 4). Directives of this sort are therefore a type of what in the last chapter we called a preemptive reason. As we saw in there, when the independent reasons that give normative force to a purported authority’s directives do not depend on her subject’s desires or aims, those directives are categorical rules for him. Thus, in cases in which A’s directives are authoritative for B in virtue of considerations which do not rest on B’s desires or aims, B is obligated to do as A says, at least on the narrow conception of obligation. 224 223 Raz (1988, 35). Usually content-independent reasons are defined as those which have no direct connection to the action for which they are reasons. I have adopted this modified definition because it is difficult to understand what it means for a reason to have a ‘direct connection’ to an action. For a detailed look at various interpretations of the traditional definition, see Sciaraffa (2009). 224 One might suspect from what I have said that requests must also constitute preemptive reasons. While I think this is right, not everyone agrees. Raz (1988, 37, 101) agrees that requests constitute reasons for action, but claims that they fail to preempt other reasons because they are not exclusionary. For him this failure distinguishes requests from authoritative directives, and thereby explains why requests fail to obligate. It seems to me, however, that in many circumstances requests do obligate. When my friend asks me to do something for her, and it is not unreasonably burdensome for me to do it, what would it say about our friendship were I to refuse? I suggest that our friendship requires of me that I satisfy requests that are reasonable in light of it. If this requirement exists then it entails that requests, when made by the right people in the right circumstances, constitute preemptive reasons. My friend’s request that I give her a lift to a meeting on my way to work in the morning provides a preemptive reason for me to give her one. It excludes a certain class of reasons I might have not to do so—including, e.g., that I would prefer to take an alternative route so that I can pick up a cup of coffee from my favorite café. (I explain the obligatory nature of this type of exclusion in the next chapter, when I place obligations in the space of requirements more generally.) One might object, however, that, along with the remaining parts of the authority theory, this account falsely implies that friends have a certain practical authority over one another, which they can exercise by requesting things. (Hershovitz (2003, 204) raises something close to this objection.) Because it seems to me that one can understand the nature of friendship only by way of these characteristics, I see the tight analogy between authoritative directives and requests as a virtue of this sort of view rather than a cost. To consider someone a friend is partly to take that person to have the standing to generate preemptive reasons for you through their requests. A friendship without this feature is, I think, no friendship at all. Because this capacity is sufficient for practical authority, I think the fact that friends have it teaches us something about friendship, rather than proving something about practical authority. 164 The normative power authorities have to obligate others through their directives is usually very limited. Governments are exceptional in this respect. For the tremendous financial and informational resources at their disposal, and for the social function they are meant to perform, governments are uniquely positioned to usefully direct their subjects’ actions in many realms, and may therefore qualify as legitimate authorities in a significant range of circumstances if they use these resources to direct their subjects well. Citizens who will act in accordance with their independent reasons better by accepting and trying to follow these directives have an obligation to do so on the Razian account. Because governmental directives are issued as laws, subjects will have an obligation to obey the law. Thus, the authority theory concludes, with respect to the realms of action in which the government constitutes a legitimate authority for any given subject, that subject will have political obligations in the narrow sense with which we are concerned. The authority theory is, therefore, clearly a theory of political obligation. 225 5.2 The Authority Theory in Detail The argument for the authority theory has two interdependent parts. The first part explains the conditions a person must meet to be a legitimate practical authority, while the second part explains an authority’s normative capacities—her ability to affect what others are to do through her directives. These parts are interdependent in that on the authority theory you 225 I should note that, because on the authority theory a government may have authority only with respect to limited realms of actions, and so may fail to generate obligations for some subjects to obey some laws, the theory’s acceptable answer to the validity question does not entail that it can answer the applicability question. See Chapter 2. 165 cannot have one without the other. Satisfying the conditions of legitimacy entails that one’s directives constitute categorical preemptive reasons for those with respect to whom one qualifies as an authority, while it follows from the fact that one’s directives constitute such reasons for others that one is a legitimate authority for them. 226 Together these claims purport to describe one circumstance in which an individual will be obligated to obey an authority, and to provide us with what we need to determine in which political communities these considerations actually exist, and to whom they apply. 5.2.1 How Practical Authorities are Justified A practical authority is, in Raz’s words, an “authority affecting what is to be done.” 227 While many different types of authority exist—the more narrow of which, such as the right to rule, some political philosophers prefer to focus on—the capacity to affect what is to be done (in a sense to be discussed below) is that feature all and only practical authorities share. It is therefore the sort of authority for which we shall try to account, which I shall call the general conception of practical authority: A practical authority is someone capable of affecting another person’s normative situation (his Hohfeldian incidents) through her directives, that is, someone with 226 However it does not entail that one is any particular type of legitimate authority. There are multiple sources of, and thus multiple types of, authority. It follows from one’s having normative capacities only that one is an authority of at least one of these types. See the following section, as well as Chapter 6. 227 Raz (1988, 29). 166 the normative power to impose obligations on another simply by expressing an intention to do so. 228 One can affect another’s normative situation in at least two different ways: by providing him with reasons for action, and by requiring him to act. 229 However not all ways of affecting what is to be done are equal. Holding a gun to your head provides you with a compelling reason to give me your money, just as walking in front of your car creates a requirement for you to stop. In both cases I affect what you are to do by creating new considerations that weigh in favor of some particular action. Yet in neither case do I have authority over you. 230 If the manner by which reasons and requirements are precipitated by an authority is unique in some way, then we must explain how. To be a practical authority it must be the case that one can obligate others to act as one directs in virtue of the fact that one has directed it. One’s directive must, in other words, itself constitute both a first-order and an exclusionary (categorical) reason for someone to act. This explains why my capacity to create a requirement for you by walking in front of your car does not make me an authority, for doing so entails nothing about my capacity to affect your normative circumstances through my directives. But what about the gunman, who does issue directives? Is he an authority? No. The reason, according to Raz, is that his 228 See Raz (2010, 292). 229 Raz (1988, 38). 230 In these sorts of cases we might say, following Raz, that what I exercise is not a normative power, properly understood, but a form of influence. See Raz (1999a, 99). 167 directive lacks an “invocation of the duty to obey.” 231 The difference between the gunman and a genuine practical authority is, on this conception, that the gunman does not invoke my duty to give him my money when he demands it. To do so he would have to take himself to have a special normative capacity which, if he is sane, he probably does not take himself to have. The gunman, after all, simply wants my money, and has in the circumstances the (non- normative) power to get it. He is not concerned to provide me with an authoritative reason. 232 The government, by comparison, does take itself to have this special normative capacity, and at least purports to relate to its subjects through it. It appeals to citizens to do as it says for the reason that it said it, and so takes itself to provide special reasons for me to hand over the loot. 233 Of course, while it would be very unusual, the gunman could claim to invoke a duty that I pay him. Perhaps he believes that he is a king, and that I owe him a share of my income as a tax, which he prefers to collect face to face. The crucial difference between this delusional gunman and the government is that the gunman would be wrong to think this. And this marks the third condition on practical authority: justification. Without the 231 Raz (1988, 25-26). 232 Of course, his implication that I may be shot if I do not hand over my money might be an authoritative reason, but it is a reason for belief, not action. His authority in this situation is epistemic—it is the authority of someone who knows better than anyone else what he himself will do. Having been threatened, I now have a reason to hand my money over, but this reason is provided by the possibility that I will be shot, and not the gunman’s directive, which merely gives me reason to believe that I will be shot if I fail to comply. See Raz (1988, 36). 233 Raz claims (1988, 27) that if it did not then it would not be an authority. For Raz, claiming to impose duties and having the coercive power to maintain its rule qualify any agent or group as a de facto authority, which, on his view, any legitimate authority must be also. 168 appropriate normative backing, the gunman can do no more than purport to be a practical authority, even if he means to invoke my supposed duty to obey. To explain this distinction more formally we need a criterion which separates legitimate from merely purported practical authorities. According to the authority theory, practical authorities exist to serve their subjects. 234 This idea is expressed in its criterion of legitimacy, which Raz calls the normal justification thesis: NJT The normal way to establish that a person has authority over another person involves showing that the alleged subject is likely better to comply with reasons which apply to him (other than the alleged authoritative directives) if he accepts the directives of the alleged authority as authoritatively binding and tries to follow them, rather than by trying to follow the reasons which apply to him directly. 235 234 For this reason Raz calls his conception of authority the service conception (1988, 56). Some important questions should be raised about this approach to the problem of legitimate authority, which is designed to show that autonomy is not inconsistent with authority. One, which I have seen nowhere in the literature, is whether a theory which eliminates this inconsistency is for that reason any more likely to be the right one. Pursuing this and related questions further would take us too far afield. 235 Raz (1988, 53). As Hershovitz notes (2003, 206), although Raz uses the term ‘comply’ here, he should, based on his own understanding of that term, laid out in Raz (1999a, postscript), really say ‘conform.’ To comply with a reason (on Raz’s 1999 usage), one must not only do what it recommends, but must do it because it recommends it. That is, to comply with a reason one must act for it. However, the thrust of the NJT is precisely that subjects of authorities should not act for their independent reasons, which authoritative directives are meant to preempt. Presumably this is just a textual oversight, and not a substantive inconsistency in Raz’s view. In other places, such as the gloss just below, Raz uses ‘conform’ to explain the idea behind the NJT. In any case, I will continue to use ‘comply’ in its ordinary sense to mean ‘act in accordance with.’ 169 To put things more simply (if less precisely), the ground of legitimacy is “relative success in getting people to conform to right reason.” 236 Success in this respect separates the delusional gunman from the government. I have essentially no independent reasons to give the gunman my money, and so even though he tries to invoke my duty to pay him, his attempt fails because his directives will not cause me to act in accordance with my reasons any better than I would without them. On the other hand, I do have independent reasons to spend some of my money on important public goods, such as infrastructure, health care, and so on. So far as I will likely act in accordance with these reasons better by paying my taxes than I would were I to attempt to support the development and maintenance of these goods directly— which to do well would take time, knowledge, and skills I do not have—the government is a legitimate practical authority for me with respect to these matters. 5.2.2 How an Authority Generates Obligations Supposing that the NJT properly specifies the normal conditions of legitimacy, how is it that a legitimate authority’s directives generate obligations? Recall our definition of obligation from the last chapter: An action is obligatory if it is required by a categorical rule, i.e. a mandatory rule which applies to its subjects not merely because adherence to it facilitates achievement of their goals. 237 According to our definition, for an authority to generate an obligation for her subjects her directive must be like a categorical rule for them; it must be like a mandatory rule the proper 236 Raz (1990, 13). 237 Raz (1977, 223). 170 application of which does not depend on their goals. Mandatory rules are, again, preemptive reasons: each is a fact that constitutes a first-order reason to φ and at the same time an exclusionary reason to refrain from acting on certain first-order reasons not to φ (or to ψ, where φ-ing and ψ-ing are incompatible). 238 The fact that I promised to drive you to the airport constitutes a first-order reason to drive you and an exclusionary reason to refrain from failing to drive you because, say, I feel like sleeping late this morning. Promises like this obligate insofar as they are preemptive reasons, and because the value promise keeping serves qualifies the rule specifying that promises are to be kept as categorical. 239 To say that this rule is categorical is to say that whether I care about keeping my promises is irrelevant to whether I must. Having made a promise to take you to the airport, I cannot escape the requirement under which I have placed myself simply because I no longer care whether I keep it. That is not how promises work. 240 The thrust of the authority theory is that authoritative directives are preemptive reasons in just this way. But while the preemptive reason I have to do as I promised, and therefore my obligation to do it, is transaction-dependent in that it depends on my having made the promise, whether I have an obligation to do as you say depends, on the authority theory, on the non-voluntary considerations specified by the NJT. We look now at how an authority can generate such requirements. 238 Raz (1999a, Chapter 2). 239 See Raz (1977). 240 For an opposing view of categorical reasons, see Solomon (1975). 171 The Dilemma of Authoritative Directives To show that authoritative directives constitute preemptive reasons we need to consider the difference they make to our practical reasoning. How can an authority’s directive provide reasons that replace those one already has? It might at first seem as though authorities, if their directives are to make a difference to how one ought to act at all, could only do so when they direct their subjects to do what they ought not do on the balance of their independent reasons. This presents a dilemma. 241 If I have most reason to φ, and you order me to φ, then it is unclear how your order could make a difference to what I ought to do. 242 However, if you order me to ψ, and I have most reason to φ, then you would be ordering me to act other than I ought, in which case it is unclear how you could satisfy the NJT, and therefore how your defective order could give me a reason to ψ. The authority theory contends that authorities satisfying the NJT affect what is to be done in both sorts of cases, as well as in a third set of cases in which I have conclusive reason to achieve some end but lack conclusive reason to achieve it by any particular means. Let me briefly explain each sort of case. 243 Consider the third sort of case first. Often an authority is in a position to establish a rule which will solve a coordination problem, such as driving on a particular side of the road. These are familiar cases and so I will not belabor them. Until a convention develops, or a 241 See Marmor (2010). 242 I use ‘order’ instead of ‘direct’ because in the context ‘direct’ could be taken to mean recommend. While some writers draw a distinction between ‘order’ and ‘command’ (e.g., Hart (1961, Chapter 2)), I use them interchangeably. 243 See Raz (1988, Chapters 2 and 3) for discussion. 172 rule is established, nobody has a reason to drive on this or that side of the road; however everyone has a reason to drive on the same side as everyone else. An authoritative directive to drive on the right side of the road gives everyone a reason to drive on the right, even though nobody previously had a reason to drive on the right instead of the left side. 244 What about the other two sorts of cases? Those in which someone satisfying the NJT orders one to do something one ought already to do, or, alternatively, something one ought not do on the balance of reasons? Does the directive affect one’s normative circumstances in these cases? Raz says yes. To explain why, it will help to first understand in what sense a directive issued by someone who satisfies the NJT both excludes and replaces the reasons a person has independently. (For simplicity I shall call directives issued by such a person ‘authoritative directives,’ though let me be clear that in doing so I do not mean to assume that they have normative force, which is what we aim to show.) When a Directive is Inconsistent with One’s Reasons The first case to consider is one in which one’s reasons are inconsistent with the authority’s directive. Suppose a lieutenant orders the men in his platoon to clean their weapons before going out on patrol. The lieutenant knows, and is thought by his men to know, better than they do what they should do with respect to military matters—that is, he satisfies and is thought to satisfy the NJT in the military realm. This does not mean that the lieutenant reasons flawlessly in every case. It means only that, with respect to this realm of action, accepting and attempting to follow the lieutenant’s directives will likely cause each of the 244 Authoritative directives have a similar effect in prisoner’s dilemmas. See Raz (1988, 50-53). 173 people in the group to act in accordance with his preexisting reasons better than he would have if he tried to identify, weigh, and act on those reasons independently. Now imagine that the platoon has just come in from a long detail, and the men are clearly tired. Private B is particularly tired, and believes that he has most reason to get some sleep rather than clean his weapon. What ought he do? If B’s lieutenant is and B believes that he is a legitimate authority who satisfies the NJT—which we are supposing ex hypothesi—then the claim is that B ought to treat his directive as preempting his own practical deliberation. That is, he ought to entertain it in place of the reasons he takes himself to have in the circumstances. Why? Because by doing so B will have a higher probability of acting in accordance with his reasons than he would otherwise. But what if B does have most reason to get some sleep in this case, and the lieutenant has simply made a mistake? How can his directive nonetheless have normative force? The explanation is this. B recognizes quite clearly that he is tired, and that he needs sleep, but the lieutenant knows this as well, and so he should have, and likely did, consider it in his decision. Thus, were B now to weigh his tiredness against the lieutenant’s directive he would likely commit the error of double counting. That is, because the lieutenant, for all the soldier knows, has likely already weighed this consideration against any others that affect what the platoon ought to do, B would likely be weighing it again. Weighing the same reasons twice produces unreliable results. Thus, given that the lieutenant is, by hypothesis, more likely on the whole than B to select the right course of action with respect to these matters, B should forgo entertaining the reasons himself in any particular case. 245 245 What if the lieutenant in fact failed to take B’s tiredness into consideration? So long as in the long run B would do better by following his lieutenant’s directives, he ought to follow them in each 174 But wait, you say: In this case there is no question that the lieutenant made a mistake. We have stipulated that he has. How can a directive carry normative force when it is wrong by hypothesis? Practical reasons are, after all, facts which count in favor of performing certain actions. Directives are reasons, then, only so far as they count in favor of performing the actions they specify. But so far as a directive counts in favor of performing the act it specifies, it must be true that independent facts exist which justify performing that action, facts which give the directive the normative force the authority theory assigns to it. Since we are stipulating in this case that the lieutenant has made a mistake, we seem to be supposing that no such facts do this. Since without justification, his directive cannot count in favor of performing the act he specifies, it cannot be a reason to perform it. I take one of Raz’s most important insights to be that the normative force of authoritative directives, which ground the NJT generally, originates in the likelihood that one will do better on the whole (in the long run) by treating an authority’s directives as authoritative than one would otherwise. In other words, whether one has reason to do what a purported authority has directed on any particular occasion can only be appropriately evaluated in a broader context. Raz defends this point succinctly: “If every time a directive is mistaken, i.e. every time it fails to reflect reason correctly [and, we should add, every time it seems to fail in this respect], it were open to challenge as mistaken, the advantage gained by accepting the authority as a more reliable and successful guide to right reason would disappear.” 246 It is a individual case. However, there is one caveat, explained below, which is necessary to account for cases in which the subject of authority knows that he knows more than the authority does on some particular occasion.. 246 Raz (1988, 61). The same point is made in DeGeorge (1985, 39). 175 mistake, Raz suggests, to restrict our question of whether the authority’s directive has normative force to any individual case because the nature of authority is diachronic. Authority involves a range of potential cases and so should be evaluated with this in mind. The reason any particular directive is a reason to do what it says is that one will do better on the whole by accepting and treating every such directive as a reason. This fact separates directives issued by authorities from those issued by everyone else. If whether one ought to do what another person directs depends wholly on whether that person’s directive comports with one’s independent reasons in that particular case, then the benefits one would enjoy by treating the authority’s directives as authoritative would be lost, for the stranger’s directive would be as “authoritative” as the genuine authority’s. The point can be brought out by extending our military example. Suppose that over the course of Private B’s time in the military he has had a chance to evaluate his lieutenant’s judgment, and recognizes that, indeed, the lieutenant makes the right decision with respect to military matters more often than he does. In other words, B recognizes that his lieutenant satisfies the NJT, and because he satisfies it he is an authority with respect to B. Although one cannot determine with precision the frequency with which the lieutenant’s judgments are superior to B’s, let us say that the lieutenant gets it right about 20% more often than B does. Now sometimes B disagrees with the lieutenant’s orders, and on some of these occasions he finds later that, in fact, the lieutenant was wrong and he was right. On this basis B decides to start using his own judgment in certain cases. What ratio of success should he expect? Assuming, as we are, that B does not have access to information that is not excluded by the lieutenant’s directives, he should expect to do worse on the whole than he would were he to follow his lieutenant’s directives, at least within the relevant realm of action. So, say for 176 instance that B makes the right decision in 3 of every 5 judgment-situations, and that the lieutenant makes the right one in 4 of every 5. The more decisions B makes on his own, the closer he will get to the 3/5 ratio, which is the lower limit. The more he follows his lieutenant’s directives, the closer he will get to the 4/5 ratio, which is the upper limit. If he follows his own judgment half of the time, and the lieutenant’s the other half, the ratio for this mixed-strategy will, ceteris paribus, be 3.5/5. His best course—what he has most reason to do—will be to treat the lieutenant’s directives as preemptive reasons in all cases. And because the lieutenant satisfies the NJT with respect to B, they will be. Adopting this strategy is the best chance B has at reaching the “upper limit” of right action. Returning to the example we considered before, one should see now the sense in which the lieutenant’s order excludes certain of those reasons B would otherwise need to consider before acting, such as his tiredness. One should see, too, the sense in which the lieutenant’s directive is preemptive, for not only does it exclude a range of other reasons, it also takes their place. (To take their place, the directive must constitute a new first-order reason to do as it directs, which fills the void left by the excluded reasons.) It is this preemptive function that explains the difference authoritative directives make to how one ought to act, all things considered, even in cases in which they direct one to do something he should not do on the balance of reasons. The general idea behind all of this is that the collective force of the independent reasons one has to perform any particular action transfers to the authoritative directive that excludes them, since by following it one will satisfy more of those reasons than one would otherwise. That is, if I have most reason to φ, ψ, and χ, and if by obeying A I will be more likely to φ, ψ, and χ, then I have most reason to obey A, at least so far as the reasons I have 177 not to obey A do not weigh sufficiently against doing so. Generally, then, I have reason both to do what the authority says in the relevant realm of action and to refrain from acting on the reasons the authority’s directive excludes (lest I double-count). The reason I have to obey the authority in this realm of action generally explains the reason I have to do what the authority says on any particular occasion, even those in which the authority is mistaken. A person who substitutes his own judgment for that of the authority decreases the likelihood that he will act well, even if on occasion it pays off. When a Directive is Redundant To this point we have been discussing those cases in which an authority’s directives are mistaken, but what we have said also explains the other horn of our dilemma, namely, why authoritative directives matter when they direct one to do something one already ought to do on the balance of reasons. The reason has everything to do with preemption. As we said in the inconsistent case, the normative force of any particular directive comes from the likelihood one will do better on the whole by obeying the authority who issued it. So far as one would, one has reason to obey the authority’s directives generally, and so for any particular directive one has reason to obey it. The question now is how a directive can make a normative difference if it recommends the same action as one’s independent reasons. For if the force of a directive ultimately traces back to facts which support the authority’s directives generally, and those facts provide sufficient reasons for performing the action it specifies, then how could the directive matter in that particular case? This objection suggests that a directive can only be normatively relevant if it recommends an action distinct from the action conclusively supported by one’s independent 178 reasons. Raz argues, however, that an authoritative directive makes a normative difference when it counts as a reason to perform an action, whatever that action is, and whatever other reasons recommend it. As Raz puts it, it makes a difference to why one ought to do what one ought to do—that is, to the reasons to which we can appeal to justify our actions. 247 Consider the same example again, only now let us suppose that private B has most reason to clean his gun, which is what the lieutenant ordered. The question is whether the order has normative significance for B. Raz claims that it does. Why? For the same reason it did in the inconsistent case: the independent reasons are preempted. Recall that preemption was explained in the last case partly by the fact that if B entertained excluded reasons then he would likely double-count them. However this is true no matter which action those reasons support. In other words, the exclusionary scope of preemptive reasons does not discriminate between competing and complimentary reasons: it excludes all of them. Again, the reason the authority’s directive counts as a reason is that, so far as there is a greater likelihood one will do better by complying with her directives in the long run, those directives weigh in favor of performing the actions they recommend. This holds whether the directive is mistaken, or whether it recommends an action one has most reason to perform already. 247 Raz (1988, 60). Note that on Raz’s view authorities ought to base their directives on the reasons their subjects have independently. He calls this the dependence thesis. The dependence thesis seems true as far as it goes. However, so long as an authority must satisfy the NJT, it is unclear to me how the dependence thesis matters to the question of authoritative legitimacy. For suppose that the reasons on which A bases her directives to some group, which includes B, has nothing to do with B’s reasons. If it turns out that A’s directives cause her to satisfy the NJT with respect to B because her interests align with B’s reasons, then A will nonetheless be a practical authority for him, according to the NJT, within the limited realm in which their reasons and interests overlap. These are of course unlikely circumstances. But they show at least that the dependence thesis, if it matters at all, does so for some reason other than that satisfying it is necessary for an authority’s legitimacy. 179 A Caveat There is an important caveat to what I’ve said, however, which speaks to a concern you may have had already: What if I know that the authority is mistaken? If I know that I ought to φ, but an authority, who is on the whole right more than I am about these things, orders me to ψ, does her directive still preempt my independent judgment on the matter? For instance, suppose B could do better than his lieutenant’s 4/5 ratio by following his lieutenant’s directives in all but those cases in which he knows (rather than simply suspects) that the lieutenant is wrong. If he can do this reliably then he will enjoy all of the benefits of the authority’s judgment while suffering fewer of the mistakes, and would thereby act rightly more frequently than predicted by what we previously thought of as the upper limit. B might overhear his lieutenant say to a corporal, for instance, that the reason he is having his men clean their weapons is that the operating manual for these weapons specifies that they be cleaned every thirty days. But B has the operating manual, and sees that it says they should be cleaned every one hundred thirty days. The lieutenant has simply misread the manual. He has made a mistake, and B knows it. Is it not true that in this case he should disregard his lieutenant’s directive? This sort of case illustrates the caveat. Disregarding the directive is, I believe, precisely what B should do, and the same is true for any case in which a subject knows a directive issued to him is faulty. 248 Raz says that he wishes to express no opinion on whether 248 The caveat actually extends a bit further than this, for even if B does not know in any particular case that his lieutenant has made a mistake, there is still a way in which he could do better than the 4/5 limit. This is if B were able to reliably identify and act on his own judgment in circumstances in which his lieutenant usually exhibits worse judgment than he does. For instance, it might be that, although overall the lieutenant’s judgment is superior to B’s, the lieutenant makes particularly poor 180 legitimate authorities are limited in these circumstances. However he seems to make space for this possibility when he distinguishes between “great” and “clear” mistakes. 249 Great mistakes are irrelevant to whether one should follow an authority’s directives, at least so long as those mistakes do not prevent the authority from satisfying the NJT. But clear mistakes, no matter how serious they are, appear to exempt one from the relevant directive’s normative force. The general force transferred to the authority by one’s independent reasons, because of the general advantage one would have in satisfying those reasons by obeying the authority, does not extend to cases in which one would certainly do better, or at least just as well, by acting on one’s own. Because such a case would then no longer be within the range of cases in which one would likely do better by obeying the authority’s directives, it would fall outside the realm of cases for which the authority is an authority. It would, so to speak, create small gaps in the authority-governed realm of action, like holes in Swiss cheese. I will argue in the Chapter 7 that for the purposes of political obligation it makes little difference whether the obligating force of authoritative directives is limited to cases in which subjects do not know definitively that they are mistaken. Where a subject frequently identifies clearly defective directives, it will in all but very rare cases turn out that the decisions on Wednesdays, owing to overnight patrols he leads the night before. B might find that, by following his own judgment on Wednesdays, he gets nearly all of the benefit of the lieutenant’s superior judgment, but avoids some of his errors. Thus, even though he does not know that his lieutenant is wrong in any particular case, B can do better than the 4/5 ratio by following his own judgment in certain identifiable circumstances in which the lieutenant’s judgment has proven to be worse. In these cases B doesn’t know what it is that he knows which makes his epistemic position superior. But it is superior nonetheless, and so far as he knows this, he cannot be criticized for disobeying the authority, at least in light of the reasons given in defense of the NJT. 249 Raz (1988, 62). 181 authority in fact fails to satisfy the NJT with respect to the subject and the realm of action in question, and in these cases the purported authority will not be an authority. If, on the other hand, the subject happens on occasion to know something or be positioned such that he identifies a directive’s clear defect, then in these cases the directive will not bind him, at least relative to the interest the authority exists to serve. This will not, however, exempt him in cases in which he does not know that the authority is mistaken, and these constitute the bulk of such cases, at least in complex, modern political communities. 5.2.3 Conclusion The conclusion the authority theory aims to have reached, then, is that authoritative directives are preemptive reasons which exclude and replace a range of reasons the subject would for that reason be irrational to entertain. As our military case suggests, these reasons need not depend on the desires or aims of the subject. In many cases there will be good reason for the members of a platoon to storm a building, or to stay up all night securing a perimeter. They may have no desire to do these things; they may prefer to sleep, or play cards or whatever, and these desires may give each of them reasons to do these other things. But they cannot nullify the reasons they have to do what their platoon ought. In these cases the lieutenant’s directives, so far as they are authoritative, constitute preemptive reasons that are categorical in nature, and therefore constitute obligations for members of the platoon, at least according to the conception of obligation we have developed to this point. As I mentioned at the beginning of this chapter, most of us are not in a position to be a practical authority for many people, and where we are the scope of our authority is 182 usually quite limited. 250 But authority is plausibly more common and extensive for governments, which have the monetary and informational resources to determine what actions ought to be taken when it would be tremendously burdensome, if possible at all, for people to do this on their own. In some cases, such as coordination problems or prisoner’s dilemmas, the government is in a position to determine right action more directly by creating through its directives the circumstances in which some otherwise unattainable goods can be achieved. I hope by this point to have provided a fairly complete picture of one important way in which authorities can be justified, as well as an explanation of why their position as authorities allows them to generate obligations for others. 5.3 Objections to the Authority Theory Despite its appeal, no part of the authority theory (as we’ve so far developed it) has gone unchallenged. 251 In this section I attempt to defend the general approach from two of the three most important objections in the recent literature. I call these the no-authority and no- obligation objections. The third of the trio—the proceduralist objection—I discuss in detail in Chapter 7. 250 The ubiquity of authority depends largely on the sorts of relationships to which it properly extends. As I mentioned in an earlier footnote, I believe that authority is in fact partly constitutive of friendships and other relationships which depend on our ability to obligate one another, within a particular realm of action, simply by expressing our intentions to do so. Some philosophers (Raz, and Marmor, e.g.) reject this analysis. Because this issue has no significant bearing on the argument of the dissertation, I leave it aside. 251 For critiques of this conception of authority as a whole, see in particular: Darwall (2006, 15 fn.29), (2009) and (2010); Hershovitz (2003) and (2010); Marmor (2005), (2010) and (2011); Shapiro (2002). For a critique of Raz’s understanding of obligation see Owens (2008). For criticism of his conception 183 The main thrust of the no-authority objection is that an authority must have a right to rule, and that, because one can satisfy the NJT without having such a right, satisfying it is insufficient to ground authority. This objection’s most immediate target is the authority theory’s understanding of legitimate practical authority. When conjoined with a few plausible presuppositions, however, it follows from the no-authority objection that the authority theory cannot explain political obligation, and therefore poses a genuine threat to the theory. The no-obligation objection targets the authority theory’s argument for political obligation more directly. To be more precise, it is a family of related objections, which together suggest that even if authority is not a necessary precondition for normative power, satisfying the NJT is not a sufficient one. The family is divided into two strands. According to the first, a person satisfying the NJT cannot generate genuine obligations for me because she cannot generate preemptive reasons, either because preemptive reasons entail certain relations which the NJT cannot explain, or because, these relations aside, there is no sense in which directives issued by someone satisfying the NJT can exclude and replace reasons I have independently. The second strand of the argument accepts that agents satisfying the NJT are practical authorities, as well as that they can generate preemptive reasons, but questions, first, whether the preemptive reasons they generate are categorical, and second, whether, even if they are categorical, this notion of obligation is indeed the one that relates to the question of political obligation. I’ll argue that both the no-authority and the first strand of the no-obligation objection fail. In the first section I explain the dilemma the no-authority objection poses for of content-independent and exclusionary reasons, see (respectively) Sciaraffa (2009) and Edmundson (1993). 184 Raz: He must either give up the notion that practical authorities have a right to rule, or explain that right in functionalist or institutional terms. I then explain the functionalist conception Raz elaborates in his most recent (unpublished) response to the no-authority objection. 252 While I do not go so far as to deny that this conception can withstand the no- authority objection, I do suggest that proponents of the authority theory would do better to take the alternative route of giving up the notion that practical authorities must have a right to rule. In the subsequent section I take up the first strand of the no-obligation objection. It is, as I’ll argue, either problematically based on implausible assumptions about the relationship between accountability and obligation, or simply a misreading of Raz. What the objections we consider in this chapter have in common is that they can all be addressed without modifying the substance of the Razian account. While I answer some of them differently than Raz does—and while the answers I provide I will consider, moving forward, as the appropriate defense and elaboration of the authority theory—the updated conception is still firmly grounded in the theoretical apparatus we have developed in this chapter. The case is different for the second strand of the no-obligation objection, which I discuss in the next chapter. As I’ll explain in detail there, the foundational conception on which the Raz’s theory is based—a relation between authorities and subjects tracing back to those subjects’ reasons for action—fails to take into account the nature of those reasons and their connection to the relations each subject bears to other members of his or her political 252 Two important points: First, Raz has, so far as I know, never described his view explicitly as “functionalist.” I call it this because his defense of his conception of legitimate authority turns on the authority’s function, as I explain below. Second, while Raz’s conception has a number of things in common with the institutional conception—defended by, e.g., Marmor (2010)—they may be substantially different. 185 community. Part of the project of the next chapter is to reconstruct the authority theory to overcome this problem. 5.3.1 The No-Authority Objection Let me briefly summarize the objection. The NJT is insufficient to ground legitimate practical authority because to have such authority one must have a right to rule, and satisfying the NJT does not entail that one has such a right. 253 Thus, if a person must have authority to obligate others through her directives, then it follows that someone who is a purported authority merely in virtue of satisfying the NJT may not be capable of obligating anyone. Since a theory of political obligation is supposed to show that a governing body can obligate their subject whenever they satisfy the conditions set out by the theory, the authority theory fails as a theory of political obligation. The Objection in Detail According to the prevailing version of the no-authority objection, 254 Raz claims that (DII) follows from (DI): (DI) If B would do better in complying with independently existing reasons were B to treat A’s directives as preemptive reasons, then A’s directives actually are such preemptive reasons for B. 253 All references to authority will be to practical authority, unless otherwise noted. 254 See Darwall (2009) and (2010), from which the following argument is drawn. 186 (DII) If B would do better in complying with independently existing reasons were B to treat A’s directives as preemptive reasons, then A has authority with respect to B. The objection is that, even if (DI) is true, it does not entail (DII); in other words, the fact that A’s directives constitute preemptive reasons for B does not entail that A is B’s authority. Therefore, the argument contends, the authority theory cannot explain how individuals satisfying the NJT can generate obligations for those for whom they generate preemptive reasons, since to generate obligations one must be an authority. The following story, which I shall call the financial advisor case, has been offered to illustrate the objection: …imagine that I have prudential reasons to provide for my retirement that are independent of any obligation I might have, say, to provide for others whom I am answerable for supporting or of any obligation to support myself. It seems obvious that I do have such reason and…that it could indeed make sense for me in prudential terms to put myself in the hands of a financial expert and simply follow her directives. Suppose, then, that I would better comply with the relevant prudential reasons if I were to do so and that there are no other reasons, or at any rate no sufficient reasons, for me not to do so. 255 We assume here that, in virtue of the circumstances, my financial advisor’s directives are preemptive for me. In such a case, the objection goes, I might have a reason to act as though the financial advisor had practical authority over me, so as to derive any benefits (in terms of reason-satisfaction) such behavior might elicit. But this does not show that my advisor 255 Darwall (2010, 259). 187 actually has authority over me. So even if (DI) is true with respect to the financial advisor and me, it does not entail (DII). Of course, it is only with respect to an account of practical authority that the entailment between (DI) and (DII) can be denied. Thus, the no-authority objection must provide at least some account of what being such an authority consists in, one that resists the entailment while conforming to our pretheoretical conceptions of what an authority is. The conception we will work with, and the one I understand Raz’s critics to have in mind when they raise this objection, provides two central conditions. For A to have practical authority over B, it must be the case that: (i) A has the right to direct B, and (ii) B is accountable to A to obey her directives. These conditions make it easier to see why my financial advisor cannot plausibly be a practical authority for me, for she appears to satisfy neither condition. First, there seems to be no sense in which she has a right that I invest as she directs. 256 The sort of right the objection has in mind here is a claim-right—a claim one person has on another to behave in a way that corresponds with an obligation the other person has to one to so behave. 257 It is a claim one has “against” a particular person to perform or not perform some act. One way to determine whether a person has such a claim on anybody is to determine whether any person can wrong her by performing or failing to perform some action. And this is precisely the sort of question we can ask in the financial advisor case. Barring an adequate excuse, 256 Darwall (2009, 151) makes the same point when he discusses Raz’s infamous “Chinese Cooking” example. 257 This is one of Hohfeld’s “jural correlatives.” See Hohfeld (1917). 188 would I wrong my financial advisor by failing to invest as she directed? Quite clearly: no. She may be disappointed in me, or think I am foolish, but I do not wrong her. So condition (i) is not met, for a right implies a possibility of wronging that does not exist in this case. (ii) then follows from (i) if (i) refers to a claim-right, in which case, for the same reason, it is hard to see how condition (ii) could be met in this case. We have yet to say what being accountable to someone amounts to, but we can get at the core of the idea ostensively. If you ask me to invest some of your money in particular stocks, and I agree, then it is clear that I am accountable to you to invest the money as you asked. At the very least I owe it to you to do so—my obligation is, as it is said, “directed” towards you. For not only would I be wrong to invest your money in what I believed to be more lucrative stocks, or, say, to buy you a new car instead of investing your money in stocks at all, but by doing so I would wrong you. This is, for us, the relevant feature in these cases. Whatever the relation of accountability ultimately amounts to, it seems to follow from my being accountable to you to do something that, absent circumstances that would excuse my behavior, I wrong you by failing to do it. But, as we already said, this is precisely the relation that appears to be missing in the case of my financial advisor. I do not wrong her by failing to invest as she directs, for I do not bear any sort of relation to her that would make such a wronging possible. I cannot therefore be accountable to her to do as she says. It follows from these two conclusions—that my advisor does not have a right that I invest as she says, and that I am not accountable to her to do so—that she cannot be a practical authority for me. And so it seems that, as the objection contends, (DI) does not entail (DII). Thus, so far as the authority theory depends on this entailment to explain political obligation, this failure is fatal. 189 Solutions to the No-Authority Objection An obvious way around the no-authority objection would be to simply reject conditions (i) and (ii); and as I’ll argue below, an account of practical authority that has been developed with attention to its role in our practical reasoning does just that. It is, according to this more plausible account, simply not true that practical authorities must have a right to rule, and after this is established, it follows that the accountability relation to which the no-authority objection refers, to whatever it amounts, is irrelevant. However Raz favors a strategy which countenances (i) and (ii). Because his solution may for this reason be considered by some commentators to be less “revisionary” than the one I’ll provide, I must explain first why we should not accept his. In what follows, then, we will need to develop two lines of thought. First, I shall consider Raz’s response to the no- authority objection, in which he tries to explain how his theory satisfies condition (i) and, in a modified form, condition (ii). I’ll argue that, while the functionalist conception of authority he invokes plausibly satisfies both conditions, it introduces other difficult and unnecessary challenges, the most important of which is that the interpretation of the normal justification thesis it forces us to endorse fails to respect the service conception it is ultimately meant to reflect. Rather than accept such a serious cost, I argue that we should deny the conception of practical authority our critics endorse. 190 Rights: A Brief Clarification The debate surrounding the no-authority objection has spawned a number of misunderstandings which I believe are due in part to a miscommunication about rights. Before moving forward, I would like to make a few clarificatory remarks. As I mentioned already, the sort of right the no-authority objection claims the NJT cannot explain is a claim-right. Claim-rights are widely thought to correlate with obligations which are directed towards the claim holder, however there are other sorts of rights to which any use of the term could refer, including, on Hohfeld’s analysis, liberties (or privileges), powers, and immunities. This is important because, while Raz accepts that a practical authority must have a right to rule, he never to my knowledge claims that such a right must be a claim-right. Practical authority, he says at one point, is “authority as involving essentially the power to require action.” 258 It is, more specifically, “the power to impose duties on others simply by expressing an intention to do so.” 259 On these occasions Raz seems to have in mind a normative power rather than a claim-right. Were this Raz’s view, however, there would be no reason for him to attempt to answer the no-authority objection by accounting for authorities’ claim-rights, for there would not necessarily be any claims for which to account. Since, as I will explain in a moment, Raz does try to explain how such a right is generated by his theory, we must assume either that he accepts that authorities must have a claim on their subjects to obey, or that, though they have only a normative power, this too is a type of right which must be explained 258 Raz (1988, 38). 259 Raz (2010, 292). 191 by reference to an interest of the right holder sufficient to ground duties in others. This second possibility is, however, a non-starter, for it misapplies Raz’s interest theory of rights, which provides an analysis of claims, not powers. We should assume, therefore, that Raz takes the sort of right authorities must have to be, or at least to include, a claim-right, as the no-authority objection assumes. In what follows we will evaluate Raz’s view to see whether the no-authority objection can be answered by a version of the authority theory which accepts that practical authority consists in (or at least necessarily includes) claims on obedience. However the cost of this approach is, as I will argue later, unappealing enough to reject it in light of the available alternatives. 5.3.2 Raz’s Solution to the No-Authority Objection Raz responds to the no-authority objection by first casting it as an attack against his view of rights, rather than his view of authority per se. He accepts that a practical authority must have a right to rule, but claims that a proper understanding of this right explains how the NJT imports it. 260 According to Raz, “there is a right if and only if some interest of some entity capable of being a right-holder is sufficient to ground a duty to protect or secure that interest in a significant way.” 261 Raz’s task, then, is to explain the interest a purported authority has in being obeyed by those with respect to whom she satisfies the NJT. 260 Raz (unpublished, 17). Henceforth in this section when I talk simply about “rights” in the context of authority I will mean claim-rights. 261 Raz (unpublished, 9). This summarizes a lengthier discussion of rights (1988, Chapter 7) in which Raz conceives of the relevant interest in terms of a person’s well-being. One of the aims of Raz’s newest (unpublished) paper is to modify his view in the face of particular objections, largely those leveled by F.M. Kamm in her Intricate Ethics. 192 But what could this interest be? 262 As we saw, the no-authority objection uses the financial advisor case to illustrate the problem. It asks: Does satisfying the NJT explain in what sense my financial advisor has, not just an interest in my investing as she says, but an interest sufficient to ground my duty to do so? Two responses are available: explain the duty or reject the case. Raz chooses to reject the case. This is the right move because to have a duty to do something one must have a categorical reason to do it, and in this case it appears that I have no categorical reason to invest as my advisor directs. Confusingly, however, this is not Raz’s reason for rejecting the case. For him the problem lies in the fact that my financial advisor is merely an expert, the correctness of whose instruction “is not based on the fact that the expert will so instruct, or has so instructed.” 263 This amounts to what those familiar with Raz’s work will rightly understand to be an unprecedented admission: “On the service account,” he says, “…no legitimate authority can be based on superior knowledge alone.” This is unprecedented because Raz’s account has standardly been understood along precisely these epistemic lines. And as I explain below, I think that in giving up this justification of practical authorities, Raz gives up too much. But for now, let’s try to reissue the objection in terms Raz would accept. Raz suggests that the right sort of justification will most commonly involve others’ reliance on it by, in addition to solving coordination problems, “enabling consumers to rely 262 Understanding Raz’s view of rights is particularly challenging at this stage because he recently revised his longstanding view that a right depends on a person’s interest, contending now that it depends on a person’s value (unpublished, 10). Despite this change, however, Raz (confusingly) discusses the no-authority objection in terms that correspond with his traditional view. To avoid getting distracted by these details, we shall talk (as Raz does) in terms of interests, ignoring, for arguments sake, any objections facing this view which could plausibly be overcome by grounding rights in value as Raz attempts to. 263 Raz (unpublished, 20). 193 on products, or on providers of services meeting certain standards, enabling road users, or others who may be affected by the conduct of others, to rely on those others meeting certain standards, and so on.” 264 As Scott Hershovitz has pointed out, however, there is no reason to think that my capacity to solve a coordination problem of which you are a part would give me a right to your obedience. Suppose, for instance, that I have worked out a scheme to greatly reduce the congestion in New York subway stations, and that everyone in the subway has good reasons both to take the subway and to attenuate the congestion. The scheme requires those waiting for their subway car to stand in one of a number of lines strategically divided across the platform. Because it would enable subway users to use the subway more efficiently, much as traffic laws enable road users to use the roads more safely, the scheme enables reliance of the sort that is constitutive of authority on Raz’s account. Do these facts entail that, when I go into the subway Tuesday morning, I have a right to show everyone where the lines should be formed and to order them to stand in them? It is hard to see how it could. It seems, then, that these sorts of cases are no more immune to the no-authority objection than those which turn on the sorts of epistemic considerations for which Raz rejects the financial advisor case. 265 A Functionalist Account A major problem in the subway case is that, even if people have weighty reasons to rely on me, the nature of the interest I have in their obedience—an interest which, by Raz’s own 264 Raz (unpublished, 21). Raz discusses coordination problems on the previous page. 265 Hershovitz (2010, 19). 194 lights, must be sufficient to ground their duty to obey—is somewhat mysterious. Raz’s explanation of this interest is in fact equally mysterious: …the rights of authorities over their subjects do not derive from the interests of the people in authority. Rather, they derive from the interests of the authority, which is to be a good authority, and whatever helps it in that is in its interest. The whole point of the service conception of authority is to explain the separation between the interests of the people in authority and the authority they hold, and to explain the latter by reference to the reasons which apply to the people under their authority. 266 I take the most plausible interpretation of this passage to be a functionalist one, which I interpret as follows. The interests we are concerned with when we are concerned to justify practical authority are not the interests of the authority qua person, but her interests qua authority. Admittedly, it is difficult to understand in what sense we can assign interests to the authority a person holds independently of that person. But it does seem right to say that whatever the interests of authorities consist in, they plausibly exclude the private interests of the people holding authoritative positions. The types of food a person prefers to eat, the sorts of performances she enjoys, the exercise routines that best suit her, and so on, are not in her interest qua authority, and so the advancement of these interests does not contribute to her legitimacy, even if her status as an authority would advance them. What I think Raz means to say here, then, is that an authority’s interests are to be understood in terms of the function that authority performs for its subjects. When we are concerned about a potentially legitimate authority, Raz seems to say, we are to look at that authority, not as an individual—a person with aims, desires, and preferences of her own, which bear on her normative circumstances—but as the function 266 Raz (unpublished, 18; my emphasis). 195 she happens to perform. That Raz means this is indicated by the fact that an authority’s function—the service it performs qua authority—is precisely what’s explained “by reference to the reasons which apply to the people under [its] authority.” With this in mind, another way of putting the question “What interest do purported authorities have in being obeyed?” is “What interest is there in having these sorts of authorities?” This is to put oneself in the position of the subject—to ask, from the first-person perspective: Why have this sort of authority? 267 The answer is that there is sufficient interest in having a particular authority to the extent that the people subject to it have sufficient reason to have that sort of function performed. 268 Thus, on the interpretation of Raz’s view I’ve offered, justifying authority amounts to demonstrating this sort of compelling interest in its functional existence. 269 Such a function is justified, Raz seems to say, when a subject’s reasons for having it are sufficiently weighty, in which case those reasons are constitutive of the authority’s interest with respect 267 See Korsgaard (1996). 268 See Hershovitz (2010, 20-22) for a competing interpretation. In my view Hershovitz puts too much emphasis on the interests of the authority’s subjects, whereas Raz seems clearly from the quoted passage to be concerned with their reasons. Further, I think that what counts as a “sufficient” grounding for the right in question—something Hershovitz claims Raz fails to adequately explain— can be explained in terms of the weight of those reasons, as I have suggested here. 269 As I mentioned already, Raz does not label his view “functionalist.” My claim is merely that this is the most plausible interpretation of what he says. If this is right, then Raz would appear to endorse a position that is at least somewhat similar to one put forward recently by Andrei Marmor (2010). Marmor claims that, while the NJT may not “answer the local question of why comply with this particular instruction on this particular occasion,” it “might be a good answer to the question of why have that kind of authority at all, and why it is good, in the long term, to have such an authority make certain kinds of decisions for us” (11). On Marmor’s institutional picture, the NJT is concerned with “the long term, overall, reasons to have the kind of authority in question” (10). Aside from the institutional nature of Marmor’s analysis, his conception is not totally unlike Raz’s; though I’ll note that Marmor’s view diverges from Raz’s in that he rejects that authorities must have a right to rule (as I do below). On this point see Marmor (2011). 196 to that subject, and so, of that authority’s right to be obeyed. In this way the NJT satisfies condition (i). What about condition (ii)? Darwall claims that to have a right to rule a purported authority’s subjects must be accountable to her. Even if Raz’s explanation delivers an interest sufficient to ground a duty for others to obey an authority, does it explain on what basis that authority has the standing to hold her subjects accountable? If his recent work is any indication, Raz does not think that it has to. In his 2010 article, he attacks Darwall’s notion of accountability, claiming that it tells us little if anything of interest about moral obligation. 270 Whether Raz is right or not, it is clear that Raz does not believe his theory of authority (indeed, any theory of genuine practical authority) needs to account for this sort of relation. What is interesting about this, however, is that in his most recent (unpublished) response to the no-authority objection, 271 Raz describes this very relation, claiming that the NJT explains how subjects of authorities are accountable to them to do as they direct. 272 This has no doubt caused some confusion. So let me say something about what might be going on here. Given Raz’s explicit rejection of the accountability relation, we should accept that he rejects it. He says that his rejection is based partly on the fact that he cannot understand what accountability could amount to in light of the problems he himself raises for it. 273 It might seem best, then, to simply reject condition (ii) as one that is relevant only to theories 270 Raz (2010, 293-95). 271 “The possibility of partiality” (unpublished). See particularly pp.18-20. 272 Raz (unpublished, 18). Hershovitz (2010) has also offered a critique of Raz’s unpublished defense. 273 See Raz (2010, 299). 197 of authority that align or presuppose moral relations of the sort Darwall’s theory includes. However I think we would move too fast by doing this. We can all agree, I think, that whatever relation one person stands in to another when he is accountable to her for something, he at least owes it to her. 274 So even if we suppose that Raz is right that this relation of accountability amounts to nothing, or that, whatever it amounts to, it has not been satisfactorily explained or is irrelevant to genuine practical authority, we can still reduce this relation to a clearer relation that it includes: that of owing. An obligation that is owed to someone is a directed obligation, which is relevant here in virtue of two claims Raz appears to endorse. First, he claims that all rights attach to obligations that justify them. 275 Second, Raz seems to accept, though perhaps less clearly, that obligations which justify rights are owed to the right holder. 276 Thus, while Raz may reject condition (ii) in virtue of its reliance on what he sees as a muddled notion of accountability, directed obligations, which the accountability relation includes, Raz accepts he must at least “in some sense” explain. This means that, while Raz rejects Darwall’s 274 Even this is not without complications. What do we say, for instance, about deathbed promises? It seems to me that the accountability relation entails the owing relation in normal cases (where, e.g., both parties are alive), so I see no objection to setting these unusual cases aside. 275 This relation is part of a definition of rights which “aims to encapsulate,” he says, “the common core of all rights” (1988, 167). See also (2010, 290). 276 “It is common ground to both of us [him and Darwall]” Raz says “…that when there are rights that have corresponding duties (and I think that, like me, Darwall thinks that all rights have corresponding duties), then those duties are duties that are in some sense owed to the right-holder” (2010, 290). Also see (1988, 26): “[De facto authorities, who claim a right to rule] have legitimate authority only if and to the extent that their claim is justified and they are owed a duty of obedience.” (Raz goes on to critique a competing view of authority for accepting that “there can be [a legitimate authority] which does not claim that it is owed such a duty.”). Also see (1988, 181), where Raz states that something relevant to seeing whether an interest is sufficient to ground a right is that premises are supplied which show that the interest is “relevant to a particular person or class of persons so that they rather than others are obligated to the right-holder” (my emphasis). 198 accountability condition, that condition includes a relation that Raz admits holds “in some sense” between authorities and subjects, and so one that he obviously cannot reject outright as simply confused. Therefore, condition (ii) remains problematic even after we strip it down to the arguably simpler owing relation. In what follows, then, we will take condition (ii) to involve the owing rather than the accountability relation, which Raz apparently believes his view satisfies. “Duty is owed to the authority,” Raz says …in that it arises out of the right of the authority to direct its subjects, which right is itself in the interest of those subjects, each and every one of them. It is in a manner of speaking a duty you owe to yourself, because—according to the service conception—the authority is your servant, and in defying it you fail yourself. 277 Though Raz’s point here shares some of the obliquity of the previous one, I take them to mean, together, that an authority’s right is based, ultimately, in the reasons its subjects have to have someone perform its general function—that is, the reasons they have to have that sort of authority—and that in virtue of this each subject owes the duty to herself. As some critics have pointed out, elsewhere Raz states that authorities should not necessarily act in the interest of their subjects. 278 Sometimes subjects’ interests conflict with their reasons, and so on Raz’s own view their satisfaction should play no part in justifying authority. I think Raz’s position is consistent with this observation despite the fact that he mentions subjects’ “interests.” From the previous passage it is clear that on Raz’s view a right to rule must be justified in terms of subjects’ reasons, not their interests. Perhaps Raz 277 Raz (unpublished, 18). 278 See Hershovitz (2010, 21); Raz makes this statement at (1988, 48). 199 means to be saying that it is in their interest to the extent that they have an interest in doing what they have most reason to do. Perhaps he simply misspoke. Either way, the point is that when subjects’ reasons are sufficient to justify the authority’s right to rule, it is in virtue of those reasons that they owe her obedience. To say that subjects owe obedience to the authority is, however, just a manner of speaking. The “interest of the authority”—the interest there is in having the sort of authority in question—is ultimately explained by reference to the authority’s subjects. Because their reasons for action are ultimately the one’s that justify the authority, in owing obedience to the authority, subjects in fact owe it to themselves. It is they who have a stake in their own obedience, for the directives they obey help them accord with their own independent reasons. They owe it to themselves, on Raz’s view, via owing it to the authority. This is, anyway, what I take to be the most plausible interpretation of the response Raz has put forward. To sum up, then, on Raz’s account the authority’s interest is just the interest there is in its function, which, when this interest is important enough—when subjects’ reasons for having this function performed are sufficiently weighty—grounds each subject’s duty to obey it, and thereby explains the authority’s right to their obedience. Put simply: an authority has a right to rule when people have good enough reasons to support the authoritative function that that particular authority provides. This explains condition (i) from the beginning of this section. And because the authority on Raz’s view is in essence simply a function subjects have good reason to have performed, subjects owe it to themselves to obey. This explains (our modified version of) condition (ii). Where subjects defy their authority, they defy themselves, for they defy their own reasons. Taken together, these considerations are Raz’s answer to the no-authority objection. 200 An Example An example will hopefully clarify the intuition supporting this seemingly complicated account. Consider Smith, a philosophy professor. We generally take professors to have practical authority regarding matters relevant to the courses they teach, and presumably regarding practical matters in the area in which they specialize generally with respect to the group of students taking those courses. Suppose Smith recognizes that only if she directs her students to do a certain amount of work will they acquire the body of knowledge they have reason to acquire during the course. But suppose that the work she assigns is poorly chosen—either much too hard or far too easy—and that the students are learning very little as a result. Suppose, in fact, that the students, who have most reason to acquire the body of knowledge the course advertises it affords, do worse in satisfying these reasons by doing as Professor Smith directs than they would by trying to learn these things on their own. Are they bound by the NJT to do as she says regarding matters of the course anyway? According to the standard (non-functionalist) interpretation, whether they are, and whether the professor is a practical authority for them, depends on whether by doing what she says it is likely that they will better conform over time with the reasons they have independently. Call this the personal justificatory interpretation, and compare it with Raz’s functionalist account. According to the latter, the NJT’s justificatory scope is wider, taking into account the functional role this sort of authority plays for the students, rather than merely the practical advantage offered by this particular authority. On one interpretation of this functional role, what matters is not whether the students will do better by acting in 201 accordance with the directives of Professor Smith, but whether they will do better by having professors at all. What a theorist takes to be the relevant function will vary, of course, and significant questions turn on it. We might ask, for instance, how the students would fair without universities, of which professors are an integral part, rather than without professors in particular. The rules by which we specify the justificatory scope will matter in that, as broader and broader functions are considered, subjects’ reasons, and thus their professors’ claims on them, may gain and lose legitimacy. Perhaps I would do better on my own than I would with professors, but do better with universities, which require professors, than I would on my own. Here the move from professors to universities may deliver authority according to the theory. However we do not have to settle this here, because the point is the same whatever rules we settle on. Let’s suppose, then, that the relevant functional role is the one played by professors generally. Because students would do better without this particular professor, the personal interpretation suggests that, so far as the professor’s authority is based in the NJT, she is not an authority for them, which seems like the wrong result. The functionalist account may therefore appear to provide a better explanation, for it claims that, while students do worse with respect to academic matters by doing as this professor says, they surely do better by having professors than they would going it alone. And so long as students do better with the relevant functional role being played (taking into account the possibility that certain professors might be quite bad) than they would without Professor Smith, but also without her role, she satisfies the NJT, and this justifies her authority over her students. 202 This realigns the authority theory with our intuitions about the case. Smith has practical authority over her students (within the realm of action in question) because the interest in having her functional role is weighty enough to ground the students’ duty to obey her directives. They owe it to her to do as she says in the sense that they owe it to themselves to do what they have most reason to do—which is to do as someone with this sort of authority tells them—for they are their reasons which ultimately justify Professor Smith’s authority. By shrugging their assignments the students fail themselves, for they act against their own reasons. This is, so far as I can see, the most compelling interpretation of the functionalist conception that fits with Raz’s response to the no-authority objection. Again, it attempts to escape the objection by explaining both the manner in which a person satisfying the NJT has a right to rule, and the manner in which subjects’ of that authority owe her obedience. These are both explained in functionalist terms, by reference to the reasons subjects have to have the sort of authority in question. Objections Is the functionalist conception compelling? I think there are a number of reasons to be skeptical. The first has to do with the notion that an authority’s interest in (and thus right to) obedience can be grounded in the reasons subjects have to have that sort of authority. There is something awkward about attributing an interest to an authoritative function independently of the person performing it. But for the moment let us set this problem aside. As Raz ultimately understands it, the interest of the authority originates in her subjects’ reasons. That is why he says that in defying your authority, you fail yourself, for you act 203 against your own reason. But if you ultimately (“in a manner of speaking”) owe it to yourself to do what you have been ordered to do by the authority, then it seems to follow that (in the same manner of speaking) you have right to your own performance. (In a manner of speaking) you rule yourself. Whatever the appeal of this sort of claim for some, Raz rejects it elsewhere. “A conventional cliché says that democracies constitute self-rule,” he says. “Even if it is true that in democracies the people rule themselves, I do not rule myself, not even in a democracy.” 279 It could be, of course, that political authority is unique from other sorts of practical authority in this way, but more would need to be said to shore up such a response. If people have a duty to do what an authority told them to do—a performance that “in a manner of speaking” they owe to themselves—then in the same sense they must have a claim on themselves for their own performance. This seems to constitute the sort of self-rule Raz rejects. One way to avoid this problem would be to reject the interest theory of rights. As an advocate of that theory, Raz must, as we saw, account for the interest grounding any right he claims exists. But the authority theory need not be committed to the interest theory of rights. We are free to develop it in whatever way we would like. It would of course be problematic if our theory did not align with any plausible theory of rights; however this is not the case. On the leading alternative view, the will theory, for instance, rights do not necessarily further interests, but rather give people some degree of control over others’ duties. Their function has not to do with one’s interests, but with the choices one has available. 280 An advocate of 279 Raz (1990, 3). 280 See Wenar (2010) for discussion. 204 the authority theory who subscribed to the will theory of rights would not, therefore, be faced with the difficult task of explaining the interest a practical authority has in being obeyed, for practical authorities need no such interest to be genuine authorities. Instead, one would merely have to show that someone who satisfies the NJT has control over some class of duties strapping those with respect to whom she satisfies it. None of this is to say, of course, that the will theory is perfectly compatible with the normative capacities the authority theory claims those satisfying the NJT enjoy. I merely note that some of the primary problems the authority theory seems to face might be overcome by rejecting the interest theory, which, Raz aside, advocates of the theory are free to do. But let us suppose that the interest theory of rights is the correct theory. The second problem is that the interest grounding one person’s right to another’s performance would seem, at least in some instances, to entail, rather than to follow from, the reasons that other person has to have the sort of authority that binds her. (Raz’s conception has things the other way around.) This becomes clearer if we apply the functionalist account to other obligation-generating institutions, such as promising. Imagine that I promise you that I will do as you say regarding matters of your bakery so long as I work there, thereby giving you a right to my performance. This makes you a practical authority over me regarding bakery matters, and gives you a right to my performance that is plausibly grounded in your interest. But what ultimately explains that interest? According to Raz’s functionalist conception, it is the reasons I have for participating in the institution that binds me in the first place. Even if I would prefer not to do many of the things you tell me to do at the bakery, and even if I have little reason (independently of my promise) to do them, I have weighty reasons to be part of an institution in which people can undertake to be accountable to others to perform 205 certain actions. This explains why I am obligated to do what I promise you I will do on any particular occasion, even when, on that occasion, I have little reason aside from the promise to do it. On the functionalist conception your right to my performance ultimately traces back to my reasons for having the functional role played by those to whom I choose to bind myself—in this case, you. 281 The analogy can be drawn more carefully this way. B will be obligated to A to φ when A directs her to φ if A satisfies the NJT with respect to B. On Raz’s analysis, this is explained by the right A has to B’s performance, which is constituted by A’s interest in B’s obedience, which traces back to B’s reasons for having the function in question performed by A. Now compare this with promising. B will be obligated to A to φ if (in a suitable context) B utters “I promise to φ” to A. (Remember: This is true in spite of the fact that B did not opt to be part of the institution of promising. He may prefer to make promise-like utterances sans the normative baggage most people attach to them.) Here too, on the functionalist analysis, B’s obligation is explained by A’s right to B’s performance, which is grounded in A’s interest in it, which, again, traces back to B’s reasons for participating in the institution of promising. What I aim to highlight here is that, when applied to promises, this analysis clearly misses the important intuition that A’s right to B’s performance exists 281 Promissory obligations are transactional and voluntary, while authoritative obligations are not, and so one might object that obligations generated by authorities are not comparable to those generated by promises. However, this misses the point of the example, which is concerned not with particular promises, each of which depends on a voluntary transaction, but with the role played by promissory authorities more generally, which exists in virtue of a categorical promising rule. Being bound by promises is, in other words, not something one must or even can opt into. Promising in suitable contexts straps one with an obligation to perform whether or not one would like it to, and so in this respect the existence of the sorts of authorities created by promising is non-transactional and non- voluntary, just like the institution of authority described by the NJT. 206 independently of B’s reasons for partaking in the relevant institution. A person who has no reason to participate in the institution of promising still seems bound by its rules in virtue of the rights he generates through his actions. His reasons for partaking in it, while relevant to certain prudential questions about how he ought to act, are not relevant to the question of whether he is morally bound to act as he promises he will. However, the functionalist conception suggests that purported rights-holders have no right to the performance of people who have insufficient reason to have them perform the function they are purportedly justified in performing. For if a purported subject has insufficient reason to join up, the purported authority can have no interest in being in authority, and so has no claim. To be clear, this is not an objection to the functionalist conception per se, but to that conception as an explanation of an authority’s right to rule. It comes undone when it attempts to ground the authority’s rights in interests which are themselves grounded in subjects’ reasons, rather than the other way around. Subjects’ reasons to participate in an institution are not the sorts of things which explain the rights on which their directed obligations depend. They are explained by it. The third and last concern I will mention is that the functionalist conception appears somewhat at odds with the service conception as it has traditionally been understood. That conception involves an analysis of authority that concentrates “exclusively on a one-to-one relation between an authority and a single person subject to it.” 282 On the functionalist conception, however, an authority does not qualify individually, qua this particular authority, based on the likelihood that its directives will help its purported subject act in accordance 282 Raz (1988, 71). 207 with his independent reasons. Rather, authorities for any particular subject stand and fall together. However this seems to chafe against the traditional understanding of a particular authority’s legitimacy, which depends, in any given case, on the benefit (in terms of reason- satisfaction) her directives generate for her subjects. A consequence of the functionalist conception is that one may find oneself the subject of a fool whose claim to authority is justified in virtue of the fact that she plays a role which is generally worth having someone play. She may be a professor, a military officer, a judge, the prime minister of a nation, or anything else. So long as one would likely do better by having that sort of authority in play, the functionalist conception legitimizes her authority. This seems antithetical to the “one-to-one” relationship that is supposedly constitutive of the service conception. The concern is not that a particular authority can have many subjects, or that any person might have multiple authorities, neither of which has an effect on the justificatory mechanism. The concern is that the legitimacy of any particular authority should be based on the relationship each subject has with her, based on the quality of her directives. The relation on the service conception is not one-to-one. The incompetent professor we considered earlier typifies the problem, which we can call the problem of bad authorities. It is unclear for what reason a professor who teaches her subject poorly is a practical authority over her students, at least insofar as her authority is grounded in her students’ reasons for following her directives as a professor. (Of course, we are ignoring here any non-academic reasons they might have. That she is entertaining, for instance, is not a reason to complete the readings she assigns.) The standard understanding of the NJT has a response: at least so far as they are justified in the “normal” way, there are no bad authorities relative to a purported subject. An authority who fails her subjects on the 208 merits ceases to be an authority. This response is appropriate in light of the fact that authorities are, on the service conception, there to help their subjects. Their legitimacy is grounded in the reasons subjects have to act in particular ways, the optimal satisfaction of which they help each subject to realize individually. This response appears, however, to be unavailable to advocates of the functionalist conception, who are interested, not in the performance of individual authorities, but in the reasons subjects have to have certain sorts of authorities generally. So while the service conception is in some sense maintained, it no longer applies to services rendered by individual authorities, meaning that the caliber of any one of them as an authority is immaterial. This seems to me to give up the tight “one-to- one” connection between an authority’s directives and her subject’s reasons that made the service conception particularly appealing. 5.3.3 A Better Response to the No-Authority Objection A solution to the no-authority objection which some writers have mentioned is to deny that practical authority presupposes a right to rule. 283 If having a right to rule is not a condition of legitimacy on practical authority, then the no-authority objection is irrelevant. Many will no doubt reject this suggestion as a revisionary conception of practical authority. 284 However, as I shall argue below, there is nothing in the idea of practical authority to suggest that to be one must have a claim on others to be an authority for them, or that they must owe one obedience. This solution to the no-authority objection is, or so I shall argue, superior to the 283 See, e.g., Darwall (2009); Hershovitz (2010); Marmor (2011, 66). Darwall and Hershovitz have suggested that this may ultimately reflect Raz’s understanding of practical authority. This obviously conflicts with Raz’s own conception of his project, at least according to his most recent work. 284 E.g., Hershovitz (2010). 209 functionalist conception Raz offers in that it explains everything that needs to be explained while avoiding the additional problems that conception generates. Earlier in this chapter we said that a practical authority is an “authority affecting what is to be done.” 285 An authority “affects” what is to be done through her directives, which, we said, are authoritative if she satisfies the NJT. What are affected are a subject’s reasons and obligations. A practical authority must be capable of affecting both, but what ultimately sets her apart is her capacity to generate obligations for others to do as she says because she said it. As Raz says, what we are concerned with here is the purported authority’s “power to impose duties on others simply by expressing an intention to do so.” 286 Earlier I called this the general conception of practical authority. Does some part of our understanding of practical authority entail that such an authority must have a right to rule, or that her subjects must owe her a duty of obedience? Advocates of the no-authority objection simply assume that practical authorities must satisfy these additional conditions. Of course, Raz’s response to that objection was an attempt to make good on these assumptions, and this explains why he casts it as an objection to his view of rights (Can we understand a right such that those who satisfy the NJT have them?) rather than his view of authority (Can we understand practical authority such that authorities do not necessarily have rights?). I believe that the second question is the better one to ask. 285 Raz (1988, 29). 286 Raz (2010, 292). 210 Raz claims that, in leveling his objections, Darwall misunderstands the sort of authority he has in mind. 287 (Darwall disagrees. 288 ) Whatever is the truth about the compatibility of these two views of authority, it does not matter for the no-authority objection, for, as I pointed out already, the assumptions the objection makes regarding both rights and the owing relation Raz and Darwall share. It is with respect to these assumptions that I suggest we (and the authority theory) part ways with Raz. It is true that Raz sometimes talks about an authority’s right to rule, 289 but no other part of the view we have developed (which is largely based on his) depends on this commitment, and so it is perplexing that he tries to make good on the assumptions the no-authority objection imports into the discussion. What one is normatively capable of and what one has a right to do are two different things. Why Claim-Rights? If I am right that the general conception of authority doesn’t include claim rights, then why would one accept that a person must have a right to rule (as we’ve conceived of it) to be an authority? To answer this we can return to Darwall, the most ardent supporter of the no- authority objection. For Darwall, the notion of a right emerges as an element of what he calls the “circle of irreducibly second-personal concepts.” 290 For our purposes we do not need to understand much about the complex moral theory according to which this is true. 287 Raz (2010, 292). 288 Darwall (2010, 260). 289 Some examples: Raz (1988, 23, 27), (1990, 2-3) and (1994, 340). 290 See Darwall (2006, 11-12 and fn.24). 211 But we must understand enough of it to see that his conception is not necessarily the one that matters for our discussion. The sort of right Darwall is interested in involves, he says, “a second-personal authority to resist, complain, remonstrate, and perhaps use coercive measures of other kinds, including, perhaps, to gain compensation if the right is violated.” 291 The notion of something’s being “second-personal” plays a defining role. Second-personal reasons are those which depend on the possibility of being addressed person-to-person. For example, the fact that I will be caused unnecessary pain is a reason for you not to push me over, but because it is a reason independently of my possibly addressing it to you, it is not second- personal in Darwall’s sense. 292 That an act will cause such pain is a reason not to perform it even if such an address is impossible. However, there is something special about the relationship between you and me, given, on Darwall’s view, my dignity as a person, which gives you another sort of reason not to push me over: a claim I have on you not to push me. This reason’s validity depends on the possibility of my addressing you with it, which presupposes, Darwall argues, other relations that must hold between us—such as accountability to and authority relations—which explain the sense in which this reason is, unlike reasons of the other sort, second-personal. Your second-personal reason not to harm me is my claim on you not to, which entails both that I have the authority to address you with it— 291 Darwall (2006, 18). 292 Darwall (2006, 6 fn.9) calls these agent-neutral reasons. 212 to demand of you that you not treat me in certain ways—as well as that you are accountable to me. Each relation can, on Darwall’s view, only be understood in terms of the others. 293 Now, what is important to see here is that the sort of right Darwall has in mind—a claim-right—goes well beyond what is needed to require action in a manner consistent with the general conception of practical authority. That conception picks out a normative power or power-right rather than a claim. A power-right is a capacity to modify the normative circumstances of another in the way we discussed at the beginning of this chapter. (In Hohfeldian terms, a person has a power just in case she can modify another person’s “Hohfeldian incidents” within some set of rules.) Raz more narrowly describes the sort of power involved in practical authority as “the power to impose duties on others simply by expressing an intention to do so.” 294 This way of putting things is important, for it indicates that not any sort of power-right will do. For instance, claim-rights always give their holders a narrow normative power as well. Take again the claim I have on you not to push me over. I can release you from your obligation at any time, and since by doing so I would thereby change your normative circumstances in the relevant way (I would modify one of your Hohfeldian incidents—namely, your obligation), I also have a normative power. The difference between this more limited normative power and the sort of normative power we are interested in is that a person who has a normative power in virtue of having a claim-right cannot impose duties on others by expressing an intention to do so. I can demand of you that you not push me over, thereby pointing out to you that you are under an obligation to me 293 For a more comprehensive explanation, see Darwall (2006, Chapter 1). 294 Raz (2010, 292). 213 not to, or I can waive that right and eliminate your obligation. 295 But neither of these amounts to imposing an obligation through a directive, for the obligation is already in place. Whether I actually demand it of you or not, you have an obligation not to push me over. The most I can do with my directive is waive my claim and eliminate your obligation. 296 So while Darwall focuses on claim-rights when he presents the no-authority objection—a condition which, I’ve claimed, Raz is wrong to accept—we need not. Not only are these rights distinct from one another, they do not even overlap in many cases. The sort of power that is constitutive of practical authority neither entails nor is entailed by the limited power to remove obligations that one has whenever one has a claim-right. The Standing to Demand One might object that, even if practical authority amounts to normative power, one must still have a claim on others to exercise it. (Something like this would explain why Raz, in spite of his apparent endorsement of the normative power picture of practical authority, accepts conditions (i) and (ii) put forward by the no-authority objection, which deals in claim-rights.) Indeed, this may be Darwall’s concern. Darwall likens a practical authority’s right to direct a subject to φ to the right you have to demand that I get off your foot if mine 295 This is not to say, of course, that all rights can be waived. You may not be able to waive your right not to be tortured, for instance. But some rights can be waived, and my claim on you not to push me over is surely one of them. 296 One might suggest that, having waived my claim for some indeterminate duration, I may reinstitute it at a later time, thereby imposing an obligation on you which, at least quite recently, you did not have. The problem with this suggestion is that the obligations which existed before I waived them exhausted the obligations I could later “impose.” 214 is on top of it. 297 You (like all other members of the moral community) are in a position, or have the standing, as Darwall says, to demand that I remove my foot from yours. Is there something special about a person’s moral status which entails that they have this sort of normative standing? We make demands of people all the time. Some have normative force, others don’t. I might demand of my dean that he assign me the best office in the philosophy building, just as I might demand of my student that he turn his paper in on time. The fact that only the second demand has normative force does not mean that the first is not a genuine demand. 298 It is a demand—only one I’d be foolish to make. Orders, commands, directives, and so on, function similarly. What is at stake in the debate about the nature of practical authority is not the range of speech acts one is capable of executing, but the normative force those speech acts carry with them. To obligate others through my directives, I must be an authority. To see this more clearly it might help to compare obligations with reasons. Everyone will agree that I need no special normative standing to have my utterances count as reasons. When my dentist tells me I have a cavity that needs to be filled soon, I take that to be a reason to believe that I have a cavity and that it needs to be filled soon. Her capacity to create these reasons for me requires no special standing on her part. 299 One might suggest that she has such a standing because I have hired her to make this sort of assessment. But suppose that I did not hire her. Suppose that I was napping on the train, and have a bad 297 Darwall (2006), (2009), and (2010). 298 I argued for this in Chapter 1. 299 The fact that the reasons we have in the case of the dentist are reasons for belief is immaterial. The question is not about the difference between theoretical and practical reasons, but about reasons and obligations. 215 habit of sleeping with my mouth open. A woman in my car noticed the obvious (to her) rot on my tooth and let me know when I woke that she is a dentist, and that I have a cavity that should be filled soon. Surely she has no more standing than you—a philosopher also sitting in my car—to tell me as much. The difference is that her statement gives me a reason for belief, while, other things equal, yours probably does not. This has nothing to do with her normative standing, as Darwall understands that term. She has no right that her directives constitute reasons. Whether they do has to do with other facts—her education, her knowledge about my personal dental situation, and so on. The same applies to obligations. If anyone can issue directives, and no standing is necessary for one’s directives to create reasons, then why would someone need a standing— a right or whatever—to have her directives create obligations? One might suggest that obligations generated by authorities’ directives are unique in that you owe them to the authority, and that this would be impossible to explain unless the authority had a claim on your obedience. But this objection begs the question. To bring the point home, consider a case like voluntary arbitration. To settle a disagreement between them, B and C contract with one another to obey A, who thereafter tells them both to φ. Supposing that neither B nor C had an obligation to φ already, they both have this obligation for the first time now. They now owe it to one another to φ as the result of A’s directive. That is, A imposed an obligation on each of them simply by expressing an intention to do so. A has that normative power in virtue of the agreement B and C made. Darwall would claim that the real practical authorities here are B and C, and that they are practical authorities for one another. But this position is, or so I claim, based on an interpretation of practical authority which ignores the unique role authorities play in our 216 practical reasoning. To be a practical authority is to have the capacity to affect what others are to do simply in virtue of your intention that they do or abstain from doing it. From this perspective it seems plainly false that, of the three people who might be a practical authority in our scenario, A is not the leading contender. A fact that perhaps confuses things is that B and C are right-holders in that each is accountable to the other to do as A says, such that each will be wronged by the other’s nonperformance. (More about this in the next chapter.) But, with the exception of the limited power B and C have to waive their right to the other’s performance, they are essentially powerless. Because neither B nor C has the capacity to generate new obligations for the other through their directives, neither can affect what the other is to do in the relevant sense. For these reasons it is much more plausible that A is the practical authority. A is, after all, the one who has the normative capacity to generate new obligations for the others, to direct what they are to do through her directives. A is also the one we would most intuitively identify as the authority in the situation, not, again, because she has a right to direct B and C—I could just as well try to direct them—but because she is positioned, given the way things have gone in the world, to create obligations for them in the right way. Again, this does not entail that A has a right to rule, where that right constitutes something beyond a normative power. And it certainly does not mean that she has the power to punish. What she has is the power to affect the normative situation of B and C in the manner we have described throughout this chapter. This is the feature she shares with the wide variety of practical authorities we find in the world. 217 A Brief Disclaimer To be clear, I do not claim that there is not a technical use of ‘practical authority’ in the philosophical literature which is concerned with the right to rule, where that includes something beyond a normative power. Nor do I think that a right (a Darwallian standing, an entitlement) is never sufficient to ground one’s practical authority, or that many practical authorities do not also have a claim on their subjects to obey. If A’s normative power results from a contract B and C made with her, rather than with one another, then she has a claim on each of them to do as she says. In this case her practical authority is buttressed by her right to rule (with respect to the actions specified in the contract). The important point is that practical authority, as we have understood it here, and as Raz and, I believe, most people understand it, is consistent with the general conception as I explained it above. It amounts to the power to generate obligations for others simply by directing them in certain ways. 300 This does not presuppose a right to direct them, or a right that others not interfere with one’s attempts to issue such directives, or a right that one’s directives generate obligations for them. The concept, most fundamentally, concerns the special practical reasons one individual can create for another. 5.3.4 Conclusion Let me briefly summarize what we’ve done so far in this section. The no-authority objection claims that the NJT cannot explain political obligation because one must have practical authority to obligate others through their directives, and satisfying the NJT cannot generate 300 Here I am denying Hershovitz’ claim that “authority is not a feature of the world apart from our practices” (2010, 44). As I have presented it, practical authority is precisely that. 218 this sort of authority. To have practical authority over others, the objection goes, one must meet two conditions: first, one must have a right to direct them, and second, they must be accountable to one to do as one directs. In an attempt to explain how the NJT meets these conditions, Raz argued for what I labeled the functionalist conception of the NJT, an interpretation according to which a purported authority’s legitimacy is grounded in subjects’ reasons to have some person perform the more general function the authority in question performs. After discussing some apparent problems with this conception, I argued that a better approach would be to reject the two conditions Raz attempted to meet. I also argued that such a conception is not revisionary, but rather that it is completely in line with our conception of practical authorities as persons who affect what we are to do, and that in this respect it is consistent with the spirit of the service conception of authority Raz is ultimately most interested in defending. A final point I would like to make before moving on is that, even if the authority theory overcomes the no-authority objection in the way I have described, its success is not the threat to Darwall’s view that he takes it to be. “[I]f the reasons with which an agent would better comply were not themselves second-personal reasons,” he says, “it would then follow that someone could acquire practical authority over her owing entirely to non-second- personal considerations.” In other words, the success of the authority theory would, Darwall thinks, cast doubt on the “circle of irreducibly” in which his broader ethical theory is grounded. 301 When understood correctly, however, denying that practical authorities must have a right to rule is consistent with Darwall’s view. Regarding the arbitration example, for 301 See Darwall (2006, 11) and (2010, 258). 219 instance, I take it that what is important to Darwall is that B and C—those he would identify as the true practical authorities—satisfy the two conditions we discussed. The theory I have defended does not deny this. 302 As right-holders, B and C have the standing to demand (in Darwall’s sense) that some claim be met. However the normative capacities of genuine practical authorities are irrelevant to this picture. If Darwall would like to call B and C “practical authorities,” he has every right to. My claim is merely that, with respect to the role authorities play in our practical reasoning, which is how we ought to understand them, A is the practical authority in this scenario. It is this sort of practical authority that will be relevant to the positive view I put forward in the next chapter. 5.3.5 The No-Obligation Objection: The First Strand To represent a challenge to the authority theory, the no-authority objection assumed that a person cannot obligate others through her directives unless she has authority. This assumption is true on the view I’ve put forward, but trivially, for on that view one’s having practical authority just is one’s having these normative capacities. But even if I am wrong about this, it is important to keep in mind that we are interested, ultimately, in political obligations, and so the conditions one must satisfy to be a legitimate authority are relevant 302 This interpretation also answers an objection raised by Hershovitz (2010, 26-29), in which he points out that Raz and Darwall are possibly “talking past one another” in much the way I have suggested Darwall and I are here. “Raz ultimately construes the right to rule as ‘the power to issue obligation—or duty—imposing directives,’ not as a claim on subjects,” he says, quoting an unpublished paper by Stephen Perry (see Perry (unpublished, 5)). Hershovitz says (27): “the problem comes in thinking that an account of a moral power to create preemptive reasons amounts to an account of authority.” “That I have a moral power to change your reasons does not,” he says, “warrant my making any demands on you, nor give you any obligation to me.” It should be clear at this point why this objection is irrelevant to the interpretation of the authority theory I have suggested. 220 only if we accept that one must be an authority to have normative powers. Denying this additional premise frees the theory from the objection. The no-obligation objection is not susceptible to this sort of response, for it maintains that, even if authority is not a necessary condition for obligation, satisfying the NJT is not a sufficient one. The objection is divided into two strands. The first strand, which I discuss below, claims that one cannot generate genuine preemptive reasons for those with respect to whom one satisfies the NJT. Because preemptive reasons are necessary for obligations, it follows from this objection that the NJT cannot explain obligations either. Because this objection originated with Darwall, I will consider his two versions of the argument. The first contends that a person’s satisfying the NJT cannot generate preemptive reasons for others because nothing secures her standing to hold them accountable, which is, he holds, presupposed by normative capacities as much as it is by authority. This is the no-accountability version of the argument. The second version of the argument sets the issue of accountability aside and claims simply that one’s satisfying the NJT is insufficient to explain her capacity to generate preemptive reasons for others. This is the insufficiency version of the argument. I argue below that both versions of this first strand of the no-obligation objection fail. The second and I think more important strand of this objection accepts that one’s satisfying the NJT explains her ability to generate preemptive reasons, but claims that these do not amount to obligations of the sort that concern us. This objection also includes two approaches, both of which I address in the next chapter. As we’ll see, overcoming these will require substantially modifying the authority theory away from the largely Razian conception we’ve developed so far. 221 No-Accountability Darwall packages the first argument in a slogan: “No preemptive reasons without the standing to hold accountable.” 303 We saw earlier that Raz is skeptical of the notion of accountability Darwall endorses. What I said was that, even if Raz is right to be skeptical, the more intuitive and perhaps simpler relation of owing is plausibly part of the accountability relation, and that the same objection can be leveled using it. The same applies to preemptive reasons. If I cannot generate preemptive reasons for others without those others owing it to me to satisfy them, then this is indeed a crushing objection, for the NJT cannot explain why subjects owe it to authorities to act in accordance with the preemptive reasons their directives constitute. The question, however, is precisely whether preemptive reasons really do presuppose such an owing. Can a person generate preemptive reasons for people who owe her nothing? I argued already that practical authority does not entail an owing relation. Because practical authority includes the capacity to generate preemptive reasons, the same argument shows why these capacities do not entail the owing relation either. In making that argument I cited a case of arbitration in which B and C agree with one another to do as A says. Each new directive A issues will put B and C under a new obligation to perform whatever act A specifies. If A says, “B, you will allow C to use the family cabin for half of the year, and have a key made for him; and C, you will return the minivan to B,” both B and C will have an obligation they may not have had before. Each obligation is new—making keys and returning minivans are not things B and C were obligated to do before A issued her directives. The 303 Darwall (2010, 261). 222 important point about the example, however, is that in this case B and C do not owe it to A to do as she says. She has no claim that they do it, even though the circumstances position her to generate new obligations for them through her directives. B’s and C’s obligations are directed, but they are directed to one another, not to A. 304 One might object that, if it is indeed true that in this case B and C owe nothing to A, then she must also lack the capacity to generate new obligations for them. However, given the facts of the case it seems that only an unacceptably theory laden interpretation could describe things in such a way that the requirements with which B and C are strapped are not obligations, or not new, or not attributable to A’s directives. If this is right, then, because obligations are composed of preemptive reasons, it follows that preemptive reasons do not entail an owing relation either. Darwall’s slogan—properly explicated, and reduced to the owing relation—is the universally quantified proposition: “for any preemptive reason generated by a directive, the person to whom the reason applies owes it to the directive’s issuer to satisfy it.” Our example therefore casts doubt on it. Some preemptive reasons are generated by people who are owed nothing. Insufficiency The no-accountability objection is not the end of the no-obligation argument, for one might suggest, as Darwall does, that, accountability (and owing) aside, a person satisfying the NJT simply cannot generate preemptive reasons. This is the second objection in this strand of 304 If this is not clear enough from the example, we can modify the explicit agreement between B and C: while they agree to do what A says, they do not agree to give her any new claim-rights with respect to either of them. It should be clear with such a transparent condition that, while B and C are obligated to one another to do as A says, A has no claim on them to do so, nor to punish them for violating her orders. She is merely there to direct. 223 argument. 305 The claim is that, just because B has reason to treat the directives of an alleged authority as preemptive—because doing so will likely cause him to act in accordance with his independent reasons better than he would were he to attempt to do so directly—that does not mean that those directives are actually preemptive reasons for him. 306 The NJT, Darwall claims, is plagued by the “wrong kind of reason” problem. The problem is familiar enough. That I would get a large research grant if I believed that some squares are round would be, in some sense, a good reason to (at least try to) form that belief. But reasons of this sort do not indicate whether the belief is likely to be true, which is what such reasons do. Darwall suggests that preemptive reasons face the same problem: “Why should the fact that one has reason to regard or treat someone’s directives as creating exclusionary reasons make it the case that their directives actually do create such exclusionary reasons?” 307 Suppose, Darwall says, that B has conclusive reason to get up daily at 7am, but has trouble actually getting up at that time. Being an authority-fearing man, he gets an “authority alarm clock”—an alarm clock with an authoritative voice that says “Get out of bed now!” each morning at 7am. The alarm clock turns out to be effective. Indeed, it is the only means B has found by which he is consistently moved to get out of bed at 7am. That the “agent” issuing the “directives” is in this case a clock may complicate the objection’s application to the NJT. Thus, to avoid any trouble this might cause, let us consider the second case Darwall gives, which is in all other respects identical to the first. 305 See Darwall (2010, 268-78). 306 Darwall (2010, 269). 307 Darwall (2010, 270). 224 Rather than a clock, we imagine that B hires an actual person, A, to order him to get up in the morning. Other than that “A deliberatively rehearses the reasons for and against directing B to get out of bed,” and A’s being “both wise and concerned for B’s welfare,” A does essentially what the alarm clock did before: he orders B to get up at 7am. Call this the alarm-clock case. Darwall’s central claim seems to be this. Though B has most reason to get up at 7am, and so has good reason to treat A’s directives as giving him preemptive reasons if this is what it takes to get up at that time, this does not make them actually preemptive. As Darwall puts the point, (III) does not follow from (II), even if (II) follows from (I): 308 (I) B will do better in complying with reasons if he treats A’s directives as giving him preemptive reasons. (II) There is reason for B to treat A’s directives as giving him preemptive reasons. (III) A’s directives actually do give B preemptive reasons. For this entailment to hold, Darwall says, “the reasons that speak in favor of B’s believing that A’s directives create exclusionary reasons, or B’s regarding or treating A’s directives as creating them, would also have somehow to make B’s belief or way of seeing or treating things true or correct.” 309 308 Darwall (2010, 269). 309 Darwall (2010, 269-70). 225 That B may have one sort of reason to treat A’s directive as another sort of reason makes this case somewhat complex, and so to see the force of the objection it may be helpful to consider its application to a case that does not involve preemptive reasons. Imagine that B has most reason to get to Hollywood as soon as possible. The three claims are: (I*) B will get to Hollywood sooner if he treats A’s directions as the correct directions and tries to follow them. (II*) There is reason for B to treat A’s directions as the correct directions. (III*) A’s directions actually are the correct directions. Here Darwall’s claim is clearly correct. For (III*) to follow from (II*), it would have to be the case that the reasons B has to treat A’s directions as being correct also somehow make B’s way of treating them the right way. Clearly this fails in some cases. Sometimes the reasons we have to believe something, or to treat it in a particular way, are not, as Darwall observes, “fitting.” If someone offers B an ice cream cone on the condition that B treat A’s directions as being correct, then B has a reason to (at least try to) treat them this way. In this case (II*) is satisfied. But it does not follow from this that (III*) is also satisfied, for the reason B has to treat A’s directions as being correct has no bearing on whether they are. I agree with Darwall’s point here. Nonetheless, the objection fails because the Razian picture is not committed to the entailment Darwall suggests in the first place. According to Darwall, Raz’s view is that (I) entails (II) entails (III). But in fact the claim expressed by the NJT is closer to the content of (I) than (II); and while it is true that (II) follows from (I), 226 which Darwall accepts for argument, nowhere does Raz suggest that (III) then follows from (II). I take this to be the force of Raz’s laconic response to Darwall’s objection in his most recent work: “[It] is not that one has those derivative reasons stated in (III) because it would be good to believe in them,” he observes, “It is that it is good to believe in them because they are there.” 310 In other words, Darwall has things reversed: (II) and (II*) follow from (III) and (III*), not the other way around. And this seems exactly right. Some Implications of the Alarm-Clock Case Although Darwall claims to introduce the alarm-clock case to illustrate the objection we just answered, I think it does more than that. The objection claimed that the Razian picture endorses an entailment between (II) and (III), when in fact it does not. But surely, if Raz is correct, there must be some reason B has which satisfies (II) and simultaneously entails (III). That is, for any case in which a person A satisfies the NJT with respect to another person B, there must be some fact which provides a reason for B to treat A’s directives as preemptive reasons while at the same time showing that they actually are. I think that the alarm-clock case is best understood as an attempt to show that no such fact exists. Here is how Darwall sets things up: We can imagine that he [B] hires the very person (call him A) who played the role of the authoritative experimenter in the Milgram experiment. Assume, moreover, that before he issues his directives to B, A deliberatively rehearses the reasons for and against directing B to get out of bed; as it happens A is both wise and concerned for B’s welfare. This finally gives B a permanent fix to his problem; it provides B precisely the “authority experience” he needs to get him reliably out of bed at 7am. When B is tempted to continue lying in bed, A speaks to B in an 310 Raz (2010, 299). 227 authoritative voice, “reminds” him of his (A’s) (putative) authority and that B may not even think about remaining in bed in a sufficiently persuasive way that B “obeys.” 311 This case can be interpreted in at least three ways. In the first, A actually has more information than B does about B’s situation, which might be true, say, if A were B’s high- priced personal assistant—someone whose job involved having more knowledge about and control over B’s schedule than B did. In such a case B would rely on A to keep him on schedule, which would include counting on A to wake him up at the appropriate time each morning. Call this case (U1). On the second possible interpretation, (U2), A knows no more than B does about what reasons there are for B to get up in the morning, but waking up early causes B to reason poorly. For B, reasoning about getting up while lying in bed is like reasoning about whether he should call his ex-wife while drinking whiskey—it’s unreliable. 312 Here the problem is presumably not that B fails to identify the relevant reasons; it is that in the morning he has difficulty evaluating them appropriately. The third and final way to understand the case, (U3), involves a purely motivational deficiency. While lying in bed each morning, B sees that he has most reason to get up, but simply cannot motivate himself. As in the second version of the case, he does not persuade himself that he has most reason to stay in bed, for he knows very well he should get up. He just has trouble doing it. In this case A’s function is purely motivational. 311 Darwall (2010, 273). 312 Darwall’s description of the case at times suggests this reading. After waking up, he says, B “deliberates about whether to get up and invariably persuades himself (against, let us suppose, his better judgment) that he has more reason to stay in bed a little longer” (2010, 270-71). 228 (U3) comes closest to providing an objection to Raz. But before we assess any of these cases, we should try to better understand the concern to which this example points. I believe Darwall is worried about the idea of preemption itself: What is it for one reason to preempt another? The key to understanding his concern comes in this passage: Granted, B has pragmatic reasons to respond to the alarm [who is in this case, A], on the one hand, and, on the other, not even to think about his reasons for staying in bed (since if he does consider them he is likely to act contrary to the weightier reasons and stay in bed). But this doesn’t mean that he may not legitimately think about these reasons, that it is outside his discretion, or that the latter reasons have somehow been displaced, preempted, or defeated, only that he would be foolish to consider them. 313 To have a preemptive reason is, Darwall points out, for certain of one’s reasons to be somehow “off limits”—it is for a class of reasons to be “illegitimate” for one to entertain or act on. This clarifies what he aims to accomplish with the alarm-clock case, for Darwall denies that any reasons are illegitimate for B in this way. His point, then, is that because A presumably satisfies the NJT, we have a case in which the NJT fails to explain the capacity of a purported authority to create preemptive reasons. A cannot make any of B’s reasons illegitimate. I am sympathetic to Darwall’s point here. The argument for preemptive reasons indeed entails that an authority has the capacity to make certain reasons, in some sense, “illegitimate” for subjects to entertain or act on. It will become clear in the next chapter just how central this idea of illegitimacy is to the concept of obligation, and how we can use it to overcome the most challenging objections to the authority theory. For now, however, it will be enough to say just a few things about this particular case. 313 Darwall (2010, 271-72; my emphasis). 229 Because B would, and knows he would, likely satisfy his independent reasons better by following A’s directives, the reasons excluded by those directives are “off limits” for B in that B would be irrational to entertain or act on them. In (U1) this is explained by the fact that A, who controls B’s scheduling book, has superior access to the reasons there are for B to get up at 7am than B does. Were B to simply get up when he felt like it—as he presumably did before hiring A—he would fail to act in accordance with his independent reasons and so would be worse off in this respect. He would act other than he has reason to believe he has most reason to act. (U2) is a subtler case, but it can be explained in the same way. Here the problem is not that B lacks information about his reasons, but that he weighs those reasons unreliably in the morning. In this case A is ultimately no different than a rule B has accepted and acts on successfully; he is there to reduce the errors B makes in his practical reasoning. In Chapter 4 I explained why a rule provides preemptive reasons for those who accept it, and I will not repeat those arguments here. It should be clear, however, that were B to violate the very rule he created and believes (correctly) to be helping him, he would act irrationally. The reasons on which it would be illegitimate (because irrational) for B to act are those which A ought to take into account in determining the time B is to wake up. This explains in what sense A’s directive is exclusionary. The point of having A around is to take a class of reasons that A would normally have to consider on his own and to put their consideration in the hands of someone better suited for it, whether because B is unavoidably ignorant of those reasons, irresponsible about finding them out, unwilling to spend the time 230 to treat them appropriately, or because he is incapable of reliably entertaining them. 314 Because it would be impossible for A to realize the benefits it is B’s job to provide if he insists on taking into account the reasons B was hired to take into account for A, it would be irrational, and in that sense illegitimate, for A to do so. The sense in which these reasons are illegitimate to consider is the sense in which the directive which preempts them is exclusionary. So in (U1) and (U2), the reason B has to treat A’s directive as a preemptive reason is that it is a preemptive reason—not the other way around. It is preemptive because it excludes and replaces a class of independent reasons B has in the circumstances, reasons B should not entertain directly. This explains the sense in which it is right to say that B “may not legitimately think about these reasons, that it is outside his discretion, or that the latter [independent] reasons have somehow been displaced, preempted, or defeated.” Behaving in these ways would be outside his discretion in that it would be irrational. 315 Of course, B might not mind being irrational; but this is beside the point. The sort of illegitimacy that characterizes the preemptive nature of A’s directives has to do with the reasons B has to act, and this sort of illegitimacy exists in both (U1) and (U2) even if A does not mind acting illegitimately in this way. I call this sort of illegitimacy a cost, and say more about it and its relation to requirements and obligations in the next chapter. 314 These reasons for having A around include elements from each of what Raz (1988, 75) calls “the five most common reasons capable of establishing the legitimacy of an authority.” 315 B could nullify A’s directives by taking the time to stay informed, improving his early-morning reasoning capacities, or developing an indifference to the consequences of not getting up at 7am (supposing that his failure to do so would not violate any obligations he has to others). But beyond these alternatives, it clearly is outside his discretion whether he may (in the appropriate sense) act on the reasons excluded by A’s directives. 231 I should note that Darwall says at one point that “even if there is an exclusionary reason for B in this case, it is not the kind required to partially constitute a preemptive reason of the sort that genuinely authoritative directives create.” 316 Here Darwall presumably means to reject the sort of illegitimacy we have discussed as the wrong sort for preemptive reasons. The right sort of illegitimacy would, for Darwall, turn on accountability. For him, you act illegitimately when you violate a valid claim another person has on you. However this simply begs the question. The whole point of the insufficiency version of the no-obligation objection is to argue that, whether or not preemptive reasons presuppose accountability, a person satisfying the NJT cannot generate such reasons. It cannot therefore be claimed that these reasons cannot be preemptive because there is nobody to whom one is accountable if one ignores them. Moving on to (U3), one can see that it is importantly different than (U1) and (U2) in that the considerations which explain the sense in which A’s directives exclude and replace B’s reasons in (U1) and (U2)—such that B would be irrational to entertain and act on those reasons directly—no longer apply. For instance, we can assume in (U3) that A and B always consider the same reasons, weigh them equally, and come to the same conclusion about what B ought to do. In this case A’s function is purely motivational. He is like a drill instructor at basic training whose role it is to yell at cadets to do what they are already trying to do so that they will have the adrenaline to do it better. On this interpretation, unlike the other two, Darwall is right to say that “B has the same very weighty reasons for getting out of bed at 7am, and so for responding to the directives, and the same less weighty reasons for 316 Darwall (2010, 272; my emphasis). 232 continuing to lie in bed.” 317 It is therefore difficult to see in this case how A affects the reasons B has. (U3) is unique in that B recognizes that A knows no better than he does what he ought to do. In fact, the case would be no different if B told A what he (B) ought to do, and then A simply parroted it back to him in an authoritative (and thus, for B, motivating) tone. In this case, then, B sees that A’s directive cannot be authoritative, for that would require that it count in favor of acting. However nobody, including A and B, take A’s directives to provide reasons for action. A is there simply to motivate. This is the key point. Recall that A’s satisfaction of the NJT partly requires that B act in accordance with his reasons better by accepting the directives of the alleged authority as authoritatively binding. To meet this requirement in the circumstances under consideration, B would have to be irrational, for he would have to see that A is not there to provide (normative) reasons for action, and that he does not provide them, while accepting A’s directives as such reasons. This odd feature makes one wonder whether this case is useful to consider. But let us press the issue anyway. Supposing that B does accept A’s directives as authoritative, and that by accepting and trying to act on them he will conform with his reasons better than he would otherwise, does that make A’s directives preemptive for B? In this case it appears, in fact, that it would. For were B to accept A’s directives as genuine normative reasons, it would be illegitimate for B to consider the reasons A’s directives are meant to preempt. This would be irrational for the very same reasons it would have been in (U1) or (U2), and so just as B’s independent reasons were excluded and replaced in those cases, they are excluded and 317 Darwall (2010, 273). 233 replaced in (U3). The fact that B would have to be irrational already to be in a position in which he could accept B’s directives as authoritative, although a strange idea to accept, even for argument, may not affect the point at hand. For were B to be in such a position, the fact that his acceptance of and attempt to act on A’s directives would cause him to better conform would qualify A’s directives as preemptive reasons. To accept them is to treat them as taking account of independent reasons one has to act, and so excluding those from the range of reasons one can rationally entertain. Of course, in most actual cases this will be impossible for B to do. (Telling someone what to direct you to do makes it nearly impossible to look at his directives as providing useful practical guidance.) Were he to overcome this near impossibility, however, B would have to accept the exclusionary functions those reasons perform, and in doing so he would place rational constraints on the reasons he could entertain in his practical deliberation. It seems, then, that none of (U1) – (U3) presents a counterexample to the claim that a person who satisfies the NJT can thereby create preemptive reasons. Because these are the three most plausible interpretations of the alarm-clock case, we can dispense with it altogether. And we can dispense with the two arguments forming the Darwallian strand of the no-obligation objection, the first of which failed because there are clear cases in which an individual can generate preemptive reasons for people who do not at the same time owe it to her to act on them, and the second of which failed because it depends on an entailment relation neither Raz nor the authority theory endorses or needs to endorse. 234 5.4 Conclusion The goal of this chapter was to introduce a framework for the authority theory of political obligation, which we’ll begin to develop in earnest in the next chapter. I introduced this framework by dividing its presentation into two discussions: the first regarding the justification of authority, which has primarily to do with Raz’s normal justification thesis, and the second regarding normative capacities. I then considered the no-authority and the no-obligation objections, and developed two lines of argument in response. In the first I argued that Raz’s solutions are able to withstand the strand of the no-obligation objection that Darwall proposed, but I suggested that the no-authority objection, which attacked Raz’s conception of authority, forced important modifications to his view. Whereas Raz embraced the assumption that one must have a right to rule to be an authority and attempted to show how his theory explains it, I argued that a better solution is to reject the supposition that practical authorities need a right to rule. This conception of practical authorities, not as people who necessarily provide a function which entitles them to direct others, but as people with normative capacities which play an important role in the practical reasoning of their subjects, is, or so I argued, not a revisionist conception of practical authority but rather one that captures the essence of those with this role. 235 Chapter 6: Obligation and Justification Redux In the last chapter we considered a variety of objections to the Razian conception of authority. Some we were able to refute without making changes to the theory’s basic structure. Others forced us to develop it, and in doing so we began to move away from Raz’s conception in important ways. This chapter marks a transition in the argument: from defending Raz’s view in ways that he might have to developing useful concepts that go beyond his theory in significant ways. In what follows we’ll consider a final objection that will require us to revise the conceptions of obligation and justification we developed in the last two chapters. The conceptions with which we’re left will be critical in the arguments I present in the next and final chapter. 6.1 The No-Obligation Objection: The Second Strand Recall that, according to the authority theory, an individual is a practical authority for those with respect to whom she satisfies the NJT, and that this in turn entails that she has the normative capacity to generate obligations for them. In the last chapter I responded to three important objections to this view: first, that satisfying the NJT is in fact insufficient to 236 ground her authority; second, that authority aside, she could not have these capacities without her subjects owing their actions to her; and third, that this sort of accountability aside, she would not be capable of generating preemptive reasons for anyone anyway. Something to notice about these objections is that each of them is foremost concerned with the authority herself—with whether she is a genuine authority, with what she is owed given her authoritative standing relative to those she purports to bind, and with whether that standing is sufficient to justify the claim that her directives constitute preemptive reasons for those she directs. But there is another sort of concern you might have about the authority theory, one involving, not the purported authority, but the deontic status of the reasons she creates through her directives. You may concede, that is, that a person satisfying the NJT can generate precisely the sort of preemptive reasons I’ve argued they can, while still wondering whether these reasons amount to obligations. There are in fact two related problems here. To separate them, it will be useful to have a few of the conceptions we’d discussed earlier in front of us. First, the theory of obligation we were left with at the end of Chapter 4, which we called the narrow concept of obligation: An action is obligatory if it is required by a categorical rule, i.e. a mandatory rule which applies to its subjects not merely because adherence to it facilitates achievement of their goals. 318 318 Raz (1977, 223). 237 Second, the NJT, as we stated it in Chapter 5: The normal way to establish that a person has authority over another person involves showing that the alleged subject is likely better to comply with reasons which apply to him (other than the alleged authoritative directives) if he accepts the directives of the alleged authority as authoritatively binding and tries to follow them, rather than by trying to follow the reasons which apply to him directly. 319 Finally, the general conception of practical authority, which we also discussed in Chapter 5: A practical authority is someone capable of affecting other’s Hohfeldian incidents through her directives—more specifically, someone with the normative power to impose obligations on others simply by expressing an intention to do so. 320 6.1.2 Getting Obligations Out There is an apparent inconsistency between the NJT, which refers only to reasons, and the narrow concept of obligations, which refers to reasons of a specific sort. How does the normative force of mere reasons transform into this special sort of categorical force? Am I an authority for you—someone who can generate obligations for you to act as I’ve said— simply because by accepting my directives as authoritatively binding and trying to following them you will act in accordance with your prudential reasons better than you would have otherwise? If so, then the Razian conception of authority seems much too inclusive. For it is 319 Raz (1988, 53). As I mentioned in Chapter 5, Raz uses the term ‘comply’ here when he should, based on his own understanding of that term in Raz (1999a, postscript), really say ‘conform.’ 320 See Raz (2010, 292). 238 difficult to see how the fact that A satisfies the NJT relative to B allows A to generate, not just preemptive reasons, but categorical ones. Let’s begin with a well-known example. If my interest in making money gives me conclusive reason to maximize profit on my investments, then the fact that my financial advisor knows much better than I do how I should invest my money to accomplish my goal positions her to issue directives which, for reasons we discussed in the previous chapter, will be preemptive for me. Yet these reasons are not categorical, for they do not apply to me independently of my goals. But then they do not amount to obligations even on Raz’s own conception, and so one’s ability to generate them will not amount to the sort of normative power which characterizes practical authority on the general conception. To modify a pithy complaint in the literature: If categorical reasons are not going in, it’s hard to see how we can get categorical reasons out. 321 In response to this criticism, a handful of writers have suggested (but ultimately deny) that the NJT might account for these cases by restricting itself to a specific domain of reasons in the first place. Along these lines, we might stipulate that to be a practical authority for B, A must help B act in accordance, not just with his reasons, but with his categorical reasons. The idea is that, if we guarantee that the reasons with respect to which the authority directs her subjects are categorical, then it is less of a mystery how these directives themselves might apply independently of the subject’s aims and interests. We simply import the appropriate sort of normativity right from the start. 322 321 This is modified from Hershovitz (2010). 322 See Darwall (2009) and (2010); Hershovitz (2010); Marmor (2010). The authors who discuss this solution in fact say that the class should be restricted to obligations, not categorical reasons. My 239 I find this solution to the obligation-out problem compelling. If my independent reasons are categorical already, and you help me act in accordance with them, then why wouldn’t that categorical force transfer to your directives? If my reason to accomplish something does not depend on my aims and interests, then why would my aims and interests become relevant when you provide your service as an authority? It seems to me that there is no fact which would explain this sort of change. Still, the problem as we’ve stated it does not go far enough. For what seems objectionable is not just that the NJT cannot import the sort of normativity the Razian conception of obligation demands. It is also that that conception demands the wrong sort of normativity. The obligation-out problem turns on the important observation that obligations are unlike prudential reasons partly in that we cannot escape them by changing our preferences. If I have most reason to take a job in Los Angeles because that is where I’d most like to live, and I’m offered one, then there is normative pressure on me to take it. However this pressure is responsive to my changing preferences: if my feelings for Los Angeles change, so too will my reasons to take a job there. What we’ve emphasized so far is that obligatory pressure is somewhat unique in that it doesn’t respond to our preferences in this way. But while the notion of a categorical rule accounts for this important feature of obligations, it is too coarse grained to account for obligations themselves. Some rules which seem to apply to us independently of our goals—rules of rationality or rules of etiquette, for example—clearly don’t constitute obligations. What is most distinctive and interesting about obligations, it seems, is not that they are categorical in nature, but why they are. conception of the problem reflects the two separate concerns I believe are often run together in the no-obligation objection. 240 Obligations, at least of the sort that we are concerned about when we discuss political obligations, are moral relations. This explains why they apply, like other categorical rules, irrespective of our desires and goals, and it explains why we have a special interest in the sort of pressure they put on our practical deliberations. Consider, by contrast, the rules of rationality. You may not care whether you violate them, but your indifference in this respect has no bearing on their applicability to you. Forming a strong preference for believing both P and not-P will not bring your beliefs into compliance with the rules in the way that modifying your preferences about your living conditions will affect the reasons you have to take a job in Los Angeles. The rules of rationality are categorical in precisely the same way that obligations are. This is why the proposed response to the concern above does not go far enough. For while importing the right sort of normativity may suffice to ensure that the directive issued by a person satisfying the NJT, understood to include only categorical reasons, is itself categorical, it does not ensure that the directive will be categorical in the right way. The concept of an obligation, at least of a moral obligation, seems not to be fully captured by that of a preemptive categorical reason. The problem can be expressed in two ways, each of which presents it in a slightly different light. The first is to say that, when combined with the general conception of practical authority, the Razian conception of obligation makes practical authorities out of individuals who merely help us act in accordance with reasons that apply to us independently of our goals or preferences. The other way to put it is to say that these conceptions together fail to account for the special sort of normativity carried by genuinely authoritative directives. It is not just that authorities bind us in some way that we cannot avoid simply by 241 changing our goals or preferences. It is that they bind us in a distinctively moral way. 323 By lumping different sorts of categorical reasons together, the three interrelated concepts we listed above—the NJT, the Razian conception of obligation, and the general conception of practical authority—cannot account for the sorts of reasons practical authorities characteristically create through their directives. 6.2 Requirements 6.2.1 Austin’s Insight The solution to the problem involves an idea that traces back nearly two hundred years, to the legal philosophy of John Austin. 324 Austin believed that laws and rules are a species of command which bind in virtue of some “evil” or “sanction” disobedient subjects suffer. 325 “If you express or intimate a wish that I shall do or forbear from some act,” he wrote, “and if you will visit me with an evil in case I comply not with your wish, the expression or 323 Why a distinctively moral way? Because without moral grounding, it is hard to see how an institution’s claims on you could be categorically normative. If I start a philosophy club the rules of which state that all philosophers in the U.S. have to pay dues, then you (as a philosopher in the U.S.) are required by the institution to pay. But do those rules actually bind you to pay, in any interesting normative sense? It depends, I claim, on whether it would be wrong of you not to pay—that is, whether the rules have some moral grounding which explains why it would be, not just foolish, but wrong of you not to follow them. This would be true, for example, if you freely signed up to be a member of the club, or sought out the avoidable benefits it produced. Absent these sorts of features, it is hard to see how the clubs rules could be binding on you. The same holds for authoritative directives. 324 In his The Province of Jurisprudence Determined. See Austin (2001). 325 Austin (2001, Lecture 1). 242 intimation of your wish is a command.” 326 For Austin, being liable to the evil presupposed by a command was part of what it meant to be obligated. H.L.A. Hart went on to expose a number of serious problems with Austin’s picture of legal authority, and few philosophers today are attracted to his related conceptions of law, obligation, command, and authority. 327 There is little doubt that the threat of sanction, at least in the punitive sense that Austin intended, has no place in our conception of obligation. But while Austin’s theory of law and his conception of obligation are both, in my view, fundamentally flawed, I do believe that the more abstract idea guiding his thinking was right. This is the idea that obligations necessarily involve a substantive element that outstrips their structure. I believe that this element is a wronging. Before I explain why, however, I’d like to discuss how this idea—which I shall call Austin’s insight—contributes to our understanding of a requirement, which is the genus to which the species obligation belongs. 6.2.2 The Structure of Requirements What is a requirement? Clearly, it is something over and above a reason. When I am required to do something we say that I must do it or that I am bound to do it in a way that I am not when I merely have a reason to do it. In our discussion of obligations in Chapter 4, I explained how this sort of bindingness reduces to a reason’s preemptive structure. What emerged in the last section was that obligations include more than this—that they are not just binding reasons, but reasons whose bindingness is explained in a particular way. 326 Austin (2001, 21). I’ve deemphasized some words. 327 See Hart (1961). 243 A requirement to φ is a preemptive reason to φ, a structured reason that both counts in favor of φ-ing and excludes certain other reasons for not φ-ing. It is in virtue of this structure that any person with a requirement to φ can properly said to be bound to φ relative to a particular context or index. The significance of Austin’s insight is that every requirement is partially constituted by an additional substantive element, which distinguishes it from requirements of other sorts. Just as every expanse is a shape, every requirement includes such an element. I will refer to this in the broadest possible terms as its cost. Austin identified an obligation’s cost with the punitive “evil” visiting those who fail to comply with it. However this conception seems too narrow. I can be obligated to do something even if no evil will visit me if I don’t do it. I’ll say more about the nature of a requirement’s cost in a moment. First, however, it is important to recognize that ‘cost’ is a term of art which should not be confused with the natural language term. In natural language, ‘cost’ refers to the amount one must pay for something, and connotes more generally an unwelcomed experience of some sort. But while the various substantive elements which differentiate types of requirement are usually at least to some degree unwelcomed by the person to whom the requirement applies, this need not be true. In the case of obligation, for example, it is, when taken in the sense I intend, the person bound by the obligation who bears the cost of its violation, whereas in practice it will usually be the person to whom he owes that obligation who suffers, and who, as we say in natural language, “bears the cost” of the transgression. This is all consistent with my usage. ‘Cost,’ as I mean it, denotes what is at stake for any person subject to a requirement. 328 328 It is important that we refer to what is at stake for any person under the requirement. For every violation of a requirement will occur in a different context, and so will bring with it unique 244 An example will hopefully help bring this notion of a cost into focus. As I write this, I am subject to two distinct sets of requirements: those of rationality, and those of the library in which I’m working. The requirements of rationality include that I am not to believe both P and not-P, while the requirements of the library include that I am not to smoke inside. Both requirements are the same, structurally—preemptive reasons which apply to me independently of my goals and preferences. Yet the two are obviously distinct. This follows, I claim, from the fact that they impose different costs for nonconformity. The cost of believing both P and not-P is irrationality, which follows straightforwardly from my violation of the rational requirement against it. Smoking inside the library, on the other hand, would not necessarily be irrational. While I might have reasons to do that which are weighty enough to defeat the library’s requirement, any weighty reasons I had to believe both P and not-P would not be of the right kind to affect whether my beliefs, even if I could adopt them voluntarily, are irrational. Believing P and not-P therefore imposes a different cost than violating the library’s requirement. But if the cost associated with the library’s requirement is not irrationality, then what is it? There is no doubt, after all, that the library does not just recommend, but requires that I not smoke inside. The cost of nonconformity is determined differently for different sets of requirements. Unlike the rules of rationality, for instance, which are in some sense preexisting and naturally determined, the rules of the library are artificial and institutional. If, according to the library’s rules, a person who smokes inside is subject to removal, then this is the cost of violating the requirement. consequences. The cost is what follows necessarily from the requirement’s violation in every context. Thanks to Hrafn Asgeirsson for suggesting this way of putting things. 245 Let me forestall an objection by emphasizing the sense in which the cost of noncompliance is not merely correlated with the requirement, but partially constitutive of it. 329 One might think there is a sense in which my being subject to the library’s rule against smoking is contingent. The management could simply overlook my smoking in spite of the fact that it is against the library’s rules. So long as these exceptions are, in fact, exceptions, the requirement against smoking in the library remains a requirement, and applies to me. One might object, then, that the cost of smoking in the library cannot be constitutive of the requirement against it, for it seems that the requirement can exist without the cost. The mistake here is to identify the cost of nonconformity, as I intend it, with the sanction or punishment actually imposed by the institution—in this case, being removed from the library. Punishments are unique to certain sorts of normative systems and therefore cannot be used to provide an analysis of requirements generally. Whether requirements are natural or artificial, institutional or not, what each has in common with the others is that failure to conform with it will, ceteris paribus, qualify one, from the perspective of the set of rules of which the requirement is a part, as the appropriate object of criticism, which can in some cases be expressed in the form of a punitive action which is itself allowed by the rules in question. If smoking in the library is grounds for removal, for instance, then the cost I pay for smoking in the library, and thereby violating the requirement, is that I am, from the 329 A critic might argue that the substantive element I have described is not actually part of the identity of a requirement, but rather a necessary implication that follows from something’s being a requirement. While I suspect that in the case of requirements this would be a distinction without a difference, nothing significant would follow even if the critic were correct, for in either case the element in question would describe a necessary feature of requirements which described their nature, which is precisely our aim. On the distinction in question, see Shapiro (2011, Chapter 1). 246 perspective of the library, rightfully subject to a particular form of criticism, which may be expressed punitively. 330 The criticism itself is indicative rather than constitutive of the cost. Comparing requirements to mere reasons best brings out their unique constitutive feature. 331 Consider, for instance, that even in mid century libraries, which allowed smoking, library patrons may still have had reasons not to smoke inside. It may have distracted other patrons, for instance. Yet in such a case the library could not, even from its own perspective, rightfully criticize or sanction me for smoking, for there is no sense in which I have failed to behave in as they have specified I must. Reasons not to smoke notwithstanding, there is no sense in which not smoking can be said to be required by the library in such a case, for there is no sense in which I am bound, from any perspective, not to smoke. 332 330 The “perspective of the library” is of course not to be taken literally. It is interesting to note, however, that the cost of violating artificial requirements is different than that of violating certain natural requirements in that in the latter case we do not have to specify from which perspective the cost is rightfully imposed. For instance, we do not say that by believing P and not-P you are subject to rational criticism from the perspective of rationality. There is simply no other perspective available. This difference is therefore due to the nature of the “jurisdiction” of the two sorts of requirements in question. Every person who might have a perspective on matters of rationality will herself already and unavoidably adopt the perspective in question, and so there is a sense in which rationality’s jurisdiction is unlimited. It seems, then, that we do not speak of being subject to the rules of rationality from the perspective of rationality because there is, for us, anyway, no other perspective from which to look. 331 This conception of requirements can be used to frame the debate over the pressure rationality exerts on a person with most reason to do something. Some writers believe that we are rationally required to do what we have most reason to do. Others (e.g., Dancy (2004)) think that we sometimes have sufficient reason to do any number of things without being required to do any one of them. The first view entails, as we would expect, that doing other than what you have most reason to do carries a cost, while the second view, which lacks this requirement, has no such entailment. 332 Once Austin’s insight is added to our conception of requirements, we begin to see clearly the various sets of requirements which apply to us concurrently at any given time. While not smoking in the library may not be a requirement of the library, it may be required by the rules of etiquette, for example, or by the rules of the state (say, if I am under 18, in which case the applicable requirement specifies that I may not smoke, period). The sort of cost one incurs for noncompliance distinguishes these two species of requirement. In the first case one’s behavior is rude, for which reason it may appropriately be met with particular reactive attitudes, about which I say more below. In the second 247 I should emphasize that I have yet to say anything about the connection between the perspective of an institution and the objective fact of the matter. The fact that an institution’s requirements apply to you—that they are, I will say, de facto requirements for you—does not mean that they are at the same time normative. The fact that the mafia requires everyone in the neighborhood to pay a fee for the “protection” they receive, for example, and imposes sanctions on anyone who fails to meet their requirements does not entail that you are normatively required to pay, even if you live in the neighborhood and so fall within the scope of the requirement. For that to be the case the institution and its requirements would have to be authoritative in the sense we discussed in Chapter 5. What is important for our purposes here, however, is Austin’s insight—the idea that there is a substantive element, a cost, which is partially constitutive of requirements. To incur such a cost is to be rightfully subject to criticism from the point of view of the institution to which the requirement belongs. 6.3 Wrongings 6.3.1 Three Ways to Get It Wrong Austin’s insight helps us to describe the unique sort of requirement obligations are by suggesting that to understand obligations we must look for the cost associated with noncompliance. Locating this substantive element is precisely what allows us to overcome the second strand of the no-obligation objection. Unlike rules of rationality, violation of case it is illegal, and thus one will appear from the perspective of the state to rightfully subject to the fines or other punitive measures the state specifies. 248 which entails that one is irrational, and institutional requirements, violation of which subjects one to institutional sanctions, my contention is that (moral) obligations are set apart as a specific species of requirement by the fact that violating one entails that the agent wrongs someone. This substantive element consists of two parts, each of which draws an important distinction. First, the fact that I wrong someone by violating the obligation explains an important sense in which obligations are distinct from mere reasons, which, even when they are other-regarding in the manner of obligations, do not exact this sort of deontic cost for acting other than they recommend. The fact that you will be sad if I skip the conference is a reason for me to attend, but that reason does not entail that I wrong you by skipping it. By failing to respond to my reasons I may go about something in the wrong way, that is, make some sort of mistake, but I neither act wrongly nor commit a wrong against anyone. These last two ways of going wrong are then the objects of the second distinction. For the fact that I wrong someone explains the sense in which obligations are distinct from other deontic norms, such as, perhaps, duties to protect the environment, the violation of which entails that I act wrongly but not necessarily that I wrong someone. Monadic and Bipolar Normativity This last point is important, so let me elaborate on it. There is a division between relational and non-relational—or, to borrow language from the literature, bipolar and monadic—deontic judgments and considerations. 333 Monadic deontic judgments are non-relational, and include, 333 I find the language of monadic and bipolar deonticity cumbersome, but use it nonetheless because it is accurate and by now part of the literature. See Thompson (2004) and Wallace (2012). 249 for example, that it would be wrong of me to φ, or that I have a duty to φ. Bipolar judgments are, by contrast, relational; they include that I would wrong you by φ-ing, or that I have an obligation to you to φ. The normative difference between these two types of judgment is highlighted by the personal and impersonal reactive attitudes that going wrong, acting wrongly, and wronging others appropriately elicit. 334 Acting wrongly is a monadic deontic notion which is distinct from merely going wrong (in some respect or other) in that we appropriately feel guilt or shame for having acting as we did. When we act wrongly others may feel indignant at our behavior, and they may appropriately hold us accountable by blaming us and otherwise addressing us with a sort of moral disapprobation that would not be fitting for someone who went wrong by, for example, making an error in long division. 335 Acting wrongly is then distinct from wronging someone in that only the victim of a wronging has reason to feel personally betrayed by the wrongdoer. Wrongings, to put the point another way, necessarily result in some individual’s occupying a privileged position from which to complain about the wrongdoer’s action. 336 She may be hurt or offended, and may resent the wrongdoer in a way she would not have reason to had she merely witnessed rather than suffered the wrong. She may also be uniquely positioned to seek reparations, or at the 334 The notion of a “reactive attitude” is of course from Strawson (2008), who distinguished between “participant” and “impersonal” attitudes. For an informative discussion of reactive attitudes and the normative concepts and beliefs they presuppose, see Darwall (2006, Chapter 4) and (2007). 335 Thus Scanlon observes, “to decide that what a person has done is blameworthy is in part to decide that he has behaved wrongly” (2008, 123). 336 This is not to say, of course, that they ought to complain, which is a different issue. 250 very least to demand an apology. Later, if she sees fit, she will be able to forgive the wrongdoer in a way that others are not. 337 These differences between wrongs and wrongings can be brought out by a simple case of promising. Anyone may rightly form the judgment that I acted wrongly by failing to keep my promise to you. However you have been wronged in a way that others have not. The person to whom I made my promise, you stand in the sort of bipolar deontic relation to me that makes such a wronging possible. And so while others may impersonally hold me accountable for failing to act as morality demands by, for example, addressing me with contempt, you occupy a special status which introduces a range of attitudes (resentment, forgiveness, etc.) available to nobody else. 338 Whatever their quarrel, others cannot demand that I apologize to them, for they cannot claim that I have treated them objectionably. That moral space is available only to you, the person with whom I created the bipolar deontic relationship that made such a wronging possible in the first place. Some philosophers believe that, while bipolar deontic considerations entail monadic ones, the reverse is not necessarily true. 339 These people believe that some of our obligations are undirected, that is, that we can have such obligations without owing them to anyone. The cases they cite usually involve the destruction of things which many people consider valuable 337 See Darwall (2007, 62). It is worth noting that the distinction between the standings of the victim and the observer of a wrong are reflected in civil law, wherein only a victim, someone who bears what is legally analogous to a bipolar deontic relation, can bring charges against a wrongdoer. 338 As Darwall (2006, 62) points out, personal reactive attitudes can sometimes be addressed on the victim’s behalf by those who identify with her. It is unclear to me that such vicarious judgments are not bipolar in virtue of a wrong the agent on some level feels she has suffered as a member of a group to which she and the victim both belong. In such a case she would, in some limited sense, be a victim of a wrong as well. 339 Thompson (2004) and Raz (1988), e.g. 251 in themselves—filling the Grand Canyon with cement, for instance—or disregarding valuable institutions in a manner that makes nobody worse off. 340 If these philosophers are right, then it is easier to act wrongly than it is to wrong someone, since the set of possible wrongful acts subsumes the set of possible wrongings. Even if this is right, however—and I remain agnostic about whether it is—I will eventually rely on the stronger notion of bipolar deonticity to ground our political obligations, so nothing in my argument hangs in the balance. 6.3.2 Indexing So far I have suggested that obligations are a species of requirement constituted by a particular normative structure, coupled with a substantive element defining the unique flavor of bipolarity that sets moral obligations apart from other requirements (Austin’s insight). I have also claimed that this substantive element, the cost of a violation, is a wronging. Below I will say more about the nature of wrongings, but because we have carved out this set of normative considerations in terms of deontic judgments, it will be important, first, to be clear about the particular types of facts to which those judgments apply. Some philosophers have suggested that we need a sort of normative indexing system to separate moral and non-moral applications of monadic and bipolar deontic judgments. 341 As Michael Thompson argues, this sort of indexing allows us to shift these judgments “into a particular gear” which represents the “logical shape” of the corresponding thoughts in our 340 See Owens (2012, 67). I should say that I am skeptical that Owens’ second example—a lie told to someone who knows one is lying—constitutes a wrong without a wronging. It strikes me as a wronging without a harm, which is a sort of case I discuss below. 341 See, in particular, Thompson (2004, 340-43). 252 mind. Thompson’s worry is that the language and structure of these judgments are not exclusive to moral contexts, that, absent a predefined context (represented in logical form by a modal subscript) we cannot know exactly what it means that “it was wrong of you to φ” or that “you wronged Smith by φ-ing.” These judgments are, he says, “empty logical forms admitting many radically different interpretations or schematizations, only one of them specifically moral.” 342 The unindexed judgment that it is wrong not to keep your promises is, on this view, no different than the judgment that it is wrong not to submit the paperwork your boss requested, or that it is wrong to move your rook diagonally, the idea being, as I understand it, that these claims are structurally similar but substantively underdetermined, so that none corresponds with a particular class of normative relations. I have already argued that there is an important structural similarity between the sorts of normative relations we bear to one another in various contexts with respect to, for instance, what the law, or my boss, or morality requires of me. However I would like to resist Thompson’s implication that wrongs and wrongings are not exclusive to the moral realm. My concern is not with the terms we use to refer to particular normative phenomena actualized by the rules and practices with which we’re all familiar—families and universities, friendships and states—but rather with the sorts of moral properties that seem actually to attach to our actions within these institutions. As I understand Thompson’s view, it is that requirements share the same monadic and bipolar substantive elements—wrongs and wrongings—but that these come in various “gears” or “keys” which depend on the type of 342 Thompson (2004, 341). 253 system to which they belong. 343 Reactive attitudes (assuming, as I see no reason not to, that they exist and are appropriate across normative contexts) must therefore be put in the same normative gear. For example, an appropriate response to A’s wronging moral B by violating his obligation moral to her would be B’s blaming moral him. But we can also speak of A’s duty legal to B versus her duty moral to B, whereby the first sort of duty calls for a different set of reactive attitudes than the second. What is striking about Thompson’s attempt to explain how “customs, practices, and institutions of quite various sorts can give sense to our bipolar linkages,” however, is just how unclear many of these linkages actually are. Thompson never gives concrete examples. But while one can reasonably speculate about certain contexts in which Thompson’s claim might hold—for instance, we might have some sense of what it would mean for B to blame legal A—in other and perhaps most contexts his suggestion is obscure. What would it mean, for example, to say that B would appropriately respond to A’s wronging rules of chess her by blaming rules of chess him, or that A should feel guilty rules of chess in this case? What is unique about moral obligations, I have suggested, is that they create the possibility of wrongs and wrongings that are distinct, in a manner I have yet to explain, from the substantive normative phenomena constitutive of non-moral requirements—such as being rude in the case of etiquette, or being irrational in the case of rational requirements, each of which carries along the appropriate reactive attitudes. If Thompson is right, however, then wrongs and wrongings are common to all monadic and bipolar deontic judgments, which are, as he says, “empty logical forms.” But if this is right then there must 343 Thus Thompson (2004, 342) approvingly cites Hume’s view that competitive games such as chess and checkers “may be said to attract thoughts in these same logical shapes.” 254 be something beyond the fact of one’s having acted wrongly which separates, say, moving one’s rook diagonally and breaking one’s promise, for on Thompson’s view both are wrongs (and even wrongings) in their respective “gears.” But what could this mean? While it may be common to distinguish between one’s duty legal and duty moral , as I said above, it is not clear how these deontic concepts extend to etiquette, or games, or the Catholic Church, each of which is, at a different degree of abstraction, constituted by sets of rules and requirements which govern the behavior of those to which they apply. 344 This seems even more apparent for wrongs and wrongings, which are, as I have argued, and as Thompson seems to accept, the substantive elements which distinguish moral obligations from other requirements. What does it mean to wrong rules of grammar someone? What, even, does it mean to wrong legal someone (a concept which, because of the close resemblance between legal and moral principles, should resonate if any related deontic concepts do)? While there is indeed a normative structure shared by legal directives, moral norms, rules of games, and other institutions which purport to guide our behavior, wrongs and wrongings (as distinct from merely going wrong) appear to be tightly woven into the fabric of morality. This is not to say that we cannot co-opt deontic language to refer to the costs imposed by requirements of non-moral systems. We clearly can. But in such cases we do not import the special sort of normativity genuine wrongs and wrongings carry with them. You may wrong the rules of the philosophy club me by breaking rules which apply to you because you’re a philosopher, but which have no normative significance for you because you did not sign up 344 To be clear, the claim is not that there are no requirements (or, if you insist, obligations) of etiquette, or chess. It is that violations of these requirements are not, absent some moral context that maps onto it, wrongs or wrongings. 255 or seek out benefits. We can call this a wronging if we like, but the question remains whether such a “wronging” carries any normative weight. 345 6.3.3 The Nature of Wrongings As I indicated above, I seek to ground political obligations in bipolar deontic considerations—wrongings as a subset of the more general monadic category of wrongs. To do this, however, we must have some sense of what wrongings are. Wronging as Harm One suggestion is that wrongings are a particular type of harm. This suggestion may be plausible on the right interpretation of ‘harm.’ To be harmed by an action is, following Joel Feinberg, for that action to set back one’s interests, directly or indirectly, from where they would have been had the action not been taken. One’s interests are then to be understood as those (more or less abstract) things in which one has a stake—my bus arriving on time, the crime rate in my city, or my spouse’s professional development, for example—such that one’s well-being is dependent to some extent on their status, continued existence, flourishing, and so on. 346 On this broad conception of harm there are many harmful actions which are obviously not wrongings. My interests are at least somewhat set back if you take the last seat on the bus, for example, though you do not wrong me taking it. The broad class of things in 345 A potential concern about the view I’ve put forward is that it commits us to claiming that we cannot have obligations or duties to refrain from acting in ways that won’t wrong anyone. I address this below. 346 Feinberg (1984, 31-37). 256 which we have interests entails, however, at least on this interpretation, that essentially all wrongs are harms. 347 Even those wrongs which lead to a net benefit for the victim are still harms in that they set back some of her interests. If I broke into your house to clean it, for example, my actions would be harmless, materially speaking, and may even constitute a net gain for you, since you’re better off with a clean house than a dirty one. Yet I have arguably set another of your interests back—that in having the sort of proprietary control over your property which ensures that other people do not use or trespass on it without your permission. Thus even a setback which results in a gain, all things considered, constitutes a harm on this conception. Though it is not entirely clear that all wrongings can so clearly be counted as harms, or injuries to interests, I should emphasize that whether these broad conceptions of harms and interests are the right ones is not the issue at stake here. The problem is rather that of determining which harms are also wrongs when we agree to use ‘harm’ in this weak sense. A common approach to this problem is to conceive of wrongings in the context of rights, such as the right you have to proprietary control over your property. On Thomas Nagel’s view, for instance, rights provide us with agent-relative reasons to act or abstain from acting in particular ways, independently of the total harm our actions would cause. Such reasons “prohibit us from doing certain things to anyone but do not require that we count it equally a reason for action that it will prevent those same sorts of things being done to someone, but 347 As Owens (2012, 88-90) points out, some wrongings do not set back our interests relative to our current circumstances. For instance, I might wrong you by unfairly giving to you less than your share based on some appropriate distribution. In such a case you are wronged, not because you did not receive a benefit, or because you received less than you would have received had I distributed nothing at all, but because you received less than you would have received on an appropriate distribution. The definition of wrongings on which I ultimately settle is consistent with this point. 257 not by oneself.” 348 For example, if I have a right not to be murdered, then you have an agent-relative reason not to murder me, one that would not be overridden by the fact that by doing so you could prevent two other murders. On this view rights seem to generate obligations in the sense we have been discussing—that is, structured reasons with the proper substantive element—and so they would seem to provide some insight into the nature of wrongings. Wrongings, according to this picture, are simply those harms which violate one’s right. 349 But while a loose correlative relationship between rights and wrongings is undoubtedly both correct and significant to moral theory, it tells us little about the nature of wrongings themselves. The problem is that, while most philosophers agree that to have a right is to have a valid claim that gives others something like an agent-relative reason not to inflict certain sorts of harms, harms are constituted, as I said above, by setbacks to one’s interests, and no part of the concept of a right specifies which setbacks we wrong others by inflicting. It might at first appear that this problem can be overcome by paying closer attention to the sorts of considerations theories of rights pay attention to. On the interest theory, for instance, one has a right just in case one has an interest sufficient to place some other person under a duty; so we might say that wrongings are setbacks to these same interests (that is, to right-protected interests). 350 348 Nagel (1995, 88); I’ve deemphasized ‘doing’ and ‘anyone.’ 349 The broad thesis that wrongings can be understood in terms of set-back interests appears to be accepted by many theorists, and is most obviously aligned with Raz’s interest theory of rights (1988). 350 This seems to be a natural implication of the theory of rights found in Raz (1988, Chapter 7). While my conception of wronging does not depend on it, it seems to me that Raz’s theory employs a nonstandard conception of welfare in order to account for interests which would be better accounted for by the sort of conception I put forward below. Briefly: the Razian conception is that X has a right “if and only if X can have rights, and, other things being equal, an aspect of X's well-being (his interest) is a sufficient reason for holding some other person(s) to be under a duty” (1988, 166). The 258 However this interpretation is more germane to theories of rights than it is to theories of wrongings. We have already said that duties and obligations are constituted in part by the wrongings to which their violations give rise. Thus, the proposed conception would amount to the unhelpful claim that to wrong someone is to set back an interest of his which is significant enough that one would wrong him by setting it back. Wronging as a Failure to Respond to Value A more promising approach may be to understand a wronging, not as any sort of setback to a certain sort of interest, but as a certain sort of setback to any sort of interest. Wrongings, I would like to suggest, are unlike other harms in that they are more responsive to the circumstances of performance—including the actor’s beliefs and intentions, as well as other aspects of the context in which he acts—than to the effect his action has on the victim’s welfare. Two related observations support the view that the circumstances of an action’s performance play a pivotal role in determining whether the resulting harm constitutes a wronging. First, whether an action which causes a setback to an interest constitutes a problem is that there are surely some aspects of one’s well-being which do not seem to ground duties in others, but which seem more significant than other aspects that do. For instance, at a certain level of poverty my right to political participation might be somewhat irrelevant, and certainly less relevant than twenty thousand dollars would be, both in my opinion and as a matter of fact. Yet others do not necessarily have a duty to provide it, even if they do have duties deriving from my right to political participation. A natural response to this sort of concern is to offer a theory of well-being which prioritizes interests which we consider to be necessary to one’s living well even when securing them does little if anything to improve his situation in any measurable way. However this conception seems, at least to me, unnecessarily revisionary. It seems more accurate to suppose that the duties others have to us with respect to some action or object flow, ultimately, not from our interests, but from our value as persons, which should be respected even in cases in which the duties to which that value gives rise do not further one’s interests in a measurable way. 259 wronging depends in part on the reasons there are for the actor to perform the action, as well as their role in motivating the actor to act. Suppose, for instance, that a police officer prevents a religious leader from speaking at a public event. The speaker’s interest in determining when and with whom he shares his religious views is significant enough to impose a duty on others, including the officer, not to set it back. It is, in other words, precisely the sort of interest the officer would harm the speaker by setting back on the view proposed above. Yet whether the officer wrongs the speaker seems to depend on further aspects of the case. For instance, she would not wrong him by pulling him from the stage in order to prevent an assassination attempt. This is true even if the officer knew that the assassination was planned for the very end of the speech, after the speaker had communicated the message he had an important interest in communicating at his discretion. The officer’s intentions and the circumstances on the ground, it seems, play some role in determining whether the harm suffered by the victim was also a wrong against him. The second observation is that my action can wrong someone even in cases in which it is unclear that it has set back anyone’s interests. The Russian anthropologist Mikluko- Maklay provides us with a useful example of this sort case. 351 While visiting the Malayan archipelago, Maklay retained the services of a Malayan native on the express condition that Maklay not take his photograph, which the native believed would take something out of him. Maklay never took the native’s photo, but the opportunity, which arose late one night as the native slept, presented Maklay with the opportunity to commit what seems very much like a 351 Foot (2001, 47); originally from Peter Kropotkin’s Memoirs of a Revolutionist, 229. 260 harmless wronging, or as it is sometimes called, a “bare wronging.” 352 Again, we are interested, not in whether the native would have suffered a genuine harm had Maklay photographed him, but in explaining the apparent fact that such a harm (if it is one) constitutes a wronging. Philippa Foot’s analysis of the case is instructive. “It would have been deeply disrespectful of Maklay,” she says, “to take advantage of his servant, particularly as the latter saw it as so important not to be photographed.” 353 Advantage taking is, here, just one way of showing disrespect, which is what the action’s wrongfulness consists in according to Foot. “It matters,” she says, “…that at some basic level humans should have mutual respect. It matters, not just what people do, but what they are.” The idea that people should have respect for one another is of course neither new nor restricted to virtue theorists. What I would like to propose, however, is that understanding disrespect in the right way allows us to distinguish wrongings from other sorts of normative phenomena. A wronging is an act which is disrespectful in that it constitutes a failure to respond appropriately to the reasons to which another’s value gives rise. Understanding a wronging in this way makes sense of the two sorts of cases we discussed above, as well as a third sort of case I’ll discuss later. It makes sense of the officer’s behavior because, while it sets back the speaker’s interest in sharing his religious views with his audience, and for that reason harms him (as per the Feinbergian conception we’ve adopted), it is nonetheless an appropriate response to the speaker’s value, which provides conclusive reason to save his life even when that entails silencing him on an important 352 Owens (2012, Section 23). 353 Foot (2001, 48). 261 occasion. The Maklay case is the mirror image. Arguably no harm will be caused if Maklay photographs the native. Yet doing so would be, as Foot points out, deeply disrespectful. It is disrespectful because the native has value which, along with ethical principles applicable to him in virtue of it, makes him the appropriate object of a particular sort of treatment, namely, that which conforms to the agreement he made with Maklay. Value While the conception of wronging and thus the conception of obligation I’ve proposed presupposes that people have value, and that their value provides reasons to treat them in particular ways, no part of either of these conceptions seriously constrains which theory of human value we may accept. Still, to account for the various sorts of behaviors which constitute wrongings, not all of which are a direct affront to others’ value, an acceptable theory must explain how this value provides the reasons we have to treat people in the various ways we have reason to in various contexts. It must explain, for instance, how my reasons not to take your belongings without your permission, or to keep my promises to you or pull you from a burning building if it’s no trouble for me, each of which depends on a practical principle occupying a unique level of generality, are grounded ultimately in facts about you that make you an appropriate object of such treatment. This sort of normativity is transitive in that the chain of justification from broader to narrower principles entails that the explanation for any wronging which references those principles will ultimately trace back to one’s value. An advocate of such a theory must accept that, and explain how, your value percolates upwards, making sense of the specific obligation I have not to take some particular object you own, for example, by reference to a more general principle against 262 taking others’ things, which is itself grounded in an even more general principle that justifies our private conception of ownership, and so on. Eventually this chain of explanations must include the fact that you, like others like you, have value, and that your value grounds the reasons they and I have to perform or refrain from performing the specific actions at the start of the chain. 354 The principles which seem plausibly to be supported by persons’ value will play a central role in the argument for political authority in the next chapter. I will say more about their content there. The central point to take away from our discussion here is that disrespect travels up the chain, from violations of specific applications of principles to our actual circumstances to, eventually, the value others have as persons. It is, as I’ll explain in the next chapter, this sort of transitivity which helps explain the obligations we have to others to obey the law in certain types of political institutions. 6.4 Updating the View We are now in a position to reformulate some of the conceptions with which we started this chapter in a manner that will allow us to answer the two worries contained in the second strand of the no-obligation objection. The first concerned the inconsistency between the NJT and Raz’s narrow concept of obligation. On this account obligations are categorical reasons which, according to the general conception of practical authority, an authority’s 354 Raz describes this normative phenomenon, as it applies to reasons, as nesting. “The more general reasons apply as a standard background to our activities, and are less affected by changing circumstances, whereas the more specific reasons that nest in them tend to apply during shorter periods and depend on conditions that are often liable to change” (2009, 152). 263 directives are supposed to provide. Yet it is unclear how helping someone act in accordance with her prudential reasons could position a person to provide categorical reasons. Even if it could, we are faced with another concern, which is that preemptive categorical reasons do not amount to the normative phenomena that concern us anyway—that is, moral obligations. The new conceptions we developed in the preceding sections allow us to address both problems. Here, first, is our updated conception of the NJT: NJT* One person has authority over another if the alleged subject is likely better to comply with the (moral) obligations which apply to him (other than the allegedly authoritative directives) if he accepts the directives of the alleged authority and tries to follow them, rather than by trying to satisfy those obligations directly or by following the directives of any other alleged authority. 355 NJT* allows us to address our first concern by importing obligations right from the start. It is not the case, according to NJT*, that my financial advisor’s capacity to help me act in accordance with the prudential reasons I have to satisfy my desire to make as much money as possible justifies her authority over me, for my prudential reasons do not contain the right sort of normativity. By restricting our conception to obligations, which are already 355 The final clause addresses the problem of conflicting authorities. Though its purpose and justification are intuitive, I provide an explicit defense in the next chapter when I describe the “bridging principle” according to which the normative force of our preexisting obligations transfers to authoritative directives. 264 categorical, we restrict cases of authority to those in which categorical reasons are already in play, rendering it less mysterious how someone satisfying NJT* could generate them. We then address the conceptual problem by providing a more accurate analysis of obligations: Obligation A (moral) obligation is a requirement—that is, a preemptive reason an inappropriate response to which imposes a cost—which is unique from other requirements in that its cost is a wronging. On this analysis, due, as we saw, to Austin’s insight, an obligation is not simply a preemptive categorical reason, but includes the sort of substantive element necessary to explain, not simply why moral obligations are requirements, but why they have the special normative force that they do. It is the fact that we will wrong others by violating our obligations which makes sense of their categorical nature, and which explains, as we would expect it to, the appropriateness of the sorts of reactive attitudes we address to those who have violated them. Two Final Concerns There are two concerns I would like quickly to address. First, does the conception of obligation I’ve put forward commit us to claiming that we cannot have obligations or duties to refrain from acting in ways that won’t wrong anyone? I have attempted to remain agnostic about whether obligations can be non-directed, and owed to nobody. But while it is indeed true that the conception I’ve put forward is meant to capture the special sort of normativity directed obligations carry, I believe my view can be made to account for non-directed 265 obligations as well. What is essential, ultimately, is that we are able explain the special normativity of moral obligations created by authoritative directives, so as to bypass the no- obligation objection. And while an action which wrongs someone is importantly different than one which is merely wrongful, the latter acceptably captures that normativity, separating non-directed obligations from other requirements in precisely the way we must. On an account which groups these two sorts of wrongs together, we can cast obligations as preemptive reasons we would be wrong to transgress. So far as these wrongs can be understood as a unique form of disrespect (though not necessarily disrespect to anyone in particular) their structure would allow us to accomplish everything I hope to in the next chapter with their bipolar brethren. The second concern is whether our account allows for obligations to oneself. While I again take no stand on whether such obligations exist, the conception I’ve endorsed is clearly capable of explaining them. To have such an obligation would be to stand in a relation to yourself such that by failing to act on a particular preemptive reason you would disrespect and thus wrong yourself. In such a case you might hold yourself responsible, addressing yourself with blame while simultaneously feeling guilty for your behavior. Such feelings are, I think, familiar to most of us. 6.5 Conclusion My aim in the next and final chapter is to use the conceptions of obligation, justification, and practical authority we have developed to this point to construct three arguments for political obligation, which together constitute my take on the authority theory. The first argument 266 takes the standard Razian approach to the problem, but updates the NJT to NJT* and adds an essential “bridging principle” to explain how normativity transfers from duties to directives. The second argument addresses two concerns facing the first one, utilizing the conception of obligation we developed in Chapter 4, as well as the hypothesis, developed in this chapter, that disrespect is transitive. The third and final argument is a novel attempt to explain political obligations in cases in which authorities are not legitimate, and, even, in cases in which states are not authorities. As we’ll see, this sort of explanation is possible only in virtue of the conceptions of authority, obligation, and wronging we’ve developed throughout Part II. 267 Chapter 7: Three Paths to Political Obligation Until recently, nearly all theories of political obligation espoused a single justificatory ground. Consent theorists did not (and still often do not) consider the role fairness has to play in explaining political obligations; fairness theorists ignored the significance of our natural duties to one another; and natural duty theorists paid little attention to the prescriptive roles we play and may be obligated to play within our social and political institutions. Political philosophers of one stripe would, in other words, often ignore other potentially binding considerations which might have complemented rather than conflicted with their theories. Much of this has begun to change, however, as theorists working on the problem of political obligation have come to recognize the explanatory benefits of multiple principle (MP) theories. MP approaches can take a variety of forms, each of which illuminates its author’s understanding of the problem. George Klosko, whose theory of fair play we discussed in Chapter 2, offers a sort of tapestry, a patchwork of justificatory considerations each covering a separate class of legal requirements. Though he argues that the principle of fairness takes center stage, grounding our most important political obligations, he allows that other principles, in particular those rooted in samaritan duties of the sort we concentrated on in 268 Chapter 3, fill in the explanatory gaps left over. 356 But there are other ways to be an MP theorist. Rather than develop separate, hierarchical principles that join together to explain the range of political obligations we are thought to have, as Klosko does, Christopher Wellman weaves seemingly competing considerations directly into the two principles on which his theory is based. Thus, while Klosko’s MP theory is, we might say, like a quilt of patches, each of which maintains its shape independently but which fit together to reveal a larger picture, Wellman’s is like a loaf of bread, in which each ingredient mutually reinforces the others to produce a final product in which no ingredient is independently identifiable. In what follows I offer the final formulation of what I have called the authority theory. An authority theory is, again, a theory according to which the authoritative relation between authorities and subjects plays a necessary part in an explanation of those subjects’ obligations. My particular version of this theory is made up of three arguments for political obligation, which I shall refer to as the standard, proceduralist and subjectivist arguments. Though distinct from one another in important ways that will become evident as we move through the chapter, each argument follows directly from the conceptual apparatus we have constructed in Part II, and depends in the requisite way on the authoritative relationship citizens have with states. I believe it will be clear from the various normative considerations on which the theory as a whole depends that my approach embraces the MP methodology. However it should also be clear that the theory diverges in important ways from the other MP theories we’ve discussed. Unlike Wellman’s samaritan approach, the explanatory force of the 356 See Klosko (2004b) and (2005, Chapter 5). Klosko calls this patchwork function cumulation. 269 arguments I offer does not depend on the synergy of the various considerations contained in them. Each argument stands on its own, providing a well-defined justificatory scope. Unlike Klosko’s fairness theory, however, the arguments I offer do not line up neatly, like a patchwork tapestry. Rather, they overlap to provide layers of justification. Not only does this expand the class of laws and people to which the theory applies, but it contributes to an explanation of the various strengths of our political obligations across different laws. What I hope to explain, ultimately, is not just the obligations we have to obey the law, but also how these obligations depend on the authoritative relationships that tie citizens to governments, and through that, citizens to one another. I begin in Section 7.1 with the standard argument, so-called because I take it to be the natural extension of the standard interpretation of the Razian view as we’ve developed it. For reasons I’ll explain, the standard argument alone does not meet the satisfaction conditions of comprehensiveness or generality (Chapter 1)—the justification it provides is, as I will say, gappy—and so it cannot serve as a complete theory of political obligation on its own. I’ll argue, instead, that it plays an integral subsidiary role in our broader MP theory of political obligation by helping us explain why different legal directives appear to have different degrees of deontic force. The proceduralist argument I go on to develop in Section 7.2 is a response to an important objection to the standard argument which has emerged in the recent literature on legal authority. 357 That objection, in a nutshell, is that the Razian conception as we’ve developed it “is blind to the possibility that an authority’s legitimacy may depend as much on 357 Different authors have different perspectives on the problem. See Christiano (2004); Hershovitz (2011); Marmor (2005); Shapiro (2002). 270 how it makes decisions as on what those decisions are.” 358 Though we’ll need to formulate the objection more clearly before addressing it, it seems plainly to expose a dimension of our presumed political obligations which the standard argument cannot explain—namely, that democratic regimes seem to have special claims on their citizens that other regimes do not. However the solution, as I’ll argue below, is not to give up the authority theory, as critics have suggested, but to account more carefully for the range of duties we owe to one another and the special way the interests on which those duties are grounded can be exercised and thus respected in democratic states. I’ll argue that democratic procedure matters insofar as it is an expression of citizens’ equal status as political agents who, in virtue of their value as persons, have a claim on others to respect their interest in shaping their shared society to an extent compatible with the others’ interest in doing the same. In Section 7.2 I’ll explain how we can account for this observation with the concepts we developed in Chapters 4, 5, and 6. In Section 7.3 I offer the subjectivist argument, which I believe represents a new approach to the problem of political obligation, one that emerges naturally from the authority theory’s patient focus on the nature of obligation and authority. Rather than extending the deontic reach of the standard and proceduralist arguments, however, which together meet the satisfaction conditions on comprehensiveness and generality, the subjectivist argument identifies an overlooked source of political obligation—additional reasons for which it is wrong to disobey the law—which layers onto the other two arguments I will discuss. The subjectivist argument therefore provides a more 358 See, in particular, Hershovitz (2011, 3); Marmor (2010); Shapiro (2002). 271 comprehensive picture of the normative political landscape than competing theories of political obligation have been able to. 7.1 The Standard Argument Recall how the Razian argument is supposed to go. The service conception of authority maintains that authorities provide their subjects with a service by helping them act in accordance with their reasons. The normal justification thesis (NJT) is then meant to provide the normal conditions under which such authority is legitimate—which, to be clear, is just to say that when the NJT is satisfied the directives issued by the purported authority have a particular sort of normativity. When they do, she is a legitimate authority according to the normal conception: she can bind her subjects to act as she specifies simply by expressing her intention to bind them. Because a purported authority is most likely to help her subjects in this way when her directives respond to the reasons which apply to them independently (the dependence thesis), this is what she ought to do. As I argued in Chapter 5, it follows from the service conception, and in particular the NJT, that those who satisfy its requirements have the power to generate preemptive reasons for their subjects. Yet, as we saw in Chapter 6, we have no reason to think that these preemptive reasons are categorical in nature, or that, even if they were, transgressing them would constitute a wronging and thus imply that what we were dealing with was a genuine moral obligation. Our solution was, first, to import the necessary sort of normativity right from the start, leaving us with NJT*, and second, to provide a more accurate conception of obligations, one that explains, not just their categorical nature, but also the moral 272 considerations underlying it. What we discovered was the constitutive relationship between moral obligations and wrongings, and thus the special connection between wrongings and authority. Someone is an authority for me if, by simply expressing an intention to do so, she can change my normative circumstances so that I wrong someone by performing or refraining from performing a particular action. This updated account is the basis of the standard argument. 7.1.1 The Validity Question With our conceptual scheme in place, we are finally in a position to see whether those who satisfy NJT* can do what I’ve claimed they can do. To put the challenge more precisely, we are in a position to determine whether the apparatus developed in the previous chapters provides us with an answer to the validity question, which we first posed (along with the applicability question) in Chapter 3. That question, again, is whether the theory under consideration plausibly describes a genuine source of obligation—that is, whether there are any circumstances in which obligations are generated for the reasons described by the theory. Consider the following case, which I modify from an earlier discussion: 359 Neighbor Two neighbors, Smith and Wesson, are decent people who nonetheless cannot get along. Wesson complains that Smith frequently has large parties at her house, which make it difficult for him to sleep, and that she aggravates the problem by playing loud music and leaving her floodlights on throughout much of the night. Smith says that Wesson is 359 See Raz (1988, 41) and (1980,139). 273 unreasonably sensitive about the light and noise, which she claims surely cannot not bother Wesson anyway, since his bedroom is on the far side of his large house. Smith also complains that Wesson frequently comes over to her home uninvited to complain about noise, even when there is little or no noise to complain about. From our discussion in Chapter 6 we assume that the particular sort of value Smith and Wesson have explains why they are the proper subjects of a series of hierarchical action guiding principles, the more narrow of which ultimately relate to the way neighbors should treat one another. It follows from their value and various contextual factors that Smith and Wesson are obligated to treat each other in some particular way. The problem is that, though they are both decent people who would like to behave in the right way, they disagree about what way that is. Imagine, then, that the municipal government, perhaps frustrated with the time it wastes dealing with their complaints, assigns to Smith and Wesson an arbitrator whose job it is to determine more detailed rules by which they must abide. There are a variety of reasons to think that, by accepting and trying to act in accordance with the rules laid down by the arbitrator, both Smith and Wesson will more likely act in accordance with their obligations to one another than they would were they to continue behaving as they see fit. One is that the arbitrator will likely have access to information the complainants do not. Another is that their unbiased perspective on the descriptive and normative facts increases the likelihood that she will be able to evaluate their behavior objectively. Can Wesson really hear Smith’s music from his bedroom on the far side of his house, or is he simply upset that Smith plays 274 her music so loud even after he has repeatedly told her that it prevents him from sleeping? If he can hear it, is it in fact unreasonably loud, or is Wesson unreasonably sensitive? Could these problems be solved with a pair of light earplugs and some window shades, which Smith could perhaps pay for? On the other hand, does Smith have a sufficiently good reason to have her music as loud as she does? Does she have too many people over too often? And so on. After sorting through the details, the arbitrator produces a set of rules for Smith and Wesson to follow. Among these is that Smith is to keep her floodlights off during the night and to play her music below some prescribed level after 10pm during the week. Wesson, for his part, is to call Smith on the phone to voice any complaints before going over to her house, and to permit Smith to have large gatherings at least once each month, at which time she may play her music after the prescribed curfew and above the agreed decibel limit. The question that remains for us is whether Smith and Wesson now have an obligation to do as the arbitrator says. To be clear, they clearly did not have an obligation to comply with these specific rules before. Their obligations underdetermined the specific actions they owed one another, and could therefore have been acceptably discharged in any number of ways. 360 In our response to the second strand of the no-obligation problem, I indicated that the reason-generating force of the NJT would translate seamlessly into the obligation- producing force of NJT*. However one might wonder how the arbitrator’s directive can shrink the class of actions Smith and Wesson can perform to discharge their more general 360 Raz (2009b) suggests, plausibly, that justice similarly underdetermines which laws we must adopt. 275 obligations to one another. In order to claim that they are now obligated to do as the arbitrator says, we must say something about how the deontic force of the obligations individuals have to one another transfers to the directives issued by practical authorities. The DYB Principle To see how the deontic force of our natural duties transfers to authoritative directives, we’ll have to take a short detour. Consider what, if anything, is wrong with my behavior in the following case, which does not involve authority: Bus I have promised to pick your children up from school. In order to make it there in time, I must take the number 10 bus at 2pm. So I stand in line and purchase a ticket, which happens to be the last one available. The person standing behind me in line, disappointed that he was not able to buy a ticket for this bus, proposes a bet. We’ll roll a six-sided die. If it lands even, he gets my ticket. If it lands odd, he’ll give me $5. How does my obligation to collect your children relate to my choice to take this bet? One possibility is that the two have only a potentially causal relation. If the die comes up even, then I will not be able to get on the bus and so I will ultimately fail to satisfy my obligation to you. If it comes up odd, I can satisfy my obligation while making a bit of money. Either way, whether I have gone against my obligation depends solely on whether I make it to the school to collect your children at the prescribed time. However I think this way of looking at things oversimplifies the nature and role of obligations as normative intermediaries between individuals. As I explained in Chapter 6, obligations are preemptive reasons to which I stand 276 to wrong others by responding inappropriately. Now, it seems to me that whether I was ultimately able to pick your children up or not, you would be quite right to be upset if you learned about my bet. You would be warranted in blaming and even reprimanding me for making it. One way of making sense of the appropriateness of these attitudes is to concede that by taking this sort of risk I respond inappropriately to the reasons I have to get on the bus, reasons I have in virtue of the promise I made to you. Since these reasons are ultimately grounded in your value as a person, and the disrespect I show for them is transitive, it follows that by taking the bet I show a lack of respect for that value, and therefore wrong you. This explains the appropriateness of the attitudes with which you would seem warranted in addressing me were I to take the bet, for wrongings are the normative phenomena which justify these sorts of attitudes. I believe that the underlying principle at work in this case is the same as the one at work in Neighbor. I shall call it the Do Your Best or DYB principle: DYB If B has an obligation to A to φ, then, if B’s failing to ψ (or ψ-ing) would, from a reasonable person’s perspective, appreciably decrease the likelihood that B will satisfy his obligation, and ψ-ing (or refraining from ψ-ing) is not unreasonable, then B’s obligation to A includes ψ-ing (or refraining from ψ- ing). 277 To put the point more simply (though slightly less accurately): An obligation to φ includes doing one’s best, within reason, to satisfy it. 361 According to the DYB principle, I violate my obligation (and thus wrong you) by gambling with my bus ticket because I see (or should see) that by taking the gamble I appreciably decrease the likelihood that I make it to the school in time to collect your children, and because forgoing the expected gain of taking that gamble is not unreasonable. Even if I win the gamble and make it over to the school in time to collect your children, I have failed to do all that my obligation to you required of me. 362 My obligation is, in this important respect, inclusive. 363 Let’s now reconsider Neighbor, which diverges from the Bus only in that it involves authority. Smith and Wesson struggle to identify and respond to the obligations they have to 361 For brevity, I will continue to say that what B’s obligation demands is that B do his best within reason. However, there are two things to note. First, the principle, properly stated, does not actually require that the agent do her best, but only that she does not take appreciable risks without good reason. Second, it is important to keep in mind that the relevant perspective is not the obligated agent’s, but the reasonable person’s. 362 Note that the problem here need not be cashed out in terms of intentions. While I believe that intentions and other mental operations are in fact relevant to obligations we have to one another— such that we can wrong others by merely intending to do certain things or even by entertaining excluded reasons (Chapter 4)—that is a stronger claim than the one I’m making, which is that obligations involve restrictions on how we may act which extend beyond the narrow performance they most clearly require. 363 This idea of inclusivity extends to other moral requirements as well, which sometimes overlap with the inclusive nature of obligation I’m endorsing. Thomas Christiano (2004, 270) argues for a weak publicity constraint on actions required by justice. There is something “defective,” he says, about doing all that justice narrowly requires in a manner that cannot at least in principle be verified. (I discuss this constraint in greater detail below, in connection with political autonomy.) Thus, on Christiano’s view, justice narrowly requires that I pay you $10 when I owe it to you, but also that that I pay it in a way that you could verify if you tried. Justice-regarding constraints on my actions therefore include performances (or abstentions) which are not narrowly associated with the debt per se. 278 one another. They lack certain descriptive information that the arbitrator has superior resources to discover, and harbor biases which likely prevent them from fairly evaluating the normative facts available to them. Further, Wesson, acting in good faith, has some reason to believe that Smith will change her behavior to take advantage of Wesson’s compliance. And the same is of course true for Smith. Wesson might rearrange his work schedule so that he goes to bed only after 10pm, by which time Smith is supposed to have her music off. Since Smith has some reason to believe that Wesson will make changes of this sort, she has good reason to believe that playing her music late at night will be even more of a nuisance for Wesson than it would have been before. Wesson’s well-being, in other words, now depends more on Smith’s cooperation than it did before, and visa versa. Smith and Wesson’s option to abide by the arbitrator’s rules is, then, like my option to refrain from gambling with my bus ticket. The rules are based on a more descriptively informed view that is at the same time less vulnerable to bias, and their very existence will reasonably modify each person’s behavior in a way that depends at least somewhat on the other’s cooperation. Failing to abide by the rules will therefore appreciably decrease the likelihood that Smith and Wesson satisfy their obligations to one another. And so, because it is not unreasonable to abide by those rules, which are modest, the DYB principle entails that Smith and Wesson’s obligations to one another include obeying the arbitrator’s directives. If the DYB principle captures the moral phenomenon I believe it does, then it bridges the normative gap between the obligations we put into NJT* and the obligations we get out, and therefore describes the way in which practical authorities satisfying NJT* are a 279 genuine source of obligations in at least some cases. The standard argument therefore answers the validity question. 364 7.1.2 The Applicability Question That a person who satisfies NJT* can in fact generate obligations for others is a moral contention which, if true, invites us to rethink the nature of many of our social relationships. But the moral claim only answers the first explanatory hurdle. To be an adequate theory of political obligation, we must also show that states satisfy NJT* with respect to their constituents. This involves showing two things: first, that citizens have preexisting obligations which states can plausibly best help them satisfy, and second, that states plausibly do help in this respect. If either of these cannot be shown, or they cannot be shown to be sufficiently general, then even if the standard argument provides an acceptable answer to the validity question, it will not apply to political contexts, and so it will fail to answer the applicability question and thus to provide an answer to the problem of political obligation. Natural duties The conception of wrongings I endorsed in the last chapter presupposes that each of us has value, and that our value ultimately justifies the application to other individuals of more and less general practical principles. Making a serious case for any particular interpretation of the 364 I should note that the NJT faces a similar problem of explaining the gap between the normative force of the reasons we put into it and that of the directives on the other end. The gap goes unnoticed, however, because the principle bridging it is an obvious tenet of practical reasoning: Perform that action which will most likely accord with what there is most reason to do. The preemptive force of authoritative directives on the NJT depends on this sort of bridging principle to connect the incoming and outgoing reasons, just as NJT* depends on DYB to bridge the gap between incoming and outgoing obligations. 280 contents of these principles and their relation to our value is a substantial project in normative ethical theory, and far beyond what I can undertake here. Still, to verify the two claims the truth of which is required for the success of my argument in this chapter, we must have some sense of the principles this value justifies. The fact that someone has value is not sufficient in itself to determine which practical reasons her value engenders for others. Fish, dogs, and people are all valuable, but we have reason to treat them in different ways, reason which we can explain only by attending to the sort or form of value they have. Which features of persons as such one takes to be relevant in making this assessment will depend to some extent on the moral theory to which he subscribes. A utilitarian will be most interested in our capacity for well-being, while a Kantian will concentrate on our capacity to set our own ends and to act on them. Because I am interested in providing a theory of political obligation which is, as far as possible, impartial between competing theories of human value, I will focus on those capacities for which I believe any plausible theory would have to account, and more importantly, the duties to which they give rise. I propose to focus on the general capacities that follow from Rawls’ conception of persons as beings with equal moral status. 365 I take it, however, that there is nothing in Rawls’ conception that is peculiar to his theory, as these are at the same time the general features of persons which make them members of the “we” to which we owe certain performances in virtue of the fact that our failure to perform them would be wrong according to some “set of principles for the general regulation of behaviour that no one 365 Rawls (2003, 99). 281 could reasonably reject.” 366 Rawls’ aim was of course narrower than ours. While he wanted to determine those features of persons which made them the proper objects of the principles of justice, our aim is to identify the form of human value which engenders the deontic reasons each of us has to treat others in the many ways we have reason to. Our aim therefore includes those reasons to which considerations of justice give rise, but also any deontic reasons human value creates for us through intermediary principles unconnected to justice. Because I believe that the capacities which explain why we are, in Rawls’ view, the proper subjects of the principles of justice are the same as those which explain the particular sort of value we have, I’ll concentrate on the same two capacities Rawls did: rationality and reasonability. 367 Rationality in the intended sense involves our ability to plan and pursue our own conception of the good. Reasonability involves our desire and capacity to apply and act on principles of justice. Rationality entails, on my interpretation, an interest in guiding our own lives in a manner we see fit (what I will call autonomy), as well as an interest in being assisted by others when we are in need (mutual aid) and in being free from harm (safety). Reasonability entails, among other things, an interest in participating in the creation of rules which govern the behavior of individuals within the society in which one lives (what I will call political autonomy). 368 366 Scanlon (1998, 153). 367 Rawls (2003, §77); also see Freeman (2008, Chapter 7). 368 The duties corresponding to these interests line up roughly with those listed by Rawls (2003, 94). Two notes. First, the interests to which these capacities give rise plausibly overlap. Second, in claiming that these interests follow from the form of value that makes us the sort of beings we are, I do not mean to suggest that their satisfaction is necessarily an intrinsic good. It is enough for my argument that our value determines the reasons others have to act. See Freeman (2008, 289). 282 I have provided no argument for the capacities I claim express the unique form of value persons have as such. Nor have I provided an argument for the interests I claim follow from them. I agree with Rawls when he says that these claims “depend solely on certain natural attributes the presence of which can be ascertained by natural reason pursuing common sense methods of inquiry.” 369 I should emphasize, however, that nothing in my argument depends on these precise capacities or interests. What matters for our purposes is that we have duties of roughly the sort I’ve described, which correlate with the interests I’ve claimed rationality and reasonability bear out—in directing our own lives, in being helped when we are in need, in being protected from harm, in contributing to the creation of rules which govern our lives together, and so on. I believe that these duties or some quite like them must follow in one form or another from any plausible normative theory. There are of course other duties which plausibly follow from the interests we have as a result of these capacities. To complete the standard argument, however, I’ll focus only on those duties which correspond with our interests in autonomy, mutual aid, and safety. Our duty to acknowledge and respect others’ political autonomy is at the heart of the proceduralist argument I propose in the next section. The Service Conception Applied The final step in the standard argument is to show that states can in fact meet the conditions of NJT* by helping citizens satisfy their duties and obligations to one another. Together with the DYB principle, it follows from any such instances that states have some range of 369 Rawls (2003, 442 fn.30). 283 authority over their citizens, and that those of us living in states have, at least on some occasions, political obligations. 370 So far as this is right, the authority theory answers the applicability question at least in certain areas of the political realm. How far this authority extends is a concern we’ll take up in the next section. Because we cannot explore the intricacies of every duty we plausibly have to one another, I will explain how the theory works in a particular case from which we can then generalize. I’ll focus on our duty not to put others at risk, which corresponds to the interest each of us has in safety. What we must explain is how following legal directives which relate to the safety of others helps us act in accordance with our duty not to put them at risk. A few preliminary points are in order. First, many laws which relate to the safety of the majority nonetheless apply to a minority. Regulations specifying which tests must be run on pharmaceuticals before they enter the market are directed specifically at pharmaceutical producers, for instance. Though our normative apparatus applies equally (and perhaps even more obviously) to requirements of this sort, we are ultimately interested in the scope of the law’s normative reach—in the service it provides as an authority for the rest of us—and will therefore focus on laws with a broader application. Second, it should be recognized that built into all general duties is an implicit condition on reasonability. Every time I leave my house I pose some risk to others; yet I am surely not duty bound to live my life indoors. My duty not to put others at risk is, in fact, a duty not to put them at unreasonable risk. What is reasonable is a function not just of costs 370 Of course, this will not be true for all states. Some considerations can undermine its authority even if in cases in which it satisfies NJT* to some limited extent. This is a complicated question which I will set aside. When I speak of states generally I mean states that meet the conditions of acceptability which apply to any potential authority. 284 and benefits associated with an action, but also of the available alternatives. Whether it is reasonable to expect my employee to work on a dangerous machine, for example, will depend in part on whether a safer machine is available. This is particularly relevant for laws which set limits on a behavior or the use of a product from a series of sequential alternatives, such as, for instance, 65mph as the speed limit on California highways. Consider two possible reactions I might have to this limit. I might agree that driving too fast puts others at unreasonable risk, but disagree that at 70mph the risk I pose is significantly greater than at 65mph. Alternatively, I might believe that 65mph is already too fast, and that only by driving at 60mph can I satisfy my duty not to put others at an unreasonable risk. In the first case, the question is not whether driving at 70mph is safer—everyone agrees that driving faster adds at least some additional risk—but whether the additional cost of driving 5mph slower than I otherwise could without putting people at a significantly greater risk is reasonable in the circumstances. Most people will concede that this sort of question is best answered by agencies, such as those created by governments, which employ groups of experts to analyze the circumstances affecting traffic accidents and other risks. So far as driving at 65mph puts people at sufficiently less risk than driving at 70mph, which the agency is there to determine, the DYB principle applies, and so my obligation not to put others at unreasonable risk transfers to the content of the more specific directive issued by the government in line with agency recommendations. 371 While before the 371 Two points. First, this is a pro tanto obligation with a limited exclusionary scope, and so certain considerations will outweigh it. If I have to get my dying relative to the hospital, and the street is empty, it is highly plausible that the speed limit will not morally bind me (though, of course, it will legally bind me). Second, in setting the speed limit the government will inevitably account for considerations beyond safety, such as noise and road longevity, which they employ other agencies to study. In these circumstances the legal speed limit will reflect the considered judgment of the 285 content of my general duty not to put others at unreasonable risk was colored by my own elementary understanding of the risks associated with driving at this or that speed, and so might have been satisfied by driving at any number speeds, I am now bound to drive no faster than 65mph. Now consider the other case. If I rightly believe that my more general duty can only be satisfied by driving slower than the speed the law specifies—that is, if the law gets it wrong in that it doesn’t go as far as it should—then it is true that the law may not apply to me. But in this case the sequential nature of the limit renders the point moot. 372 The obligation I take myself to have will consist in doing even more than the law specifies, and so the reasons for which we are concerned about political obligation will be satisfied. What we want to know, again, is whether individuals must comply with the laws governing our political community. A feature of laws which involve partially arbitrary points chosen from various options along a sequence is that each of us will be obligated to do at least as much as the law says, whether the law goes far enough or not. In most cases, however, the question is not whether the law has selected the correct point on some sequence, leaving one with the binary option of doing more or doing less, but whether there are ways of satisfying one’s general obligation to others which are inconsistent with what the law prescribes. We saw earlier, for instance, that before the arbitrator got involved, Smith and Wesson could have discharged their obligations to one another in ways legislature as informed by all of these groups. However these more complicated circumstances only strengthen the standard argument. 372 Certainly in practice, though in theory it may create a gap in the argument’s application. I discuss this below when we consider objections. 286 that would not have satisfied the particular rules the arbitrator went on to prescribe. It was not a matter of doing less or more but of doing something altogether different. The most compelling answer to the applicability question will therefore be one which demonstrates the application of NJT* to this sort of case, such as, for example: Waste There is a legal regulation against the improper disposal of hazardous waste, such as motor oil. The law prescribes that such materials be collected in a government provided container and placed on the street so that they may be collected by a hazardous waste disposal service. For what reason am I bound to act as this law requires? Let’s take the questions in steps. Each of us has a duty not to put others at unreasonable risk. This duty is grounded in our mutual interest in being protected from others’ actions, which is explained by the form of value we express through our two moral powers. Now, what constitutes a reasonable risk depends largely on what resources and information we have available. I am not unlike most other citizens, I presume, in knowing next to nothing about the possible effects hazardous waste can have on humans and other living things if disposed of improperly. I don’t even know which types of waste are hazardous. Some can safely be run into the sewer but cannot safely be buried in my backyard. Some can be burned. Some are even more dangerous if burned than if buried, but more dangerous still if they are run into the sewer. The state satisfies NJT* in at least two complementary and overlapping ways. First, it has agencies which specialize in hazardous waste disposal. They determine which chemicals are hazardous, in what way they are hazardous, and how they can be most safely disposed of. Second, the state coordinates efforts to make superior disposal techniques possible. Without 287 the coordinated efforts of many people, leaving my hazardous waste in a container in front of my house would be an irresponsible way to dispose of it. First, there would be no weekly collection service, and without one, it wouldn’t be a way of disposing of it at all. Second, even if a non-state agency provided such a service, complete with its own disposal requirements, it would not likely be as effective as the public service unless it coordinated with other nearby providers. 373 Without coordination, each neighborhood’s method would likely be safer than having each resident go it alone but still more risky than a public service. 374 Local services would therefore have fewer resources to do the research necessary to determine which materials ought to be disposed of in which ways, and a more difficult time coordinating with other neighborhood services to ensure that their efforts were not in conflict. Either fact would likely be sufficient to ground the government’s practical authority in this matter, though together they raise the likelihood substantially. I now have a choice. I can dispose of my hazardous waste as the law prescribes, or I can ignore the law and dispose of it as I see fit. Employing a critical line we discussed earlier in connection with Bus, one might think that whether I do everything that is required of me depends, not on whether I do as the law says, but on whether I dispose of the waste in a manner that puts others at no unreasonable risk. If, as it turns out, burying my motor oil in my backyard is perfectly safe, then burying it there is, on this view, one way of satisfying my obligation. As I suggested, however, our obligations demand more than first meets the eye. 373 If it was as effective, or more effective, then there is good reason to think that the government would allow or even encourage it. More importantly, however, to know whether this alternative service was more effective, I would have to already know quite a bit about hazardous waste, which by hypothesis, I don’t. 374 For instance, two services might unknowingly dump their waste products into a landfill which is sufficiently large for either of them individually, but not large enough for both. 288 They are preemptive reasons to which I stand to wrong others by responding inappropriately. We need to ask, then, whether I have such a reason to comply with the law in Waste. Bus illuminates the sort of wrong I can potentially commit in Waste. Recall that in that example I had an obligation to pick your children up from school, but was presented with the option of gambling with my necessary means of reaching them in time. Taking the gamble seemed inappropriate and blameworthy, and would provide you with a special standing to complain even in the event that I won the bet and made it to the school in time to collect your children. This was explained by the related facts (i) that my behavior would be an inappropriate response to the reasons my promise to you engendered, reasons which trace back (transitively) through the norms governing promises to the value you have as a person, which ultimately grounds them, and (ii) that such behavior therefore constitutes a wronging, the normative phenomenon which provides one with a special standing to complain. We used the DYB principle to explain more formally the way in which obligations make such wrongings possible, the idea being that, in virtue of the role they serve as intermediaries between individuals, obligations include, beyond some required performance, that we do our best within reason to discharge them. The foregoing considerations apply equally in Waste. My obligation to others not to dispose of my hazardous waste improperly is, according to the DYB principle, not satisfied simply because I happen to bury it in a manner which is, as it turns out, no more dangerous than the state’s method. By burying it when I could have disposed of it as the state required, I take what is from a reasonable point of view an appreciable risk, and for two separate, individually sufficient reasons: first, because the state likely knows better than I do, and 289 second, because even if this would have been the best option before the state got involved, its coordinative efforts enable cooperative strategies which open new and better possibilities which I can take advantage of only by doing things as it specifies. By disobeying the law I therefore fail to respond appropriately to the reasons engendered by others’ value—reasons which, even before the law was in play, justified my duty not to put them at unreasonable risk. The law in this case is authoritative because following its directives constitutes my best chance of satisfying my preexisting obligation, which, according to the DYB principle, explains why the normative force of my duty transfers to the content of the legal directive. Because obeying the law is not in this case unreasonable, I have an obligation to obey it. 7.1.3 Objections I discuss the advantages of the standard argument towards the end of this chapter. First, however, we must deal with two important objections. The first is that the argument has justificatory gaps. The second is that it does not account for the special claims democracies have on their citizens—or the proceduralist objection. The Gappiness Problem As we discussed in Chapter 1, political obligations are supposed to be both general (or, better, universal), meaning that they bind most (or all) of the state’s inhabitants, and comprehensive, meaning that they bind with respect to all laws. The standard argument fails to meet these criteria in two ways. First, some people will systematically outperform the state when it comes to identifying how best to satisfy certain classes of obligations on the whole, and will recognize this. For these people the political authority will be no authority at all, at least with 290 respect to the realms of action in question, for it will not perform a service for them. By doing as the state says they will systematically do worse than they would were they to go it alone. Thus, NJT* does not apply and the standard argument cannot get a foothold. Such people constitute gaps in the seamless web of authority the law purports to string. 375 With respect to these realms of action, they are sovereigns unto themselves. Beyond this class of people there is another class which cannot systematically outperform the state’s directives, but which will occasionally recognize that the law has made a mistake (Chapter 5). In these instances NJT* cannot explain how their obligations are grounded, for with respect to these laws, people, and occasions, the authority again fails to provide a service. These cases diverge from those in the first class because they do not undermine the state’s authority in any particular realm of action. (Authorities can, as we discussed previously, make mistakes.) Yet because they are clear mistakes they remain instances in which the law lacks obligatory force. Since political obligations are nevertheless thought to bind even in cases in which the law is wrong and citizens know it is wrong, this again creates justificatory gaps in the theory. 376 375 A similar point is made in Christiano (2004, 267). 376 Two points. First, this second sort of gappiness is one manifestation of a much more general problem facing what might be called two-level theories—those, like rule utilitarianism, which prescribe action at one “level” in light of justificatory considerations provided at another. Second, some people might think that a law which is known to be wrong does not actually bind. If this is right, then this objection does not apply to the standard argument. Still, most theorists believe that a proper conception of authority provides for the binding force of directives even in cases in which subjects know an error has been made. For this reason I try to account for it. 291 The Proceduralist Objection The second primary challenge to the standard argument is the proceduralist objection. 377 The plausible idea behind this objection is that procedure may matter to political authority in a way that the standard argument (and the authority theory more generally) simply cannot capture. Unfortunately, it is not always clear from the literature exactly what the objection is supposed to be. Before moving forward, it will be important to be clear about the problem. Here are two potential interpretations (adapted to address our updated conceptions of the NJT and moral obligation), drawn from Scott Hershovitz’ recent essay: (i) The standard argument wrongly rules out the possibility that considerations besides NJT*, and in particular considerations about procedure, can justify authority. 378 (ii) The standard argument is not sufficient to justify authority. Political institutions are genuinely authoritative only if they are democratic, and no part of NJT* ensures that the purported authority is democratic. 379 377 The objection has been presented in different ways in Christiano (2004); Hershovitz (2011); Marmor (2005); Shapiro (2002). 378 “Democracies, on Raz’s view, have a leg up on other political arrangements only to the extent they produce better results. We might think, however, that well-constructed democracies can claim legitimacy on other grounds.” Hershovitz (2011, 3). 379 “[T]he normal justification thesis is blind to the possibility that an authority’s legitimacy may depend as much on how it makes decisions as on what those decisions are.” Hershovitz (2011, 3). 292 I have already given my argument against (ii). Its advocate either wrongly rejects the normal conception of authority (Chapter 5), misunderstands the nature of obligations (Chapters 4 and 6), or overlooks the fact that deontic normativity transmits from natural duties to authoritative directives in the manner explained by the DYB principle (7.1.2). Moreover, I am skeptical of the implications of this version of the objection. Hershovitz claims, for instance, that “the capacity to help subjects conform to reason does not ground political authority because political authorities have other roles to play.” 380 This way of putting things strikes me as doubly problematic. First, it suggests that a double success may be a failure— that a government cannot be a genuine authority in virtue of its ability to help its subjects comply with their reasons because it might serve other important functions as well. My view, which I will continue to develop in the next section, is that authority can be grounded in a variety of ways, some of which overlap. A double success is, on the authority theory, simply a double success. Second, this version of the objection suggests that no role outside that a position, institution, or practice is most centrally designed to play can explain its legitimacy. Yet this seems false too. A person can consistently maintain, for instance, both that the primary role of promising is coordinative and that the obligations promises generate have nothing to do with their coordinative function. Similarly, a democratic regime might, in virtue of being democratic, play a variety of important roles none of which ultimately grounds its authority. The first version of the objection is also problematic. Its primary flaw is that it falsely implies that the authority theory rejects other potential justifications of authority. As 380 Hershovitz (2011, 1). 293 Hershovitz says, “[o]ur theory of authority should not rule out the possibility that legitimacy depends on process, in addition to (or instead of) substance.” 381 To be fair, Hershovitz’ objection is aimed at Raz, who indeed seems to equivocate on this point. 382 If Raz means to claim that the NJT offers an exhaustive justification—one that rules out the possibility that the legitimacy of a governing body might in some cases depend on other things, such as process—then showing that alternative considerations can ground authority would be one way of undermining it. But the claim that NJT* offers the only possible justification for authority can and should be rejected. The most that can be claimed, and all that I have claimed in previous chapters, is that NJT* is a sufficient justification in particular contexts. It seems plainly false, in fact, that it might be the only one. Consent, for instance, would in many cases be sufficient to bind citizens to states, whether or not the government helped them on the whole to act in accordance with their reasons. I will therefore set aside those versions of the proceduralist objection which, like (i), aim to reject the authority theory on the basis that the justification of authority it provides is not the only possible one. But while we should be open to the possibility that NJT* is not a necessary part of any justificatory story, we should closely scrutinize the claim that “well-constructed democracies can claim legitimacy on other grounds.” 383 Certainly there are other considerations, some of them procedural, which may contribute to an authority’s legitimacy. But this does not entail that satisfying something like the NJT is not necessary. Procedure 381 Hershovitz (2011, 3) 382 Raz’s original presentation of the NJT (1988, 53) explicitly states that “[t]his way of justifying a claim that someone has legitimate authority…is not the only one.” Yet later he suggests that he intends (and has always intended) his general justificatory account to be “exhaustive” (2009a, 132). 383 Hershovitz (2011, 3). Here I take legitimacy to include authority. 294 alone cannot legitimate a (nonvoluntary) institution that produces legal norms which are consistently at odds with reasons citizens have independently. There is no doubt that in some instances the value of making decisions in accordance with a particular procedure outweighs the costs of the suboptimal decisions it might produce. But this is true only within some permissible range. In most circumstances, including the political realm, outcomes matter a great deal. 384 Notwithstanding the shortcomings of these various interpretations of the objection, it seems clear that the general proceduralist point is on target. I believe the best expression of this point is that “democracies have special claims on the allegiance of their citizens” which cannot be accounted for on the standard picture. 385 Instead of challenging the idea that NJT* is necessary or sufficient for practical authority, the objection, on this interpretation, squarely addresses our argument’s scope. The problem is not that the standard argument claims to provide the only possible justification, or that it has no part to play in justifying authorities, it is that NJT* applies to wise and benevolent tyrants just the same as it does to their democratic counterparts. 386 This is, I believe, the most illuminating version of the proceduralist objection: 384 A point with which Hershovitz (2011, 4) ultimately agrees. 385 Hershovitz (2011, 3). 386 Christiano (2004, 278) objects to the NJT on the grounds that it legitimates “ferociously unjust regimes.” However I find the two sorts of cases on which he bases his argument problematic. In the first a person can, by taking a job the Nazi regime has assigned him, better comply with the reasons he has to save Jews. This is possible only because, as a slow and comparatively incompetent worker, he will not well serve the reasons for which the regime assigned him the job. In the second case, a person can do better by doing as the purported authorities say but merely because they “implicitly threaten morally terrible consequences if their subjects do not comply with commands that require them to participate in evil activities” (279). The first case relies on a particular sort of deviance— parallel to Gettier cases in epistemology—in which the authorities have made a mistake (from their 295 The proceduralist objection. The standard argument’s narrow focus on a particular sort of instrumental value fails to account for a feature of democratic political communities which explains one important dimension of the political obligation binding their members. 7.2 The Proceduralist Argument To be clear, both the gappiness problem and the proceduralist objection hold even if the standard argument succeeds in telling an important part of the justificatory story. Even if many of us have an obligation to obey large swaths of law, and for the reason that a genuine practical authority promulgated them, important things are left unaccounted for. In this section I shall argue that a single solution, constructed from the resources available as part of the theoretical apparatus we’ve built so far, addresses both problems. The general idea is that we can better satisfy one of our more important natural duties to others by obeying democratic laws, and that, because the feature of these laws which explains this fact is content-independent, we have this special reason to obey all laws which are created through an acceptably democratic process. Since all laws share this feature, our special obligation to point of view) which happens ultimately to align with the subject’s independent reasons. At most this shows that an additional premise is needed to specify the sort of relation which must hold between authoritative directives and a subject’s independent reasons. (Raz’s Dependence Thesis (1988, 42) will not do because it is not a condition on the status of legitimate authorities, but a prescriptive principle intended to guide the directives they issue.) The second case is nothing more than a threat, and as such the consequences of failing to comply should not be included in the class of independent reasons we use to determine whether the authority is legitimate. Some may feel that excluding this class of cases is ad hoc; however it is perfectly in line with our conception of authorities as people or institutions which perform a service. A purported authority which on the whole causes people to act worse than they would were there no authority at all, or a different sort of authority, as is the case in both of Christiano’s examples, would perform no service at all, and should not therefore count as an authority on our conception. 296 obey them satisfies the proceduralist objection while at the same time ensuring that the service conception meets our constraints on generality and comprehensiveness. By addressing the proceduralist objection, we therefore plug the gaps in the theory. 7.2.1 Political Autonomy In outlining the standard argument we focused on duties which correspond to our interests in autonomy, mutual aid, and safety. 387 These interests are outcome oriented—that is, primarily concerned with certain outcomes being achieved—and can therefore distinguish forms of government only by reference to their instrumental value. 388 Since, as it is commonly thought, much of the value of democratic governance is attributable to its process rather than its outcome, to account for it we will have to identify a significant, process-oriented interest people have which is served specially by democracies. One such interest—political autonomy—I mentioned earlier as one that follows plausibly from our capacities as individuals with equal moral status. So far as this is true, one answer to the proceduralist objection and the gappiness problem involves explaining the special way in which democratic authorities help citizens satisfy their otherwise underdetermined duties to respect others’ process-oriented interest in political autonomy. 387 Autonomy here refers to personal autonomy, and thus the conditions under which I can direct my own life. It includes, for instance, an interest in not having my freedom restricted by being placed in jail without cause. The potentially intrinsic value of fair processes in making decisions about how society will be structured is captured by political autonomy, which I discuss presently. 388 I believe this partially explains why, as Thomas Christiano (2004, 277 fn.10) has observed, important contemporary discussions of authority such as Joseph Raz’s The Morality of Freedom and Leslie Green’s The Authority of the State, both of which employ a version of the standard argument, do not discuss democracy as a possible source of authority. 297 The value of both political autonomy and personal autonomy is rooted in the idea that people should direct their own lives—that they should, in some respect, participate in the “ideal of self creation.” 389 But the two concepts are distinct. I am personally autonomous when my judgments and decisions determine the shape my life takes. I am politically autonomous when I have the capacity and opportunity to help create the rules we use to govern our lives together, when my judgments and decisions determine the shape our life takes, collectively. 390 There are three contextual constraints that must be met before our duty to respect others’ interests in political autonomy becomes relevant. These are constraints on expression, effectiveness, and publicity. First, the governing institution must ensure that each person has an available means by which to express her views about the rules governing public life. Voting is of course the most familiar mechanism, however this too must be supplemented by further conditions. A one-candidate election, for instance, lacks the features of a genuine choice situation which make a person’s vote an acceptable expression of her views. Second, political autonomy requires that the means of appropriate expression the government ensures are, at least in principle, efficacious. Protest is a clear means of expressing one’s views about legal norms, but because there is no institutional connection between protest 389 Raz (1988, 369). 390 In this respect the concept of political autonomy is empty outside of social contexts, and tied up with other normative social concepts. It may be, for instance, that we cannot coherently conceive of political autonomy without at least an implicit conception of what is fair, or of our “status as political equals.” (See Gutmann and Thompson (1996, 18).) One concern, then, might be that this is not the sort of independent interest which grounds duties that authorities are meant to help us satisfy. However I think this concern reflects a misunderstanding of the features of our independent interests and duties which are relevant on the authority theory. What matters is not that we have them independently of particular social contexts, such as a political community, but that we have them independently of the authoritative directives issued within such a community. 298 and law—no assurance that protest will ever effect legal change—it is not sufficient on its own to meet this condition. 391 One lacks political autonomy if one has the capacity to express oneself politically but not the power to make a political difference, just as one lacks personal autonomy if one has the capacity to deliberate and decide but not the power to form intentions that reflect those decisions. Third, our interest in political autonomy includes a weak constraint on publicity. 392 It is not enough that we express our views about the rules of public life in a manner that actually affects them; we must also have some reasonable way of knowing that it does. The reason for this final constraint is simply that a regime which did not meet it would leave its citizens in the dark about the meaning of their own political participation. Much of political philosophy is dedicated to the study of these conditions, and particularly to questions of whether and how democratic systems satisfy them. Our interest, however, is in whether the authority theory can explain why legal directives issued in democracies bind citizens in ways other regimes cannot. I will not, therefore, provide an argument for “the moral necessity of the state,” 393 nor will I defend democratic states as such. Whether democracies have some special claim to legitimacy is not our concern here. I will instead simply assume that democratic institutions exist and that those we are concerned with (such as the U.S.) meet the three contextual constraints I’ve identified—that they (i) 391 Christiano (2004, 276) makes what I take to be a similar point when he says our right to participate in the process of political adjudication must include a right to participate in the process of decision. 392 Christiano (2004). 393 Christiano (2004, 281). Recall that the moral necessity of the state was an important component of the arguments provided by both Klosko and Wellman. 299 provide citizens with an appropriately expressive (ii) means of changing legal rules which (iii) allow them to see that the first two conditions are in fact met. So far as the democracies we have in mind meet these constraints, they provide a mechanism through which their citizens can exercise their process-oriented interest in political autonomy, and therefore provide the circumstances in which the question of how each of us can best satisfy our duty to respect that interest becomes relevant. 7.2.2 The Argument How does a democratic state help citizens act in accordance with their duty to respect others’ political autonomy? One final premise is necessary to make the argument. I call it the fact of functioning political communities, and it includes three related observations. First, almost everyone alive today was born into, and lives now as a part of, some political community. Second, each political community has a functioning political institution. Third, the authority which each such institution claims does not permit competing institutions of political authority. 394 Some states are so tyrannical that their inhabitants are unable to partake in meaningful political expression. Respecting another person’s political autonomy in these circumstances is like respecting her personal autonomy when she’s chained to the ground: the circumstances preclude the possibility. In political communities of this sort a number of political questions arise. Is the regime legitimate? Does its lack of legitimacy entail that it lacks authority as well? Do its citizens have a duty to rebel? Because our concern is with 394 This last point is emphasized in many places by Raz and others. 300 western liberal democracies, the questions we want to ask are different. What we want to know is whether these democratic institutions—which (by hypothesis) provide an acceptable outlet for political expression while prohibiting competing forms of political authority, and thus, eliminating other mechanisms of political expression which meet the three contextual constraints above—provide special binding reasons to obey the law. The proceduralist argument is intended to explain these binding reasons in terms laid out in the standard argument, to explain how democratic states help us act in accordance with those of our duties which correspond to our fellow citizens’ process-oriented interest in political autonomy, just as the standard argument explains the service the state provides relative to the outcome-oriented interests others have. The premises we’ve added to account for the complexities of democratic states are sufficiently detailed to warrant a formal statement of the proceduralist argument: (1) Each of us has an interest in political autonomy sufficiently important to ground a duty for others to respect it (assumption). (2) The fact of functioning political communities entails a. that each of us is unavoidably a part of a political community, b. that the community of which we are a part has a functioning political institution, specifically, in the cases we’re interested in, a democracy, 395 and 395 Democracies function in many different ways. Again, my purpose here is neither to provide a theory of democracy nor to defend any particular democratic regime, but merely to point out that many of us live in a democratic state which is for that reason commonly thought to provide an acceptable mechanism through which to exercise our political autonomy. 301 c. that the authority each such democracy claims does not permit competing institutions of political authority. (3) By creating law through democratic processes, the democracies in which we find ourselves meet the three contextual constraints—on expression, effectiveness, and publicity—which must be met before we can satisfy our duties regarding political autonomy (assumption). (4) Democratically created law is therefore the one and only extant means through which citizens of these states can exercise their interest in political autonomy (from 2c and 3). (5) One’s duty to respect others’ political autonomy includes doing one’s best to satisfy it (the DYB principle). a. This entails that one would wrong others by doing other than one’s (reasonable) best, for in such a case one would disrespect the distinctive form of value they have as persons (from the conception of obligation developed in Chapter 6). (6) While there are multiple ways in which one could genuinely attempt to meet one’s duty to respect others’ interest in political autonomy, the only way others can actually exercise this interest is through the democratic process that exists in their state (from 4), and so respecting the outcome of that process is the only way of respecting it. (7) If there is only one way to do something then it is the best way (assumption). (8) Obedience to law in a democracy is the best way its inhabitants have to respect others’ political autonomy (from 6 and 7). (9) We therefore have an obligation to obey the law (from 5 and 8). 302 Let me now try to put the same argument less formally. We begin by making two assumptions. The first is that others have an interest in political autonomy sufficiently important to ground a duty for us to respect it. The second is that the democracies in which we find ourselves meet those conditions which must be met by a political institution for its members to acceptably exercise their political autonomy. The most common way of doing this is of course through the franchise, which is expressive, efficacious, and (as a collective practice) public. As the fact of functioning political communities suggests, however, states do not allow competing authoritative institutions. Thus, well-functioning democracies exhibit two important characteristics: They allow us to exercise our political autonomy through a collective decision-making process that meets minimum standards of acceptability, and they prevent the emergence of any competing process which meets these same standards. The decision each of us is therefore faced with is whether to (i) respect the one and only means others have to exercise their political autonomy, or (ii) try to satisfy our duty to respect others’ political autonomy in another way. Because our obligations to others to respect their political autonomy includes doing our best to satisfy them (the DYB principle), I’ve suggested that we would (other things equal) wrong others by trying to meet our duty in a manner that is less reliable than obeying the law. This is true for the same reason that I should not bury my hazardous waste in my backyard, even if I believe that it would be perfectly safe. While it might turn out that I did not put my neighbors at any greater risk than I would have otherwise, the fact that, among other things, the government is positioned to know better makes my decision an unnecessary risk, one which indicates a particular sort of disrespect for my neighbors’ value. Similarly, because the political procedures constitutive 303 of a democracy are the only means citizens have of exercising their political autonomy, obeying the law is a more reliable way to respect this exercise than any other. We are therefore bound to obey it. 396 7.2.3 Answering the Proceduralist and Gappiness Objections I said before that the proceduralist argument provides an answer to both the proceduralist and gappiness objections. According to the proceduralist objection, the standard argument cannot account for the “special claims” democracies have on their members. 397 The gappiness problem highlights the fact that the argument does not apply to citizens who either know better than the government in certain realms of action or recognize that a law is mistaken in a particular case. The argument’s normative scope is therefore neither comprehensive nor sufficiently general. Let’s begin with the proceduralist objection. To address the concern that the standard argument cannot account for the special ways in which democracies bind their citizens, the proceduralist approach is to layer another argument on top of the standard one. Structurally, the two arguments are the same—they both use DYB as a bridging principle to explain how the deontic force of our natural duties transfers to legal directives. Rather than focus on the duties we have relating to autonomy, mutual aid, and safety, however, as the standard argument does, the proceduralist argument focuses on our duty to respect others’ 396 The claim I’ve made regarding the relation between this function of democratic authorities and our two moral powers echoes a passage in T.H. Green’s lectures on political obligation: “The value then of the institutions of civil life lies in their operation as giving reality to these capacities of will and reason, and enable them to be really exercised” (1927, 32). 397 Hershovitz (2011, 3). 304 political autonomy. This difference is important, for, as I indicated already, our interest in political autonomy has as its object a particular sort of process. What is distinctive about a democratically enacted law, according to the argument, is not its content, but the fact that it results from the exercise of our interest in political autonomy. Since participation in the democratic process is the only means anyone in a functioning democracy has available to genuinely exercise this interest, our duty, along with the DYB principle, entails that we wrong others by disobeying the law, for disobedience shows (other things equal) that we are insufficiently concerned with others’ interest in exercising their political autonomy. It follows that we have an obligation to obey the law for the reasons I explained already. The key point to notice about the proceduralist argument is that political institutions which fail to meet the contextual constraints we discussed earlier cannot satisfy their members’ interest in political autonomy. The upshot is that political institutions that do meet them—which is to say, for the most part, sufficiently just democracies—bind their citizens in ways that other political institutions cannot. These institutions are unique in that they create the context in which members of the community can respect the political autonomy of their compatriots—something that democracies, but not dictatorships, enable. This answers the proceduralist objection, for it explains the special way in which democratically enacted laws bind us. 398 398 Importantly, the proceduralist argument does not entail that citizens of non-democratic states do not have obligations to obey the law. Whether they do will depend on a variety of considerations, one being whether there are any realms in which the state satisfies NJT* with respect to its purported subjects, or in other words, whether the standard argument applies in those states. This is one of the advantages I discuss below. 305 Let me now explain how the argument answers the gappiness problem. Two of the theses we established previously are central. The first is once again that our interest in political autonomy is process-oriented. The second is, as I explained in Chapter 6, that duties and obligations are partially constituted by wrongings, or failures to respond appropriately to value. From these two theses we see that all three types of gaps we previously discussed close. First, we close the comprehensiveness gap because every law is just as much a result of the democratic process as any other; thus, no law is more or less an exercise of our political autonomy than any other. If we have an obligation to obey any law because it constitutes an exercise of political autonomy, then we therefore have such an obligation with respect to all laws. 399 Similar reasoning applies to the second gap: generality. Since the duty to respect others’ interest in political autonomy applies to everyone, and everyone has but one best (reasonable) way of exercising that duty, every person is bound to obey every law which is a genuine exercise of that interest. Finally, because, from the perspective of the proceduralist argument, what it important about each law is that it constitutes the exercise of political autonomy, the question of whether any particular subject of the law “knows better” is irrelevant. Unlike the standard argument, the proceduralist argument does not consider the consequences of any law’s content. There is therefore no way in which a person could, with respect to others’ interest in political autonomy, (i) systematically know a better way to act than the one specified by the directive, or (ii) recognize that the law made a “mistake.” 400 399 It might be objected that an enacted law which I do not personally support cannot be thought to be an exercise of my political autonomy. I address this objection below. 400 Again, this is not completely accurate. See below. 306 Because doing as the law specifies will constitute his best effort at satisfying his duty, the deontic force of that duty transmits to the law. 7.2.4 Objections The most important concern I would like to address is that the proceduralist argument presents nothing but an empty shell of the service conception—a principle which is trivially true because it assumes what it aims to prove. Before I do, however, I would like quickly to dispatch three concerns one might have either as a result of misunderstanding the argument, or because I’ve not sufficiently explained the theory’s moral underpinnings. The first concern one might have is with my claim that all laws in a sufficiently just democracy bind universally. So far as this is taken to mean that all laws bind conclusively, it is simply a misunderstanding of my view. My claim is not that we have an all-things-considered obligation to do as the law says, but rather that, for the reasons given in the proceduralist argument, everyone within the relevant community has a pro tanto obligation to obey those laws which are produced in accordance with the legislative norms of their sufficiently just democratic state. In fact, it is a virtue of the proceduralist objection, I believe, that this may be a relatively weak obligation. Unlike those produced in accordance with the standard argument, the force of which vary across the various duties we serve by obeying particular laws, obligations which are owed in virtue of our duty to respect others’ political autonomy are both uniform across laws as well as the only obligations we have to obey legal directives which are morally binding for no other reason. The proceduralist argument explains, in other words, why we have obligations to obey even those trivial, obnoxious laws that serve no clear purpose. We are obligated to obey them because the reasons we have to obey have the 307 appropriate structure—they are, while weak, preemptive—and because we would wrong others, however slightly, by disregarding them. It seems to me that we would, with respect to many such laws, expect our obligations, if they exist at all, to be relatively weak, and this is precisely what the proceduralist argument delivers. But such a broad defense of the obligation generating potential of democratic regimes might raise a second concern: If our obligations in democratic states are both general and comprehensive, then does it follow that no such state can produce a law immoral enough to dissolve our obligation to obey it? Shouldn’t an acceptable theory reject the possibility that immoral laws bind? One possible response would be to say that the existence of such a law entails that the democracy which produced it is not sufficiently just to meet the contextual constraints employed by the argument, and that obedience to law does not therefore constitute one’s best means of respecting others’ political autonomy. We might simply claim, then, that the argument does not apply in these cases. While concise, however, I’m afraid that this response is implausible. Reasonably just democracies sometimes produce immoral laws, and in such cases those laws would seem to have as much claim to the intrinsic justifiability of democratic procedure as any other. A second and better response, I think, is to point out that this objection poses a genuine problem only for a person who maintains that one cannot have an obligation to do something immoral. However this is not something I maintain. It is a virtue of the conception of obligation we put forward in Chapters 4 and 6 that we can have obligations to do immoral things. Consider an unexceptional example. I have agreed to give the keynote speech at a women’s healthcare event this Saturday. However, on Friday I learn that the purpose of the event is to raise money for ultrasound machines women will be required to 308 use before receiving an abortion. I think that this practice is immoral, and I believe (rightly, let’s assume) that I have an obligation to local women not to aid this movement. Still, I made a promise to speak at the event, and it is my own fault that I did not pay closer attention to the materials the event organizer sent me weeks ago, which laid out the event’s purpose. It seems to me that in this case I have conflicting obligations, one of which is to do something immoral, and that determining what I ought to do will involve weighing them against one another. My purpose here is not to defend this position but to point out that accepting it makes this second concern regarding the proceduralist argument irrelevant. 401 Finally, one might be concerned about the proceduralist argument’s dependency on democratic means of exercising one’s interest in political autonomy. It cannot be said, the objection goes, that laws with which I disagree are nonetheless an expression of my political autonomy. But if they are not then it cannot be that others owe it to me to obey them. Thus, according to the critic, they cannot have political obligations on this ground. Two responses are again available. The first and stronger response is to deny the objection’s first premise. Indeed, much democratic theory is seriously confused if any particular law is justified, if it is at all, only by the will of those who endorse it. Surely it follows from any satisfactory defense of democracy that whichever considerations regarding a citizen’s will justify a law of which she approves also justify laws of which she does not. To explain which considerations these are, we would need to develop a theory of democracy itself. This is far more than we can possibly do here. This first response must therefore be a conditional one: If the authority 401 This view of obligations has an elegant explanation for how we should weigh such obligations against one another: Obligations to perform immoral actions simply cannot exclude the moral reasons we have not to do them. The implication is that there are sometimes bad reasons not to do bad things. 309 theory cannot answer this concern, then democracy itself remains perilously gappy. 402 However a second, weaker response is also available. This is to accept the premises while denying that the conclusion follows from them. Even if it is true that laws with which I disagree are not exercises of my political autonomy, someone in my political community—and often, as democracies are structured, the majority—will endorse them. In such cases detractors will owe it to the law’s supporters to obey. At first this response may seem suspect, for it restricts the class of people to whom each of us owes obedience to an ever- changing subset of the population—those who endorse the law in question. On reflection, however, it is not clear why this should be a problem. First, it is of course possible to owe an obligation to a member of group without knowing which member it is. Second, there is nothing in such a scenario to limit the obligation’s force. Thus, even in the unlikely case that the will of a law’s detractor in no way contributes to that law’s binding force, our obligation to obey it for the reasons described in the proceduralist argument still applies. An Empty Argument? I said earlier that the standard and the proceduralist argument have the same structure. In both cases the deontic force of the general duties members of society owe to one another transfers to the more specific directives issued by the authority, and for the reason that, in each case, following the authority’s directive constitutes the subject’s best (reasonable) effort 402 It is possible, however, that some dysfunction within a normally well-functioning democracy could undermine one’s claim to de jure authority within the institution, and thus to that person’s role in helping citizens exercise their political autonomy. Election fraud is a plausible example. In such a case any laws enacted by the fraudulent legislator would likely fail to constitute an exercise of public political autonomy, and would therefore lack binding force. This is, however, a different sort of case than the one under discussion, and one that poses no problem for (but is in fact an explanatory virtue of) the theory. 310 at satisfying his duty. According to this account, authorities have instrumental value: they help us act in accordance with a special class of our reasons. The two arguments diverge only with respect to the type of service authorities provide. On the standard argument, the service depends on the directive’s content, which guides the agent in the realm of action to which that content relates. I have an obligation to dispose of my hazardous waste as the law directs, and the state has authority over me regarding such disposal, because I will likely better protect other people from harm if I dispose of it as directed rather than as I think I should independently. If such directives are consistently mistaken, then the state will not have authority, and so they will not be able to bind their inhabitants in the way the standard argument claims. The sort of service the authority provides according to the proceduralist argument is slightly different. It relates, not to a directive’s content, but to the fact that the directive constitutes a genuine exercise of other citizens’ political autonomy. Thus, according to our MP approach, a genuinely democratic law straps me with two distinct obligations—two ways in which I stand to wrong others by disobeying. Returning to Waste, the first follows from the fact that I would likely put others in more danger than is reasonable by not doing my best to dispose of my hazardous waste safely, which entails disposing of it as directed. The second follows from the fact that, by disobeying the law, I do not do my best within reason to respect my compatriots’ political autonomy. In both cases by failing to obey the law I fail (other things equal) to respond properly to the reasons to which their value gives rise, and thus wrong them. The different types of service on which these arguments turn are necessary to explain the full scope of the layered obligations I’ve claimed we have. The first explains (as I 311 elaborate below) the range of reactive attitudes which seem appropriate for different sorts of lawbreaking. The second explains, as I already argued, the general (indeed, universal) and comprehensive scope of our duty to obey, as well as the special sorts of obligations to which democracies properly lay claim. The emptiness concern is that the proceduralist argument provides this additional coverage only by abandoning the service conception’s instrumental approach, that the service conception …blunts the proceduralist objection only at the expense of making the normal justification thesis nearly empty. To decide whether the thesis is satisfied, it is no longer enough to determine whether compliance with authoritative directives will allow a subject to conform better to the reasons she has on the matter the directive regards…We must also consider, in this case and in every other, whether a further criterion of legitimacy (not specified by the normal justification thesis) gives rise to a duty for the subject to comply, irrespective of the instrumental value of the authority’s directives. So understood the normal justification thesis ceases to be a competitor with other theories of legitimacy; instead, it subsumes any theory of legitimacy that turns out to be true. 403 This concern boils down to a question about the true direction of the proposed explanation. According to the authority theory, the fact that a person helps others in the manner described in the proceduralist argument explains that person’s authority. The government provides a service, the argument claims, by helping us act in accordance with our duty to respect others’ political autonomy. The emptiness objection complains that the explanation is in fact the other way around: Citizens satisfy their obligations by following directives because someone in authority issued those directives. And if this is the case, then of course 403 Hershovitz (2011, 5). Hershovitz’ criticism is aimed at a pair of responses Raz (2009a) gives to the proceduralist objection. The proceduralist argument diverges from Raz’s responses. Nonetheless, the two lines have enough in common that it is not unreasonable to think that the emptiness objection might still apply. It is therefore worth addressing. 312 the theory hasn’t explained anything. For on this picture the authority must already be an authority before any of the purportedly justificatory work can be done. The authority therefore satisfies NJT*, but only trivially. Whether or not this objection identifies a flaw in Raz’s view, I don’t believe it gains traction with the proceduralist argument as I’ve presented it. The question to ask would seem to be whether the argument can explain why we have an obligation to obey the law without presupposing that the democratic state of which we are members is a legitimate authority. If it can, then it is not empty in the manner described by the objection, for it is not the legislator’s authority, presupposed to be grounded in some other way, which explains subjects’ obligation. Let’s suppose that it is an open question whether our democratic state has authority. We accept two propositions: first, that we have a duty to others to respect their political autonomy (an assumption regarding the form and significance of the value we have as persons), and second, that if the law helps us act in accordance with the obligations we have independently of the potential authority of the lawgiver, then that lawgiver is an authority (NJT*). The question is therefore whether the state can help us in this way. For the reasons provided in the proceduralist argument, I believe it can, and does. But—and this is the important point—this is not because the lawgiver in the argument is already presumed to have authority. The argument turns on the fact (i) that obeying the law is the best of various possible ways in which we can try to respect others’ political autonomy, and (ii) that our obligations include doing our best (within reason) to satisfy them. It is for these facts that the deontic force of our underdetermined duty transfers to the more specific action the law directs us to perform. Since the deontic force transfers, we have an obligation to obey. And 313 because what it is to be a practical authority just is to be in a position to create obligations for others to behave in particular ways merely by expressing an intention to that effect (the general conception), the government in this example is a practical authority. Clearly, then, this is something which follows from the argument, not something we presuppose to make the case. 404 Robust Content Independence Let me address one final point in relation to the emptiness concern which might explain its initial plausibility. Something unique about the directives issued by a person whose authority depends on the proceduralist argument (whose authority is, as I will say, proceduralist) is that they are robustly content-independent. Recall from Chapter 5 that a content-independent reason is one that does not depend entirely on the independent reasons a person has. While that definition is correct, it neglects to mention that there are different ways in which a reason can be content-independent. The defense of the NJT I offered in Chapter 5 entails that an authority satisfying it can give you reasons to do things that you don’t have independent reasons to do, but only because on the whole you satisfy your independent reasons better by doing as the authority says than you would otherwise. So while no particular directive’s force depends directly on the compatibility between it and the reasons you have independently, those reasons play a key justificatory role nonetheless. 404 What this reveals is that the argument shows two separate things simultaneously, neither of which must be shown before the other: first, that we have political obligations, and second, that democratic states have at least some degree of authority in virtue of the service they perform, as per the proceduralist argument. 314 But the connection between one’s independent reasons and the directives authorities issue is less direct in some cases. For example, if I promise to do whatever you say with respect to housework while I’m staying with you, and my promise creates a genuine obligation to that effect, then you have the power to produce binding directives for me with respect to housework. If I have no independent reason to sweep the floor, and you tell me to sweep it, the deontic force of my promise transfers to the directive, leaving me with an obligation. The same would be true, however, if you told me not to sweep the floor, or to do something else entirely. And it would remain true no matter what you told me. The interesting feature of the sort of authority you have in this case is that it is not true that on the whole my independent reasons must line up with the content of your directives to some sufficient extent. Your track record, so to speak, doesn’t matter. In this case your directives are robustly content-independent. The directives issued by a person whose authority traces back to the standard argument (whose authority is, as I will say, standard) are content-independent, but not robustly so. As we saw, the government’s directives would carry no obligatory force in Waste if I knew that I had no independent reasons to bury my hazardous waste as they specified, for in such a case complying with their directive would not constitute my best reasonable effort to satisfy my duty to others. The government’s track record doesn’t matter because, as we discussed in Chapter 5, their expertise will certainly not help me in this particular case. Not so for proceduralist authority. Any law which is produced by a well-functioning democratic institution is for that reason binding on its subjects—at least to a limited extent— irrespective of the independent reasons they have. Whatever those reasons are, the exercise of political autonomy remains relevant. 315 It may be this robust form of content independence that raises critics’ suspicions about the proceduralist argument. There is, I admit, an apparent tension in the view. For how could it be that a person has this sort of authority—authority which appears totally disconnected from the independent reasons subjects have to act in particular ways—in virtue of the fact that by acting in accordance with her directives her subjects will satisfy obligations they would likely not otherwise satisfy as reliably? Does this not imply that the authority’s directives, whose content independence depends on their independence from the subject’s reasons, depends on those reasons after all? One way to bring out the error in this objection is to point out that authorities justified through the standard argument can occasionally produce robustly content- independent directives as well, though by a different mechanism than proceduralist authorities. This occurs, for instance, when such an authority is best positioned to solve coordination problems or prisoner’s dilemma cases. When I have a coordination problem, I have no conclusive, independent reasons to φ or ψ unless others are going to act in one way or the other. Since they will act in concert only if the authority tells them to, if the authority says to φ, I will then have a reason to φ. But I had no independent reason to φ before the authority issued her directive, and I would have had a reason to do the opposite had the authority said to, for it was her directive which gave me the reason. 405 In this case, then, there 405 One might say that this directive is not robustly content-independent because before the directive was issued I did have an independent reason to do as other people did if they acted in concert. Appealing to conditional reasons, however, seems only to confuse the issue. What is unique about directives of proceduralist authorities is that they give us reasons which are disconnected from our independent reasons at a local level. At a higher level, all directives which are genuinely normative will somehow connect up with our independent reasons. Without such a connection no directive would be normative. 316 is no risk of my knowing better than she does, for what she directs me to do could not be wrong. The question, then, is whether this sort of case is covered by the standard argument, or whether that argument loses its justificatory force in these instances unless it presupposes the sort of authority it claims to explain. I see no reason to think that the argument fails in these cases. For what remains true is that the authority is an authority because she helps me act in accordance with the (underdetermined) obligations I had prior to her directive. Since this is what justifies her authority, it is difficult to see why the fact that her directives are robustly content-independent is relevant. And if it is not, then it is hard to see why that should be relevant to our analysis of proceduralist authority. 7.3 The Subjectivist Argument The subjectivist argument is the last argument we’ll consider. I should note upfront that it is substantially different from the last two arguments we’ve considered. Whereas the standard and the proceduralist arguments depend on facts about the service the state actually provides to citizens and the beliefs a citizen ought reasonably to have about the likely consequences of that service being provided, the subjectivist argument looks at citizens’ actual beliefs about these services, irrespective of whether they are true or even justified. The argument begins with the conception of obligation we developed in Chapter 6: A (moral) obligation is a requirement—that is, a preemptive reason an inappropriate response to which imposes a cost—which is unique from other requirements in that its cost is a wronging. 317 We also know from Chapter 6 that a wronging is an act which is disrespectful in that it constitutes a failure to respond appropriately to another’s value, or to the reasons to which that person’s value ultimately gives rise. We saw, too, how the deontic force of our duties to others can transfer to the directives issued by authorities in accordance with the DYB principle, such that we wrong others by failing to obey the law. The point of the standard and proceduralist arguments was to show that the government plausibly provided the service necessary to qualify it as an authority on the service conception, and thus, that by disobeying the law we wrong others. But what if the state is not an authority according to NJT*? Can we still conceivably be bound to obey the law? I think that in some cases, surprisingly, we can. This is what the subjectivist argument is designed to show. Let’s concentrate first on wrongings. A wronging is, as we’ve seen, the substantive quality that sets moral obligations apart from other requirements. To wrong someone is to disrespect her by responding inappropriately to her value, either directly, or indirectly by failing to see or respond to the various action-guiding principles her value justifies. But what is it about this sort of disrespect that is significant to us? As we discussed already, it is not that it’s harmful. Things might turn out better for me if you decide not to keep your promise to meet me. In some cases, I might not even know that I’ve been stood up. Still, you wrong me by deciding to brush me off simply because you felt like doing something else, and I have standing to rebuke you for it. The disrespect, and thus the wrong, is independent of the any harm that follows from it. What’s troubling about your decision to brush me off is what it implies about your valuation of me. You treat me as though I have less value, or a lesser sort of value, than I actually do. 318 This grounds political obligations from the standard and proceduralist perspective because by disobeying the law I fail to respond to the reasons your value in fact creates. But if what is troublesome about wrongings is what they entail about one person’s valuation of another, must the victim’s value in fact engender reasons that the wrong-doer fails to notice or disregards? The central proposal on which the subjectivist argument turns is that the disrespect constitutive of wrongings can occur even in deviant cases in which a person has false beliefs about what reasons there are. I argued earlier that obligations are inclusive—that they include performing certain actions failure to perform would appreciably increase the risk that the obligated party will fail to perform the more narrow action specified in the obligation’s content. This idea of inclusivity explained your reactive attitudes in Bus, the case in which I chose to gamble with my only means of picking your children up at the school. To simplify that case, imagine that the person who offered me the gamble instead simply offers to buy my ticket outright for $10—double the price I paid for it. Surely I would wrong you by selling the ticket for a $5 profit if I would wrong you by gambling with it for the same, for by doing so I guarantee that I will not collect your children as I promised. But now let us imagine something new: that in fact I have made a scheduling error. I am supposed to pick your children up tomorrow, not today. What follows? Well, if I am supposed to pick them up tomorrow, then, though I do not know it, there is no reason for me to go to the school today. Thus, there is no reason to which I could fail to respond appropriately by selling my bus ticket. It seems, therefore, that I cannot wrong you by selling it, for a wronging just is such a failure. Call this case Bus*. 319 The problem is that I seem to behave at least as badly towards you in Bus* as I do in Bus. Surely you would be warranted in being upset with me in either case. It seems that I would owe you an apology for selling the ticket I believed I needed to make it to the school by 2pm, even if my belief was false. How can we reconcile this fact with the fact that there was no reason for me not to sell the bus ticket? 7.3.1 Having: Reasons and Obligations Having Reasons Let’s draw a distinction between there being a reason, and one’s having a reason. 406 Consider a simple case: Dinner My brother and I both love meatballs, which my mother is making tonight for dinner. However only I know this. There is a sense in which my brother and I both have a reason to go to dinner: dinner is meatballs. But there is another sense in which only I have a reason to go to dinner, since only I know that dinner is meatballs. This second sense seems, indeed, to be a sense of having a reason. First, those fully informed about the situation can expect that, unless there is something else to do, I, but not my brother, will go to dinner. Second, I, but not my brother, can fairly be criticized for not going to dinner if there is not something else to do. And third, if I do go to dinner, we will not think that I went for no reason at all. We will think, rightly, 406 See Schroeder (2008), from which I borrow heavily in the following two paragraphs. Subjective obligations, which I discuss below, are not part of Schroeder’s work. 320 that I went for the reason that dinner was meatballs. These three facts are, as Mark Schroeder says, earmarks of having a reason in this second sense. 407 Now let’s expand the case: Dinner* My sister hates meatballs, but loves lasagna. She believes, falsely, that my mother is making lasagna tonight instead of meatballs. It seems that the same three earmarks of having a reason that apply to me in Dinner apply equally to my sister in Dinner*. First, given that she has the false belief that she does, we can expect that she’ll go to dinner. Second, we can fairly criticize her, given what she believes, for not going if there is not something else to do. And third, if she goes, we can say, rightly, that she went for a reason. Still, there is an important difference between my sister and me. The reason I have to go to dinner is also a reason for me to go to dinner. Dinner is, after all, meatballs. But since dinner is not lasagna, there is no analogous fact that is a reason for my sister to go to dinner. She appears to have a reason to go to dinner which is not a reason for her to go to dinner. Having Obligations The distinction between there being and one’s having a reason sheds light on the connection between Bus and Bus*. Bus is analogous to Dinner in that the obligation there is for me to pick your children up is also an obligation I have. My situation in Bus*, however, seems like my 407 Schroeder (2008, 61). 321 sister’s situation in Dinner*: I have an obligation to pick your children up at 2pm which does not correspond to an obligation there is for me to pick them up at that time. The earmarks of having an obligation are similar to those of having a reason. What divergence there is results from the sort of reasons obligations are. First, just as a fully informed bystander would expect my sister to go to dinner if she had nothing else to do, he would expect me to pick your children up so long as I did not have something else to do which was not excluded by the obligation I believed I had. The difference is in how we can expect a person to weigh the reasons he or she has, given the sort of reasons they are. We can expect my sister to go to dinner because, with nothing else to do, there is nothing to go on the opposite side of the “scale” on which she weighs the purported fact that dinner is lasagna. Since reasons are connected to motivation, we can expect her to go to dinner when the scale tips in that direction. Similarly, with no nonexcluded reasons on the scale, the purported fact that I promised to collect your children tips the balance in favor of picking them up, and so it is reasonable to expect that I will. It seems true, of course, that conclusive prudential reasons exert a more reliable kind of motivational force on people than conclusive obligations do. However the point is that a fully informed bystander’s credence that I will pick your children up from school should be the same in both Bus and Bus*. We can, in other words, expect the same behavior from someone who has an obligation that there isn’t as we can from someone who has an obligation that there is. Second, just as my sister can rightly be criticized in Dinner*, I can be rightly criticized in Bus*. The two relevant differences again turn on the sorts of reasons in play. First, as we would expect, the conditions under which I would be open to criticism in Bus* are broader than those under which my sister would be open to criticism in Dinner*. For while my sister 322 can only be criticized for not going to dinner if her desire for lasagna outweighs whatever reasons she has to do something else, I can be criticized for not collecting your children even if my first order reasons, taken together, weigh in favor of doing something else. Because obligations exclude first order reasons, I will be free from criticism for choosing not to collect your children only if I have sufficient nonexcluded reason to do something else. So, whereas coming upon opera tickets for tonight would free my sister from any criticism she might have otherwise been subject to for skipping dinner, coming upon similar tickets for an afternoon performance would not free me from criticism, for my desire to do something besides pick your children up is excluded from consideration by the obligation I have. 408 Finally, and most importantly, neither my sister nor I acted for no reason at all. If you asked my sister why she intended to go to dinner, she would tell you that dinner was lasagna. If you asked me why I was catching the 10 bus to the school, I would tell you that I have an obligation to pick your children up there at 2pm. So far as she has a reason, it seems that I have an obligation. A question we might ask at this point is just what these reasons are supposed to be. If lasagna is not for dinner, and the promise I made was to pick your children up tomorrow, then what could be the reason that both my sister and I have? This is an important question about practical reason, but nonetheless one that I will not try to answer here. (Though I’ll say that I’m sympathetic to the pragmatic answer Schroeder defends. 409 ) For our purposes, 408 And in fact the exclusionary nature of obligations opens me to criticism for a variety of other actions besides failing to perform the narrow action specified by the obligation. However these details are not important here. 409 Schroeder (2008). 323 the two key takeaways are, first, that there is indeed a sense of having a reason which distinguishes my sister and me in Dinner and Dinner* from my brother in Dinner, and which at the same time applies to my circumstances in Bus*, and second, that, save for the exclusionary nature of obligations, the earmarks of having an obligation are the same as those of having a reason. 7.3.2 Validity The upshot of the distinction between there being and one’s having an obligation is that it helps us see what is objectionable about selling my bus ticket despite the fact that there is no reason for me to take the bus today. While there may not be a reason for me to take the bus, I have a reason to take it, and that reason is of the special preemptive sort promises generate. The central insight is that I stand to wrong you in precisely the same way whether there is such an obligation or whether I merely have one. That is why you have the same standing to rebuke me and why I have the same reason to apologize to you in Bus and Bus*. In both cases your decision says something objectionable about your valuation of me. Whether your beliefs about what you are obligated to do are true or false, by selling your bus ticket you act as though I have less value, or a different sort of value, than I actually do. Because this is ultimately what troubles us about the sort of disrespect constitutive of wrongings, not just in this case, but in any case, and because wrongings are the substantive feature which sets obligations apart from other requirements, we arrive at the following principle: Principle of deontic convergence. A person’s having an obligation is of similar if not identical normative significance to his being under one. 324 To this we can add a principle specifying the conditions under which one has an obligation: Subjectivism about obligations. I have an obligation to φ just in case I have a belief the content of which, were it true, would constitute a necessary, normative component of a practical derivation the conclusion of which is that there is an obligation for me to φ. 410 To put the point more simply (though less accurately): I have an obligation to φ just in case I have a belief such that, were it true, there would be an obligation for me to φ. 411 410 By practical derivation I mean an instance of practical reasoning. Philosophers disagree about the nature of the conclusions of such derivations. One popular view is that such a conclusion is simply the formation of an intention. Another is that it is a deontic statement of some sort, such as that “I ought to (am obligated to) φ.” The thesis I’ve endorsed comports with the second conception; however, it could be restated along the lines of the first. (On different conceptions and their plausibility, see Raz (1978).) There is also a question about what premises fill in the rest of the derivation. Are they propositions that are in fact true? Propositions that a reasonable agent would believe to be true? Propositions the agent actually does believe to be true? I believe they are propositions of the second and third kind: those the agent believes to be true or would so believe were he reasonable. 411 This thesis is less accurate because it doesn’t specify the nature of the belief one must have or its role in the derivation. To take the second point first, say that I made a promise to meet you at noon, then forgot about it. Say also that I believe that grass is blue. Were my belief true then there would be an obligation for me to meet you, since, because I did in fact promise, there is such an obligation anyway. But that means, according to the definition, that this is an obligation that I have. And of course, this seems wrong. The more complicated statement of the principle avoids this result by specifying that the conclusion (“That I am obligated to φ”) actually follows from the believed proposition. Returning now to the first point, the nature of the belief seems important. Say that I promise to clock you in at work if you’re running late, but then forget about this promise. Later, I come to believe that you’re running late. If that’s true, then there is an obligation for me to clock you in. However it seems wrong to say that I have this obligation given that I don’t remember the event which is, as I’ll say, normatively responsible for it. To put the point in terms that Jonathan Dancy has, the belief, if true, must bear a favoring relation to the conclusion, rather than providing an enabling condition. See Dancy (2004b). 325 Together these principles allow us to answer the validity question. They say, first, that a person has an obligation to do something when she has certain sorts of beliefs, whether those beliefs are true or not, and second, that having an obligation of this sort is, given the substantive element which sets obligations apart from other requirements, of the same normative significance as there being such an obligation. Bus* then bears this out. 7.3.3 Applicability The final question we must ask is whether we can have subjectivist political obligations. There are actually two questions here, one normative and the other empirical. The normative question concerns the connection between the content of one’s belief and the obligation one allegedly has. Must one see what would follow from the truth of her belief in order to have the obligation that follows from it? The descriptive question is then simply whether people have beliefs which have these entailments. The subjectivist response to the normative question is that it is sufficient for a person to have an obligation to obey law L that she believe (i) that her state has the authority to make law, (ii) that L was created through the legislative process prescribed by that state’s political institution, and (iii) that L applies to her. The response to the descriptive questions is that, while the question is empirical and we do not have empirical data to answer it, it is highly plausible that most people have these three beliefs. 326 The Normative Question There are two reasons to be worried about our answer to the normative question. The first concerns the nature of practical inferences generally. The second concerns disagreement about the contents of normative concepts. Earlier I suggested that my sister, who loves lasagna, has a reason to go to dinner because she believes that dinner is lasagna. But this may have been an oversimplification. If she has a broken jaw, and so cannot eat solid foods, then the fact that dinner is lasagna doesn’t seem to be a reason she has to go to dinner. In such a case we won’t expect her to go, or criticize her for not going, and if she did go, it would not make sense if she said that she went because dinner was lasagna. For dinner’s being lasagna to be a reason my sister has to go to dinner, then, it seems that she must also believe, or it must be reasonable for her to believe, that if she goes she will be able to eat the lasagna. Similarly, there are beliefs it seems she must lack to have this reason. For instance, my sister hates sardines. So if my mother put sardines in the lasagna, my sister won’t eat it. Thus, for the fact that dinner is lasagna to be a reason my sister has to go to dinner, she must lack the belief that the lasagna has sardines. We see, then, that there are some beliefs my sister must have (or that it must be reasonable for her to have) and others she must lack (or that it must be reasonable for her to lack) in order for the purported fact that dinner is lasagna to count as a reason she has to go. Which beliefs these are is a difficult question, and not one I will try to answer here. 412 It 412 The concern is that some practical inferences from partial reasons to practical conclusions could not ever be complete, and thus, that there cannot be truly complete reasons. (The concern can be restated without the term “complete reasons.” See next footnote.) To my knowledge Raz, whose language of complete reasons I’ve borrowed, never answers this concern. Thus he says: “The fact that p is a complete reason to φ for a person x if, and only if, either (a) necessarily, for any person y 327 seems to me, however, that nothing integral to my argument rides on it. Inferring complete reasons from partial reasons is in fact something that we do extraordinarily well. If you ask what reason my sister has to go to dinner, then I can simply say that dinner is lasagna. Were you a stranger, I would potentially have to add the fact that my sister loves lasagna. Everyone, however, will assume that implicit in the complete reason, of which the fact that dinner is lasagna is supposed to be a part, is the fact that, if she goes, my sister will be able to eat the lasagna, and that, given the fact that she will not eat anything with sardines, the lasagna does not have sardines. 413 I will not provide a complete list of the beliefs a citizen must have, beyond the three I listed above, for the partial reasons she has to amount to a complete practical reason of the obligation sort. Such a list may be (practically) impossible to generate. I trust, however, that which reasons those are will be intuitively clear. The second worry concerns the fact that people don’t agree about what follows from certain normative concepts. We might agree that Andy has authority over Bill, but disagree about just what his authority entails. Part of the problem can be solved simply by being more who understands both the statement that p and the statement that x φ’s, if y believes that p he believes that there is a reason for x to φ, regardless of what other beliefs y has, or (b) R(φ)p,x [“the fact that p is a reason for x to φ”] entails R(φ)p,y which is a complete reason” (1999a, 24). (Also see Raz (1978, Chapter 1).) A problem with this way of understanding the notion of a complete reason is that y might have beliefs which conflict with those in his practical derivation. Since y can have these beliefs in any case, it follows that in no case will x have a complete reason to φ—viz., one that satisfies either (a) or (b). But this seems counterintuitive. 413 Of course, different philosophers have different theories about the nature of these other facts or beliefs. Raz (1999a, Chapter 1) and Broom (2004) suggest, as I understand them, that they are reasons which come together with other reasons (such as the one generated from the agent’s belief) to form a “complete reason” on which the partial reason that is their belief is derivative. Schroeder (2007, Chapter 2) argues that they are “background conditions” against which the propositional content of my belief is a full-fledged (rather than a partial) reason. Dancy (2004b) argues that these are “enabling conditions” which support the favoring relation held by what I earlier described as the proposition which is normatively responsible for the complete reason. Because my theory is not dependent on the superiority of any one of these views, I leave this debate aside. 328 specific about the concept we have in mind. If we agree that Andy’s authority is limited to a particular project Bill is working on, then we’ll likely agree that Andy’s authority doesn’t extend to whom Bill dates, though we may still disagree about whether it extends to what Bill wears to work. This problem is particularly acute for the subjectivist account, which holds that the obligations one has depend on what follows from the truth of one’s beliefs in the context. If Bill believes that Andy is an authority for him regarding matters involving this project, and Andy tells Bill that he needs to wear a suit to work, does Bill have an obligation to wear one? Does the answer depend on what Bill thinks his concept of authority includes, or simply what it does include? Three variations of the case seem relevant: (1) Bill believes that Andy has authority, and that Andy’s authority extends to his wardrobe. (2a) Bill believes that Andy has authority, but not that Andy’s authority extends to his wardrobe. However, were Bill’s belief about Andy’s authority true, then Andy’s authority would in fact extend to Bill’s wardrobe in the context. (2b) Bill believes that Andy has authority, but not that Andy’s authority extends to his wardrobe. If Bill’s belief about Andy’s authority were true, then Andy’s authority would not extend to Bill’s wardrobe in the context. The subjectivist answer is that Bill has an obligation in (1) and (2a), but not in (2b). What matters is not what Bill believes about the implications of his belief, but what his belief 329 would imply if true (in the context). 414 Similarly, what matters for the question of political obligation is not whether citizens believe that the state’s authority entails that its laws bind them morally, but whether, if their belief that the state has authority were true, its laws would bind them. Almost everyone, theorists and non-theorists alike, would agree that it would, for to have this power is, on just about everyone’s conception, precisely what it is to be an authority (the general conception). The Descriptive Question The final question is whether people actually believe the three propositions I mentioned. Those were, again, (i) that the state has the authority to make law, (ii) that the (purported) law in question was created through the legislative process prescribed by that state’s political institution, and (iii) that that law applies to them. To reject (i) in most cases amounts to being a philosophical anarchist, and I am skeptical that most non-philosophers are philosophical anarchists. To reject (ii) amounts either to accepting that the government as a whole is operating outside the bounds of the constitution, in which case (i) plausibly follows, or to accepting that a fraud at the highest levels of government has been perpetrated. The first possibility requires that one be familiar enough with the constitution to have such beliefs, which limits the class of people of whom this could be true to a tiny fraction of the total population, or that one is insane. The second possibility would require that one have inside information so groundbreaking that its publicity would shake the population’s trust in the government itself—or, again, that one is insane. Presumably, the total number of people 414 A similar entailment relation holds in language. If I say “All police are pigs” while under the impression that only uniformed police are police, I have nonetheless said something offensive about plain clothes police as well, even though that is not what I meant to say. 330 satisfying either of these conditions is minute. To reject (iii), finally, would simply be to misunderstand the law. For whether a particular law applies is not an evaluative question, as (i) is, or a question about the complex inner-workings of legislative practice and its relation to constitutional provisions, as (ii) is, but a simple question about the scope of law. Thus, no Californian believes that the liquor laws in Nevada apply to them while they are in California, just as they do not think that California’s liquor laws apply to them when they are visiting Nevada. People have, I believe, a general understanding of how laws apply. 7.4 Advantages of the Authority Theory The three arguments I canvassed above together constitute the authority theory. The standard argument is an updated version of the standard line taken by Razians, supplemented by an important bridging principle and applied to the question of political obligation. The proceduralist argument is then a unique application of that argument to a particular set of interests relevant to democratic states. The subjectivist argument provides, finally, an additional layer of justification for state authority, one that does not depend on the service states actually provide but rather on the nature of obligations and the concept of authority. I believe that this three-layer approach has a variety of advantages over competing theories of political obligation, particularly the two non-voluntary views I discussed in Part I. One of the clearest and most general of these advantages is that, by making authority relations a central feature of the view, the theory captures what most legal theorists today take to be an important feature of the “special relationship” that exists between states and citizens: the idea that laws are authoritative and their authoritative nature is to be understood 331 in terms of the special sorts of reasons citizens have to comply. 415 Theories of fair play are generally blind to these relations, focused, as they are, on the obligations we have not to ride free on others’ productive efforts. On views of this sort the rule of law is no different than a plan to build a drainage ditch—it merely specifies the rules through which cooperative action productively comes off. Samaritanism does somewhat better by making the government’s role a substantive element of the theory. But on this view its role amounts to fairly distributing coercive burdens—a function consistent with an outdated Austinean conception of legal institutions as mere power brokers. The authority theory, in contrast to these other two, is concerned first and foremost with the relationship between citizens and states. It is, according to the standard and proceduralist arguments, the service states provide to citizens which explains the obligations they have to one another to obey the law. Even on the subjectivist argument, it is the conceptual relationship between authorities and their subjects which partially explains why my belief in the authority of the state binds me (at least subjectively) to obey it. A natural duty theory at its core, the authority theory also enjoys an important explanatory advantage over the theory of fair play, which faces the difficult challenge of explaining, not just the application of a particular sort of preexisting normativity to the political context, but also its source. Both Simmons and Klosko tried to meet this challenge by espousing some form of acceptance condition—Simmons an actual/objective form, Klosko a hypothetical/subjective form. For both of them, however, the importance of acceptance lay in the fact that it could be connected to a source of normativity everyone 415 Simmons (1981, 34). 332 accepts: the subject’s informed, voluntary choice. We discussed the problems these theories faced in Chapter 2, and so I will not rehash them here. The point is that the authority theory avoids these problems by presupposing the sort of normativity necessary for the view. By starting with natural duties most people believe exist, we do not have to explain where the relevant normativity comes from, but only how it can be redirected towards obedience to law. Of course, samaritanism is like the authority theory in precisely this way. But because it neglects the authoritative role of governments, it cannot explain how this sort of redirection takes place. “[I]t is,” as Raz noted, “a melodramatic exaggeration to suppose that every breach of law endangers, by however small a degree, the survival of the government, or of law and order.” 416 For this reason it is, as we saw, difficult to see how the samaritan analysis answers the particularity concern. Why must our contribution to the state’s preservation take the form of obedience to law? And why must we pay particular attention to “our own” state? Wellman attempted to answer these questions by citing the discretion to which we unfairly help ourselves by contributing elsewhere, or by means other than obedience to law. Besides being inconsistent with the samaritan theory’s criticisms of the fair play argument, the notion that we take such discretion by choosing to meet our natural duties in ways we select ways presupposes the truth of the samaritan theory (Chapter 3). The authority theory, by contrast, grounds our political obligations in natural duties without giving up the tools necessary to answer the particularity concern. First, our own state is in a superior position, both epistemically and with respect to coordination, compared 416 Raz (1988, 102). 333 to other states, and it is therefore in the best position to help us satisfy our obligations to our compatriots. Since most people act as though the laws of their state apply to them, we can expect that they will act in accordance with those laws over others, just like we can expect the Californian to obey the laws of California rather than Nevada. Obeying the laws of another state would be counterproductive in most instances for this reason alone, even if it were true that, if everyone acted in accordance with the laws of a foreign state, everyone would be better served. The particularity concern is answered by the fact that people do better on the whole by doing as their state says than they would if they obeyed other states. 417 Second, because it is clear that the laws of my country are supposed to apply to me, and that my compatriots have an obvious interest in my recognizing them as such, it is only in my own state that I am positioned to respect others’ political autonomy. Why? Because it is only through the extant political process—one whose democratic process meets the contextual constraints on the genuine exercise of political autonomy—that they can exercise that interest. To venture an analogy: The language of democracy, as it happens to exist in our time, simply doesn’t have a vocabulary which allows citizens of one country to speak legislatively to citizens of another. 418 We express ourselves politically through “our” laws, which we believe apply to those within “our” country. As such, we could through no domestic action respect the political autonomy of members of foreign nations. It should be 417 What if this weren’t true? In that case the standard argument would fail to tie citizens to “their” states. However, I’m not convinced that this is not the result we want in any case in which one’s state was sufficiently substandard that another state (or institution of any sort) had produced laws which were clearly superior to one’s own state’s. 418 In fact, the language of statehood itself does not have such a vocabulary. The world does not today operate as part of a single cosmopolitan political institution. As this changes, the scope of our political obligations should be expected to change with it. This is, happily, something the authority theory permits. 334 clear, then, that this argument is grounded in existing practice, which is necessary to determine both the boundaries of our political expression and the domain of our potential interlocutors. Taking our political practice as it is, rather than as it could or should be, the best a person can do to respect a foreigner’s political autonomy is to respect his laws when she is in his country. (This is a secondary feature of political obligation that the authority theory is able to explain. 419 ) The third and final point follows directly from the last one. The structure of our political process shapes the way we actually think about nations and their authority. We do not have subjective obligations to obey foreign laws because we do not believe that the laws of foreign nations are authoritative for us (unless we happen to be visiting one of them). Most of us do, however, countenance the authority of the laws of our own country, admitting both that they are valid and apply to us. I admit that these three answers to the particularity concern—each of which corresponds to one of the three arguments making up the authority theory—reflect practical and contingent features of the position we, as citizens, happen to occupy relative to our compatriots and our governors. There is, I claim, no necessary or intrinsic relation between my state and me which explains why I should support it, or why that support should take the form of obedience to law. 420 That is not to say, of course, that there is no interesting sense in 419 The fact that we are expected to obey the laws of foreign countries when we are in those countries is something Locke thought a theory of political obligation should explain. Contemporary theorists no longer try to explain this phenomenon, however I think they should, and I count it as a virtue of the authority theory that it can. 420 Intrinsic relations of this sort would plausibly include, for instance, familial ties. I agree with the numerous critics who have pointed out that such relations do not provide a compelling model for political obligations. 335 which my state is indeed mine. But on the authority theory this is merely a descriptive fact and cuts no normative ice. This will no doubt dissatisfy some critics. Dworkin, for instance, claims that arguments of this sort fail to “capture the intimacy of the special duty” associated with political obligation. 421 I find this objection difficult to understand, however, and perhaps even backwards. In what way is an “intimate” duty distinct from other duties? The authority theory’s answer to the particularity concern is, at least with respect to the application of the duty, structurally similar to the explanation we would provide for why I have a duty to rescue you if I happen to find you drowning in a pond. The reason I am obligated, and the reason I am obligated to perform this action, is, I believe, practical and contingent in precisely the way the obligation I have to you to obey the law is practical and contingent. Does my duty to save you for this reason lack intimacy? Does our explanation for it somehow diminish the much closer relationship you and I have as members of the same moral community? I don’t see why it would. A duty is a duty, it seems, whatever explains it. I say that the suggestion is perhaps even backwards because it is the intimacy which theories such as Dworkin’s have attached to our associations that cast those theories into doubt. As Jonathan Wolff and others have argued, what associative theories seem to capture is the sense of identity members of groups feel, not the obligations those members have to one another. The authority theory’s subjectivist argument explains how these feelings can have normative significance. I do not owe you a political obligation (subjective or otherwise) because I feel like you and I are in fact a we, or because I feel like the state is in fact mine, as associative theories maintain. I owe it to you because I feel like the state is authoritative and 421 Dworkin (1986, 193). 336 because it is authoritative over us. The authority theory pairs my application of us with a normative concept that, if true, entails that there is an obligation for me to obey the law. Who counts as us? Radicals and jingoists aside, most of us deploy the political use of that term along strictly nationalist lines. I look at other Americans as being part of us in a way I do not look at Germans or Peruvians, and in doing so I look at Americans as being subject to the same laws I am. It seems to me that associative theories tell a plausible causal story but confuse it with a justification. We believe the things we do about the state’s authority because we meet the conditions associativists identify—we feel an important connection to those we see through the lens of our social practices. But it is our beliefs about authority and other normative concepts that explain the obligations we have to those people. In short, the authority theory explains the (practical and contingent) sense in which we are we, and so captures the important fact that we identify with our compatriots in a special way, without making the mistake of identifying this feeling with obligation itself. Another related advantage of the authority theory follows from the sorts of actions it counts as transgressions. What matters is not whether your disobedience to law gives you an unfair advantage (it often will not), or contributes to the state’s dissolution (it almost certainly will not), but whether it is disrespectful to your fellow citizens in light of the more general duties you owe them as valuable beings with particular sorts of interests. Fair distributions of goods and freedom from peril are different from respect in that they are, as I explained before, the objects of outcome-oriented interests. Thus, in cases in which fairness would be promoted were I to disobey the law, the theory of fair play should recommend that I disobey it, and in cases in which the state’s peril-preventing project would be promoted by disobedience, samaritanism should say the same. Both results are, of course, highly 337 counterintuitive. Whether one person disrespects another depends, by contrast, in large part on the circumstances of an action’s performance. Views which turn on it are therefore invulnerable to counterexamples in which disobedience leads to fairer relations between citizens, more stable states, or other consequences which competing theories should, when followed to their logical ends, support rather than proscribe. Finally, I believe that the layered nature of the authority theory’s MP approach helps us explain why different reactive attitudes strike us as appropriate responses to different instances of lawbreaking. This is not a topic I have seen discussed anywhere in the literature on political obligation, but I think it is an important observation that a complete theory of political obligation should explain. Why is it that some violations of the law we shrug off, while others we take seriously? Because it does not seem that I act more or less fairly by breaking one law rather than another, this fact will be difficult to explain on both fairness and samaritanism accounts (the latter of which relies on considerations of fairness for distributive purposes). One possibility is that, while we take the same attitude toward all law breaking per se, our various responses reflect what we take to be the likely consequences of different sorts of legal violations. While there is something to be said for this explanation, it nonetheless seems implausible. Something’s being the law seems clearly to affect how we look at it. For instance, most people (presumably) believe that drug smuggling is a serious crime. But to smuggle something is just to illegally transport it, and it seems likely that, even without reliable information indicating that these drugs would cause less harm overall, most people would care significantly less about this form of transportation if the drugs were legal (as the Mexican government is currently urging for some substances). How can we explain 338 this while maintaining that it is merely the consequences of one’s legal violation that explain our various attitudes? It seems to me that the authority theory’s various justificatory layers can begin to account for these differences. The standard argument tracks important duties whose more and less flagrant violations seem to justify more and less serious reactions. Since it is the deontic force of these duties which underlies the laws we are obligated to obey, we would expect more flagrant violations to evoke more serious responses. The samaritan duty I have to you is arguably weightier than the duty I have to respect your autonomy, which is arguably weightier than the duty I have not to put you at unreasonable risk. Thus one might believe that selling automatic weapons to felons is a more serious transgression than auto theft, which is a more serious transgression than running a stoplight at an empty intersection at 4am. On the other hand, something does seem to remain constant across all legal violations. The authority theory offers the proceduralist argument to explain this. Since every law is an expression of our political autonomy, every violation is, at least to some extent, a form of disrespect. The severity of the resentment a person addresses to a lawbreaker, even for this reason, usually turns on the attitude she takes toward the law in question. Abortion and free speech laws strike emotional chords that traffic laws often do not, but the important point is that some level of resentment can intelligibly be addressed to a lawbreaker in any circumstance. “Who do you think you are?” is a coherent question a citizen might address to someone who breaks even trivial laws, for by doing so the lawbreaker suggests that the rules we use to govern our lives together do not or should not apply to him. Of course, we don’t often take this attitude, but the fact that we can seems to verify—as much as one could hope, anyway—that our interest in political autonomy finds expression in every law. This 339 fact is captured by the authority theory more obviously than it is or could be by any competing theory of political obligation. 340 Conclusion I had two principal aims in this dissertation. I undertook, in Part I, to establish the need for an adequate theory of political obligation by highlighting the challenges facing (what I consider to be) the two leading contemporary theories. I then attempted, in Part II, to meet this need with a novel theory of my own, one which attended to important normative considerations which many theorists working on the problem have ignored. While I maintain that the authority theory more plausibly meets the success conditions we established in Chapter 1 than competing theories have, and so more convincingly answers the question of political obligation we set out to answer at the beginning of this essay, the most philosophically gainful feature of the theory is, from my perspective, the use it makes of authority in addressing the problem. Authorities are, on my view, natural features of the world around us. They are servants at our disposal, instruments we use to meet the demands morality imposes on us. What this entails, however, is that the authority theory can plausibly be reconciled with anarchism itself. It is, I believe, an anarchist theory of political obligation— one that is consistent with anarchist principles and which therefore cannot be rejected from the anarchist perspective. Properly understood, political obligations are no more the anarchist’s foe than the more mundane obligations we incur in everyday life. They are merely 341 one sort of moral requirement—one type of demand we ought to address to ourselves in virtue of what we owe others. This is, anyway, how I have come to understand the problem, and its solution. 342 References Albert, Michael. 2004. Parecon: Life after capitalism. Verso Books. Anscombe, G.E.M. 1958. “Modern moral philosophy.” Philosophy 33: 1-19. Aquinas, St. Thomas. 2006. Aquinas: Political writings. Edited by R.W. Dyson. Cambridge University Press. Arneson, R. J. 1982. “The principle of fairness and free-rider problems.” Ethics 92, no. 4: 616–633. Austin, John. 2001. The province of jurisprudence determined. Cambridge University Press. Austin, J. L. 2000. How to do things with words. Harvard University Press. Bach, Kent and Robert Harnish. 1979. Linguistic communication and speech acts. MIT Press. Beran, Harry. 1987. The consent theory of political obligation. Routledge & Kegan Paul. ———. 1972. “Ought, obligation and duty.” Australasian Journal of Philosophy 50, no. 3: 207– 221. Bookchin, Murray. 1986. Post-scarcity anarchism. Montreal, Canada: Black Rose Books. Brandt, R. B. 1964. “The concepts of obligation and duty.” Mind 73: 374–393. Bratman, Michael. 1987. Intention, plans and practical reasoning. Harvard University Press. Broom, John. 2001. “Are intentions reasons? And how should we cope with incommensurable values?” In Practical reason and preference: essays for David Gauthier, edited by Christopher Morris and Arthur Ripstein, 98-119. Cambridge University Press. 343 ———. 2004. “Reasons.’ In Reason and value: Themes from the moral philosophy of Joseph Raz, edited by R. Jay Wallace, Michael Smith, Samuel Scheffler, and Philip Pettit, 28–55. Oxford University Press. Carter, Alan. 2001. “Presumptive benefits and political obligation.” Journal of Applied Philosophy 18, no. 3: 229-243. Christiano, Thomas. 2004. “The authority of democracy.” Journal of Political Philosophy 12: 266-290. Dagger, Richard. 2010. “Political obligation.” The Stanford Encyclopedia of Philosophy (Summer 2010 Edition), Edward N. Zalta (ed.), URL = <http://plato.stanford.edu/archives/sum2010/entries/political-obligation/>. Dancy, Jonathan. 2004. “Enticing reasons.” In Reason and value: Themes from the moral philosophy of Joseph Raz, edited by R. J. Wallace, P. Pettit, S. Scheffler, & M. Smith, 91-118. Oxford: Clarendon Press. ———. 2004b. Ethics without principles. Oxford University Press. Darwall, Stephen. 2006. The second-person standpoint: Morality, respect, and accountability. Harvard University Press. ———. 2007. “Reply to Korsgaard, Wallace, and Watson.” Ethics 118: 52-69. ———. 2009. “Authority and second-personal reasons for acting.” In Reasons for action, edited by D. Sobel and S. Wall, 134-154. Cambridge University Press. ———. 2010. “Authority and reasons: Exclusionary and second-personal.” Ethics 120: 257- 278. DeGeorge, Richard. 1985. The nature and limits of authority. University Press of Kansas. Dworkin, Ronald. 1986. Law's empire. Belknap Press. Edmundson, William. 1993. “Rethinking exclusionary reasons: A second edition of Joseph Raz's Practical Reason and Norms.” Law and Philosophy 12, no. 3: 329–343. ———. 1999. The duty to obey the law: Selected philosophical readings. Lanham, MD: Roman and Littlefield. Estlund, David. 2005. “Political authority and the tyranny of non-consent.” Philosophical Issues 15 (on Normativity): 351-367. ———. 2008. Democratic authority. Princeton University Press. Feinberg, Joel. 1984. Harm to others. Oxford University Press. Flathman, Richard. 1972. Political obligation. New York: Atheneum. 344 Foot, Philippa. 1972. “Morality as a system of hypothetical imperatives.” The Philosophical Review 81, no. 3: 305-316. ———. 2001. Natural Goodness. Oxford University Press. ———. 1978a. “Are moral considerations overriding?” In Virtues and vices and other essays in moral philosophy. University of California Press. Frankfurt, Harry. 1969. “Alternate possibilities and moral responsibility.” The Journal of Philosophy 66: 829-839. Freeman, Harrop A. 1996. “The right of protest and civil disobedience.” Indiana Law Journal 41: 228–254. Freeman, Samuel. 2008. Rawls. New York: Routledge. Friedman, David. 1973. The machinery of freedom. New York: Harper and Row. Gans, Chaim. 1992. Philosophical anarchism and political disobedience. Cambridge University Press. Gilbert, Margaret. 2006. A theory of political obligation: membership, commitment, and the bonds of society. Oxford University Press. Green, Leslie. 1988. The authority of the state. Oxford: Clarendon Press. ———. 1999. “Who believes in political obligation?” In The duty to obey the law, edited by William Edmundson, 301-318. Maryland: Rowman & Littlefield. Green, T. H. 1927. Lectures on the principles of political obligation. London: Longmans, Green and Co. LTD. Greenawalt, Kent. 1987. Conflicts of law and morality. Oxford University Press. Grice, Paul. 1991. Studies in the way of words. Harvard University Press. Gutmann, Amy and Dennis Thompson. 1996. Democracy and disagreement. London: Belknap Press. Haidt, Jonathan. 2006. The happiness hypothesis: finding modern truth in ancient wisdom. New York: Basic Books. Hart, H.L.A. 1955. “Are there any natural rights?” The Philosophical Review 64: 175–191. ———. 1961. The concept of law. Oxford: Clarendon Press. ———. 1982. Essays on Bentham. Oxford University Press. Hershovitz, Scott. 2003. “Legitimacy, democracy, and Razian authority.” In Legal Theory, 9: 201-220. ———. 2011. “The role of authority.” In Philosophers’ Imprint 11, no7: 1-19. 345 Hobbes, Thomas. 1985. Leviathan (Penguin classics). Middlesex, England: Harmondsworth. Hohfeld, Wesley. 1917. “Fundamental legal conceptions as applied in judicial reasoning.” In The Yale Law Journal, 26: 710-770. Jacobs, James. 1986. Socio-legal foundations of civil-military relations. New Jersey: Transaction Publishers. Kant, Immanuel. 1996. Groundwork of the metaphysics of morals. In The Cambridge Edition of the Works of Immanuel Kant—Practical Philosophy, edited & translated by Mary J Gregor. New York: Cambridge University Press. Kavka, Gregory. 1983. “Review of Moral Principles and Political Obligations, by A.J. Simmons.” Topoi 2: 227-30. ———. 1986. Hobbesian moral and political theory. Princeton University Press. Klosko, George. 1987a. “The principle of fairness and political obligation.” Ethics 97, no. 2: 353–362. ———. 1987b. “Presumptive benefit, fairness, and political obligation.” Philosophy & Public Affairs: 241–259. ———. 1992. The principle of fairness and political obligation. Lanham, MD: Rowman & Littlefield. ———. 2003. “Samaritanism and political obligation: A response to Christopher Wellman’s ‘Liberal theory of political obligation.’” Ethics 113: 835-40. ———. 2004. “Duties to assist others and political obligations.” Politics, Philosophy, and Economics, 3. ———. 2004b. “Multiple principles of political obligation.” Political Theory, 32. ———. 2005. Political obligations. Oxford University Press. Korsgaard, Christine. 1996. The sources of normativity. Cambridge University Press. Ladd, John. 1970. “Legal and moral obligation.” In Nomos XII: Political and Legal Obligation, edited by J. W Chapman and J.R. Pennock, 3–35. Atherton Press. Lehrer, Jonah. 2009. How we decide. New York: Houghton Mifflin Harcourt. Lewis, David. 1997. “Finkish dispositions.” The Philosophical Quarterly 47, no. 187: 143-158. Locke, John. 2003. John Locke: Political writings. Edited by David Wootton. Indianapolis: Hacket Publishing Company. Lloyd, S. A. 1992. Ideals as interests in Hobbes's Leviathan: The power of mind over matter. Cambridge University Press. 346 Manley, David and Ryan Wasserman. 2008. “On linking dispositions and conditionals.” Mind 117: 59-84. Marmor, Andrei. 2005. “Authority, equality and democracy.” Ratio Juris 18, vol. 3: 315-345. ———. 2010."The dilemma of authority.” University of Southern California Legal Studies Working Paper Series. Working Paper 66. http://law.bepress.com/usclwps/lss/art66. ———. 2011. The philosophy of law. Princeton University Press. Martin, C. B. 1994. “Dispositions and conditionals.” The Philosophical Quarterly 33: 1-8. McCloskey, H.J. 1969. Meta-ethics and normative ethics. The Hague: Martinus Nijhoff. McPherson, Thomas. 1967. Political obligation. Routledge & Kegan Paul. Mill, John Stuart. 2008. On Liberty and other essays. Edited by John Gray. Oxford University Press. Murphy, Mark. 1999. “Surrender of judgment and the consent theory of political authority.” In The Duty to Obey the Law, edited by William A. Edmundson, 319-346. Maryland: Rowman & Littlefield. Nagel, Thomas. 1986. The view from nowhere. Oxford University Press. ———. 1995. “Personal rights and public space.” Philosophy and Pubic Affairs 24: 83-107. Nozick, Robert. 1974. Anarchy, state, and utopia. Basic Books. Owens, David. 2008. “Rationalism about obligation.” European Journal of Philosophy 16, no. 3: 403-431. ———. 2012. Shaping the normative landscape. Oxford University Press. Pink, Thomas. 2004. “Moral obligation.” In Modern Moral Philosophy: Royal Institute of Philosophy Supplement 54, edited by Anthony O’Hear, 159-185. Cambridge University Press. Prior, Elizabeth. 1985. Dispositions. Aberdeen: Aberdeen University Press. Rawls, John. 1964. “Legal obligation and the duty of fair play.” In Law and philosophy, edited by S. Hook. New York University Press. ———. 2003. A theory of justice (revised edition). Cambridge MA: Belknap Press. Raz, Joseph. 1974. “Reasons, Requirements, and Practical Conflicts.” In Practical reasoning, edited by S. Korner, 22-35. Oxford University Press. ———. 1977. “Promises and Obligations.” In Law, morality, and society: Essays in honour of H.L.A. Hart, edited by P.M.S. Hacker and Joseph Raz, 210-228. Oxford University Press. 347 ———. 1978. “Introduction.” In Practical reasoning, edited by Joseph Raz, 1-17. Oxford University Press. ———. 1979. The authority of law. Oxford University Press. ———. 1988. The morality of freedom. Oxford University Press. (First paperback printing.) ———. 1990. “Authority and justification.” In Authority, edited by Joseph Raz, 115-141. New York University Press. Originally published in Philosophy and Public Affairs, vol. 14, No. 1 (1985). ———. 1994. Ethics in the public domain. Oxford University Press. ———. 1999a. Practical reason and norms. Oxford University Press. ———. 1999b. “The obligation to obey: revision and tradition.” In The duty to obey the law, edited by William Edmundson, 159-175. Maryland: Rowman & Littlefield. Originally published in the Notre Dame Journal of Law, Ethics & Public Policy, vol. 1 (1984), 139- 155. ———. 2009a. “The problem of authority: Revisiting the service conception.” In Between authority and interpretation. Oxford University Press. Originally published in Minnesota Law Review, vol. 90 (2006), 1003-1044. ———. 2009b. “On the authority and interpretation of constitutions: Some preliminaries.” In Between authority and interpretation. Oxford University Press. Originally published in Constitutionalism: Philosophical foundations, edited by Larry Alexander, 152-193 (Cambridge university press, 1998). ———. 2010. “On respect, authority, and neutrality: a response.” Ethics 120: 279-301. ———. Unpublished. “The possibility of partiality.” Downloaded October, 2010. josephnraz.googlepages.com/ThePossibilityofPartiality-temp.pdf Ross, W.D. 1930. The right and the good. Oxford University Press. Scanlon, T.M. 1998. What we owe to each other. Belknap Press. ———. 2008. Moral dimensions. Belknap Press. Schroeder, Mark. 2007. Slaves of the passions. Oxford University Press. ———. 2008. ‘Having reasons.’ Philosophical Studies 139, no.1: 57-71, 2008. Philosopher’s Annual. Sciaraffa, Stefan. 2009. “On content-independent reasons: It’s not in the name.” Law and Philosophy, 28: 233-260. Sen, Amartya. 1979. “Equality of what?” Stanford University’s Tanner Lecture on Human Values, May 22. Available at http://www.dse.unifi.it/sviluppo/doc/Subramanian1.pdf. Sahlins, Marshall. 1972. Stone age economics. Hawthorn NY: Transaction Publishers. Shapiro, Scott. 2002. “Authority.” In The Oxford Handbook of Jurisprudence and Philosophy of Law, edited by Jules Coleman and Scott Shapiro, 382-439. Oxford University Press. ———. 2011. Legality. Harvard University Press. 348 Simmons, A. J. 1981. Moral principles and political obligations. Princeton University Press. ———. 1987. “The anarchist position: A reply to Klosko and Senor.” Philosophy & Public Affairs 16: 269–279. ———. 2001. “Fair play and political obligation: Twenty years later.” In Justification and legitimacy: Essays on rights and obligations. Cambridge University Press. Simmons, A.J. and Christopher Wellman. 2005. Is there a duty to obey the law? Cambridge University Press. Singer, P. 1973. Democracy and disobedience. Clarendon Press. Smith, M.B.E. 1999. “Is there a prima facie obligation to obey the law?” In The duty to obey the law, edited by William A. Edmundson, 75 – 106. Maryland: Rowman & Littlefield. Solomon, William David. 1975. “Moral Reasons.” American Philosophical Quarterly 12, no. 4: 331-339. Strawson, P. F. 2008. “Freedom and resentment.” In Freedom and resentment and other essays, 1- 28. London: Routledge. Originally published in Proceedings of the British Academy, vol. XLVIII (1962). ———. 1987. The possibility of cooperation. Cambridge University Press. Tavris, Carol, and Elliot Aronson. 2007. Mistakes were made (but not by me): Why we justify foolish beliefs, bad decisions, and hurtful acts. Orlando, Florida: Harcourt. Thompson, Michael. 2004. “What is it to wrong someone? A puzzle about justice.” In Reason and value: Themes from the philosophy of Joseph Raz, edited by R. Jay Wallace, Philip Pettit, Samuel Scheffler, and Michael Smith, 333-384. Oxford University Press. Thomson, Judith Jarvis. 1971. “A defense of abortion.” In Philosophy and Public Affairs 1, no.1: 47-66. Waldron, Jeremy. 1993. “Special ties and natural duties.” In Philosophy and Public Affairs 22: 3- 30. Wallace, R. Jay. 2012. “The deontic structure of morality.” Available at http://philosophy.berkeley.edu/file/2/Deonticstructure.pdf. Last accessed February 6, 2012. Wasserstrom, Richard. 1999. “The obligation to obey the law.” In The Duty to Obey the Law, edited by William A. Edmundson, 17–47. Maryland: Rowman & Littlefield. Wellman, C. H. 1997. “Associative allegiances and political obligations.” Social Theory and Practice 23, no. 2: 181-204. 349 ———. 2001. “Toward a liberal theory of political obligation.” Ethics 111, no.4: 735-759. ———. 2004. “Political obligation and the particularity requirement.” Legal Theory 10: 97- 115. Wenar, Leif. 2010. "Rights." The Stanford Encyclopedia of Philosophy (Fall 2010 Edition), Edward N. Zalta (ed.), URL = <http://plato.stanford.edu/archives/fall2010/entries/rights/>. Wolff, Jonathan. 2000. “Political obligation: A pluralistic approach.” In Pluralism: The philosophy and politics of diversity, edited by M. Baghamrian and A. Ingram, 179 – 196. Routledge. Wolff, Robert Paul. 1970. In defense of anarchism. Harper & Row. Zerzan, John. 2002. Running on emptiness. Los Angeles: Feral House.

Abstract (if available)

Abstract I have two principle aims in this dissertation. I attempt in Part I to establish the need for an adequate theory of political obligation by highlighting the challenges facing what I consider to be the two leading contemporary theories: George Klosko’s theory of fair play, and Christopher Wellman’s theory of samaritanism. I then attempt, in Part II, to meet this need with a novel theory of my own—what I call the authority theory—which attends to important normative considerations many theorists working on the problem have ignored. While I argue that the authority theory more plausibly satisfies the conditions on a successful theory of political obligation, the most philosophically gainful feature of the theory is, from my perspective, the use it makes of the concept of authority in addressing the problem.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Describing authority

PDF

Reasons, obligations, and the structure of good reasoning

PDF

Units of agency in ethics

PDF

Contract, from promise to commodity

PDF

Vagueness, legal content, and legal interpretation

PDF

The mental states first theory of promising

PDF

The boundaries of the intentional

PDF

Distributive equality and labor burdens

PDF

What it is to be located

PDF

Political identity and political participation: China's post-80's generation

PDF

The politics of transformation: mass media and the Northern Ireland peace process

PDF

Taking orders from Brussels? external actors and regional organizations in the developing world

PDF

Habits and policy: the socio-cognitive foundations of foreign policy stability

PDF

The political representation of Kurdish, Kemalist, and conservative Muslim women in Turkey (1990-2010)

PDF

The virtue of reasonableness: on the normative epistemic implications of public reason liberalism

PDF

Responding to harm

PDF

A deontological explanation of accessibilism

PDF

Responsibility and the emotional structure of relationships

PDF

Changing political attitudes and behavior in a diverse America: incorporating individual and contextual determinants

PDF

Intrinsic and extrinsic values of distributional equality

Asset Metadata

Creator DeClark, Kory (author)

Core Title On being bound: law, authority, and the politics of obligation

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Philosophy

Publication Date 06/13/2012

Defense Date 05/24/2012

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag authority,exclusionary reasons,Joseph Raz,normal justification thesis,OAI-PMH Harvest,obligation,political obligation

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Watson, Gary L. (committee chair), Keating, Gregory (committee member), Marmor, Andrei (committee member)

Creator Email kdeclark@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-47997

Unique identifier UC11290438

Identifier usctheses-c3-47997 (legacy record id)

Legacy Identifier etd-DeClarkKor-887.pdf

Dmrecord 47997

Document Type Dissertation

Rights DeClark, Kory

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA