Coercion and the Nature of Law [First ed.] 9780198854937

Table of contents :
01. Title_Pages
02. Preface
03. The_Coercion_Thesis
04. Methodology_and_the_Nature_of_Law
05. A_Prima_Facie_Case_for_the_Coercion_Thesis_Sanctions_as_a_Paradigmatic_Feature_of_Municipal_Law
06. What_Law_Must_Be_Able_to_Do_The_Coercion_Thesis_and_the_Need_to_Keep_the_Peace
07. Three_Conceptual_Problems_of_Legal_Normativity_The_Logical_Space_of_Reasons
08. The_Coercion_Thesis_and_the_How_Problem_of_Legal_Normativity
09. The_Coercion_Thesis_and_the_Order_Problem_of_Legal_Normativity
10. The_Coercion_Thesis_and_the_Content_Problem_of_Legal_Normativity
11. Coercive_Sanctions_and_International_Law
12. Can_There_Be_Law_in_a_Society_of_Angels
13. Appendix_Can_There_Be_a_System_of_Municipal_Law_with_Only_Compensatory_Damages
14. Index

Citation preview

OX F O R D L E G A L P H I L O S O P H Y Series editors

TIMOTHY ENDICOTT JOHN GARDNER (1965–2019) LESLIE GREEN

Coercion and the Nature of Law

OX F O R D L E G A L P H I L O S O P H Y Series editors Timothy Endicott, John Gardner (1965–​2019), and Leslie Green Oxford Legal Philosophy publishes the best new work in philosophically-​ oriented legal theory. It commissions and solicits monographs in all branches of the subject, including works on philosophical issues in all areas of public and private law, and in the national, transnational, and international realms; studies of the nature of law, legal institutions, and legal reasoning; treatments of problems in political morality as they bear on law; and explorations in the nature and development of legal philosophy itself. The series represents diverse traditions of thought but always with an emphasis on rigor and originality. It sets the standard in contemporary jurisprudence. ALSO AVAILABLE IN THE SERIES Faces of Inequality: A Theory of Reason and Restitution: Wrongful Discrimination A Theory of Unjust Enrichment Sophia Moreau Charlie Webb Property Rights: A Re-​Examination JE Penner Right of Redress Andrew Gold A Theory of Legal Personhood Visa AJ Kurki Imposing Risk: A Normative Framework John Oberdiek Law and Morality at War Adil Ahmad Haque Ignorance of Law: A Philosophical Inquiry Douglas Husak

Allowing for Exceptions: A Theory of Defences and Defeasibility in Law Luís Duarte d’Almeida The Ends of Harm: The Moral Foundations of Criminal Law Victor Tadros Corrective Justice Ernest J. Weinrib Conscience and Conviction: The Case for Civil Disobedience Kimberley Brownlee The Nature of Legislative Intent Richard Ekins Why Law Matters Alon Harel

Coercion and the Nature of Law K E N N E T H E I NA R   H I M M A

1

3 Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Oxford University Press 2020 The moral rights of the author have been asserted First Edition published in 2020 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Crown copyright material is reproduced under Class Licence Number C01P0000148 with the permission of OPSI and the Queen’s Printer for Scotland Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2019957407 ISBN 978–​0–​19–​885493–​7 Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

Preface This is the second volume of a planned four-​volume series on the metaphysical nature of law. The first volume, Morality and the Nature of Law, argues that the metaphysical nature of law allows for both the existence of legal systems without content-​based constraints on what counts as valid law and the existence of legal systems with content-​based constraints on what counts as valid law and thus defends an inclusive positivist view against anti-​positivist and exclusive positivist views. The third volume, Authority and the Nature of Law, explicates the metaphysical nature of practical authority as it is exemplified in systems of law. The final volume, Methodology and the Nature of Law, defends a modest approach to conceptual analysis that seeks to identify the metaphysical nature of law as our conceptual practices determine it. This second volume defends the Coercion Thesis, according to which it is a conceptually necessary feature of a legal system that some mandatory legal norms governing non-​official behavior are backed by the threat of a coercive sanction. According to the Coercion Thesis, it is a necessary condition for something to count as a legal system, as far as our conceptual practices are concerned, that it includes norms governing official behavior that authorize courts to impose coercive sanctions as a response to non-​compliance with mandatory norms governing non-​ official behavior; a mandatory legal norm L governing non-​official behavior counts as being backed by the threat of a coercive sanction, as a conceptual matter, only insofar as there is a legal norm governing official behavior that authorizes courts to impose a coercive sanction for non-​compliance with L. The issue of whether law is coercive in this respect has deeply engaged me since I first encountered legal philosophy while a second-​year law student at the University of Washington. While it seemed plausible to me that there could be legal norms not backed by authorized coercive sanctions, I was dumbfounded by the idea that there could be legal systems that utterly lacked authorized sanctions: although I could not articulate the objection at the time, it seemed obvious to me that it is just not possible—​in any practical sense that matters—​to regulate the behavior of passionately self-​interested subjects like us in worlds of acute material scarcity like ours without backing some mandatory norms prohibiting violence with the threat of a coercive sanction sufficiently severe to deter such acts enough of the time for us to live together in a state of comparative peace. Everything I had seen in the world about how law works depended on its being enforced by coercive sanctions.

vi Preface The idea that law is essentially coercive in this respect has a history suggesting that my initial reaction to the conceptual possibility of a legal system without sanctions is not unnatural. Positivists Jeremy Bentham and John Austin held the Coercion Thesis, grounding it in the ideas that it is a conceptually necessary feature of an obligation that, as Bentham put the matter, it defines a “liab[ility] to be punished” and that it is a conceptually necessary feature of a legal system that its norms give rise to legal obligations.1 Hans Kelsen argued that law is distinct from all other institutional normative systems in virtue of being coercive by nature: If we confine our investigation to positive law, and if we compare all those social orders, past and present, that are generally called “law,” we shall find that they have one characteristic in common which no social orders of another kind present. This characteristic constitutes a fact of supreme importance for social life and its scientific study. And this characteristic is the only criterion by which we may clearly distinguish law from other social phenomena such as morals and religion. This criterion is coercion.2

The Coercion Thesis was likewise universally accepted among classical natural law theorists of note. William Blackstone adopted the Coercion Thesis eleven years before Bentham did and nearly sixty years before Austin did: Legislators and their laws are said to compel and oblige; not that by any natural violence they so constrain a man, as to render it impossible for him to act otherwise than as they direct, which is the strict sense of obligation: but because, by declaring and exhibiting a penalty against offenders, they bring it to pass that no man can easily choose to transgress the law; since, by reason of the impending correction, compliance is in a high degree preferable to disobedience. And, even where rewards are proposed as well as punishments threatened, the obligation of the law seems chiefly to consist in the penalty: for rewards in their nature, can only persuade and allure; nothing is compulsory but punishment.3

1 Jeremy Bentham, A Fragment of Government (Cambridge University Press 1988) 109. John Austin puts it in predictably similar terms: “[b]‌eing liable to evil from you if I comply not with a wish which you signify, I am bound or obliged by your command, or I lie under a duty to obey it.” John Austin, The Province of Jurisprudence Determined (Cambridge University Press 1995) 22. 2 Hans Kelsen, General Theory of Law and State (Harvard University Press 1949), 15. It is not obvious that the availability of something properly characterized as a sanction is what distinguishes the familiar systems of municipal law from systems of religious or moral law: the law of every mainstream religious tradition, as far as I can tell, creates a liability to punitive detriment in the form of expulsion from the church for violations of various norms; likewise, the law of every system of moral norms, conventional or otherwise, creates a liability for non-​compliance in the form of justified social blame, censure, and punishment. Whether or not these liabilities constitute coercive sanctions depends on what is meant by the notion—​which is explicated in Chapter 1. 3 William Blackstone, Commentaries on the Laws of England, Volume 1 (University of Chicago Press 1979) 57.

Preface  vii What difference there is between Blackstone’s conceptual views here and those of Austin and Bentham seems to consist in Blackstone’s view that the true sense of obligation consists in the “compulsion” of “natural violence,” by which he might mean violence imposed by a deity for disobedience of divine law. But, either way, it is clear that Blackstone accepts the Coercion Thesis. Contemporary natural law theorists also hold this view. In Natural Law and Natural Rights, John Finnis argues that “law needs to be coercive (primarily by way of punitive sanctions, secondarily by way of preventive interventions and restraints).”4 Although Finnis believes that certain features of a legal order would be necessary in a society where coercive sanctions are not needed to deter non-​ compliance, he seems to view such a society as giving rise to something that closely resembles a “legal order,” rather than something that counts as a legal system. Ronald Dworkin builds the Coercion Thesis into the foundation of his constructivist views, holding that the very point of legal regulation is to provide a moral justification for the coercive enforcement of legal norms: “The law of a community on this account is the scheme of rights and responsibilities that . . . license coercion because they flow from the past decisions of the right sort.”5 Law is not capable, in the relevant sense, of licensing or justifying coercion unless judges are authorized to impose coercive sanctions for non-​compliance. Although he would probably not put the matter this way, Dworkin takes the view that it is a conceptually necessary condition for the existence of a legal system that it backs some mandatory legal norms governing non-​official behavior with the threat of a coercive sanction. So obviously ubiquitous and central are coercive sanctions to empirical legal practice as we understand it that some commentators have criticized the very project of conceptual jurisprudence on the ground that the authority of the courts to impose coercive sanctions for non-​compliance does not play any explanatory role in conceptual theories of law. As Frederick Schauer puts the point: That a feature of law that is so important and salient to almost everyone . . . is so marginal to the jurisprudential enterprise says something about the enterprise of philosophy of law—​jurisprudence—​itself. And thus a running subtext of this book is a challenge to a prevalent mode of jurisprudential inquiry.6

Schauer regards it as a defect of the project of conceptual jurisprudence, and not as a defect of any particular conceptual theory of law, that it does not include the authorization of coercive sanctions as a theoretically significant feature of law. While he is incorrect in thinking that any of this is even remotely equipped to call into

4

John Finnis, Natural Law and Natural Rights, 2nd edn (Oxford University Press 2011) 266. Ronald Dworkin, Law’s Empire (Harvard University Press 1986) 93. 6 Frederick Schauer, The Force of Law (Harvard University Press 2015) Preface. 5

viii Preface question the viability of conceptual jurisprudence, the only way that a conceptual theory of law could give expression to a “feature of law that is so important and salient to almost everyone” is by adopting the Coercion Thesis as a centerpiece.7 Despite its abiding influence in conceptual jurisprudence, the Coercion Thesis began to be questioned in the early 1960s. While observing the ubiquity of coercion in the legal systems of nation-​states, Hart questioned its conceptual necessity on the ground that there is an international normative system lacking a formal system of authorized coercive sanctions that is plausibly characterized as a system of law; as Hart puts the point, “[t]‌here are no settled principles forbidding the use of the word ‘law’ of systems where there are no centrally organized sanctions, and there is good reason (though no compulsion) for using the expression ‘international law’ of a system, which has none.”8 Hart’s claim here is not that our system of international regulation obviously counts as law despite lacking sanctions. It is rather that the canons of ordinary usage are indeterminate with respect to whether this system counts as law despite lacking sanctions: if the canons of ordinary usage are indeterminate with respect to whether our system of international regulation comprises a legal system despite lacking sanctions, then they are indeterminate with respect to the Coercion Thesis. Joseph Raz offers a stronger argument against the Coercion Thesis grounded in what he takes to be obvious features of our conceptual practices with respect to the term law.9 Whereas Hart argues that our conceptual practices are indeterminate with respect to the Coercion Thesis, Raz rejects this view, arguing that the canons of ordinary usage governing the use of law are unambiguously consistent with the conceptual possibility of a sanctionless legal system in a society of beings who are 7 Not surprisingly, the Coercion Thesis also enjoys widespread support among laypersons whose views condition our lexical conventions for using the term law. While there are no general sociological polls of competent speakers, which is unsurprising given the obvious content of our lexical conventions on the point, I commonly ask law students whether they think that an institutional normative system with all the features of the legal system of the society in which they live except for the coercive enforcement practices authorized for non-​compliance is, on their views, properly characterized as a system of law. I estimate that at least three-​fourths of the students respond that a system like that does not count as a legal system; the remainder are split between not being sure and believing that such a system is properly characterized as one of law. 8 H.L.A. Hart, The Concept of Law, 2nd edn (Clarendon Press 1994) 199. Hart eventually endorsed the Coercion Thesis—​at least with respect to municipal legal systems—​and revised his theory of legal obligation to require the authorization of coercive enforcement mechanisms: “This revised account of obligation under a simple regime of custom-​type rules can be extended to the more complex case of a developed legal system where legal obligations are imposed by statute-​type rules enacted by a legislature. Such enacted rules may not be generally accepted by ordinary members of a society but are recognized as valid rules of the legal system by courts. . . .Such enacted rules imposing obligations need not be and frequently are not supported by general social pressure but are supported by ancillary rules permitting or requiring officials to respond to deviation with demands and coercive measures to secure conformity. These responses will not be merely predictable consequences of deviation (and indeed may not always be predictable) but will be legitimate responses to deviations, since deviations are permitted or required to make them.” Luis Duarte d’Almeida, James Edwards, and Andrea Dolcetti (eds.), “An Interview with Hart,” in Reading H.L.A. Hart’s The Concept of Law (Oxford and Portland, OR, 2015) at 283. Emphasis added. 9 Joseph Raz, Practical Reason and Norms (Oxford University Press 2002) 159–​60.

Preface  ix conclusively motivated always to do what the norms of the system require; insofar as sanctions are not needed in a such a society to induce compliance, on his view, there can be law without sanctions. Contemporary positivists who recognize the central role of coercion in existing legal practice sometimes appear to carve out a middle position. Leslie Green seems to take a conceptual position that ostensibly rejects the Coercion Thesis while building the salience of coercion back into the metaphysical nature of law: [T]‌here are conceptually possible legal systems that do not deploy any coercion. That leaves open other possibilities. . . . But here is an important one: because the authority of law is comprehensive, there are no legal systems that lack norms capable of imposing sanctions if necessary.10

The idea here seems to be, as the matter is sometimes put, that legal norms are, by nature, coercion-​apt in the sense that they can be enforced. While there is nothing in the concept of law that precludes the possibility of a legal system in which the peace is efficaciously kept without the application of authorized sanctions, it is a conceptual truth, on his view, that anything properly characterized as a system of law “contains norms capable of imposing sanctions.” Although Green might seem to be taking a more nuanced position that denies the Coercion Thesis, his view is more plausibly construed as consistent with that thesis because the only plausible interpretation of the idea that there are no possible legal systems “lack[ing] norms capable of imposing sanctions if necessary” expresses the claim that judges must be authorized, as a conceptual matter, to impose coercive sanctions for non-​compliance with some mandatory norms governing non-​official behavior. Every norm properly characterized as governing the behavior of someone in virtue of directing her to do something she can do is capable, as a practical matter, of being enforced by the imposition of a coercive sanction; moral norms and norms of etiquette are no less capable than any other norm of being enforced by the imposition of a coercive sanction if there is someone around to impose that sanction on those who fail to comply. Someone who accepts the Coercion Thesis as obvious might think that a book-​ length argument is not needed to defend it insofar as most theorists and laypersons already accept it; but even if it is true that most—​or even all—​persons accept it, that would not alleviate the need for an argument. Since people can be mistaken about what they characterize as obvious at any given moment, an argument is needed to defend even claims that seem obvious. But it is false that the Coercion Thesis is so obviously true that there is no need to mount a rigorous philosophical defense of it. If it were, theorists like Hart and Raz 10 Leslie Green, “The Forces of Law: Duty, Coercion and Power” Ratio Juris, vol. 29 (2016): 164–​81; emphasis added.

x Preface would not be able to give principled reasons for rejecting that thesis; it is enough to warrant rejecting the claim that something is obvious that there are persons who give plausibly principled reasons to regard it as either false or indeterminate. The existence of informed skeptics is reason enough to justify undertaking the project of this volume. This book expresses the fruits of my efforts, such as they are, to say something more intelligent in defense of my intuitions about law’s coercive nature than I was able to when first exposed in law school to the various views on the matter. I make what I take to be a comprehensive case for the Coercion Thesis that is grounded in the canons of ordinary usage but goes beyond them to rigorously explicate shared philosophical assumptions underlying our conceptual practices about what legal systems must be capable, as a practical matter, of doing in order to succeed in efficaciously regulating behavior. It also seeks to rebut skepticism about the Coercion Thesis that is grounded in considerations having to do with international law and societies of angels. I would like to thank a number of people who have commented on some of the ideas developed in this volume or have otherwise encouraged me in preparing this monograph:  Edgar Aguilera, Alexia Michelle Araujo, Brian Bix, Luka Burazin, Thomas Bustamante, Celia Canislupus, Kenneth Ehrenberg, Imer B.  Flores, Michael Green, Milena Jovanović, Lorenz Kähler, Matthias Klatt, Mario Krešić, Marija Knežević, Robin McCall, Lucas Miotto, Francisco M.  Mora Sifuentes, Zrinka Mrkonjić, Damhepreti Nawameen, Bojana Nedeljković, Vitaly Ogleznev, Ashley Robles, Corrado Roversi, and Wilfrid Waluchow. I would also like to thank the Faculty of Law at the University of Zagreb for its generous support of my work over the years. In particular, I would like to thank Dean Igor Gliha and Vice-​Deans Nina Tepeš, Tereza Rogić Lugarić, and Marko Jurić; it has been an honor and pleasure to work with them over the years. Finally, I want to thank my über-​boo, Maria Elias Sotirhos. She has read and commented on everything I have published in conceptual jurisprudence during my career despite finding the topics uniformly uninteresting, irrelevant, frustrating, tedious, and generally not worth the time and effort that either of us put it into them—​something she loudly reminds me of with clock-​like regularity. If there is anything worth reading in this volume, it has as much to do with her love, encouragement, support, comments, and painstakingly detailed copy-​editing as it does with anything having to do with me. I cannot fathom how anyone gets through life without having a Maria of their own; everyone should be so fortunate in finding a lifetime partner. I utterly adore her.

1

The Coercion Thesis This book is an essay in what is sometimes called—​somewhat pejoratively by academic lawyers who have an interest in conceptual issues concerning law but lack formal training in philosophical method—​metaphysically driven conceptual analysis.1 It defends what I call the Coercion Thesis, according to which it is a conceptually necessary condition for the existence of a legal system that it backs some mandatory legal norms prohibiting breaches of the peace with the threat of detriment that rationally competent subjects have a practical reason to avoid. While the discussion begins with something resembling a sociological analysis of our lexical practices with respect to the relevant concept-​terms, it is nonetheless concerned with the metaphysics of law and is unapologetically philosophical in character. This introductory chapter explicates the content of the Coercion Thesis. It begins by distinguishing a number of various usages of the concept-​terms law and coercion as a means of clarifying the content of the Coercion Thesis. It then explains the role that moral norms play in determining which acts must be coercively prohibited by anything that counts as a legal system and logical relationship between the Coercion Thesis, as well as the implications of the Coercion Thesis with respect to what motivations rationally competent subjects characteristically have, as a descriptive matter of contingent fact, to obey law. The chapter concludes with an overview of the arguments of the book and a brief description of how they work together to provide what I take to be a comprehensive case in favor of the Coercion Thesis.

1.  Ambiguities of the Concept-​Term Law The concept-​term law is ambiguous in a number of ways. To begin, the term can be used to refer both to legal norms and to legal systems—​and these two objects are conceptually related in an obvious way: the existence of one depends, as a conceptual matter, on the existence of the other. There are no conceptually possible legal norms that are not laws of some conceptually possible legal system; and there are no conceptually possible legal systems that do not include some legal norms. But they are also conceptually distinct in obvious ways. Something properly characterized as a legal norm is a propositional object metaphysically capable of



1

Sigh. As if there is some other kind.

Coercion and the Nature of Law. Kenneth Einar Himma, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198854937.001.0001

2 

The Coercion Thesis

guiding behavior, but something properly characterized as a legal system is not a propositional object metaphysically capable of guiding behavior. Legal systems are metaphysically capable of efficaciously regulating behavior in virtue of being partly constituted by a collection of norms governing behavior; institutions regulate behavior through the governance of norms metaphysically capable of guiding behavior. The ontological relationship between the two reflects this difference. While the existence of legal norms partly constitutes something as a legal system, the existence of a legal system does not partly constitute something as a legal norm; a legal norm is a proper part of the legal system of which it is a member and hence partly constitutes the system of which it is a member as a legal system, but the legal system of which the legal norm is a member is not a proper part of the norm and hence does not even partly constitute something as a legal norm. Otherwise put, legal systems are partly manufactured out of legal norms, but legal norms are not partly manufactured out of legal systems; legal norms are manufactured by and in legal systems. This should not be surprising, as this metaphysical relationship holds between anything properly characterized as an artifact and its constituent parts. Something that starts a car is a car-​starter in virtue of being reasonably contrived to start a car and is hence a proper part of some car it functions to start; however, while there cannot be something that starts a car unless there is a car of which it is (or can be) a proper part, a car is not a proper part of a car-​starter any more than a legal system is a proper part of a legal norm. The relevant usage for purposes of this volume picks out legal systems and not legal norms. The Coercion Thesis asserts that it is a conceptually necessary condition for something to be properly characterized as a system of law that it backs certain mandatory legal norms governing non-​official behavior with the threat of a sanction.2 Properly understood, the Coercion Thesis neither asserts nor entails that every legal norm is backed by the threat of a coercive sanction. Further, the concept-​term law, as it denotes legal systems, is ambiguous between a wide usage and a narrow usage. The wide usage includes not only the legal systems that define municipal and international entities but also systems of religious law and systems that define institutional memberships, such as the system of rules creating and governing a chess association. The narrow usage encompasses only those systems of international and municipal law that comprise the focus of study in law school.

2 The term official behavior refers to those acts making, changing, and adjudicating law; official behavior is, as a conceptual matter, governed by the rule of recognition. The term non-​official behavior refers to all other acts; non-​official behavior is, as a conceptual matter, governed by legal norms that are valid under the requirements of the rule of recognition governing official behavior.

Ambiguities of the Concept-​T erm Law  3 The Coercion Thesis, as it is defended here, applies only to the narrow usage and says nothing about systems counting as law that fall outside the reference class of the narrow usage. While systems of religious law have much in common with systems of municipal and international law, the former differ from the latter in theoretically significant ways. To pursue an explication of a general notion of law that applies equally to each would suppress theoretically significant features of one that distinguish it from the other and would hence result in a theory that under-​explicates  both. None of this should be taken to disparage the project of explicating the general concept that applies equally to municipal law, religious law, and the rules of a chess association. The linguistic conventions we adopt define a conceptual framework that not only structures the world of our experience but also says something important about us. That our ordinary usages permit the application of the term law to systems as different from one another as municipal law, religious law, and chess law tells us something important about who we are and what we value, but I have no urgent interest in the metaphysical nature of religious law or chess law. What I am interested in is the metaphysical nature of the type of law studied in law schools and practiced in systems of municipal and international law—​and this is my exclusive concern here. Finally, the term law is ambiguous between a purely descriptive usage and a thick usage that has both descriptive and evaluative content. The purely descriptive usage identifies empirically verifiable existence conditions for law without making any evaluative judgments about its moral qualities. The thick usage incorporates the content of the purely descriptive concept but also has evaluative content. Whether something that counts as law on the purely descriptive usage also counts as law on the thick usage depends on whether it is a good example of its kind according to certain moral standards that express values that should be satisfied by anything that antecedently satisfies the existence conditions defined by the descriptive usage. The evaluative content of the thick usage expresses the idea of what counts as law in its fullest or ideal sense—​i.e. what the law should be given some understanding of its point or aspirations. The relevant usage here is the purely descriptive usage that picks out systems of what is sometimes called “positive law”—​i.e. institutional normative systems that belong to the same class as the familiar paradigms comprising the focus of legal study and legal practice. Properly construed, then, the Coercion Thesis asserts that everything that counts as a legal system in the purely descriptive sense includes legal norms governing official behavior that authorize the judicial imposition of coercive sanctions for some violations of mandatory legal norms governing non-​ official behavior; mandatory legal norms governing non-​official behavior can be backed by a threat of a coercive sanction only insofar as legal recognition norms governing official behavior authorize courts to impose coercive sanctions for non-​compliance.

4 

The Coercion Thesis

2.  The Coercion Thesis The Coercion Thesis, expressed at the most general level, asserts that it is the metaphysical nature of a legal system that it is coercive in some sense—​i.e. that the properties constituting something as a legal system include that it backs some mandatory legal norms with a threat of a coercive sanction. In particular, the Coercion Thesis asserts that it is a conceptually necessary feature of a legal system that it includes norms authorizing the judicial imposition of sanctions as a formal response to non-​compliance with both court orders and some mandatory legal directives governing non-​official acts. This section explicates more fully the content of the Coercion Thesis to clarify what it does and does not express.

2.1  What Is Coercion? The term coercion, like the term law, has a purely descriptive usage as well as a thick usage incorporating evaluative and descriptive elements. To say that a directive or norm is coercive on the descriptive usage is to say that it is reasonably contrived to deter non-​complying behavior in virtue of being backed by a threat of detriment that counts as a coercive sanction; the characterization of something as coercive in this sense neither asserts nor implies a moral judgment of any kind with respect to its moral quality.3 To say that a directive is coercive on the thick usage is to say both that it is coercive in the descriptive sense and that the underlying threat of detriment for non-​compliance is morally problematic; the thick usage thus expresses a negative moral judgment about whether the measure characterized as coercive is permissible. A paradigmatic example of coercion on both the descriptive and thick usages would be an armed robber’s threat to kill me if I do not surrender my money. The robber’s demand for my money is coercive in the purely descriptive sense that it is backed by the threat of detriment that rationally competent subjects are likely, as a descriptive matter of contingent fact, to regard as a defeasible practical reason to comply because they should, as an objective matter of normative practical

3 Oxford Online Dictionary defines this purely descriptive usage as follows:  the concept-​ term coercion means “[t]‌he action or practice of persuading someone to do something by using force or threats”; see:  https://​www.lexico.com/​en/​definition/​coercion. Merriam-​Webster defines the concept-​term coerce as “to achieve by force or threat”; see:  https://​www.merriam-​webster.com/​dictionary/​coerce. It is important to note that it is enough to constitute something as coercive, according to the canons of ordinary usage, that it is backed by the threat of something reasonably contrived to increase the probability of compliance; it need not involve the threat of physical force. Threatening to reveal some personal secret unless some sum of money is paid is coercive despite not involving a threat of physical force.

The Coercion Thesis  5 rationality, regard it as such.4 It is also coercive in the evaluative sense partly constituting the thick usage that the demand coupled with the threat is morally wrongful. The point of coercion, on both the descriptive and thick usages, is to increase the probability that a rationally competent subject will comply by providing something she is likely to regard as a defeasible prudential reason to comply because she should regard it as such. The robber’s demand coupled with the threat is coercive in both respects: it attempts to increase the probability that I comply with her demand to surrender my money by illegitimately backing it with a threat of harm I am likely, as a descriptive matter of contingent fact, to regard as something I should avoid because I should, as an objective matter of normative practical rationality, regard it as such. The most salient conceptual feature of coercion, on the descriptive usage, is that the consequence threatened for non-​compliance is reasonably contrived to deter non-​compliance in virtue of its being likely to be experienced by a rationally competent subject as detriment that she would prefer to avoid, as a descriptive matter of contingent fact, because she should, as an objective matter of normative practical rationality, regard it as such; it is this feature of coercion that increases the probability that the subject will comply with the directive. P cannot coerce Q to do something by “threatening” a consequence that rationally competent subjects would, as a descriptive matter of contingent fact, characteristically regard either with indifference or as desirable; if Q is practically rational and like most people, P cannot coerce Q to do something by threatening to make Q eat something Q would very much like to eat and has no reason not to eat because P has not threatened anything that rationally competent subjects with the preference structure of Q should, as an objective matter of normative practical rationality, regard as something they have reason to avoid.5 The conceptual point of coercion is to deter non-​compliance by threatening the imposition of detriment as a response.

4 An agent P is rationally competent, as I intend the notion, if and only if P satisfies three conditions: (1) P’s perceptual beliefs about the basic characteristics of the empirical world generally conform to those of most other persons; (2) P is epistemically rational in virtue of being largely responsive to shared standards of epistemic justification; and (3) P is practically rational in virtue of being largely responsive to shared standards of practical rationality. The point of condition (1) is to exclude persons whose sensory beliefs are generally formed in response to what we take to be physical hallucinations. Someone deceived by a Cartesian evil demon might form beliefs, given her sensory input, in a manner largely responsive to the applicable norms of epistemic justification and might make decisions about what to do, given that sensory input, in a manner largely responsive to the applicable norms of practical rationality, but she is not properly characterized as rationally competent insofar as her behavior cannot be reliably regulated by systems of law. 5 This does not preclude the possibility of attempts to coerce that misfire but are nonetheless properly characterized as attempts to coerce compliance. The case where P shoots Q because Q failed to comply with P’s expressed threat to kill Q unless she surrenders her money is properly characterized as an unsuccessful attempt to coerce: insofar as the threatened consequence of non-​compliance is properly characterized as detriment, the threat is coercive and the making of the threat is an attempt to coerce compliance; insofar as the threat fails to elicit the desired behavior, it is properly characterized as an unsuccessful attempt to coerce.

6 

The Coercion Thesis

The claim that a directive is coercive, as I use the term, implies little with respect to the normative force of the threat. The normative force of a threat is partly determined by the objective and subjective (i.e. perceived) undesirability of the threatened detriment. In some cases, such as in the case of an armed robber’s threat to kill the victim if she does not comply, the undesirability of the threatened detriment is characteristically regarded as conclusive, as a descriptive matter of contingent fact, because it should be regarded as conclusive, as an objective matter of normative practical rationality: since premature death is, and should be, regarded by rationally competent subjects as among the worst fates that can befall a person, a credible threat of premature death is likely to be regarded as having significant normative force because it should be regarded as such. But, in other cases, such as in the case of a small financial penalty for parking illegally, the undesirability of the threatened detriment is characteristically regarded as falling short of conclusive, as a descriptive matter of contingent fact, because it should be regarded as falling short of conclusive, as an objective matter of normative practical rationality. In a case where the cost of a parking fine qualified by the perceived probability of incurring it is outweighed by the cost of taking the time to find a legal parking place qualified by the perceived probability of finding one, I will take the risk of being fined because I should do so: if I will lose my job if late to a meeting and cannot find a legal parking place without being late to the meeting, then it seems clear that I should park illegally—​as long as I do not park in front of a fire hydrant or in a space reserved for emergency vehicles. The normative force of a threat backing a directive is also partly determined by the perceived probability that the threatened detriment will be imposed on persons who do not comply with the directive. In the case of an armed robber’s threat, the probability of dying prematurely is properly treated, absent truly exceptional circumstances, as approximating 1. In the case of a municipality’s threat to impose a parking fine, the probability of being fined is, if my extensive experience parking illegally is any indication, less than .05. Both threats are properly characterized as coercive insofar as they are reasonably contrived to deter non-​compliance enough

But it is a conceptually necessary feature of coercion that P’s threat be grounded in a reasonable apprehension of what Q will regard as detriment. A reasonable apprehension of Q’s preferences can be grounded in knowledge of certain general facts about what rationally competent subjects characteristically regard as detriment because they should regard it as detriment—​at least insofar as one does not have specific information about Q that indicates that Q’s preferences diverge significantly in relevant respects from that of most people: absent reliable information that Q is suicidal, it is reasonable to presume that Q would regard the threat of being killed as detriment that would be likely to induce to her to comply with an armed robber’s threat. Nevertheless, it can also be grounded in specific knowledge about what Q idiosyncratically regards as detriment: if P knows that Q has a paralyzing fear of kittens and threatens to unleash a large litter of them in Q’s living space unless Q gives her $5.00, then it is reasonable to presume that the threat of having a litter of kittens released in her living space makes it more likely that Q comply with P’s demand for $5.00. Either way, if P threatens detriment that P has reasonable grounds to believe is likely to induce Q to comply with P’s demand, then the demand is properly characterized as coercive.

The Coercion Thesis  7 of the time to bring about the desired state of affairs; the desired state of affairs in the case of the robber is the victim’s surrendering her money while the desired state of affairs in the case of parking fines is the continuing availability of parking places for residents and visitors to the area, including first-​responders providing emergency services. It is enough, then, to warrant characterizing a threat as coercive that the threatened detriment is likely to deter non-​compliance with the norm n it is intended to enforce enough of the time to enable n to minimally succeed in bringing about the desired state of affairs—​regardless of whether the threatened detriment is regarded as severe. A parking law intended to ensure that certain spaces are available for people who reside in or visit the area must be backed, to count as coercive, by a fine sufficient to deter enough violations among would-​be offenders to ensure that these spaces remain available enough of the time for that law to have solved the problem it is intended to solve. Indeed, it is not implausible to characterize a threat to punish a child with a time-​out if she does not behave properly as coercive; although the threatened detriment might seem trivial to adults, it is regarded by the child as sufficiently unpleasant that it has a reasonable probability of inducing the desired behavior and is hence reasonably contrived to do so. What constitutes a coercive threat as likely, as a descriptive matter of contingent fact, to deter non-​compliance among rationally competent subjects enough to enable the law to solve the problem that it is intended to solve is that it should be, as an objective matter of normative practical rationality, regarded by rationally competent subjects as sufficient to deter non-​compliance in enough of the relevant situations. Law can efficaciously regulate the behavior of rationally competent subjects only insofar as officials can successfully predict how subjects are characteristically likely, as a descriptive matter of contingent fact, to respond behaviorally to the threatened detriment on the basis of how they should, as an objective matter of normative practical rationality, respond behaviorally to it. While rationally competent subjects sometimes respond idiosyncratically to legal norms, the practices constituting something as a system of law are not reasonably contrived to anticipate every possible idiosyncratic response; the best that officials can do by way of reliably predicting what will deter non-​compliance is to ascertain what should, as an objective matter of normative practical rationality, deter non-​compliance and assume that subjects of the system will characteristically do what rationally competent subjects should do.

2.2  What Is a Sanction? While the notion of coercion is conceptually liked to deterrence, the notion of a sanction, unlike the notion of coercion, is conceptually linked to punishment. Deterrence and punishment are conceptually related but distinct. Coercion is

8 

The Coercion Thesis

wholly forward-​looking; it seeks to deter non-​compliance by threatening detriment. Punishment has both forward-​and backward-​looking elements:  it is forward-​looking in virtue of being contrived to prevent violations by deterring them but is also backward-​looking in virtue of being contrived to respond after a violation by imposing the threatened detriment. A  sanction is coercive in virtue of seeking to deter violations, but it is punitive, in effect if not expressed intent, in virtue of being imposed in response to what is taken to be wrongful non-​compliance. Punishment should be distinguished from retaliation. Punishment incorporates a retributive element that presupposes it is justified under some set of norms that are not prudential in character because it is deserved under norms of that set. These norms belong to the same system of norms that includes the substantive norm dictating what the subject must do to avoid punishment: punishment authorized by law is legally justified while punishment permitted by morality is morally justified. Either way, something counts as punishment only insofar as part of the justification for imposing the threatened detriment makes reference to its being deserved under the relevant norms for non-​compliance culpable under those same norms. Retaliation, in contrast, is not necessarily norm-​governed in this respect. An act a is properly characterized as one of retaliation only insofar as the imposition of detriment is motivated by a desire for revenge rather than by considerations having to do with whether doing a is morally or legally justified. This does not preclude the possibility of something that counts as retaliation being justified under some set of norms: while it might be true that retaliatory acts are morally wrongful in virtue of being unmotivated by considerations of justice, the concept of retaliation, like the concept of punishment, is purely descriptive; depending on the content of the relevant moral or legal norms, something done as an act of retaliation, as far as the content of the concept is concerned, might be morally or legally justified. If the retaliatory imposition of detriment, for instance, is morally deserved because it is a proportional response to moral wrongdoing, then it might, as far as just the content of the concept is concerned, be morally justified. Unlike punishment, retaliation lacks a necessary forward-​ looking dimension. Insofar as punishment is norm-​governed, as a conceptual matter, in virtue of being justified under some set of norms that rationally competent subjects are presumed to know, the detriment is intended and reasonably contrived to deter because threatened by norms that those subjects are presumed to know. Insofar as retaliation lacks this norm-​governed aspect,6 it is possible to retaliate for acts

6 As Merriam-​Webster defines it, to retaliate is “to return like for like; to repay in kind”; see: https://​ www.merriam-​webster.com/​dictionary/​retaliate; as Oxford Online Dictionary defines it, to retaliate is “to make an attack in return for a similar attack”; see: https://​www.lexico.com/​en/​definition/​ retaliate.

The Coercion Thesis  9 that rationally competent subjects are not plausibly presumed to know will elicit retaliatory detriment; if you find the color red deeply offensive and you hit me for wearing red in your presence, the act of hitting me might count as retaliation but not as punishment—​unless I have some sort of special reason to know that you regard red as deeply offensive—​because there is nothing in the relevant facts reasonably likely to communicate the threat. Since retaliation lacks this conceptually norm-​governed dimension plausibly presumed to communicate a threat, it is not reasonably contrived, as a conceptual matter, to deter the acts triggering it and hence lacks the forward-​looking dimension of punishment. It is true that retribution, as our conceptual practices define the corresponding concept-​term, is purely backward-​looking insofar as it is exclusively concerned to remedy some past moral or legal wrong, but retribution expresses only the backward-​looking dimension of punishment.7 The forward-​looking dimension of punishment is expressed, at least with respect to institutional normative systems, by its deterrence function. While one might plausibly argue that objective moral norms governing punishment are concerned only with remedying past injustice by making sure wrongdoers get what they deserve, institutional punishment is concerned to maintain social order by deterring violations of the institutional norms that define and preserve that order.

2.3  What Is a Coercive Sanction? Something counts as a coercive sanction, for purposes of the Coercion Thesis, if it is reasonably contrived to deter and punish non-​compliance with some norm. The notion of a coercive sanction thus incorporates both the deterrence function of coercion and the punitive function of sanctions and is hence something that is reasonably contrived to perform both those functions. A coercive sanction, as I have explicated the notion here, is concerned to deter violations of the norms it backs by threatening something that counts, in effect if not expressed intent, as punishment for non-​compliance with them. The notion of a coercive sanction, as I use the corresponding concept-​term, includes both the penalties for criminal violations and the civil and criminal contempt sanctions that back court orders in both civil and criminal cases.8 Even if is true that the compensatory damages that can be imposed for breaches of duty 7 As Merriam-​Webster defines it, the term retribution means “the dispensing or receiving of reward or punishment  . . .  ; something given or exacted in recompense”; see:  https://​www.merriam-​ webster.com/​dictionary/​retribution; as Oxford Online Dictionary defines it, retribution means “punishment inflicted on someone as vengeance for a wrong or criminal act”; see: https://​www.lexico. com/​en/​definition/​retribution. 8 The civil contempt sanction is different from the paradigmatic sanctions of criminal law and the criminal contempt sanction insofar as it is intended only to induce compliance, rather than punish non-​compliance.

10 

The Coercion Thesis

under the civil law do not count as sanctions because they are not intended to deter violations of civil duties, the court orders intended to induce payment of those damages are contrived to induce compliance by deterring non-​compliance with the threat of incarceration and hence count as backed by coercive sanctions. The intent might not be punitive, as in the case of criminal sanctions, but this does not matter: the intent of an armed robber who backs her command with the threat to shoot is to induce compliance by deterring non-​compliance and not to punish non-​compliance. As long as the threat creates a prudential inducement to comply as a means of avoiding detriment, it is reasonably contrived to deter and punish non-​compliance and counts as a coercive sanction—​even if not explicitly intended as such. Thus defined, the term coercive sanction is purely descriptive. To say that a directive is backed by a coercive sanction neither expresses nor implies anything with respect to whether the imposition of detriment for non-​compliance is morally justified. It simply expresses that the directive is backed by a threat of detriment reasonably contrived to deter non-​compliance among rationally competent subjects in virtue of providing something they characteristically regard, as a descriptive matter of contingent fact, as having reason to avoid because they should, as an objective matter of normative practical rationality, regard it as such. While the notion of a coercive sanction is defined with reference to objective standards of practical rationality, the relevant standards are prudential in character. No moral judgment is expressed by the claim that something is backed by a coercive sanction. This is as it should be, and not just for the methodological reason that an account of a purely descriptive concept should express no moral judgment about the things to which the concept-​term applies; it is also because the claim that law is coercive in the ethically thick sense described above would have problematic substantive implications: the claim that law is essentially coercive in this ethically thick sense logically entails the claim that law is essentially morally illegitimate. This might be true, I suppose, but it is not a claim that should be implied by any conceptual claims about the content of the term law; the content of the term law, as defined by our empirically contingent linguistic practices, straightforwardly presupposes the conceptual possibility of a morally legitimate system of law.

2.4  The Imposition of Coercive Sanctions for Non-​Compliance Is Authorized by the Norms of the System The Coercion Thesis neither asserts nor entails that an institutional normative system counts as a legal system only if coercive sanctions are actually imposed on subjects for non-​compliance. The Coercion Thesis asserts only that it is a conceptually necessary condition for an institutional normative system to count as a system of law that it includes legal norms authorizing judges to impose something that

The Coercion Thesis  11 rationally competent subjects characteristically regard, as a descriptive matter of contingent fact, as detriment because they should regard it, as an objective matter of normative practical rationality, as detriment. Since a person can, as a conceptual matter, be institutionally authorized to do something she never does, there is nothing in the Coercion Thesis that even remotely suggests that judges sometimes impose the coercive sanctions they are authorized to impose. This is as it should be. The point of authorizing courts to impose sanctions for non-​compliance is to enable the system to deter non-​compliance as much as is consistent with other aims of the system. We know, given our experience in the world, that backing a mandatory legal norm with the threat of a non-​trivial sanction, other things being equal, significantly decreases the probability of non-​ compliance among rationally competent self-​interested subjects like us. Rationally competent self-​interested subjects who can be tempted to violate a mandatory legal norm when they believe doing so conduces to maximizing their interests can also, in virtue of being prudentially rational, be deterred from doing so by sanctions sufficiently severe to raise the obvious expected costs of non-​compliance enough to offset their obvious expected benefits. How successfully a legal system deters non-​compliance depends on a number of considerations. First, it depends on whether subjects know sanctions are authorized for non-​compliance; if rationally competent subjects do not know that non-​compliance will be met with a sanction, the threat of the sanction cannot, as a conceptual matter, deter non-​compliance. Second, it depends on whether they know how severe the authorized sanctions are; if rationally competent subjects know that the death penalty is authorized to punish jaywalking, they would jaywalk far less frequently than they would if they believe it is punished with a small fine. Finally, it depends on whether they believe the system is sufficiently reliable in detecting and punishing non-​compliance; if rationally competent subjects know officials always detect and punish non-​compliance, they will be far less likely to commit non-​complying acts. A state of affairs in which all non-​compliance is successfully deterred is exactly what legal systems in totalitarian police states seek to achieve by backing norms with the most severe detriment and deploying the most intrusive surveillance technologies available to detect non-​compliance. Some particularly repressive regimes attempt to minimize non-​compliance by conspicuously enforcing laws requiring persons to report crimes of which they have knowledge with draconian sanctions. These regimes are not always successful in deterring crime, but this is partly because their economies are not sufficiently robust to satisfy the basic needs of their subjects who frequently turn to crime out of frustrated desperation.9 The extent 9 Per capita annual income in North Korea, which express the average purchasing power of residents, was estimated in 2013 to be $1,800 (compared to $32,400 in South Korea and $49,720 in the U.S.). See “Datablog:  Facts are Sacred,” The Guardian; available at:  https://​www.theguardian.com/​world/​ datablog/​2013/​apr/​08/​south-​korea-​v-​north-​korea-​compared. See also https://​tradingeconomics.com/​

12 

The Coercion Thesis

to which a legal system can reduce the occurrence of assaults against persons and property depends as much on whether there are adequate material resources to satisfy the most urgent needs for them as on these other factors; the fewer urgent needs that go unsatisfied among subjects of a legal system, the fewer prudential incentives there are to aggress against persons and property. Although there are no existing legal systems that have completely succeeded in deterring non-​compliance, there is no reason to think that rationally competent self-​interested subjects like us could not be deterred from doing anything that might be prohibited by the relevant legal norms in a world where the worst effects of material scarcity have been minimized by the development of technologies and social institutions that maximize and distribute vital resources more equitably. The more severe the sanctions and the better the mechanisms for detecting non-​ compliance, the more successful the system will be in deterring non-​compliance—​ provided that the system distributes material resources in a manner ensuring that the most vital material needs are satisfied. The idea that something counts as a system of law only insofar as it does not deter all non-​compliance lacks any remotely plausible rationale. But even if there were reason to rule out the conceptual possibility of a legal system that successfully deters all non-​compliance, that tells us nothing about the Coercion Thesis. The Coercion Thesis, as I have explicated it, entails neither that such a legal system is conceptually possible nor that such a legal system is conceptually impossible. Whether or not such a legal system is conceptually possible rests on considerations that do not bear on whether or not it is a conceptually necessary condition for the existence of a legal system that it authorizes the imposition of coercive sanctions for non-​compliance with mandatory legal norms governing non-​ official behavior.

2.5  Sanctions Are Authorized as Judicial Response to Non-​Compliance with Some, and Not Necessarily All, Mandatory Norms of the System The Coercion Thesis neither asserts nor implies that there must be a coercive sanction authorized for every violation of something with the form of a mandatory legal norm. Although this might be true of every existing legal system with which we are familiar, the claim that it is a conceptually necessary condition for the existence of a legal system that it backs every mandatory legal norm with the threat of a sanction

united-​states/​gdp-​per-​capita. The disparity between per capita income helps to explain why the homicide rate per 1,000,000 people in North Korea (15.2 homicides per 100,000 people per year) is consistently higher, despite its deployment of draconian detection methods and sanctions, than that of the U.S. (4.2 homicides per 100,000 people per year). Ibid.

The Coercion Thesis  13 implies the facially implausible claim that an institutional normative system resembling that of the U.S. in every salient respect except that one of its mandatory legal norms is not backed by the threat of a sanction does not count as a system of law. But the problem is worse than this: the idea that it is a conceptually necessary condition for the existence of a legal system that it backs every mandatory legal norm with the threat of a coercive sanction would call into question whether we are correct with respect to what we regard as paradigmatic instances of legal systems that ground conceptual theorizing about law. Insofar as we characterize the institutional normative system in the U.S. as a system of law without knowing whether every mandatory norm of the system is backed by a coercive sanction, we are not epistemically justified in attributing that status to the system. Given that our conceptual practices determine what counts as a legal system for us, this claim is absurd for that reason: I have no idea whether every mandatory legal norm in the U.S. is backed by the threat of a sanction, but I am quite certain that the U.S. has a legal system. While the Coercion Thesis is consistent with the existence of a legal system in which all mandatory legal norms are backed by the threat of a sanction, it expresses no more than that it is a conceptual truth that every legal system backs some mandatory legal norms with the threat of a coercive sanction. The Coercion Thesis is not only agnostic with respect to the existence of a legal system in which all mandatory norms are backed by the threat of a sanction; it is also agnostic with respect to whether content with the logical form of a mandatory norm (i.e. in virtue of deploying the relevant deontic operators expressed by the terms “must” and “shall”) but not backed by a sanction counts as a mandatory norm.

2.6  Sanctions Need Be Authorized Only with Respect to Non-​Official Behavior The Coercion Thesis asserts that it is a conceptual truth that legal systems back some mandatory legal norms governing non-​official behavior with the threat of a coercive sanction and hence asserts nothing with respect to whether it is a conceptual truth that legal systems also back some mandatory norms governing official behavior with the threat of a coercive sanction. Insofar as a conceptual claim about the content of some mandatory norms governing non-​official behavior logically implies nothing with respect to the content of mandatory norms governing official behavior, the Coercion Thesis takes no position, either explicitly or impliedly, with respect to whether mandatory legal norms governing official behavior are, as a matter of conceptual necessity, backed by such threats.10 10 The relevant behaviors are properly characterized as non-​official in both systems of municipal law and in systems of international law, but the subjects of these constraints on non-​official behavior differ according to whether the system is one of municipal law or is instead one of international law. In systems of municipal law, the subjects are both natural persons, such as human beings, and non-​natural

14 

The Coercion Thesis

It is not implausible to think that mandatory recognition norms are, as a conceptual matter, backed by something that counts as a coercive sanction. Insofar as it is a conceptual truth that officials take the internal point of view towards the rule of recognition, it is also a conceptual truth that officials characteristically express disapproval of behavior that does not conform to its requirements. Further, insofar as the expressed disapproval by one’s peers towards one’s behavior is likely, as a descriptive matter of contingent fact, to be regarded as detriment that defines a prudential deterrent to non-​compliance because it should, as an objective matter of normative practical rationality, be regarded as detriment, it is not implausible to think that the prospect of incurring such expressed disapproval counts as a threat of a coercive sanction. But there is nothing in the rule of recognition that authorizes officials to publicly express disapproval in response to non-​complying behavior. The rule of recognition is a social rule manufactured by practices that include the public expression of disapproval for non-​compliance. The rule of recognition counts as a social rule because officials take the internal point of view towards the norm, which is constituted, in part, by a propensity to criticize deviations from the norm; in the absence of that propensity, there is no social rule of recognition. The propensity is thus built into the construction of the norm itself and is hence not authorized by the rule of recognition.11 A rule of recognition might include norms explicitly authorizing the coercive removal of wayward officials, but it could do so only in virtue of officials converging on taking the internal point of view towards a norm governing official behavior that authorizes the coercive removal of officials. Constitutions, which play a role in determining the content of the rule of recognition, typically include norms explicitly authorizing removal procedures. Article II, Section 4 of the U.S. Constitution expressly authorizes Congress to coercively remove the President from office for

entities that are treated as persons for purposes of the law, such as corporations; in international law, the subjects are nation-​states. Although something properly characterized as a system of international law regulates the behavior of nation-​states by regulating the behavior of persons who act as officials on behalf of those nation-​states, the persons who serve as officials of the nation-​state do not necessarily serve as officials in the system of international law. The rule of recognition giving rise to something counting as a system of international law defines duties that bind only those persons who serve as officials of the international system in virtue of being empowered to recognize, apply, or enforce the valid norms of that system of international law. Since that rule of recognition neither empowers nor binds officials of the subject nation-​states, the valid norms of international law govern only non-​official behavior. 11 This suggests another important conceptual distinction: coercive sanctions backing mandatory legal norms governing non-​official behavior are formal in virtue of being explicitly authorized by law while coercive sanctions backing a social rule of recognition are informal in virtue of not being explicitly authorized by law. While the sanctions backing the rule of recognition are neither authorized nor formal, they are nonetheless legally permissible; the difference between what is authorized and what is permissible is that doing what is authorized requires authority to do it while doing what is merely permissible does not require authority to do it. These issues will be discussed in more detail in Kenneth Einar Himma, Authority and the Nature of Law (forthcoming 2021).

The Coercion Thesis  15 certain offenses:  “The President, Vice President and all the civil officers of the United States, shall be removed from office on impeachment for, and conviction of, treason, bribery, or other high crimes and misdemeanors.” Insofar as the content of the U.S. Constitution determines the content of the rule of recognition governing official behavior, officials in the U.S. converge on practicing a norm that requires them to abide by the Constitution’s formal requirements for removing officials.12 But it is not a conceptual truth that everything that counts as a legal system contains norms explicitly authorizing the coercive removal of officials. There are clearly conceptually possible legal systems with sovereign rulers who cannot be removed by any legal mechanism and hence conceptually possible worlds in which legislators cannot be removed by any legal mechanism. If it is also conceptually possible for there to be a legal system without a norm explicitly authorizing the removal of misbehaving judicial officials, then it is not a conceptually necessary feature of a legal system that it includes legal norms authorizing the removal of any persons who serve as officials of the system. Either way, the Coercion Thesis, as I have expressed it, is agnostic with respect to whether it is a conceptual truth that the rule of recognition includes norms authorizing the imposition of coercive sanctions for non-​compliance with mandatory legal norms governing official behavior.

2.7  Sanctions Are Authorized for Some Mandatory Norms Prohibiting Acts that Breach the Peace The Coercion Thesis asserts that it is a conceptual truth that every legal system authorizes coercive sanctions for some non-​official acts that breach the peace. Municipal legal systems must, as a conceptual matter, prohibit enough acts of aggression to permit subjects to live together and enjoy the benefits of social cooperation in a community of natural persons. International legal systems must, as a conceptual matter, prohibit enough acts of aggression to permit nation-​states to exist peacefully together as members of a community of nations. Legal systems regulate the behavior of members of something that counts as a community; there can be no community in conditions defining a war-​of-​all-​against-​all such as occurs in a Hobbesian state of nature. The Coercion Thesis should not be taken to imply that it is a conceptually necessary condition for an institutional normative system to count as a legal system that it prohibits every act that breaches the peace. It is clear there are conceptually possible legal systems in which some acts breaching the peace are permitted; history is replete with examples of institutional normative systems that count as systems of law despite allowing some morally wrongful acts of aggression against 12 Insofar as these removal procedures count as coercive sanctions, it is because they are reasonably contrived, in effect, to deter and punish non-​compliance; they need not be intended as punishment.

16 

The Coercion Thesis

disfavored minorities—​and, indeed, are regarded as paradigm instances of legal systems. There were federally enacted norms in the U.S. in the nineteenth century that permitted the kidnapping, incarceration, and forced labor of black people by white people who were treated by the law as their “owners,” and there is simply no question that the U.S. had something properly characterized as a legal system.

3.  The Coercion Thesis and Morality The breaches of the peace prohibited by law in every existing legal system in our world are, as a descriptive matter of contingent fact, generally thought by subjects to be morally wrongful. There is no existing society with a municipal legal system in which the acts of homicide and theft prohibited by law are not generally thought to be wrongful by rationally competent subjects of the law; likewise the acts of aggression against nations prohibited by our system of international law are generally thought to be wrongful by rationally competent officials and residents of those nations. The core prohibitions of the law are everywhere regarded, as a descriptive matter of contingent fact, as incorporating mandatory moral norms. This is not an accident. An institutional normative system is not reasonably contrived to regulate behavior unless the peace is kept enough for subjects to live together in a community; in our world of acute material scarcity, this can be done by a legal system only insofar as it includes mandatory legal norms prohibiting acts commonly thought to violate moral norms protecting bodily integrity and property. Beings like us are far more likely, other things being equal, to respond with force to aggressions we believe impinge on these morally protected interests than to aggressions we believe do not impinge on such interests. Since law can be efficacious in keeping the peace only insofar as it deters enough violence to enable us to live together, it must prohibit aggressions that are highly likely to escalate into conflicts breaching the peace. These acts include assaults on persons and on property—​despite the fact that it is not implausible to think that some property crimes do not involve violence that directly constitutes a breach of the peace.13 The reason has to do with the prudential importance of our interests in property relative to other interests. I would rather that you break my arm than steal my life savings; the state of affairs in which I cannot use my arm for an extended period of time is preferable to me than the 13 One might object to characterizing property crimes as assaults; on this line of objection, crimes against property like theft do not count as assaults because only things that have interests can be the subject of an assault, and property has no interests. This is a fair point, but I use the term only to simplify the exposition; what I characterize as an assault on property might be more accurately characterized as an assault on the property interests of the person who owns the property. Nothing in this volume turns on whether I characterize these acts as assaults or use more words to express the matter without using the term. Similar issues arise with respect to the use of violence; but I think it is plausible to think that breaking-​and-​entering counts as an act of violence.

The Coercion Thesis and Morality  17 state of affairs in which I suddenly find myself without material resources. Either way, it should be clear that passionately self-​interested beings like us in worlds of acute material scarcity like ours cannot live together in a community without efficacious legal prohibitions on such assaults. The Coercion Thesis, for this reason, includes assaults on property that arguably do not directly breach the peace in addition to the assaults on persons that do. Systems of law in our world prohibit these acts, then, not only because we believe they are morally wrongful assaults on interests that are vital to our well-​being, but also because they can escalate conflicts in the form of a cycle of violence involving increasingly larger circles of subjects. How much any particular assault on persons or property threatens the peace in a community depends on how victims respond. A reasonably self-​contained fist-​fight posing no appreciable risks to innocent persons can easily get out of hand and become something approximating a riot posing substantial risks to the well-​being of other persons; a reasonably self-​contained armed conflict between nations posing no appreciable risks to non-​combatant nations can easily get out of hand and become a global war. The point of law is to prevent the sparks that can ignite the forest and make it uninhabitable. This presupposes that beings like us characteristically care about the requirements of morality, but that is not a problem—​because we do. Even persons doubting the objectivity of morality will react to these assaults as though they have been morally wronged; I have never met a moral nihilist who will walk away from some perceived slight on the strength of a conviction that there is no right and wrong. Similarly, even persons who believe they have been wrongly convicted of a criminal offense, when the facts of the crime are not disputed, will argue passionately that they were morally justified in doing what they were convicted of doing and that their convictions were morally wrongful for this reason. So much does morality matter to us that we treat someone who does not care at all about its requirements as suffering from a mental illness. It is not just that the various anti-​social personality disorders are defined by psychologists in terms of the extent to which the person cares about moral requirements; sociopathy and psychopathy are both defined in terms of a symptomology that includes a lack of due emotional regard for those requirements.14 It is also that the law defines legal competency partly in terms of moral competency; someone is insane and immune from criminal liability if it can be shown that they cannot conform their behavior to the requirements of morality either because they cannot tell the 14 Insofar as psychopaths and sociopaths are treated as accountable for their acts, they are thereby presumed clinically sane. Someone who is not treated as accountable for their acts is presumed to be criminally insane in virtue of not being able to rationally direct their behavior because, as the matter is commonly put, they do not know the difference between right and wrong. Psychopaths and sociopaths who are accountable for their behavior are presumed not to view what they know to be basic moral requirements as normative, while persons who are clinically insane are presumed not to understand these requirements. In neither case are such persons appropriately responsive to the objective requirements of normative practical rationality—​though the disability is different in each case.

18 

The Coercion Thesis

difference between right and wrong or because their behavior is psychologically compelled in problematic ways.15 Insofar as a conceptual theory of law is supposed to explicate our conceptual practices to enable us to better understand ourselves, it must take these elements into account because they characteristically inform our legal practices. Even so, it seems that there could be a legal system that does not prohibit assaults generally thought by subjects of the system to be wrongful. It is not preposterous to think that rationally competent subjects like us could have an efficacious institutional normative system that counts as a legal system even if we all became convinced of moral nihilism; legal systems are, after all, contrived to provide non-​ moral reasons that will persuade people who are not moved by moral convictions to do what law requires. Either way, insofar as people in such a society are still prone to commit acts that create conflicts likely to breach the peace, they would still need a system of law that prohibits such socially disruptive acts.

4.  The Coercion Thesis and the Motivations of Subjects The Coercion Thesis implies nothing with respect to what motivations rationally competent subjects have, as a descriptive matter of contingent fact, that explains their behavior when it satisfies mandatory legal norms governing non-​official behavior. In particular, it does not imply that it is a conceptual truth that rationally competent subjects of a legal system sometimes obey the law simply as a means of avoiding the detriment imposed by coercive sanctions for non-​compliance. The Coercion Thesis is logically consistent with a state of affairs in which subjects always comply, when they comply, with these mandatory legal norms for reasons having nothing to do with those norms. The Coercion Thesis asserts only that it is a conceptually necessary condition for something to count as a legal system that it backs certain mandatory norms prohibiting breaches of the peace with the threat of coercive sanctions. While the Coercion Thesis presupposes the claim that rationally competent subjects characteristically regard, as a descriptive matter of contingent fact, coercive sanctions as detriment they have reason to avoid because they should, as an objective matter of normative practical rationality, regard it as such, this latter claim does not imply either that rationally competent subjects characteristically comply, as a descriptive matter of contingent fact, with the law only to avoid such detriment or that they should comply, as an objective matter of normative practical rationality, with the law only to avoid such detriment. The Coercion Thesis simply acknowledges the fact that the practices constitutive of a legal system are reasonably contrived to 15 If we do not have free will, the behavior of people without severe mental health issues is also psychologically compelled.

The Coercion Thesis and the Motivations of Subjects  19 minimize breaches of the peace only insofar as they back legal norms prohibiting such breaches with the threat of a coercive sanction. Every rationally competent subject regards the prospect of having coercive sanctions imposed against her as normatively relevant—​even when it does not change her behavior. Criminals regard the prospect of having sanctions imposed against them for non-​compliance as normatively relevant even when they have decided to commit the worst acts of violence prohibited by law. Every rationally competent criminal will consider the probability that she will be caught and punished for a crime before she commits it: if she decides that the probability is significant, she will either delay committing the crime or take additional precautions to conceal her involvement; if she decides that the probability is insignificant, she will commit the crime without additional precautions. Morally decent people likewise take into account the deterrent value of the coercive sanctions that can be imposed on them for violating laws prohibiting more innocuous municipal offenses. People inclined to exceed the speed limit on a highway—​which, if my experience is any indication, includes nearly everyone who feels comfortable driving on a highway—​will refrain from doing so if they believe there is a state patrol officer nearby. My partner and I overpay our taxes each year as a means of ensuring we never have any legal problems with the IRS, and not as an expression of moral rectitude. While the IRS always, to its credit, determines that we have overpaid and refunds the appropriate amount, that is irrelevant as far as we are concerned: we are happy to pay something resembling a fee to stay out of tax trouble with one of the most aggressive arms of the federal government when it comes to prosecuting violations of law. Nor is it preposterous to suppose that the threat of coercive sanctions commonly deters good people from acting out of character and breaching the peace when they get angry. If there are people among us incapable of getting angry enough to assault another person, they are exceptional. The vast majority of us are capable of becoming, if only for a potentially life-​changing instant, a Holmesian “bad man” who can be dissuaded from committing a criminal act only by being convinced severe punishment is inevitable.16 Every rationally competent subject regards the prospect of having sanctions imposed for non-​compliance as normatively relevant; this is why the practices constitutive of law are reasonably contrived to keep rationally competent self-​interested subjects out of a perpetual condition of war of all against all. But none of this implies it is a conceptual truth that people whose behavior complies with the law must be motivated, even in part, by a desire to avoid the sanctions judges are authorized to impose on subjects for non-​compliance. While there are

16 Road rage all-​too-​commonly brings out the worst in us. An attorney in Atlanta, for example, recently became enraged and killed someone he incorrectly believed hit his car with a golf ball; it was his first violent offense. See https://​www.cnn.com/​2019/​08/​26/​us/​atlanta-​lawyer-​road-​rage-​murder-​ charge/​index.html.

20 

The Coercion Thesis

some legally prohibited acts from which I abstain, such as jaywalking in front of a police officer, only out of a desire to avoid the sanctions, the fact that I have never intentionally killed someone is not at all explained by the fact that murder is punishable by a long term of incarceration; however, if I am ever tempted to do so, I would certainly regard the fact that I would likely be sentenced to prison as normatively relevant. The Coercion Thesis should hence not be thought incompatible with the idea that most of us are morally decent people who refrain from doing bad things for reasons having nothing to do with law; however, it is the fact that there is a legal system that coercively keeps the peace that helps to explain why there are morally decent people. Being minimally decent from the standpoint of morality is a luxury made possible only by the existence of a legal system that succeeds in keeping the peace enough for rationally competent subjects to feel they need not expend all their psychic and physical resources to defend themselves against an unending series of violent threats. Just as young children require coercive discipline, which can consist entirely of minor sanctions such as the preposterously precise one-​ minute-​per-​year-​of-​age time-​out sanction,17 to learn how to become morally good persons, rationally competent self-​interested subjects in a world of acute material scarcity must be sufficiently restrained by some kind of coercive authority to fruitfully cultivate what is latently decent in them. Life in a Hobbesian state of nature does not conduce to the development of moral decency. The Coercion Thesis is not equipped to tell us anything about why any or all people refrain from violating law because the considerations that explain those motivations are contingent upon a number of circumstances having nothing to do with our conceptual practices with respect to using the term law—​such as how successfully they have learned to be kind and loving with respect to other people. A conceptual theory of law must do no more in this regard than explain how those practices constitutive of law are reasonably contrived to keep the peace among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours; a conceptual theory of law that includes the Coercion Thesis does exactly this by explaining how those constitutive practices can provide subjects with a prudential disincentive to violate the law. But it cannot tell us anything about the motivations that people actually have, as a descriptive matter of contingent fact, that explain why their behavior conforms to the law.

5.  Outline of the Arguments The remainder of the first part of this book is devoted to explicating and making a preliminary case for the Coercion Thesis. Chapter  2 outlines the 17 See, e.g., “Essentials for Parenting Toddlers and Preschoolers,” Center for Disease Control and Prevention; available at: https://​www.cdc.gov/​parents/​essentials/​timeout/​steps.html.

Outline of the Arguments  21 modest approach to conceptual analysis deployed throughout this volume, which is grounded in identifying the ordinary lexical meanings of the term law together with the underlying philosophical assumptions about the metaphysical nature of the thing to which the term refers. Chapter 3 considers the various dictionary reports of the lexical meanings of the concept-​term law and shows how the linguistic conventions determining these meanings are conditioned by and reflect our universal empirical experience with institutional normative systems we characterize as systems of law. Chapter 4 argues that the conceptual function of a legal system is to keep the peace through the governance of norms metaphysically capable of guiding behavior and that a legal system is not reasonably contrived to minimally succeed in keeping the peace among rationally competent self-​interested subjects like us in worlds of material scarcity like ours, unless it authorizes the imposition of coercive sanctions for non-​compliance with mandatory legal norms prohibiting breaches of the peace. The second part shows how the Coercion Thesis explains law’s presumed conceptual normativity. Chapter 5 distinguishes three conceptual problems of legal normativity and situates them with respect to different kinds of reasons. Chapter 6 shows how the Coercion Thesis is uniquely equipped to explain how the practices constituting something as a system of law create reasons to conform to legal norms prohibiting acts that breach the peace (the How Problem). Chapter 7 addresses the issue of whether the practices constitutive of law are reasonably contrived to create second-​order reasons not to act on some excluded class of first-​order practical reasons (the Order Problem). Chapter 8 shows how the Coercion Thesis is uniquely equipped to explicate the content of the first-​ order reason favoring compliance that law is reasonably contrived to provide (the Content Problem). Insofar as law’s conceptual normativity can be explained only by assuming that the practices constitutive of law include authorizing coercive sanctions for non-​compliance, the chapters in this section augment the arguments made in the first section. The third part addresses what are sometimes thought to be counterexamples to the Coercion Thesis. Chapter 9 is concerned to show that what we call international law in our world, regardless of whether it is really law, is not a counterexample to the Coercion Thesis because certain mandatory legal norms of the system are backed by economic or diplomatic measures that count as coercive sanctions for purposes of the Coercion Thesis. The final chapter, Chapter  10, argues that the so-​called society-​of-​angels argument depends on assumptions about the psychological features of the “angels” that are too far removed from what is possible for rationally competent self-​interested subjects like us in the conditions of acute material scarcity in which we live to tell us anything of theoretical significance about our conceptual practices and hence cannot ground a persuasive counterexample to the Coercion Thesis.

2

Methodology and the Nature of Law This volume defends the Coercion Thesis, according to which it is a conceptual truth that anything that counts as a system of law backs some mandatory legal norms governing non-​official behavior with the threat of a coercive sanction. If the Coercion Thesis is true, it is because our conceptual practices entail that only institutional normative systems backing some mandatory norms governing non-​ official behavior with the threat of a coercive sanction count as systems of law. This chapter explicates and defends the philosophical methodology deployed throughout this book.1 It argues, to begin, that a conceptual theory of a thing can be properly grounded only in an analysis of the canons of ordinary usage governing use of the associated concept-​term together with the assumptions about the metaphysical nature of the thing to which the concept-​term refers that help to determine the content of these canons. The chapter goes on to explain how conceptual analysis, thus conceived, can tell us something about the objective world despite being wholly grounded in intersubjective conventions. It ends by describing a two-​ step methodology for explicating the metaphysical nature of a thing as it is determined by the canons governing the use of the associated concept-​term.

1.  Explicating the Metaphysical Nature of a Thing: Three Types of Modal Claim Claims about the nature of a thing are claims about what properties are inherent to things of that kind.2 Properties defining the nature of some kind of thing C in virtue of being inherent to things that count as C explain why something that instantiates those properties counts as C—​because it is wholly in virtue of instantiating

1 Philosophical methodology is distinguished from sociological methodology in virtue of being concerned to determine what is true of law in every conceptually possible world rather than what is true of law in just the world of our empirical experience; sociological methodology can epistemically justify only contingent claims about what is typically true of law in this world while philosophical conceptual analysis is concerned to justify claims about what is true of law in every possible world. Only a theory grounded in this philosophical methodology counts as a piece of conceptual analysis. 2 As Oxford Online Dictionary describes the relevant usage, the concept-​term nature means “the basic or inherent features, character, or qualities of something”; see: https://​www.lexico.com/​en/​ definition/​nature. As Merriam-​Webster describes it, the concept-​term nature means “the inherent character or basic constitution of a person or thing,” or its “essence”; see: https://​www.merriam-​webster. com/​dictionary/​nature. Coercion and the Nature of Law. Kenneth Einar Himma, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198854937.001.0001

24 

Methodology and the Nature of Law

those properties that something counts as C.3 Instantiating the properties of being unmarried, being an adult, and being a male explains why something counts as a bachelor because something counts as a bachelor wholly in virtue of instantiating those properties. While anything that counts as a bachelor has an infinite number of other properties,4 none have anything to do with why it counts as a bachelor because they are not inherent properties of bachelorhood. Insofar as something counts as C wholly in virtue of instantiating each of the inherent properties of C, it follows that nothing lacking any of those properties counts as C.5 If the inherent properties of a bachelor include the properties of being unmarried, being an adult, and being a male, then nothing lacking even one of those properties counts as a bachelor; if something lacking one of these properties nonetheless counts as a bachelor, then nothing that counts as a bachelor counts as a bachelor in virtue of instantiating all of these properties. The inherent properties defining the nature of C thus define necessary and sufficient conditions for something to count as C: anything instantiating each of the inherent properties defining the nature of C counts as C in virtue of instantiating each of these inherent properties, and anything not instantiating each of the inherent properties of C does not count as C in virtue of not instantiating one or more of these inherent properties. If the inherent properties of bachelorhood are exhausted by the properties of being unmarried, being an adult, and being a male, then something counts as a bachelor if and only if it is an unmarried adult male. These necessary and sufficient conditions are sometimes (and somewhat inaptly) called existence conditions. The necessary and sufficient properties for counting as C are characterized as defining existence conditions because they define the reference class of the concept-​term “C” we use to talk about C—​and not because they bring some particular thing that counts as C into existence.6 What explains why something that counts as a bachelor exists is whatever explains why that person was born, continues to be alive after the age of majority, is male, and is not married. What explains why that existing thing counts as a bachelor—​and is hence a member of the reference class of the concept-​term bachelor—​is that it instantiates 3 The underline indicates that I am talking about the thing to which a concept-​term refers whereas the absence of an underline indicates that I am talking about the concept expressed by the concept-​ term; when I enclose something in quotation marks, it is to indicate that I am talking about the concept-​ term. On these conventions, all and only Cs fall under C and hence count as C in the sense that all and only Cs are properly characterized as “C.” These conventions are needed only to disambiguate sentences in which the use of a variable is needed because of the generality of the claim. When I am talking about specific examples in natural language, like bachelors or legal systems, the relevant differences will be obvious without being marked with unsightly underlines and quotation marks. 4 The proof is straightforward. Since every bachelor is an adult male and no two-​year-​old is an adult male, every bachelor has the relational property of being older than any human being who is one and 1/​ n years old for every natural number n. 5 Something counts as C if and only if it falls under C. Counting as C connotes that something falls under the concept because it instantiates each of the inherent properties of C. 6 The reference class of a concept-​term referring to C is the set of things to which the concept-​term properly applies; the reference class of the concept-​term “C” is hence just the set of all Cs.

Explicating the Metaphysical Nature of a Thing  25 all the inherent properties of bachelorhood that define its nature.7 These existence conditions express the nature of C in virtue of exhaustively identifying the properties inherent to everything counting as C. A claim about the nature of some C is a metaphysical claim about what is necessarily or not necessarily true of C. To say that the property of being unmarried is part of the nature of bachelorhood is to say that it is not possible for something not unmarried to count as a bachelor; likewise, to say that the property of authorizing coercive sanctions for non-​compliance with some mandatory norms governing non-​official behavior is part of the nature of law is to say that it is not possible for something not authorizing coercive sanctions for non-​compliance with some mandatory norms governing non-​official behavior to count as law. Theorizing about the metaphysical nature of some kind of thing C raises difficult issues of epistemic justification because claims about what is necessarily or possibly true are not empirical claims. Empirical claims express something that can be reliably verified—​and hence epistemically justified—​by deploying the senses to observe whether what is asserted by the claim is true; the proposition expressed by the sentence “it is snowing outside” is an empirical claim because its truth can be verified by going outside and looking to see if snowflakes are falling from the sky. In contrast, the proposition expressed by the sentence “it is not possible for it to snow unless the temperature is below 36 degrees Fahrenheit” is not wholly empirical because its truth cannot be verified simply by going outside and looking around—​no matter how many times one goes out on days in which the temperature is above 36 degrees Fahrenheit.8 Complicating these epistemic issues is the fact that there are different types of necessity operators independent of one another in the sense that no one can be fully explained in terms of the others. The modal claim expressed by “it cannot snow unless the temperature is below 36 degrees Fahrenheit” cannot be epistemically justified by the same strategies that enable us to epistemically justify the modal claim expressed by “it is not possible for 1 + 1 = 3”—​assuming that both are true and epistemically justified. 7 These inherent properties might be either conjunctive or disjunctive in character. If the inherent properties of C, which define the existence conditions for C, are exhausted by the conjunction of properties p and q, then something counts as C if and only if it instantiates both property p and property q. If the inherent properties of a C are exhausted by the disjunction of properties p and q, then something counts as C if and only if it instantiates either property p or property q (or both); in this latter case, the essential property is disjunctive in the sense that it is the property of being either p or q. Either way, the existence conditions for C are defined in terms of a set of inherent properties that are singly necessary and jointly sufficient for something to count as C. 8 This is part of what led Hume to believe that causal claims cannot be epistemically justified. Since causal claims do not express relations of ideas that can be conclusively justified intuitively by pure reason, they express matters of fact that can be justified only empirically by sense observation. But since we can directly observe only correlations between a putative cause P and its putative effect Q and not the causal mechanism thought to explain the correlation, claims to the effect that P causes Q cannot, on his view, be epistemically justified. David Hume, Enquiries Concerning Human Understanding and Concerning the Principles of Morals (Oxford University Press 1975).

26 

Methodology and the Nature of Law

The reason different strategies are needed to justify these claims is that modal claims are necessary or possible only relative to some set of favored claims. The claim expressed by “it cannot snow unless the temperature is below 36 degrees Fahrenheit” says something about what is possible given the content of the physical laws describing causal regularities in our material universe; whether or not that is true depends on the content of those physical laws. In contrast, the claim expressed by “it is not possible for 1 + 1 = 3” says something about what is possible no matter what those physical laws happen to be; whether or not that claim is true has nothing to do with the content of the physical laws governing our material universe because that claim has nothing to do with causation. The claim expressed by this latter sentence is epistemically justified relative to the axioms of Zermelo-​Fraenkel set theory and classical logic. There are at least three types of modal claim: nomological, metaphysical, and logical. A claim is logically necessary if and only if it follows deductively from some set of favored logical axioms, which include axioms of propositional and modal logic. The claim expressed by the schema “if P, then P” is logically necessary relative to some set of favored logical axioms, no matter what proposition P is taken to express, in virtue of being logically entailed by just the axioms of that set; this claim is logically necessary in virtue of being deducible by the application of valid inference rules to the relevant set of logical axioms, each of which is also presumed to be logically necessary. Since these axioms are logically necessary and the valid rules of inference are truth-​preserving in the sense that they cannot result in the deduction of something false from something true, any claim deduced from just these axioms using just these truth-​preserving inference rules is also logically necessary. A claim is metaphysically necessary if and only if (1) it follows deductively from some set of favored logical axioms together with some non-​empty disjoint set of basic claims assumed to be true no matter what general principles describing causal regularities in this world happen to be true;9 and (2) it does not follow deductively from just the favored set of logical axioms.10 The claim that nothing can be red all over and green all over at the same time is thought to be metaphysically necessary in the sense that it would be true no matter what the so-​called laws of nature had turned out to be but does not follow from just the set of favored logical axioms; it hence depends for its deduction on some set of favored metaphysically necessary claims.

9 This set defines the set of favored basic metaphysical claims. 10 Axioms of mathematical theories are neither axioms of logic nor physical laws describing causal regularities in the world. If they are necessarily true, it is in virtue of either expressing some epistemically privileged set of claims including basic metaphysical claims also thought to be necessarily true or being deducible only from sets of propositions including both the axioms of logic and these basic metaphysical claims; if it is necessarily true that 1 + 1 = 2, that proposition is metaphysically necessary, and not logically or nomologically necessary. Axioms of mathematical theories, if necessarily true, are metaphysical in character.

Explicating the Metaphysical Nature of a Thing  27 A claim is nomologically necessary if and only if (1) it follows deductively from the union of some set of favored logical axioms, some set of favored metaphysical claims, and a set of physical laws describing what we assume are necessary causal regularities in our world, such as those of physics;11 and (2) it does not follow logically from just the union of the set of favored logical axioms and the set of favored metaphysical claims. The claim that the temperature at which water freezes is 32 degrees Fahrenheit expresses a nomologically necessary truth in virtue of being deducible only from a set of claims that includes the physical laws we believe correctly describe the necessary causal regularities in our world. Claims about what is or is not necessarily true can be expressed in the possible-​ worlds talk presupposed by the semantics of modal logic: a claim is necessary relative to some set of favored claims if and only if it is true in all possible worlds in which every member of that set of favored claims is also true. A claim is necessarily true relative to the same set of favored logical axioms (i.e. logically necessary) if and only if it is true in every possible world in which each of those axioms is true; it is necessarily true relative to the set of favored metaphysical claims (i.e. metaphysically necessary) if and only if it is true in every possible world in which each of those favored metaphysical claims and favored logical axioms is true; and a claim is necessarily true relative to the physical laws describing causal regularities in our world (i.e. nomologically necessary) if and only if it is true in all possible worlds in which each of those physical laws, favored metaphysical claims, and favored logical axioms is true.12 The existence conditions of C not only explain why something counts as a C but also determine standards that epistemically justify claims that something counts as a C. Some existing person counts as a bachelor wholly in virtue of instantiating each of the inherent properties of bachelorhood; the existence conditions are thus ontological in the sense that they explain why something counts as a bachelor (or, less perspicuously put, exists as a bachelor). We can hence ascertain in an epistemically justified manner whether some person counts as a bachelor by ascertaining in an epistemically justified manner whether that person instantiates each of the inherent properties of bachelorhood; the existence conditions also define epistemic standards that enable us to distinguish persons who count as bachelors from persons who do not count as bachelors insofar as we can reliably identify whether an existing thing instantiates the inherent properties of bachelorhood.13 11 It should not be thought that the deduction must involve at least one proposition from each set. One might be able to deduce a nomological claim without the help of, say, a metaphysical claim. If so, the relevant set of metaphysical claims would be empty. 12 A claim is possibly true if and only if it is not necessarily true and not necessarily false. Contingent claims are possibly true and possibly false; empirically contingent claims are true in the world of our empirical experience and false in some other possible worlds in which the physical laws that truly describe causal regularities in our world are also true; empirically contingent claims thus could have been false in our world. 13 The existence conditions for C need not define standards we can use to determine that something counts as C because we might not be epistemically able to determine whether something satisfies

28 

Methodology and the Nature of Law

Though quite natural, the above considerations raise the question of how we can be justified in believing that the metaphysical nature of C is exhausted by some set of properties we believe are inherent to C and hence define necessary conditions for something to count as C. The ability to ascertain whether some object in the world instantiates the inherent properties of a C depends on our having the ability to reliably identify the inherent properties defining the existence conditions that exhaust the metaphysical nature of C. But it is not clear how we could be justified in believing metaphysical claims that are neither empirically observable nor deducible from some set of favored logical axioms; it is, for this reason, not obvious how we can be epistemically justified in believing claims purporting to identify the inherent properties exhausting the metaphysical nature of some kind of thing. The traditional methodology for ascertaining the inherent properties exhausting the metaphysical nature of some entity of interest is conceptual analysis, which purports to explicate the content of the concept of that entity.14 The content of a concept C expresses the application-​conditions for using the associated concept-​term “C”, and the application-​conditions for using the concept-​term “C” pick out what are thought to be the inherent properties exhausting the metaphysical nature of C. The application-​conditions for using the concept-​term bachelor determining the content of the concept are exhausted by the properties inherent to bachelorhood: something counts as a bachelor if and only if it is an unmarried adult male. Likewise, the application-​conditions for using the concept-​term law, as it picks out systems rather than norms, are exhausted by the properties inherent to systems of law; on the Coercion Thesis, something counts as a system of law if and only if (1) it backs some mandatory norms governing non-​official behavior with the threat of a coercive sanction and (2) instantiates the other inherent properties of anything that counts as a system of law.

2.  Conceptual Analysis as a Descriptive Enterprise Conceptual analysis is concerned to describe the content of some concept of interest by identifying the properties inherent to things of the relevant kind; these

those existence conditions. If one thinks that it is the metaphysical nature of the good that anything that counts as good counts as such in virtue of its causally contributing to maximizing utility, then instantiating the property of maximizing utility is an existence condition for some act to be good; however, it is reasonable to think that we lack the ability to reliably ascertain whether some act is good. Determining the long-​term consequences of an act on human well-​being is prohibitively difficult; we have no reliable methodology for predicting such consequences that would have enabled anyone at the time to accurately predict the long-​term consequences of not killing Hitler when he was a small child. 14 Some fundamental set of these conceptual claims defines the set of favored metaphysical claims from which the less fundamental conceptual claims can be deduced. See the discussion on the three types of modal claim and subsequent two paragraphs in Section 1 of this chapter at p. 26.

Conceptual Analysis as a Descriptive Enterprise  29 inherent properties distinguish things that are members of the reference class of the associated concept-​term from things that are not. Insofar as all and only things that are unmarried, adult, and male count as bachelors, the content of the concept of bachelor distinguishes things that count as bachelors from things that do not: anything instantiating all three inherent properties of bachelorhood counts as a bachelor while anything not instantiating all three inherent properties of bachelorhood does not count as a bachelor; neither married things, children, nor women count as bachelors because none instantiates all of the inherent properties defining the metaphysical nature of bachelorhood. Conceptual analysis is concerned to describe what the content of the concept is, and not what the content of the concept should be. While it is sometimes thought that a conceptual theory should also be evaluative in this latter respect, this view is problematic. Insofar as the content of a concept is defined by a list of the relevant distinguishing inherent properties, the claim that a conceptual theory should be evaluative, in essence, asserts that it should be concerned with determining whether the content of the relevant concept should be revised, however that might be done, to express different distinguishing inherent properties from those expressed by it. What these normative claims ultimately amount to is not clear. The implicit claim that we should, or should not, revise an existing concept C can be construed as grounded in either prudential, aesthetic, or moral standards. Construed as prudentially normative, the claim is that we should change the content of C because our existing linguistic practices with respect to the associated concept-​term “C” do not optimally conduce to our collective prudential interests; construed as aesthetically normative, the claim is that we should change the content of C because our existing linguistic practices with respect to “C” are not pretty enough; and construed as morally normative, the claim is that we should change the content of C because our existing linguistic practices with respect to “C” are morally problematic. None of these standards provides a plausible rationale for changing the content of a concept, which can be done only by changing the linguistic practices that define the lexical meaning of the associated concept-​term and thereby determine the canons of ordinary usage with respect to that term. Insofar as the content of a concept is determined by the lexical meaning of the associated concept-​term, it is not clear how changing the lexical meaning of a concept-​term can impact our collective prudential interests in any significant way. In the case of the concept-​ term law, it is hard to see how using the term to refer to only morally legitimate norms would change anything that even minimally impacts our collective prudential interests. Since merely changing the label we use to refer to unjust mandatory legal norms does not change anything with respect to which norms are enforced by courts against subjects, it is simply not equipped to address the problems that really matter from the standpoint of collective prudential rationality.

30 

Methodology and the Nature of Law

Nor is it clear how changing the lexical meaning of a concept-​term is reasonably contrived to produce an aesthetic benefit because it is simply not clear what it means to say that the lexical meaning of some concept-​term (and hence that the content of the associated concept) is “pretty.” While the concept-​term love picks out an emotion that one might plausibly think is pretty, it is not clear what it means to say that defining the concept-​term love to express that emotion results in something that has aesthetic value in virtue of being pretty. It makes sense to attribute an aesthetic quality to the sound of a concept-​term: many people think that the sound of the concept-​term moist is gross, but that is a different matter from thinking its lexical meaning is gross. It likewise makes sense to attribute an aesthetic quality to the referent of a concept-​term: nearly everyone thinks that the objects to which the concept-​term excrement refers are gross, but that is a different matter than thinking its lexical meaning is gross. The idea that the conventional assignment of a concept-​term to a meaning has aesthetic qualities lacks a plausible foundation in ordinary experience. It is no more plausible to think there could be moral reasons to change the content of a concept—​at least as far as the concept of law is concerned. The substantive problems arising in connection with legal practices, whether moral or collectively prudential, have to do with the enforcement of morally illegitimate norms and not with the fact that the canons of ordinary usage dictate that we call those norms law.15 As long as those norms are enforced by courts, the relevant institutional practices and norms are morally problematic—​no matter what we call them. Changing our linguistic practices does absolutely nothing to change the character of institutional practices and norms we regard as morally problematic and is hence not reasonably contrived to make the world a morally better place. Conceptual analysis is exclusively concerned with what the content of some concept of interest is rather than with some mysterious normative notion of what that content should be. The underlying idea is that the philosophical claims fleshing out the content of the concept of some kind of thing express the metaphysical nature of the thing to which the corresponding concept-​term is used to refer. Explicating the content of these philosophical claims is sufficient to explicate the metaphysical nature of the thing with which we are concerned.

3.  The Distinguishing Task of Conceptual Analysis Fundamental to conceptual analysis is the task of identifying those properties jointly instantiated by all and only things that fall under some concept C of philosophical interest and are hence members of the reference class of the associated 15 Those norms might be illegitimate in virtue of either expressing morally wrongful content or in virtue of being promulgated by a morally wrongful process.

The Distinguishing Task of Conceptual Analysis  31 concept-​term “C.” These properties not only distinguish things that count as C from things that do not count as C; they also explain why something that falls under C is properly characterized as a “C”: something is properly characterized as a “C” because and only because it instantiates the inherent properties of C that constitute anything that instantiates them as falling under C and hence as counting as C. These properties explain why something falling under C is properly characterized as “C”: anything falling under C instantiates all of the properties inherent to C and thereby satisfies the application-​conditions for using the concept-​term “C.” The relationship between instantiating all the properties jointly instantiated by all and only things picked out by the concept-​term and being a member of the reference class of that concept-​term is that of constitution and not that of causation. Being an unmarried man might have many causal effects—​such as loneliness. But being an unmarried man does not cause a person to be a bachelor; whatever it is that causally explains why that man is unmarried also causally explains why he is a bachelor: if he is unmarried because he is an asshole, then being an asshole causally explains both why he is unmarried and why he is a bachelor. But being an asshole neither constitutes him as being unmarried nor constitutes him as being a bachelor. Being unmarried constitutes a man as a bachelor in the sense that he falls under the concept of bachelor—​and hence counts as a bachelor—​wholly in virtue of being an unmarried adult male. The difference can be understood in terms of the difference between being a C and counting as a C. What explains why some particular thing is a bachelor has to do with causal considerations that explain why that something is unmarried; these considerations can differ from one bachelor to the next: one bachelor might have wanted to marry but never found someone he wanted to marry while another might have wanted never to marry no matter what he otherwise felt about some other person. But what explains why something counts as a bachelor are considerations that cannot differ from one bachelor to the next: in every case, what explains why something counts as a bachelor is that it instantiates each of the inherent properties that express the metaphysical nature of bachelorhood and thereby constitute anything instantiating all of them as counting as a bachelor and constitute anything not instantiating all of them as not counting as a bachelor. The properties constituting something as C define necessary conditions for something to count as C and hence to be properly characterized as “C.” If the joint instantiation of the properties of being unmarried, being an adult, and being a male constitutes something as a bachelor, then nothing that lacks any of those three properties is properly characterized as a bachelor; if nothing that lacks either the property of being unmarried, the property of being an adult, or the property of being a male counts as a bachelor, then (1) it is a necessary condition for being properly characterized as a bachelor that something is unmarried; (2) it is a necessary condition for being properly characterized as a bachelor that something is an adult; and (3) it is a necessary condition for being properly characterized as a

32 

Methodology and the Nature of Law

bachelor that something is male. Something counts as C if and only if it is properly characterized as “C,” and something is properly characterized as “C” if and only if it instantiates all the inherent properties of C. Conceptual analysis of C, as traditionally conceived, is concerned with identifying the properties that constitute something as C and hence with explicating the metaphysical nature of C in terms of a description of properties that are essential to C in the following sense: A property p is essential to C if and only if it is not possible for a thing to count as C without instantiating p.

A complete list of all the properties essential to C exhausts the metaphysical nature of the kind picked out by C in the following sense: something properly characterized as “C” falls under that concept—​i.e. counts as C—​only and wholly in virtue of instantiating all the essential properties of C. The metaphysical nature of a thing is nothing more than its essence, and the essence of a thing is fully explicated by a description of all its essential properties. If, for example, the essential properties of bachelorhood are exhausted by the compound property of being an unmarried adult male, then this compound property fully explicates the metaphysical nature of bachelorhood; instantiation of this compound property is thus constitutive of bachelorhood in the sense described above. The set of essential properties of C is identical with the set of its inherent properties.16 The language of essences and essential properties is inapt insofar as it suggests that things have some sort of absolute nature defined by the metaphysical fabric of the objective world in a manner having nothing to do with the fact that we exist. On this view, it is simply a brute immutable fact about our world that bachelors are unmarried adult males. One can take that position, of course, but one need not do so. Given that describing the relevant properties as essential can be interpreted as insinuating, if not entailing, a substantive commitment to that position, the term is potentially misleading and is best avoided by those not wishing to endorse that commitment. Since it is no part of the point of a conceptual project to either endorse or reject that implausible position, it is better to speak instead of properties that are conceptually necessary for membership in the reference class of the relevant concept-​term in virtue of being inherent to things of that class. While the relevant terminology should not assume that the metaphysical nature of a thing is fully determined by our conceptual practices, it should not rule that out either. The concept-​term conceptually necessary is preferable to the concept-​term essential insofar as it is agnostic with respect to the idea that things in the world have, even in part, an immutable essence that is independent of whether we exist. 16 While the two notions are hence extensionally equivalent, it is not clear whether they are intensionally equivalent insofar as they arguably differ with respect to lexical meanings.

Modest and Immodest Conceptual Analysis  33 Conceptual analysis is, by nature, metaphysically driven. Insofar as an explication of a concept C consists of claims about the inherent properties that something must instantiate in order to count as C, it is concerned to identify necessary truths that express the metaphysical nature of C. Whatever else might be a true of an analysis that is concerned to identify empirically contingent properties of a thing—​ even those that are regarded as paradigmatic or prototypical—​it is not properly characterized as conceptual analysis; conceptual analysis is philosophical, and not sociological, in character.

4.  Modest and Immodest Conceptual Analysis Conceived as a descriptive metaphysical enterprise, conceptual analysis requires a different methodology from those deployed in normative and empirical inquiries. Empirical inquiry is concerned with identifying perceptual features of this world that distinguish it from other possible worlds and hence with what is contingently true in this world. Normative inquiry is concerned with identifying certain standards dictating what our world should be like and with applying these standards to our world to guide us in changing it to better realize the values they protect. Conceptual inquiry is concerned with identifying features of this world that could not be different no matter what else might be true of this or other worlds. Whereas empirical inquiry trades in descriptive claims that are contingently true and normative inquiry trades in evaluative claims that are either contingently or necessarily true, conceptual inquiry trades in descriptive claims that are necessarily true. Frank Jackson distinguishes a modest from an immodest approach to conceptual analysis. Modest conceptual analysis (MCA) gives us insight into the metaphysical nature of a thing as it is determined by the conceptual framework we impose on the world through our empirically contingent linguistic practices: [W]‌e need to attend to what the users of the language mean by the words they employ to ask their question. When bounty hunters go searching, they are searching for a person and not a handbill. But they will not get very far if they fail to attend to the representational properties of the handbill on the wanted person. Those properties give them their target, or, if you like, define the subject of their search. Likewise, metaphysicians will not get very far with questions like: Are there Ks? Are Ks nothing over and above Js? and, Is the K way the world is fully determined by the J way the world is? in the absence of some conception of what counts as a K, and what counts as a J. 17

17 Frank Jackson, From Metaphysics to Ethics: A Defense of Conceptual Analysis (Oxford University Press 1998) 30–​1. Hereinafter FEM.

34 

Methodology and the Nature of Law

On the modest approach, it is our linguistic practices with respect to the relevant concept-​terms “K” and “J” that determine what counts as a K and what counts as a J. Given that (1) what counts as an entity of interest is determined by the inherent properties constituting its metaphysical nature; and (2) it is our linguistic practices that determine which properties are inherent to things that count as that entity, MCA gives us insight into the metaphysical nature of an entity of interest as it is determined by our empirically contingent linguistic practices. The immodest approach to conceptual analysis (ICA) is also concerned with explicating the metaphysical nature of something to which some concept-​term refers, but it seeks to explicate that nature as it really is independent of the conceptual framework that we impose on the world to understand it. Jackson illustrates ICA with a well-​known line of criticism of the view that time constitutes a fourth dimension on par with the three spatial dimensions represented by the x, y, and z axes of a graph: [M]‌any have taken this kind of consideration to show that four-​dimensionalism qua thesis about what our world is like is false. They, in effect, argue as follows: Pr. 1  Different things (temporal parts or whatever) having different properties is not change. (Conceptual claim illustrated in the case of temperatures)18 Pr. 2  Things change. (Moorean fact) Conc.  Four-​dimensionalism is false. (Claim about the nature of our world) We now have an example of conceptual analysis in what I call its immodest role. For it is being given a major role in an argument concerning what the world is like (FEM 42–​43).

This line of reasoning purports to define a constraint on the conceptual framework we impose on the world to understand it: any conceptual framework entailing the truth of four-​dimensionalism is problematic because inconsistent with the immutably brute mind-​independent metaphysical fact that things change. Insofar as this

18 As Peter Geach describes this case with temperatures: “[T]‌he variation of a poker’s temperature with time would simply mean that there were different temperatures at different positions along the poker’s time axis. But this, as McTaggart remarked, would no more be a change in temperature than a variation of temperature along the poker’s length would be. Similarly for other sorts of change.” Peter T. Geach, “Some Problems about Time” in Geach, Logic Matters (Basil Blackwell 1972) 302–​18, 304. The idea is that a difference in temperature between one point on the poker and another is not properly characterized as a change in temperature; it is just a difference between the temperature at one location on the poker and the temperature at another location on the poker, such as would be the case if one end of the poker had been placed in fire and the other end in a freezer. If, however, reality is four-​ dimensional in the sense that time is a fourth dimension on par with the three dimensions of space, then there cannot be any changes in temperature from one moment in time to the next. What appears to us in a three-​dimensional world as a change in the temperature of the poker would really be a variation of the temperature at “different positions along the poker’s time axis.”

Modest and Immodest Conceptual Analysis  35 line of reasoning is grounded in claims about what the world is like independent of the conceptual framework that we adopt to carve it up, it rests on an analysis grounded in ICA. The difference between MCA and ICA, then, is that MCA seeks to understand the metaphysical nature of a thing as it is defined by the conceptual framework we impose on the world through our empirically contingent linguistic practices, while ICA seeks to understand the metaphysical nature of a thing as it really is independent of any such conceptual framework. As Kant might put the difference, MCA is concerned with understanding the metaphysical nature of things as they appear to us, while ICA is concerned with understanding the metaphysical nature of things as they are in themselves.19 MCA and ICA also differ with respect to the character of the modal claims expressing the metaphysical nature of the relevant kind of thing. Insofar as these modal claims are logically grounded in the empirically contingent linguistic conventions defining the application-​conditions for using the relevant concept-​term, as is true of MCA, the character of the resulting modal claims is conditionally necessarily, rather than absolutely necessary; insofar as the metaphysical nature of C as we conceive it is determined by the linguistic conventions for using the associated concept-​term “C” and these linguistic conventions can change, the metaphysical nature of C as we conceive it can also change because the content of claims expressing the metaphysical nature of C as we conceive it supervenes on the content of empirically contingent claims expressing our linguistic conventions for using “C.” But insofar as these modal claims are not logically grounded in the linguistic practices defining the application-​conditions for using the relevant concept-​term, as is true of ICA, the character of the resulting modal claims is more aptly described as absolutely necessary, rather than conditionally necessary; on this view, the nature of C is determined by objective features of the world that do not depend on the content of our empirically contingent linguistic conventions for using the associated concept-​term and hence does not change as the content of our conventions changes. Despite these differences, the methodologies of MCA and ICA are both partly empirical in character insofar as each adopts ordinary intuitions about the metaphysical nature of C as the starting point of an analysis of the corresponding concept; an intuition is ordinary only insofar as it is commonly shared among members of the relevant community of speakers, and which intuitions are commonly shared among members of that community is an empirically observable feature of the world. Insofar as the methodologies of both MCA and ICA require the 19 Kant distinguished between things as they are in themselves and things as they appear to us mediated through the categories of space and time through which all our sense perceptions are processed. Immanuel Kant, Critique of Pure Reason (Cambridge University Press 1998) in Paul Guyer and Allen Wood (eds), The Cambridge Edition of the Works of Immanuel Kant (Cambridge University Press 1992–​1998).

36 

Methodology and the Nature of Law

identification of ordinary intuitions as the starting point of conceptual analysis, both are grounded in empirical observation. But if the two approaches agree on starting from ordinary intuitions, they disagree on why we should start there. Insofar as MCA grounds conceptual theories in the canons of ordinary usage defined by our empirically contingent linguistic conventions, it requires that we consider ordinary talk as a touchstone for evaluating theories about the metaphysical nature of the relevant thing, and ordinary talk reflects ordinary intuitions conditioned by our empirically contingent linguistic conventions for using the relevant concept-​term. Accordingly, MCA begins from an understanding of the lexical meanings of words defined by the canons of ordinary usage; these canons form the starting point for an investigation into what deeper philosophical commitments are logically implied or presupposed by the linguistic practices defining them. In contrast, since ICA does not ground conceptual theories in our empirically contingent linguistic conventions for using the relevant concept-​term, there is nothing in ICA that requires we take the lexical meanings of words defined by the canons of ordinary usage as a starting point for conceptual theories; insofar as ICA seeks to understand what the world is like independent of our empirically contingent linguistic practices, there is no reason to think that ordinary intuitions are even helpful, much less necessary, in understanding what the world is like. The two approaches differ for this reason with respect to how much epistemic weight should be assigned to the corresponding intuitions. MCA takes these intuitions as providing the ultimate standard for evaluating a conceptual theory because MCA assumes that the object of conceptual analysis is to uncover the metaphysical nature of the world as we define it through the conceptual framework that our empirically contingent linguistic conventions impose on it. ICA, in contrast, takes ordinary intuitions to be nothing more than a guide to understanding the metaphysical nature of a thing; our intuitions or ordinary talk are not conceived as defining the metaphysical nature of the thing.20 Both approaches are properly conceived as concerned with claims that can be justified a priori. The claim that a proposition p can be justified a priori does not assert that p can be justified independently of all empirical experience; it asserts only that p can be justified independently of any empirical experience not needed to understand the lexical meanings of the concept-​terms used to express p. While ICA is obviously concerned with explicating claims that can be justified a priori, MCA is also concerned with explicating such claims insofar as it relies on empirical observation only to determine the lexical meaning of the relevant concept-​term as a

20 It is not clear why ordinary intuitions would even be relevant from the standpoint of ICA; there is nothing in the idea that conceptual analysis articulates the nature of things as they are independent of our thoughts and practices that even gestures in the direction of starting from the canons of ordinary usage or the philosophical intuitions that ground them.

Modest and Immodest Conceptual Analysis  37 means to identify the philosophical assumptions about the metaphysical nature of the thing to which it refers that ground our empirically contingent linguistic practices that determine the concept-​term’s lexical meaning. The putatively a priori character of both MCA and ICA also supports the idea that conceptual analysis is metaphysical in character. Insofar as conceptual analysis seeks to produce claims that are justified a priori, a true conceptual claim p can be justified with no further empirical observation than needed to understand the lexical meanings of the relevant terms in the sentence expressing p. Insofar as no further empirical observation than needed to understand the meaning of a sentence expressing a conceptual claim is required to confirm or disconfirm that claim, the considerations confirming or disconfirming that claim are the same in every metaphysically possible world. Purely descriptive claims about the nature of a thing are metaphysical precisely in virtue of being true in all metaphysically possible worlds if true in any metaphysically possible world—​regardless of whether the relevant necessity modality is absolute or conditional in character. Although it is natural to think that metaphysical inquiry is concerned to explicate the nature of something as it really is independent of how we perceive or understand it, ICA is not a methodological approach that beings like us can fruitfully adopt to understand the metaphysical nature of things. If ICA requires being able to understand the metaphysical nature of C independent of the canons of ordinary usage that determine the application-​conditions for using the associated concept-​term “C,” then it requires something beyond our epistemic abilities; the only epistemic access that we have to the metaphysical nature of C is through the linguistic conventions we adopt that define the application-​conditions for using the concept-​term “C.” We can apprehend the metaphysical nature of things only as they appear to us mediated through the concepts we deploy to organize and make sense of our experience; we have no reliable way to apprehend things as they are utterly independent of the concepts through which we organize the materials of our experience.21 It is an exercise in futility to attempt to understand C as it really is independent of the empirically contingent linguistic practices that enable us to talk about Cs to begin with. The only epistemically viable approach to conceptual analysis is MCA. MCA is epistemically viable because it is concerned with explicating the metaphysical nature of an entity of interest as it is determined, in part, by our empirically contingent linguistic practices; these practices include not only those defining the canons 21 Insofar as Geach’s criticism rests on the claim that four-​dimensionalism is inconsistent with some mind-​independent brute metaphysical fact that things change, it is epistemically problematic because it presupposes that we can know that things change independent of the conceptual framework we impose on the world. The most that can be said in criticism of four-​dimensionalism, on the view I am arguing for here, is that it is inconsistent with any conceptual framework we could fabricate to make sense of the world because we cannot reconcile it with our sense observations of the world; as far as we can tell given our perceptual abilities, time could not define a fourth-​dimension on par with the three spatial dimensions we characteristically perceive visually.

38 

Methodology and the Nature of Law

of ordinary usage but also practices picking out certain axioms of logic that we adopt because they strike us as defining epistemically justified standards of argumentation; these axioms are favored because they are ours. MCA purports to tell us something about how our conceptual activities structure the world—​and not about how the world is structured independent of these conceptual activities that enable us to talk about something and hence to understand it. MCA is not just the only epistemically viable approach to conceptual analysis; it is the only one of the two approaches that, despite Jackson’s taxonomy, counts as conceptual analysis. Insofar as ICA purports to identify the metaphysical nature of some entity of interest as it really is independent of the linguistic practices that enable us to talk about it, ICA is not really conceptual analysis at all. ICA is, rather, an attempt to identify the entity’s metaphysical nature without recourse to the canons of ordinary usage that determine the content of the only concepts to which we have epistemically reliable access—​namely ours.22 Insofar as the methodology for ICA does not require recourse to the empirically contingent linguistic conventions partly determining the content of our concepts (i.e. the only ones to which we have epistemically reliable access), it does not require analysis of a concept and is hence not properly characterized as conceptual analysis.

5.  The Relevant Philosophical Assumptions Are Ours Insofar as MCA attempts to determine the metaphysical nature of some entity of interest as it is determined by our linguistic practices, it requires recourse to both the lexical meaning of the concept-​term we use to refer to things of that kind and the philosophical assumptions that ground the empirically contingent linguistic practices that determine its lexical meaning. As I have described it, a modest analysis of a concept begins from an explication of the canons of ordinary usage governing the use of the associated concept-​term, but its point is to identify, explicate, and theorize the deeper philosophical assumptions about the metaphysical nature of the thing to which the concept-​term refers that ground these canons. These deeper assumptions are ours in the same sense that the empirically contingent linguistic conventions determining the canons of ordinary usage are ours. These linguistic conventions are ours in virtue of being determined by social practices that are also ours; the concept-​term bachelor means “unmarried adult male” because we converge on practicing a social convention that maps the former to 22 This is not to suggest that our empirically contingent linguistic practices ontologically explain the existence in logical space of the abstract propositional content that constitutes a concept. It might well be true that all possible concepts exist in logical space; in this case, the canons of ordinary usage determine which proper subset of all possible concepts comprises the set of our concepts by mapping the relevant concept-​terms to the lexical meanings that pick out which among existing concepts are expressed by those terms.

The Relevant Philosophical Assumptions Are Ours  39 the latter. But these deeper philosophical assumptions are also ours in virtue of helping to explain why we converge on practicing the particular convention that defines bachelor as “unmarried adult male”; we converge on practicing these conventions because we converge in believing—​impliedly and latently if not explicitly and occurently—​that bachelors are, by nature, unmarried adult males.23 This should not be taken to entail that we would all accept these beliefs if made aware of them—​although it is plausible to think that most of us would in the absence of exceptional circumstances. The analysis here is concerned with identifying metaphysical commitments that are plausibly attributed to us in virtue of being entailed by views that are plausibly attributed to us in virtue of being entailed by our empirically contingent linguistic practices; it is not concerned with trying to ascertain exactly what all or most people believe by taking a poll. Insofar as people’s intuitions on possible cases matter here, it is because they function to guide, rather than exhaust, the analysis. Because these philosophical assumptions grounding the canons of ordinary usage are ours, they need not conform to what is generally regarded by people in the academic philosophical community as being the best theory about the thing that is the subject of the relevant philosophical assumptions. It might be, for example, that what is regarded as the best philosophical theory of morality denies that there are any mandatory moral norms objectively binding people regardless of what they believe; but that is irrelevant with respect to explaining our conceptual practices with respect to law—​even if we assume that views about morality figure into the relevant conceptual practices. What matters is what people believe about the character of these mandatory moral norms; if ordinary views about the character of morality figure into these conceptual practices and those ordinary views entail that mandatory moral norms objectively bind people regardless of what they believe, those are the views that matter in explicating our conceptual practices. The relevant philosophical assumptions might for this reason be objectively incorrect. The empirically contingent linguistic practices determining the lexical meaning of phlogiston were grounded in the philosophical assumption that it is metaphysically impossible for a material object to burn unless it contains a fire-​ like substance; the underlying idea was that the only way to explain how material objects can combust is to suppose that this fire-​like substance is activated and released into the atmosphere when the object burns.24 But this assumption turned out to be false: the empirical evidence showed there are no fire-​like substances in

23 Given that most people, being untrained in philosophy, would be unable to explain what it means to say that it is in the metaphysical nature of law that it is backed by coercive sanctions, these views are more likely latently held—​if held at all. 24 See, e.g., Michael Weisberg, Paul Needham, and Robin Hendry, “Philosophy of Chemistry” Stanford Encyclopedia of Philosophy (Edward Zalta, ed.); available at: https://​plato.stanford.edu/​entries/​ chemistry/​#Phl.

40 

Methodology and the Nature of Law

our world of the sort posited by phlogiston theory, thereby refuting the assumption that material objects could not combust unless they contained such objects. Although the relevant philosophical assumptions are false, the conceptual framework of the time cannot be properly explicated under MCA without including the philosophical assumptions about the metaphysical nature of material things that ground phlogiston theory and the prevailing contingent linguistic practices with respect to the concept-​term phlogiston. It does not matter that these assumptions turned out to be false; the content of the relevant conceptual practices cannot be explained without recourse to these assumptions because the practices that matter under MCA are theirs—​and not some set of practices that better conforms to, as an objective matter, the metaphysical nature of the world as it is determined independent of the conceptual framework we impose on it. The fact that these assumptions are false warrants no more than that we should reject that framework as helpful in explaining material combustion; it cannot warrant rejecting an explication of their conceptual practices that includes those assumptions. We cannot explicate the conceptual practices of some group with respect to some concept-​term without exhibiting the philosophical assumptions that ground them—​regardless of whether we find them plausible or not. The same is true of explicating the concept of law under MCA. What matters with respect to a philosophical explication of our empirically contingent linguistic practices as they determine the lexical meaning of the concept-​term law are the philosophical views that, as a descriptive matter of contingent fact, condition and inform our linguistic practices because it is what we believe or assume that explains why we converge in cutting up the world conceptually the way we do and why we create the institutions we do. Regardless of whether these philosophical views are ultimately true from some God’s-​eye perspective we cannot achieve, they explain why we distinguish institutional normative systems that count as law according to our empirically contingent linguistic practices from institutional normative systems that do not count as law according to these practices. One interesting example of a philosophically contentious assumption that helps to explain what we do with what we call law is the idea that we have free will in some sense compatible with its being morally appropriate for the law to evaluate our behavior under mandatory legal norms and to punish us with the imposition of coercive sanctions for non-​compliance with certain mandatory legal norms. Insofar as the notion of punishment has a backward-​looking retributive element, as argued in Chapter 1, characterizing certain coercive sanctions as punishment presupposes that subjects at least sometimes deserve them for non-​compliance; but punishment can be deserved, as far as our conceptual practices (i.e. our empirically contingent linguistic practices properly fleshed out to expose the underlying philosophical assumptions about the metaphysical nature of the thing picked out by the relevant concept-​term) are concerned, if and only if the agent’s non-​compliance is explained by her having freely chosen to act in a manner that does not comply with

The Relevant Philosophical Assumptions Are Ours  41 the norm.25 Insofar as the notion of punishment has a forward-​looking deterrence element, as also argued in Chapter 1, characterizing these measures as punishment presupposes that rationally competent subjects can be rationally induced by the threat of punishment to comply with norms they might otherwise violate; but a rationally competent subject can be rationally induced to do something if and only if she has the capacity to freely decide what to do based on her assessment of the relevant reasons. Academic philosophers disagree about the metaphysical nature of free will, as well as about whether beings like us actually have free will in a sense logically consistent with our assumptions about the conditions under which rationally competent subjects may, as an objective matter of morality, justifiably be held accountable for their acts. So-​called incompatibilists and compatibilists disagree on whether the claim that every event in our universe has a cause is consistent with our having free will; the compatibilist claims that causal determinism is compatible with our having free will while the incompatibilist denies that claim. Whether or not we have free will, then, depends on at least two issues: (1) whether causal determinism is true of our world;26 and (2) whether our decisions count as free in a sense that is logically consistent with causal determinism. Trying to ascertain whether the assumption that we have free will in a sense compatible with the moral propriety of holding us legally accountable for non-​compliance is a sticky enterprise. Notice, however, that there is nothing in our conceptual practices that really turns on this issue, though they clearly presuppose that we have free will in a sense logically consistent with the moral appropriateness of punishing non-​compliance. A court’s authority to impose coercive sanctions for non-​compliance is reasonably contrived to reduce the probability of non-​compliance either way. Only the mechanism by which these practices reduce the probability of non-​compliance is different: if we have free will, the relevant mechanism is rational inducement in the form of persuasion; if we lack free will, the relevant mechanism is causal inducement in the form of creating a new causal factor that contributes, among countervailing factors, to mechanistically determining whether we comply with law. Either way, we cannot fully understand our conceptual practices with respect to the notion of law—​and, for that matter, other social practices involving evaluating the behavior of people against other classes of norms—​without explicating the underlying presupposition that it is appropriate to evaluate our behavior because it is freely chosen. Whether we have free will makes a difference with respect to

25 The claim here is not that the agent’s intention was to violate the legal norm or that she knew that the act would violate the norm; it is rather that the agent’s act was freely chosen. 26 There are two possibilities here: determinism might be true of our world in virtue of its being true of all metaphysically possible worlds in which case it would express a purely metaphysical claim; alternatively, determinism might be true of our world in virtue of physical laws of our material universe that are metaphysically contingent (i.e. not true of all metaphysically possible worlds) in which case it would express a claim that is at least partly nomological in character.

42 

Methodology and the Nature of Law

whether our contingent legal practices are morally justified;27 however, it does not make a difference with respect to whether a conceptual theory of law that incorporates the assumption that we have free will is true. To understand what we do with law, we must understand why we do what we do with law. The ultimate point of MCA is to explicate the world of our experience as we construct it socially—​and not, as ICA presupposes, to explicate the world of our experience as it is independent of the conceptual framework that we create to understand it. The operative reason here is not that we cannot understand the metaphysical nature of the world independent of the conceptual framework we impose on it—​although that is obviously true. The operative reason is rather that a modest approach is contrived ultimately to enable us to understand something about ourselves. Even if we could ascertain the metaphysical nature of law as it is independent of our conceptual practices, we would still have reason to want to understand the metaphysical nature of law as it is determined by our practices. An explication of the metaphysical nature of the world as it is determined by considerations independent of our conceptual practices might otherwise be equipped to tell us something important about the objective world; however, it is not equipped tell us anything about who we are and what we value—​and that is ultimately the point of MCA. One could, I suppose, see ICA and MCA not as rival methodologies, as Jackson seems to, but as methodologies defining complementary endeavors. On this line of reasoning, we want to understand both the metaphysical nature of some C as it is independent of the conceptual framework that we create through our empirically contingent linguistic practices and the metaphysical nature of C as it is determined by that conceptual framework. To put the point in Kantian terms, we want to understand both the world as it appears to us and the world as it is in itself—​ perhaps for the purpose of reflectively adjusting our conceptions and practices to the extralinguistic realities they seek to describe. The point of MCA would be to enable us to understand the world as it appears to us while the point of ICA, assuming it is epistemically viable, would be to enable us to understand the world as it is in itself; once a modest and immodest analysis of the relevant concept are given, we could adjust our linguistic practices to ensure that they conform to what ICA tells us, on this line of reasoning, about the relevant entity of philosophical interest. But, either way, the exclusive concern of this volume is to understand what we do with what we call law—​and why we do what we do as a means of better understanding who we are in a material world that exists independently of us and our

27 It makes a difference in the following way. If our behavior is freely determined, we are really subject to the requirements of an objective morality. If our behavior is mechanistically determined, we are not really subject to the requirements of an objective morality—​though our behaviors may include those comprising something that counts as defining and applying mandatory moral norms.

Modest Conceptual Analysis and Objective Truth  43 conceptual practices—​and only MCA is equipped to enable us to do that. Even if ICA were an epistemically viable approach, it is just not equipped to answer those questions. The only methodology equipped to explain our practices and hence to answer those questions is MCA, but the relevant philosophical assumptions about the metaphysical nature of law are ours and might turn out to be inconsistent with what is regarded by academic philosophers as the best philosophical theory of the thing with which those assumptions are concerned and, indeed, might well turn out to be objectively false.

6.  Modest Conceptual Analysis and Objective Truth One might argue that MCA is no better equipped, as an epistemic matter, to tell us anything about the objective world than ICA. On this line of reasoning, a conceptual theory grounded in MCA tells us nothing about the objective world because the truth-​value of a conceptual claim under MCA is determined not by mind-​independent considerations having to do with whether the claim correctly describes the objective world, but rather by mind-​dependent considerations having to do with whether it conforms to the empirically contingent linguistic conventions defining the canons of ordinary usage. Since the truth-​ determinants of conceptual claims under MCA are thus intersubjective rather than objective, a modest conceptual theory cannot tell us anything about the objective world. This line of criticism is problematic. It is true that purely conceptual claims cannot tell us anything about the objective world beyond explicating the content of the empirically contingent linguistic conventions we adopt to impose a conceptual framework on it and the philosophical assumptions grounding those conventions, but MCA purports to do no more than that. Conceptual problems arise only insofar as these conventions are either indeterminate or facially inconsistent with respect to some entity of philosophical interest. The point of a conceptual theory under MCA is to explicate the underlying philosophical assumptions grounding these linguistic conventions with sufficient depth and detail to resolve the indeterminacy or inconsistency motivating the inquiry. But it does not follow that MCA cannot tell us anything about the objective world. Notice, in this connection, that anyone can be mistaken about either the canons of ordinary usage or the unstated philosophical assumptions that ground or condition them. I might mistakenly believe that the Pope is not a bachelor either because I mistakenly believe that canons of usage governing the concept-​term bachelor explicitly preclude its application to men occupying some religious office or because I mistakenly believe that the philosophical assumptions grounding these canons of usage entail that it is the metaphysical nature of a bachelor that anything properly characterized as bachelor is institutionally eligible for marriage in

44 

Methodology and the Nature of Law

some way that the Pope is not in virtue of the norms governing the religious office he occupies. This is the hallmark of objectivity. To say that a claim is objective is to say that its truth-​value is determined entirely by considerations that are mind-​independent and hence that its truth-​value does not turn, even partly, on what any one person believes with respect to its truth-​value: that is, a claim p is objective if and only if whether p is true does not depend on whether anyone believes or desires that p is true; whatever considerations determine whether p is true, those considerations do not include any person’s believing that p is true. But insofar as the truth-​value of an objective claim does not depend at all on any person’s beliefs or preferences about the relevant matter, any person can be mistaken about the truth-​value of an objective claim; scientists could, though this is presumably quite unlikely, be mistaken about whether the thing to which we attempt to refer with the concept-​ term earth really instantiates the property to which we refer with the concept-​term spherical.28 In contrast, insofar as the truth-​value of a claim is subjective in the sense that it is fully determined by what the speaker believes about the claim, no person can be mistaken about a subjective matter; if I sincerely believe I have a headache and I am not confused about the lexical meaning of headache, then I have a headache. Conceptual claims are no less objective insofar as they express mind-​ independent facts about the content of the canons of ordinary usage. Conceptual claims express peculiar facts about the objective world—​facts that depend on conventional (or intersubjective) mappings of words to meanings that are grounded in certain philosophical assumptions about the metaphysical nature of the things to which the words refer—​but that does not make those claims any less objective. The truth-​value of the claim that the concept-​term bachelor means “unmarried adult male” depends on competent speakers converging on practicing an empirically contingent linguistic convention dictating that the concept-​term bachelor is properly applied only to unmarried adult males. But the truth-​value of a proposition purporting to report those facts is not determined by whether every competent speaker converges on believing that proposition: any competent speaker can be objectively mistaken about whether bachelor means “unmarried adult male” because general competence with a language does not require knowledge of every term of the language; if it did, then none of us would be properly characterized as competent speakers of English. Likewise, any competent speaker can be objectively mistaken about the content of the philosophical assumptions that ground the canons that determine the lexical meaning of the relevant concept-​term; the philosophical assumptions grounding a linguistic 28 A growing number of people are skeptical with respect to whether the earth really is spherical. Only 66% of millennials, according to a recent poll, confidently believe that the world is spherical. See Craig Foster and Glenn Branch, “Do People Really Think the Earth Might be Flat?” Scientific American (August 21, 2018); available at:  https://​blogs.scientificamerican.com/​observations/​do-​people-​really-​ think-​earth-​might-​be-​flat/​.

Modest Conceptual Analysis and Objective Truth  45 convention are determined by both empirical considerations and considerations of logic. Regardless of whether the facts expressed by conceptual claims are peculiar in some sense, those facts are objective insofar as any competent speaker can be mistaken about them. There is one respect in which conceptual claims differ from other sorts of claim about the objective world. While it is true that everyone can simultaneously be mistaken about the shape of the earth—​indeed, there might have been a time when everyone believed the earth was flat—​it is not true that everyone can simultaneously be mistaken about a conceptual claim from the standpoint of MCA. If everyone believes that bachelors can be married, then there can be married bachelors. But the concept-​term bachelor does not mean “unmarried adult male” under this assumption; since lexical meanings are conventionally determined and there is no convention dictating that bachelor means “unmarried adult male,” the concept-​ term bachelor would not mean “unmarried adult male.” While conceptual claims cannot tell us anything about the extra-​linguistic objective world, they nonetheless contribute in an epistemically essential way to our ability to know things about the extra-​linguistic objective world. We can know that certain propositions about the extra-​linguistic objective world are true only insofar as we can express them in a language comparatively free of ambiguity: we cannot know anything worth knowing about the earth without having a conceptual framework that includes a concept-​term that refers to what we take to be that object. Concepts are tools we employ to assist us in arriving at epistemically justified beliefs about the extra-​linguistic objective world—​and the content of a belief is only as reliable as the content of the concepts we deploy to express that belief is clear. It might be true that we can come to know a good bit about the extra-​linguistic objective world without conceptual analysis because our pre-​theoretic conceptions of the relevant concepts are good enough; but there is something to be said for the value of clarity to which explicating the content of our concepts conduces. This should not be taken to imply anything with respect to whether we are epistemically justified in believing that the structure of the objective world as it really is mirrors the structure that we impose on it through the conceptual framework we create to understand it. We might not be in a position to discern whether, according to the conceptual framework we impose on the world, the thing that is picked out by the concept-​term earth instantiates, as an objective matter of fact, the property picked out by the concept-​term spherical. Indeed, we might have no way of knowing (in the sense of having an epistemically justified belief to this effect on the best available normative theory of epistemic justification) whether the ontological presupposition that underlies our use of the concept-​term earth—​namely that there exists a material object in the universe as it really is that is picked out by the concept-​term earth—​is true, as a matter of objective fact. But that is an epistemological concern that applies to both conceptual and non-​ conceptual claims about the objective world. The basic problem giving rise to

46 

Methodology and the Nature of Law

normative epistemology is to explain how we can be epistemically justified with respect to our beliefs about the empirically contingent features of the objective world. But this problem arises regardless of the content of our conceptual practices; regardless of how we carve up the world and of whether we do it with a language that defines a conceptual framework, it is simply not clear what would epistemically justify us in thinking that our subjective perceptions of the objective world accurately reproduce what the objective world is really like. Either way, the project of normative epistemology presupposes that our beliefs about the mind-​independent world express claims that are objective in the sense of having truth-​makers that are mind-​independent in the relevant sense. If it matters to us whether our beliefs about what the objective world is like are epistemically justified, that is because the statements expressing our beliefs purport to be objectively true; the claim that the earth is really spherical, as an objective matter of fact, can be epistemically justified only insofar as its truth-​value is determined by mind-​ independent considerations—​and not by whether or not most people believe that it is spherical. The very project of normative epistemology presupposes that our beliefs about the world as they are formulated in a language defined by our conceptual practices express claims that have mind-​independent truth-​makers and are hence objective in character.

7.  The Logical Structure of a Modest Conceptual Analysis It is sometimes thought that conceptual analysis, conceived as grounded in the conventional linguistic practices defining the canons of ordinary usage, involves little more than giving what amounts to a dictionary definition, but this is mistaken. Conceptual analysis begins from shared intuitions about the application-​ conditions of the concept-​term of interest that are conditioned by the canons of ordinary usage defining its lexical meaning but goes on to identify, explicate, and theorize the deeper philosophical commitments these intuitions imply or presuppose. There are thus two steps to a modest analysis of a concept. The first is to explicate the lexical meaning of the corresponding concept-​term as it is determined by the canons of ordinary usage governing its use; these canons are roughly but accurately expressed by any reputable dictionary definition. The second is to identify the philosophical assumptions about the metaphysical nature of the thing to which the corresponding concept-​term refers; under MCA these philosophical assumptions contribute as much as the canons of ordinary usage, as they are expressed in a dictionary definition, to determining the content of the relevant concept. The first step is wholly empirical while the second is not. The first step requires identifying the conventional practices determining the application-​conditions for using the concept-​term; this is an empirical task that bears a superficial

The Logical Structure of a Modest Conceptual Analysis  47 resemblance to what lexicographers do in virtue of being concerned to identify the concept-​term’s lexical meaning but is distinct from what they do in virtue of presupposing that lexicographers have antecedently explicated its lexical meaning in a manner satisfying the professional standards governing lexicography. But the second step requires a logical analysis that is no more properly characterized as empirical than is the project of proving some mathematical theorem in a manner satisfying the favored logical axioms—​despite the fact that which logical axioms we favor is also empirically contingent; while it might be grounded in something resembling an empirical survey of competent speakers on some possible case, the point of eliciting ordinary responses is to ensure that the non-​empirical philosophical analysis adequately coheres with the shared philosophical views that ground the lexical meaning. Both steps are necessary to fully justify any contentious conceptual claim because a claim is likely to be contested only insofar as the canons of ordinary usage governing use of the relevant concept-​term are epistemically indeterminate with respect to that claim. Even a convincing showing that the Coercion Thesis is logically entailed by the lexical meaning of the concept-​term law constitutes no more than a prima facie case for it that can be rebutted by a showing that it conflicts with the deeper philosophical assumptions grounding the lexical meaning without being expressly incorporated into the relevant canons of ordinary usage. As discussed above, the canons of ordinary usage with respect to the concept-​term bachelor straightforwardly entail, given that the Pope is an unmarried adult male, that the Pope counts as a bachelor; yet many competent speakers share the intuition that the Pope is not properly characterized as a bachelor because one must be institutionally or psychologically eligible for marriage to count as a bachelor—​and the Pope is neither. On this line of reasoning, the Pope instantiates all the properties expressed by the relevant canons of ordinary usage but is not properly characterized as a bachelor because the dictionary reports of these canons are not fully accurate insofar as they do not express that something must be institutionally or psychologically eligible to marry to count as a bachelor. The remainder of this book is concerned to provide a modest analysis of the concept of law as it picks out systems, rather than norms, of law. It begins in the next chapter with an explication of the canons of ordinary usage as they express our empirical experience with institutional normative systems we characterize as law and then goes on to explicate the underlying philosophical assumptions about the metaphysical nature of law that ground these canons. The book concludes with a consideration of two putative counterexamples to the Coercion Thesis.

3

A Prima Facie Case for the Coercion Thesis: Sanctions as a Paradigmatic Feature of Municipal Law Chapter 2 described the modest approach to conceptual analysis (MCA) adopted in this volume. MCA seeks to explicate the metaphysical nature of a kind of thing as it is determined by our conceptual practices—​and not as it really is independent of our conceptual practices. MCA is thus concerned to explicate the content of concepts that are ours in virtue of being determined by linguistic conventions that are also properly characterized as ours in virtue of being adopted, as a descriptive matter of contingent fact, by us to talk about things of that kind. Insofar as these empirically contingent linguistic conventions are grounded in philosophical assumptions about the metaphysical nature of the thing to which the associated concept-​term refers, MCA requires a two-​step analysis. The first is to explicate the content of the canons of usage we adopt to talk about that kind of thing, while the second is to explicate the philosophical assumptions about the inherent properties expressing its metaphysical nature that ground these canons; these philosophical assumptions are plausibly characterized as ours in virtue of being logically presupposed or entailed by the canons of ordinary usage that are also ours in virtue of being defined by what we do with language. The first step defines a prima facie case that can be rebutted by the second step. Since the canons of ordinary usage are grounded in implicit philosophical assumptions about the metaphysical nature of thing, a satisfactory modest analysis of a concept must harmonize with those philosophical assumptions. While the sounds, scribbles, and definitions we adopt as a language might be arbitrary, the conceptual framework we impose on the world through these choices reflects shared judgments of salience conditioned by views that are ultimately philosophical in character.1 Since these underlying philosophical judgments explain why we carve up

1 One can see this dynamic at work in evolving conventions with respect to using gendered language. The use of he to refer to both men and women as well as the use of mankind to refer to all persons reflect philosophical judgments about the respective roles and comparative value of men and women that are increasingly considered morally problematic by competent speakers. Our linguistic practices with respect to such usage are changing in response to changing views about the propriety of these assumptions. Coercion and the Nature of Law. Kenneth Einar Himma, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198854937.001.0001

50  A Prima Facie Case for the Coercion Thesis the world conceptually the way we do, a modest analysis of a concept must cohere with those judgments as they are expressed in the canons of ordinary usage. This chapter undertakes the first of these two steps by showing that the Coercion Thesis better coheres than its negation with the canons of ordinary usage as they reflect our universal experience with systems of municipal regulation we designate as law.2 Every existing system of municipal regulation we characterize as law backs some mandatory legal norms governing non-​official behavior with the threat of incarceration. Since our experience with respect to distinguishing things counting as law from things not counting as law conforms to what one would expect given dictionary reports of the relevant canons of usage, these reports accurately express the linguistic conventions we adopt for using the concept-​term law and thereby define a prima facie case for the Coercion Thesis.

1.  The Meaning of the Term Law as Defined by the Canons of Ordinary Usage Conceptual analysis is concerned to explicate the metaphysical nature of the thing C picked out by some concept-​term “C” of interest.3 The metaphysical nature of a C is exhausted by the inherent properties distinguishing things counting as C from things not counting as C. These inherent properties constitute something as C in the sense that the instantiation of each is conceptually necessary for something to count as C and the instantiation of all is conceptually sufficient for something to count as C. Something counts as C wholly in virtue of instantiating all of these properties: anything lacking even one of them is not properly characterized as “C” because it lacks one of the inherent properties distinguishing things that count as C from things that do not count as C and hence fails to fully manifest the metaphysical nature of C. There are, as discussed in the last chapter, two methodological approaches to explicating the metaphysical nature of C. The immodest approach to conceptual 2 The system of international regulation defined by the U.N. Charter is considered in Chapter 9. The reason it is omitted here is not that it is contentious whether there can be an institutional system of international regulation properly characterized as a system of law. Neither the canons of ordinary usage nor the philosophical assumptions grounding those canons preclude the conceptual possibility of a system of law regulating the behavior of nations, as the matter is commonly but misleadingly put; otherwise the term international law would be oxymoronic. The reason is that our empirical experience with such systems is too limited and contentious, unlike our empirical experience with systems of municipal law, to constitute a class of paradigm cases of law. Regardless of whether the U.N. Charter system is properly characterized as a system of law, it is not a paradigm case of law, unlike the systems of municipal law considered in this chapter, that motivates our conceptual practices. 3 I use the underlined letter, C, to denote the thing to which a concept-​term is used to refer; I enclose that letter in quotation marks to denote the concept-​term, “C,” that picks out the concept C and refers to Cs. See Chapter 2, Note 3. Unless necessary for disambiguation, I use italics when discussing some particular concept-​term like law and use quotation marks when I am quoting something from another source.

Law as Defined by the Canons of Ordinary Usage  51 analysis (ICA) is concerned to articulate the metaphysical nature of C as it is determined independent of the conceptual framework that we impose on the world through our linguistic practices to talk about things that count as Cs. MCA, in contrast, is concerned to articulate the metaphysical nature of C as it is determined by the conceptual framework that we impose on the world through these empirically contingent linguistic practices to talk about Cs. Because we have no epistemically reliable way to understand the metaphysical nature of the world independent of the conceptual framework we impose on it, this volume adopts MCA and therefore takes the canons of ordinary usage with respect to the concept-​term law as the starting point for theorizing about the metaphysical nature of law. Since the canons of ordinary usage express the lexical meaning of the concept-​term law, the appropriate starting point in defending the Coercion Thesis under MCA is to consider the lexical meaning of the term law as it is reported by dictionaries. At first glance, these dictionary reports seem to disagree with respect to whether it is a conceptually necessary condition for the existence of a legal system that some mandatory norms are backed by the threat of a coercive sanction. On the one hand, the concept-​term law is sometimes defined in a manner suggesting the conceptual possibility of law without sanctions: Merriam-​Webster defines the concept-​term law as (1) “a rule of conduct or action prescribed or formally recognized as binding or enforced by a controlling authority”;4 the disjunction suggests that there could be something properly characterized as law wholly in virtue of being formally recognized without being backed by the threat of a sanction. On the other hand, the term law is also defined as (2) “the principles and regulations established in a community by some authority and applicable to its people, whether in the form of legislation or of custom and policies recognized and enforced by judicial decision”;5 (3) “the system of rules which a particular country or community recognizes as regulating the actions of its members and which it may enforce by the imposition of penalties”;6 and (4) “a rule or set of rules, enforceable by the courts, regulating the government of a state, the relationship between the organs of government and the subjects of the state, and the relationship or conduct of subjects towards each other.”7 The appearance of disagreement is misleading insofar as the term law is ambiguous in at least two ways. The term is ambiguous, to begin, between referring only to systems that count as law and referring only to norms that count as law; and

4 Merriam-​Webster Online Dictionary; available at:  https://​www.merriam-​webster.com/​dictionary/​law. 5 Dictionary.com; available at: https://​www.dictionary.com/​browse/​law. Emphasis added. 6 Oxford Online Dictionary; available at:  https://​en.oxforddictionaries.com/​definition/​law. Emphasis added. 7 Collins English Dictionary; available at:  https://​www.collinsdictionary.com/​us/​dictionary/​ english/​law. Emphasis added.

52  A Prima Facie Case for the Coercion Thesis the definitions above differ with respect to the relevant referent: whereas definition (1) is quite clearly concerned with the lexical meaning of law qua norm, definitions (2), (3), and (4) are no less clearly concerned with the lexical meaning of law qua system. These two sets of definition are not in conflict: the claim that not every conceptually possible legal norm is backed by the threat of a sanction is consistent with the claim that every conceptually possible legal system backs at least some mandatory legal norms governing non-​official behavior with the threat of a sanction. The concept-​term law is also ambiguous between a wide usage applying to systems of religious and chess law in addition to the usual systems of municipal and international law comprising the focus of law-​school study and a narrow usage applying only to the latter systems. While definition (2) is not as obviously concerned as definitions (3) and (4) with only the narrow usage of law, which excludes the rules of a chess association and systems of religious law, it clearly supports the Coercion Thesis; law, on this report of the term’s lexical meaning, is comprised of norms “recognized and enforced by judicial decision.” As definitions (3) and (4) are clearly concerned with the narrow usage and define law as including norms backed by the threat of a sanction, these definitions also support the Coercion Thesis: definition (3) defines law as a “system of rules . . . which it may enforce by the imposition of penalties” while definition (4) defines law as “a rule or set of rules [] enforceable by the courts.” These dictionary reports line up with the Coercion Thesis in two crucial respects. To begin, apart from definition (2), these reports neither assert nor entail that the sanctions authorized for non-​compliance are ever imposed in any particular legal system; the lexical meaning of law, as defined by these reports, is compatible with the existence of a legal system in which authorized sanctions are never imposed for non-​compliance because no one ever violates the law. That is as it should be; as discussed in Chapter 1, there is nothing in our conceptual practices with respect to the term law that precludes the conceptual possibility of a legal system in which everyone always obeys the law because all prohibited behavior has been successfully deterred by the threat of sanctions.8 While the existence of a legal system that always succeeds in deterring prohibited behaviors among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours might be highly improbable, it is conceptually possible. Either way, even if there were something in our conceptual practices precluding the possibility of something properly characterized as a system of law in which sanctions are never imposed because subjects always obey the law, this is not a problem for the Coercion Thesis. Insofar as it is a conceptual truth that sanctions 8 This example defines a radically different state-​of-​affairs from that of the society of angels, as described by Joseph Raz. Insofar as the law changes the behavior of subjects in the example above, the assumption is that non-​compliance was deterred by the threat of sanctions—​and not motivated by some improbably peculiar commitment always to defer to the judgments of legal authorities as they are expressed in law. The society-​of-​angels argument will be discussed in Chapter 10.

Law as Defined by the Canons of Ordinary Usage  53 are sometimes imposed, it is also a conceptual truth that officials of the system are authorized to impose them. While the Coercion Thesis does not affirm the claim that it is a conceptual truth that sanctions are sometimes imposed, it does not deny that claim. If definition (2) expresses a stronger conceptual claim about law’s coercive nature than is expressed by the Coercion Thesis, it nonetheless presupposes the Coercion Thesis and hence provides prima facie support for it. Definition (3) clearly expresses that it is part of the metaphysical nature of a legal system that it authorizes the judicial imposition of coercive sanctions for non-​ compliance with some mandatory legal norms. Definition (3) asserts that the relevant norms of a legal system “may” be enforced by the courts and hence entails that the courts are authorized in the relevant sense to enforce those norms. While the term may in this context connote only what is legally permitted, an official must be legally authorized to do something not generally permitted by law. Insofar as it is the job of judicial officials to do what non-​officials are not legally permitted to do, those officials need legal authorization (i.e. authority) to do those things. Definition (4) asserts, less perspicuously, that the relevant norms are “enforceable by the courts.” While one might construe definition (4) as meaning only that it is logically, metaphysically, or nomologically possible for the courts to enforce the relevant norms, this construction says nothing that needs to be said: it is logically, conceptually, and nomologically possible for courts to enforce anything with the form of a norm—​including moral norms, norms of etiquette, and chess rules. Ordinary principles of semantic interpretation dictate against construing definition (4) as asserting that vacuous claim because doing so renders the words expressing it as superfluous. It might be helpful here to contrast definition (4) with others in the same dictionary having to do with words referring to other types of norms. Collins English Dictionary defines morals as “principles of behavior in accordance with standards of right and wrong” and rules as “instructions that tell you what you are allowed to do and what you are not allowed to do.”9 The fact that these definitions do not report that the relevant “principles of behavior” or “instructions” are enforceable suffices to show that its definition that law is “enforceable” should be construed as conveying that the courts are authorized to enforce those norms. Ordinary principles of interpretation dictate, then, that the proper construction of the term expresses that legal norms are enforceable in the sense that courts are authorized—​by law—​to impose them for non-​compliance with the relevant norms. Although the dictionary reports of the lexical meaning of law support the Coercion Thesis, the prima facie case for the Coercion Thesis cannot rest on just these dictionary reports. If there is something in our world lacking authorized sanctions that we converge, as a descriptive matter of contingent fact, in 9 Collins English Dictionary; available at:  https://​www.collinsdictionary.com/​us/​dictionary/​ english/​morals; and https://​www.collinsdictionary.com/​us/​dictionary/​english/​rule.

54  A Prima Facie Case for the Coercion Thesis characterizing as a system of law, the dictionary definitions misreport the relevant canons of usage—​as would be the case if the Pope is not properly characterized as a bachelor. Since the lexical meanings of words are determined by our empirically contingent linguistic practices with respect to using them, any dictionary definition inconsistent with those practices is problematic for that reason: dictionary definitions do not establish the correct meaning of terms; they report them—​and any such report must be consistent with the observable linguistic practices they purport to accurately report. The following sections are concerned to more directly examine our linguistic practices with respect to distinguishing things we characterize as systems of law from things we do not characterize as systems of law. Since systems of municipal law constitute the paradigm cases conditioning the canons of ordinary usage with respect to the concept-​term law reported in dictionary definitions, the remaining sections of this chapter are concerned with our linguistic practices with respect to distinguishing existing systems of municipal regulation we characterize as law from existing systems of municipal regulation we do not characterize as law.

2.  Types of Coercive Sanction in Modern Municipal Law We cannot ascertain whether our experience distinguishing things counting as systems of law from things not counting as systems of law conforms to the content of the canons of ordinary usage described by the dictionary definitions above unless we have some idea of what authorized enforcement mechanisms are deployed in existing systems we characterize as systems of law. This section is concerned to determine whether the various enforcement mechanisms typically deployed in existing systems of municipal law are properly characterized as coercive sanctions. The compound concept of a coercive sanction—​for purposes of understanding the Coercion Thesis—​was explicated in Chapter 1. Something counts as coercive if backed by something reasonably contrived to deter non-​compliance by threatening the imposition of something rationally competent subjects characteristically regard, as a descriptive matter of contingent fact, as a prudential deterrent to non-​compliance because they should regard it, as an objective matter of normative practical rationality, as such. Something counts as a sanction if reasonably contrived to punish non-​compliance, in effect if not expressed intent, with something rationally competent subjects characteristically regard, as a descriptive matter of contingent fact, as detriment they prefer to avoid because they should regard it, as an objective matter of normative practical rationality, as detriment they prefer to avoid. The compound concept of a coercive sanction thus incorporates both the forward-​looking deterrence function of coercion and the backward-​looking retributively punitive function of sanctions:  something counts as a coercive

Types of Coercive Sanction in Modern Municipal Law  55 sanction, then, for our purposes, if and only if reasonably contrived to perform both of these functions. This conforms to the canons of ordinary usage as they are described by the corresponding dictionary definitions. Oxford Online Dictionary defines the concept-​term coercion as meaning “[t]‌he action or practice of persuading someone to do something by using force or threats” and the concept-​term sanction as meaning “[a] threatened penalty for disobeying a law or rule.”10 Merriam-​ Webster defines coerce as meaning “to achieve by force or threat” and the concept-​term sanction as meaning “the detriment, loss of reward, or coercive intervention annexed to a violation of a law.”11 Extrapolating from these definitions, a coercive sanction is, by nature, reasonably contrived to deter and punish violations insofar as rationally competent subjects characteristically regard it, as a descriptive matter of contingent fact, as something reasonably contrived to deter and punish because they should regard it, as an objective matter of normative practical rationality, as something reasonably contrived to deter and punish. One might think that the Oxford Online Dictionary definition of sanction does not fully conform to the analysis of the notion given in Chapter 1 insofar as it presupposes that the explicit intent of backing a norm with a sanction is to threaten punishment, but this does not matter for our purposes. This volume is concerned to show that a modest analysis of the concept of law entails that every conceptually possible legal system backs some mandatory legal norms governing non-​official behavior with the threat of something that counts as a coercive sanction as I explicate that compound notion—​and not necessarily as this notion is defined by the canons of ordinary usage. This volume is concerned to give a modest analysis of only the concept of law as it picks out systems of law; while I think that the analysis given here of the concept of a sanction conforms to the canons of ordinary usage, this volume is not concerned to give a modest analysis of that concept. It is enough for our purposes that what counts as a sanction largely conforms to the relevant linguistic conventions. The analysis of Chapter 1 surely meets that standard even if we assume that the Oxford Online Dictionary definition, construed hyperliterally, better reports the canons of ordinary usage as they pertain to the term sanction than my analysis; I could, after all, simply affix an asterisk to sanction to indicate I am assuming a non-​standard notion for the sake of expository simplicity. While expressing the Coercion Thesis using the special term sanction* might have the advantage of avoiding this worry, there is no need to do this; once it is understood what I mean to express by the concept-​term coercive sanction, there

10 Oxford Online Dictionary; available at: https://​www.lexico.com/​en/​definition/​coercion and https://​www.lexico.com/​en/​definition/​sanction. 11 Merriam-​Webster Online Dictionary; available at: https://​www.merriam-​webster.com/​dictionary/​coerce and https://​www.merriam-​webster.com/​dictionary/​sanction.

56  A Prima Facie Case for the Coercion Thesis is no reason to further clutter the discussion with yet another homely symbolic convention. Accordingly, the following subsections discuss the various enforcement mechanisms typically authorized in existing legal systems by criminal and civil legal norms and distinguish those that count as coercive sanctions, on the analysis given in Chapter 1, from those that do not count as coercive sanctions.

2.1  Criminal Punishment and the Court’s Authority to Hold a Party in Contempt There are two classes of enforcement mechanism deployed in every existing legal system with which our experience acquaints us. The first class consists of those deployed to punish violations of the criminal law and might include, depending on which legal system we are talking about, community service, punitive fines, imprisonment, execution, and torture. Each of these mechanisms is imposed on subjects against their will for the express purpose of punishing violations and is reasonably contrived to do so in virtue of inflicting detriment that subjects characteristically avoid, as a descriptive matter of contingent fact, because they should avoid it, as an objective matter of normative practical rationality. Since, as a conceptual matter, something is reasonably contrived to punish non-​compliance only if also reasonably contrived to deter non-​compliance, each of these mechanisms is also reasonably contrived to deter; insofar as rationally competent subjects prefer to comply as a means of avoiding detriment contrived to punish non-​compliance, it is also reasonably contrived to deter non-​compliance. Insofar as each is reasonably contrived to deter and punish violations, each is properly characterized, on any plausible conceptual explication of the notion, as a coercive sanction. The second class of enforcement mechanism is defined by the court’s authority to deploy the so-​called contempt sanction. There are two types that differ according to whether deployment is authorized by criminal or civil law. The criminal contempt sanction is authorized under criminal law and is explicitly punitive in character: the point is to punish subjects for defying a court order or for otherwise impugning judicial authority in a manner that interferes with a court proceeding. The civil contempt sanction can be punitive but is more generally deployed to induce compliance with a court order. Both the criminal and civil contempt sanctions may be deployed by judges at the trial and appellate levels to enforce any court order within the court’s legal authority; subjects of such orders can include both officials and non-​officials. Each enables a court to enforce its orders by backing them with a threat of incarceration—​whether by punishing non-​compliance or by inducing compliance. While a court might also be authorized by law to perform either of those functions by ordering financial penalties, its authority to do any of these things depends on

Types of Coercive Sanction in Modern Municipal Law  57 its ability ultimately to forcibly incarcerate recalcitrantly non-​complying parties; anything that counts as an enforcement mechanism, according to the canons of ordinary usage, involves a deployment of something that counts as force12—​and the court’s ability to enforce anything ultimately rests on its authority to incarcerate.13 The authority to incarcerate is the most fundamental of enforcement mechanisms; without this authority, there is nothing that a court can do, as far as the world of our experience is concerned, to enforce its directives. Criminal penalties and both forms of the contempt sanction are properly characterized as coercive sanctions, on any remotely plausible analysis of the notion that conforms to ordinary talk. The point of both is to inflict something experienced as detriment against the subject’s will; financial penalties, imprisonment, execution, and torture are characteristically regarded, as a descriptive matter of contingent fact, by rationally competent subjects as detriment that functions as a non-​trivial but defeasible prudential deterrent to non-​complying behavior because they should, as an objective matter of normative practical rationality, regard it as such; these measures are thus reasonably contrived to deter non-​compliance to some non-​trivial extent.14 But something is reasonably contrived to deter non-​ compliance, as a conceptual matter, only if also reasonably contrived to punish it:  someone who is deterred by the threat of detriment is deterred from non-​ compliance because she would experience its infliction as punishment—​regardless of whether it is explicitly intended to punish non-​compliance. It is true that criminal penalties differ from certain uses of the contempt sanction in that detriment is imposed in the former case after a violation takes place while it is imposed in the latter case to induce compliance, but this difference is of no importance. While criminal punishment might be imposed after the fact as retribution for culpable non-​compliance, the authorization of criminal penalties is as much contrived to induce compliance by discouraging non-​compliance as it is to respond retributively after a violation has occurred. In both cases, the deployment

12 As Oxford Online Dictionary defines the term, enforce means “cause (something) to happen by necessity or force.” Not surprisingly, the etymology of the term confirms the centrality of the notion of force to the notion of enforcement: “mid-​14c., ‘to drive by physical force’ . . . late 14c. as ‘exert force, compel’ . . . from Old French enforcier . . . ‘use force (on) [or] offer violence (to)’. . . .” Online Etymology Dictionary; available at: https://​www.etymonline.com/​word/​enforce. 13 The more painful measures involved in execution and torture all rest on the court’s ability to forcibly detain—​or incarcerate—​a subject against her will: obviously, enforcement officials cannot execute or torture someone not in their physical custody. The notion of incarceration, as intended here, involves no more than the forcible detention of subjects—​regardless of the means or conditions under which the subject is detained; such measures include prison sentences but also include detention not linked to criminal convictions or prison sentences. 14 This should not be construed to entail any claims about a party’s actual motivation for refraining from behaviors prohibited by law. See Chapter 1, Section 4. The claim that a rationally competent agent regards something as a deterrent to non-​compliance expresses no more than that the subject would regard it as a defeasible prudential reason to comply if antecedently motivated to do something the law prohibits. I have never been tempted to kill anyone; however, if I were, I would regard the prospect of being punished with life imprisonment as a defeasible prudential reason not to kill.

58  A Prima Facie Case for the Coercion Thesis of these enforcement mechanisms is intended and reasonably contrived to induce compliance by providing a prudential deterrent to non-​compliance in the form of detriment rationally competent subjects characteristically prefer, as a descriptive matter of contingent fact, to avoid because they should prefer, as an objective matter of normative practical rationality, to avoid it. These mechanisms are properly characterized as coercive sanctions, on any remotely plausible explication of the concept, including that articulated in Chapter 1.

2.2  Do Compensatory Damages Count as Coercive Sanctions? The civil law also makes characteristic use of something one might think is properly characterized as a coercive sanction. Civil defendants who violate some duty under tort law in a manner that proximately injures a plaintiff can, as a matter of law, be ordered by courts to pay money damages in the form of punitive or compensatory damages. Unlike punitive damages, the expressed rationale for imposing damages characterized by the law as compensatory, rather than punitive, is to compensate the plaintiff for injuries wrongfully caused by non-​compliance—​and not to punish non-​compliance. The expressed rationale, unlike that for imposing damages characterized by the law as punitive, is to restore the plaintiff, insofar as money damages are equipped to accomplish this, to the position she would have been in but for the injury she proximately sustained as the result of the defendant’s breach. Both types of damage award inflict something likely, as a descriptive matter of contingent fact, to be regarded by rationally competent subjects, other things being equal, as detriment because they should, as an objective matter of normative practical rationality, regard it as such. Unless a person is unconscionably wealthy or instantiates an idiosyncratically pathological set of preferences, she is likely to regard the prospect of being ordered by a court to pay another person a substantial sum of money as stressful and unpleasant—​regardless of whether these damages are characterized by law as punitive or compensatory and regardless of whether the expressed rationale for imposing them is to punish non-​compliance, rather than to compensate for injuries caused by non-​compliance. Both types of damage award, thus, satisfy one conceptually necessary condition for being properly characterized as a coercive sanction insofar as they are reasonably contrived to be regarded by defendants as detriment. Punitive damages are, of course, properly characterized as coercive sanctions. Punitive damages are intended to express judicial disapproval of conduct and may legally be imposed by courts only for the expressed retributive purpose of punishing non-​compliance that is “wanton, willful, or egregious.”15 They are reasonably



15

See, e.g., Lompe v. Sunridge Partners, LLC, 818 F.3d 1041 (10th Circuit, WY 2016).

Types of Coercive Sanction in Modern Municipal Law  59 contrived to punish wanton, willful, or egregious non-​compliance insofar as they inflict non-​trivial detriment that rationally competent subjects characteristically regard, as a descriptive matter of contingent fact, as punishing non-​compliance because they should, as an objective matter of normative practical rationality, regard it as such. Since something is reasonably contrived to punish if and only if reasonably contrived to deter, it follows that punitive damages are reasonably contrived to deter non-​compliance—​assuming that the detriment they impose is reasonably contrived to punish non-​compliance.16 The issue of whether compensatory damages, in contrast, are properly characterized as coercive sanctions depends on factors varying from one area of civil law to the next. The norms authorizing compensatory damages for breach of contract are not, as a general matter, reasonably contrived to deter or punish non-​compliance since they require no more than that the breaching party pay the value of the performance required by the contract; since the party paying the damages will be made no worse off by paying the ordered damages than she would have been had she complied, the threat of having to pay compensatory damages for breaches of contract is not reasonably contrived to either deter or punish non-​compliance. Something is reasonably contrived to deter or punish non-​compliance, as a conceptual matter, if and only if reasonably contrived to leave the offending party worse off than she would be if she complied.17 Insofar as the compensatory damages of contract law are not reasonably contrived to do so, they are not properly characterized as coercive sanctions. But compensatory damages imposed for breaches of tort law are reasonably contrived to deter and punish—​even if not expressly intended to do so. This is obvious with respect to intentional torts because intentional acts of any kind can be deterred by threatening anything rationally competent subjects characteristically regard, as a descriptive matter of contingent fact, as detriment to be avoided because 16 Not surprisingly, punitive damages are explicitly intended by the courts to deter similar misconduct by the defendant and other persons. See, e.g., Philip Morris USA v. Williams, 549 U.S. 346, 352 (2007). 17 The party being ordered to pay compensatory damages might be better off paying those damages where the breach is motivated by the idea that performance would cost the breaching party more than the value of the amount that she is ordered to pay. In many instances, breach of a civil legal duty is motivated by the prospect of saving money. One especially egregious example of such cost-​benefit thinking was the decision of Ford Motor Co. to delay recalling the Ford Pinto to correct a design defect it knew would cause numerous deaths on the ground that the cost of repairing the defects exceeded the expected sum of compensatory damage awards in cases where persons were killed or injured because of the defect. The jury awarded more than $100 million in punitive damages in the expressed hope that product manufacturers would take notice that the expected costs of such behavior might include unexpectedly large punitive damage awards; the idea was either to provide an incentive for manufacturers to make a more ethical and civic-​minded decision as a means of avoiding worst-​case punitive damage awards that could not be reliably predicted or to raise the predictable expected costs enough to ensure that manufacturers took greater care to avoid having to pay them. See, e.g., Roy J. Harris, “Jury in Pinto Crash Case: ‘We Wanted Ford to Take Notice’” Washington Post, February 15, 1978; available at: https://​ www.washingtonpost.com/​archive/​politics/​1978/​02/​15/​jury-​in-​pinto-​crash-​case-​we-​wanted-​ford-​to-​ take-​notice/​996a9aa5-​8f48-​4541-​8553-​19c63f666830/​?noredirect=on.

60  A Prima Facie Case for the Coercion Thesis they should, as an objective matter of normative practical rationality, regard it as such. But even unintentional torts can be deterred by threatening the imposition of detriment because carelessness can be deterred: if I know I will have to pay for a person’s injuries caused by my carelessness, I will try harder not to act carelessly. Insofar as these damages are reasonably contrived to deter non-​compliance, they are also reasonably contrived, as a conceptual matter, to punish—​even if not imposed for retributive purposes. Regardless of whether or not the punitive and compensatory damages authorized by civil law are properly characterized as coercive sanctions under one’s favorite theory of the corresponding concept, there is an indisputably coercive aspect to civil law practice—​even in systems of law resembling that of the U.S. that, as a descriptive matter of contingent fact, prohibit incarcerating defendants for failure to pay a civil debt.18 Courts in the U.S. are authorized to order employers to garnish an employee’s wages in order to ensure payment of a civil debt to a third party, and these court orders are backed by the court’s contempt sanction. Although any legal norms governing official behavior authorizing courts to sanction employers who fail to comply with a court order to garnish an employee’s wages would be coercive with respect to the employer, the threat of forced garnishment is coercive with respect to those subject to a civil judgment requiring the payment of compensatory damages; having one’s wages garnished against one’s will is reasonably contrived to both punish and deter non-​compliance with mandatory legal norms governing non-​official behavior that authorize the imposition of compensatory damages. Either way, the court’s practical ability to enforce any mandatory legal norms, whether criminal or civil, with anything properly characterized as a coercive sanction for my purposes ultimately depends on the cooperation of armed officials of the system who regard those orders as legally binding and comply with them for 18 It is, of course, conceptually possible for a system of law to authorize courts to order the incarceration of defendants who fail to pay a civil debt; however, this is not permitted, as a descriptive matter of contingent fact, in the U.S. Federal statutes prohibit incarcerating someone for failure to pay a civil debt, regardless of whether the failure is the result of unintentional indigence. It should be noted that a person can deliberately render herself indigent to avoid having to pay a civil debt or judgment in a number of ways. She can, for instance, quit her job; although this seems problematic, as an objective matter of normative prudential rationality, a defendant might feel sufficient animus toward the plaintiff to be willing to subject herself to material hardship in order to deny the plaintiff the amount of the judgment—​something that sometimes happens in angry domestic disputes involving child support. In addition, more affluent defendants can hide assets or move them to a foreign entity over which a court lacks jurisdiction; doing so limits what the plaintiff can reach by other means, such as garnishment. The Constitution prohibits incarcerating someone for inability to pay a criminal penalty that is not intentionally created by an act motivated by a desire to avoid having to pay the penalty. Bearden v. Georgia, 461 U.S. 660 (1983) (“A sentencing court cannot properly revoke a defendant’s probation for failure to pay a fine and make restitution, absent evidence and findings that he was somehow responsible for the failure or that alternative forms of punishment were inadequate to meet the State’s interest in punishment and deterrence, and hence, here the trial court erred in automatically revoking petitioner’s probation and turning the fine into a prison sentence without making such a determination.”); available at:  https://​supreme.justia.com/​cases/​federal/​us/​461/​660/​. I  am grateful to Maria Elias Sotirhos for pointing this out to me and to Michael Green for helping me work through some of the issues of U.S. law here.

Types of Coercive Sanction in Modern Municipal Law  61 that reason.19 A court cannot make a party to a judicial proceeding do anything without the assistance of other officials who, in our world, have the practical ability to force the party to do what the court has ordered because, unlike judges, they are legally authorized to use guns—​within limits all-​too-​frequently ignored—​to ensure compliance with their directives. Our ability to live together in a community in any existing legal system of which we know depends to a demoralizing extent on the proper behavior of officials authorized by law to use lethal weapons to force people into their custody. Courts in the world of our experience cannot do anything to enforce any directive, norm, or order without the cooperation of those officials.

2.3  Is Non-​Enforcement of Improperly Formed Instruments a Coercive Sanction? A remaining question with respect to areas of civil law has to do with what John Austin characterized as the “sanction of nullity.” Courts are typically required by mandatory legal norms governing official behavior to abstain from enforcing certain agreements and commitments, like those frequently memorialized in written contracts or wills, that are not properly formed under the relevant legal norms of civil law governing non-​official behavior. The question is whether the non-​ enforcement of improperly formed instruments counts, on the explication of the notion articulated in Chapter 1, as a coercive sanction. The Austinian idea that this required abstinence on the part of courts counts as a coercive sanction is not implausible. Insofar as people typically create such instruments as a means of ensuring courts enforce them, the court’s legal duty not to enforce improperly formed instruments defines a prudential incentive to ensure that instruments are properly formed. One can plausibly argue, then, the court’s required abstention from enforcing an improperly formed instrument is reasonably contrived to deter and punish non-​compliance with the relevant legal norms governing non-​official behavior and hence counts as a coercive sanction, on any plausible analysis of the notion. There are a couple of problems with this line of reasoning. The first arises from the fact that the relevant civil norms governing these instruments do not require anything by way of non-​official behavior in the sense that norms of criminal law

19 This is why it is implausible to think that the courts ultimately have, as a conceptual matter, sovereign lawmaking authority in a legal system in virtue of being exclusively empowered to decide what content will be enforced. Without the cooperation of non-​judicial enforcement officials with guns, courts cannot succeed in enforcing their orders by imposing sanctions. The idea that the courts, by nature, have sovereign lawmaking authority is as problematic as the idea that courts are, by nature, supposed to interpret the law and not make it—​which ignores the fact that the so-​called common law in legal systems is ultimately created by judicial decision.

62  A Prima Facie Case for the Coercion Thesis and norms of tort law do. Consider the norms of contract law: if two parties wish to make an informal agreement because they want to leave the unfriendly armed machinery of the law out of their affairs, they are free to do so as far as contract law is concerned. It is true that their agreement does not satisfy those formal requirements, but that does not constitute their behavior as non-​compliance with the norms of contract law any more than drinking a cup of coffee constitutes non-​ compliance with those norms; the relevant norms of contract law define conditions that an agreement must satisfy—​and not conditions that their behavior must satisfy. There is no plausible sense in which either party has violated any applicable mandatory legal norms because the norms governing contract-​formation do not require anything of them. The relevant civil norms are more plausibly thought of as governing the official behavior of the courts. Thus construed, the norms defining what counts as a properly formed contract define constraints on what courts are legally authorized to do: courts are legally required, other things being equal, to enforce instruments satisfying the relevant conditions and legally required, other things being equal, to abstain from enforcing instruments not satisfying the relevant conditions. When judges do what these norms require in deciding contract disputes, their conduct both satisfies the relevant norms and complies with them; however, when the parties take steps to ensure that their contracts are properly formed, they neither satisfy nor comply with the relevant norms of contract law—​because, properly construed, those norms do not govern their behavior. But the notions of satisfying a mandatory norm and complying with it can come apart even with respect to the behavior of those persons it governs. My eating a sandwich does not satisfy the requirements of a mandatory legal norm requiring me to pay my taxes by April 15 but does not count as a non-​complying behavior. My eating a sandwich defines an instance of non-​compliance only with respect to mandatory norms that either directly or indirectly require me to abstain from eating that sandwich: if my eating the sandwich entails eating an ingredient the law prohibits eating, then my eating it defines an instance of non-​compliance; if it does not, then my eating it does not define something that counts as non-​compliance. Non-​satisfaction of a norm by a person whose behavior it governs does not necessarily constitute non-​compliance with that norm. The converse is not true:  any behavior satisfying the requirements of a legal norm constitutes compliance, as a conceptual matter, with that norm. My eating something counts as satisfying only norms requiring, either directly or indirectly, that I eat that something: if the law requires me to be properly vaccinated against polio but the only way to receive the vaccine is to eat a sugar cube containing it, then my eating a sugar cube with the vaccine satisfies the norm requiring me to be vaccinated against polio; I am properly characterized as complying with the norm insofar as my behavior satisfies the requirements of the norm. One can comply with a norm, as a conceptual matter, only by doing something that satisfies its

Types of Coercive Sanction in Modern Municipal Law  63 requirements, but one does not necessarily violate a norm by doing something that does not satisfy its requirements: the two notions are logically related but conceptually distinct.20 Something counts as a coercive sanction for purposes of the Coercion Thesis only if reasonably contrived to deter and punish non-​compliance with some mandatory legal norm. It might be true that the “sanction” of nullity, such as it is, is reasonably contrived to incentivize rationally competent subjects to ensure their instruments are properly formed; however, doing what is necessary to create a properly formed instrument does not count as compliance with the relevant norms insofar as not doing what is necessary to create a properly formed instrument does not count as non-​compliance; it is a straightforward conceptual truth that an agent’s performing an act when required counts as compliance with a norm if and only if the agent’s abstaining from that act when required counts as non-​compliance with that norm. Not every norm defining a prudential incentive to do a defines a coercive sanction for abstaining from doing a. Promising a reward for capturing an armed robber creates a prudential incentive to do something psychologically healthy people lacking the requisite training are not likely to do because they should not do that something; it is obviously a bad idea, absent exceptional circumstances, to attempt to capture a violent criminal without having the skills to do so safely and legally. But that does not make not giving the reward to someone who does not capture the robber a coercive sanction for non-​compliance—​regardless of whether that person did not capture the robber because she did not try or whether she did not do so because she tried and failed. Sanctions, by nature, threaten to impose detriment and do not merely “threaten” to withhold a benefit; for this reason, the “sanction” of nullity does not count as a coercive sanction.21

20 It is not implausible to think that, strictly speaking, the two concept-​terms apply to different classes of things: behaviors, and not people, satisfy requirements, as a conceptual matter, while people, not behaviors, comply with requirements. While I would argue that norm-​conformity is like norm-​ satisfaction in this regard, nothing of import turns on any of this; I will speak as though behavior and persons can both comply with and satisfy requirements, even if conceptually inapt, when doing so simplifies the exposition without causing any obvious problems. 21 Nor does it promise, for what it is worth, anything properly characterized as a reward. While it is a conceptually necessary condition for something to count as a reward that it confers a benefit, it is not conceptually sufficient for something to count as a reward that it does so. As far as the canons of ordinary usage are concerned, it should be clear that the judicial enforcement of a contract in favor of a plaintiff is not properly characterized as a reward for properly forming the contract insofar as the defendant against whom the contractual terms are being enforced is equally responsible for doing so. The so-​called sanction of nullity occupies a conceptual space somewhere in between the notion of a sanction and the notion of a reward. There is also something misleading, for what it is worth, about characterizing the legal norms requiring judicial enforcement of a properly formed instrument as conferring a benefit—​though doing so might not be, strictly speaking, incorrect—​because those norms entitle a person to have such instruments enforced by the court. Legal entitlements are somewhat misleadingly characterized as conferring benefits insofar as the term connotes that the desired outcome is conferred as a matter of privilege, rather than right; this is what I take to be meant by the claim that health care is a right and not a privilege.

64  A Prima Facie Case for the Coercion Thesis Even so, the fact that the so-​called sanction of nullity is reasonably contrived to provide a prudential incentive to ensure that one’s instruments are properly formed tells us something of importance about how legal systems in our world are equipped to regulate behavior. Insofar as self-​interested beings like us are sometimes disposed to do things that can adversely impact the prudential interests of other people as a means of benefitting ourselves, law is best equipped to make a difference with respect to what beings like us do by deploying mechanisms that bear in the appropriate ways on the prudential interests of subjects. This can be done by either imposing detriment or creating entitlements. Law is a carrot-​and-​stick approach to regulating behavior—​though it is characteristically, if not necessarily, more stick than carrot. None of this entails that there could not be a legal system backing legal norms governing the proper formation of an instrument with the threat of a coercive sanction. If the law of contracts punished parties to an improperly formed agreement with a term of imprisonment, then the relevant legal norms would be backed by the threat of something properly characterized as a coercive sanction for non-​ compliance: both parties would be liable to punishment for non-​compliance unless it could be shown in some legally satisfactory manner that only one party is culpable under the relevant standards. While it is hard to imagine a state of affairs in which doing so would accomplish anything worth accomplishing, such a practice is conceptually possible. But there is a deeper problem with the idea that non-​enforcement of a norm counts as a coercive sanction. It is a conceptually necessary condition for something to count as a coercive sanction, as I have explicated the notion, that it is deployed against a party or imposed on a party; however, the so-​called sanction of nullity as it functions in our world is not properly characterized as either deployed or imposed in the relevant sense. To be properly characterized as something that is deployed against or imposed on someone in the relevant sense here, the agent doing the deploying or imposing must be granted authority, or authorized, by the law to do so. It mischaracterizes the situation to say that courts in our world have been granted authority not to enforce improperly formed instruments; the reality of the situation is that courts are required not to do so and hence lack authority to do so. It is no more plausible to characterize a court’s abstaining from enforcing what it lacks legal authority to enforce as something deployed against, or imposed on, the party wanting it enforced than it is to characterize a court’s abstaining from punishing an act that is prohibited by morality but permitted by law as something deployed against, or imposed on, the party wanting it punished. But if it is not properly characterized as deployed against or imposed on someone, it is not properly characterized, for my purposes, as a coercive sanction. This should not be construed to deny that there could be a coercive sanction constituted by non-​enforcement. If a court has the authority to enforce a

Types of Coercive Sanction in Modern Municipal Law  65 properly-​formed contract against one party P for non-​compliance with a mandatory legal norm m requiring P to honor the contract but declines to do so on the ground that the opposing party Q has violated some other mandatory legal norm n, the court’s refusal to enforce is not preposterously characterized as a sanction. But in this case, the court’s refusal to enforce m against P is legally discretionary because not legally required and is putatively justified under the law as a means of enforcing n against Q; despite the fact that it involves declining to enforce one law, the court’s refusal is properly characterized as enforcing the other with something that might count as a sanction. Insofar as non-​enforcement in this case is properly characterized as deployed against, or imposed on, Q, it is properly characterized as a coercive sanction. Something like this happened in Riggs v. Palmer.22 Elmer Riggs stood to take a substantial sum under his grandfather’s will; when his grandfather remarried, Elmer murdered his grandfather out of fear that he would change the terms of his will to exclude Elmer. When Elmer completed his term of imprisonment for the murder, he petitioned the court to take under the will. The court declined to allow Elmer to take his gift under the will to prevent him from profiting from his grandfather’s murder—​and not to punish him or deter others. But it is not preposterous in this case to characterize the court’s non-​enforcement with respect to Elmer’s gift as an additional coercive sanction for murdering his grandfather; the decision was likely experienced as punishment by Elmer and would have, in some sense, deterred him from killing his grandfather had he known doing so to take under the will would be futile. While it is not clear that non-​enforcement in this case is properly characterized as a coercive sanction, the Riggs case seems to confirm that there could be a coercive sanction constituted by non-​enforcement. Either way, Austin was incorrect in thinking that non-​enforcement of improperly formed instruments counts as a coercive sanction. Non-​enforcement of an instrument counts as a sanction, for purposes of the Coercion Thesis, only insofar as the court has the authority to decline to enforce a properly formed instrument as a means of punishing a party to that instrument for violating some other legal norm governing her behavior; that might include norms permitting a court to refuse to enforce an instrument in favor of a party who behaved in bad faith with respect to the creation of that instrument, as is true of contract law in legal systems like ours. So-​called sanctions of nullity might thus be conceptually possible; however, what Austin characterized as a sanction of nullity in legal systems like ours does not count as a coercive sanction, as I have explicated the notion.



22

Riggs v. Palmer, 115 NY 506 (1889).

66  A Prima Facie Case for the Coercion Thesis

2.4  Criminal Law, Civil Law, and the Coercion Thesis Regardless of whether compensatory damages and the sanction of nullity count as coercive sanctions, the enforcement mechanisms of civil law—​other than the civil contempt sanction—​are not relevant with respect to the Coercion Thesis. The Coercion Thesis asserts that it is a conceptually necessary condition for something to count as a system of law that it backs court orders and some mandatory legal norms prohibiting assaults on persons and property with the threat of something reasonably contrived, in effect if not expressed intent, to deter and punish non-​compliance. Apart from the civil contempt sanction, which is imposed for the purpose of inducing compliance rather than for the purpose of punishing non-​compliance, the only norms that a legal system must include, according to the Coercion Thesis, are mandatory legal norms that count as norms of criminal law in virtue of being expressly concerned with deterrence and punishment. It is enough to constitute something as a coercive sanction that it is reasonably contrived to deter and punish non-​compliance—​even if not explicitly intended for those purposes; however, it seems to be a conceptually necessary condition for a norm to be properly characterized as one of criminal law that the relevant enforcement mechanisms are imposed for the explicit purpose of punishing violations and for the implicit purpose of deterring them.23 Regardless of whether the civil contempt sanction is imposed for the purpose of deterring and punishing non-​ compliance, it counts as a coercive sanction because it is reasonably contrived to deter and punish non-​compliance. The Coercion Thesis is facially agnostic with respect to whether it is a conceptually necessary condition for the existence of a legal system that it includes any norms properly characterized as civil—​apart from those authorizing courts to impose the contempt sanction for the purpose of inducing compliance with court orders. The rationale is this: it is clear, on the one hand, that (1) the law must efficaciously prevent enough acts breaching the peace for us to be able to live together in a community and that (2) law is reasonably contrived to do this among passionately self-​interested subjects like us in worlds of acute material scarcity like ours only insofar as it backs some mandatory norms prohibiting such acts with a threat of incarceration; but it is not clear, on the other hand, whether there is anything law must be reasonably contrived to do to enable us to live together

23 There is one difficult case, however, that will be considered in the Appendix: suppose there is an institutional normative system that is as much like that of the U.S. as is compatible with the following: instead of punishing acts that are criminalized in the U.S. with incarceration, execution, or punitive fines, those “criminal” norms are enforced by imposing only compensatory damages. The issue is whether it counts as a legal system. I consider this question in the Appendix because both answers are compatible with the Coercion Thesis.

The Ubiquity and Centrality of Coercive Sanctions  67 in a community that requires the inclusion of civil norms like those of tort and contract law. Either way, the court’s effective authority to enforce its orders with the threat of incarceration grounds the practical capacity of a legal system to enforce anything with the threat of a sanction; if a court, or some other tribunal, cannot efficaciously order officials to take a party convicted of a punishable offense into custody, the system is not reasonably contrived to enforce any norms. The legal power of the courts to punish non-​compliance with mandatory legal norms governing non-​official behavior is ultimately grounded in its legal power to induce compliance with mandatory legal norms requiring officials to comply with court orders by imposing the contempt sanction. But this power also depends on the general cooperation of people serving as officials; the court cannot efficaciously order the incarceration of one official for contempt without the cooperation of other officials. Law might be an inherently coercive endeavor, even with respect to regulating the behavior of officials, but it is essentially grounded in both the general cooperation of those who serve as officials of the system and the acquiescence of those whose non-​official behavior is governed by the norms of the system.

3.  The Ubiquity and Centrality of Coercive Sanctions in Existing Municipal Legal Systems The prima facie case for the Coercion Thesis is grounded in the central role that coercive sanctions play in every system of municipal law in our world. Every such system, past or present, has backed court orders and mandatory legal norms prohibiting assaults on persons and property with the threat of a coercive sanction. Authorized coercive sanctions are hence ubiquitous among systems of municipal law constituting states and other domestic municipal entities in our world. But it is not just that authorized coercive sanctions are ubiquitous among existing systems of municipal law; it is also that they play a central role in legal practice in each of these legal systems. Nothing is more essential to criminal practice in existing legal systems than the fact that non-​compliance with mandatory legal norms prohibiting acts breaching the peace can be punished with the imposition of coercive sanctions. Although a court must obviously decide whether someone is guilty before reaching the further question about the appropriate legal remedy, the most important issue for both the prosecution and the defense will be whether the defendant will be punished and, if so, how severely. This is not to suggest that criminal defendants do not care about a guilty verdict; they do. But the reasons for this are generally instrumental in the sense that

68  A Prima Facie Case for the Coercion Thesis they make reference to the social consequences of a guilty verdict. A  criminal conviction exposes a person to other undesirable consequences than just the coercive sanctions imposed as punishment: it also exposes her to social ostracism that can affect her in ways that significantly impact her well-​being. Employers are often afraid or loathe to hire someone convicted of a felony who has served time in prison—​regardless of how otherwise qualified she might be.24 But while there are many good instrumental reasons to care about a guilty verdict, none of the feared consequences matters more, or should matter more, to a defendant in a criminal case than the prospect of being on the business end of the severe detriment that courts are authorized to impose as punishment on persons convicted of felonies. The authorization of coercive sanctions as a response to non-​ compliance with mandatory norms prohibiting assaults on persons and property is both ubiquitous among existing legal systems and central to criminal practice in every municipal legal system in our world. Coercive sanctions play an equally central role in the civil law practices of every existing legal system of which we have ever known. While it is true that parties might attempt to settle a case prior to trial, the point of proceeding to trial is to secure a favorable judgment with respect to whether and how the court will deploy its authority to impose coercive sanctions: in the case of the plaintiff, the point is to secure an order that will be enforced against the defendant by the threat of the contempt sanction; in the case of the defendant, the point is to have the case dismissed without an order that can be enforced against her to her detriment. Either way, the court’s authority to issue a contempt sanction to induce compliance with a court order looms nearly as large in the mind of the parties to a dispute under civil law as its authority to impose punishment does in the minds of the parties to a dispute under criminal law. This is not to suggest that the court’s substantive holding is of no importance to the parties in a civil dispute. Not every plaintiff is necessarily seeking money damages in a lawsuit. A plaintiff might be primarily concerned to establish new law or to have her position publicly validated by a court. Someone who wins a civil lawsuit but is awarded only a token sum as money damages might still feel wholly satisfied with the result in virtue of having her legal position vindicated; however, as a general matter, plaintiffs pursue litigation to get an award of money damages that can be enforced by the contempt sanction while defendants attempt to persuade the court not to award such damages.

24 The reluctance to hire persons who have served time is exacerbated by the unfortunate fact that the inhumane conditions most inmates experience while incarcerated are considerably more likely to increase the probability they reoffend upon release than to decrease it by rehabilitating them. A U.S. Bureau of Justice study tracking 401,288 persons released from state prisons in 2005 found that about 68% of them were arrested within three years, 79% within six years, and 83% within nine years. See “Recidivism,” National Institute of Justice Report; available at: https://​www.nij.gov/​topics/​corrections/​ recidivism/​Pages/​welcome.aspx#note1.

The Ubiquity and Centrality of Coercive Sanctions  69 The court’s ability to coerce compliance with its orders hence defines the most prominent motivation for pursuing either criminal or civil litigation. A prison sentence means nothing, from a purely prudential standpoint, unless the court is willing and able to force you into that cell. The prospect of being held liable for a certain amount of money damages likewise means nothing, from a purely prudential standpoint, unless the court is willing and able to force you to pay those damages by credibly threatening something you are likely to regard as leaving you worse off than if you simply pay them. A judicial holding that cannot be enforced might still matter to a party but not as much as it does if it can be enforced. So central is the role of coercive sanctions to existing legal practices that the idea that the point of litigation is to secure an enforceable judgment is built into the concept of an actionable grievance. A grievance is defined as “actionable” insofar as it furnishes “a legal ground for an action,”25 which is constituted by a “fact, or a state of facts, to which a law sought to be enforced against a person or thing applies.”26 The nature of an actionable grievance or cause, as this notion operates in systems of municipal law resembling that of the U.S., is partly constituted by the availability of coercive sanctions as part of an authorized legal remedy. The empirical centrality and ubiquity of legal practices permitting courts to impose coercive sanctions on persons who violate court orders or criminal norms prohibiting assaults on persons and property is unsurprising. Law is needed in this world to prevent disputes among passionately self-​interested beings like us from erupting into violence that can threaten the well-​being of everyone in the affected community. It is true, of course, that the threat of sanctions is not always enough to deter rationally competent self-​interested subjects from violating mandatory legal norms prohibiting assaults on persons and property; despite the severe sanctions authorized worldwide for violations of laws prohibiting homicides, an estimated 385,000 were committed worldwide in 2016.27 But, as long as we lack the political will to address the root causes of violent crime, the best we can do to minimize it is to deter it with threats of force: we know of no reasonably large society in which people live peacefully together without the mediation of the law with its coercive enforcement mechanisms. It might be nomologically possible for beings like us to live in peace without law’s coercive governance; however, as a practical matter, we could do so only under circumstances that have not been realized in any existing society of which we know.

25 “Actionable” Black’s Law Dictionary, 11th edn (West Publishing Company 2019). 26 Ibid; emphasis added. See also, e.g., Federal Rule of Civil Procedure, 12(b)(6), which authorizes the dismissal of a pleaded cause of action for “failure to state a claim upon which relief can be granted[.]‌” 27 Kevin Rawlinson, “Global Homicide Rate Rises for First Time in More Than a Decade” The Guardian, December 7, 2017; available at:  https//​www.theguardian.com/​world/​2017/​dec/​07/​ global-​homicide-​rate-​rises-​first-​time-​decade-​venezuela-​jamaica.

70  A Prima Facie Case for the Coercion Thesis

4.  Sanctions as a Paradigmatic Feature of Law: A Prima Facie Case for the Coercion Thesis The prima facie case for the Coercion Thesis under MCA is grounded in the canons of ordinary usage as they are defined by our empirically contingent linguistic practices and our empirical experience with respect to what we characterize as systems of law in our world. Ordinary dictionaries, to begin, converge in reporting the lexical meaning of the relevant usage of law as encompassing only systems of norms authorizing the imposition of coercive sanctions for non-​compliance with some mandatory legal norms. These dictionary reports, further, conform to our experience with respect to what we characterize as law in our world: there is nothing in our world that we characterize as a system of law that lacks norms authorizing courts to impose sanctions for non-​compliance with at least some mandatory legal norms governing non-​official behavior. The authority of courts to impose coercive sanctions for non-​compliance is ubiquitous among existing legal systems. But the prima facie case for the Coercion Thesis does not just rest on the determinacy of our linguistic practices with respect to characterizing as systems of law only systems backing some mandatory norms governing non-​official behavior with the threat of a coercive sanction; it also rests on the utter centrality of the role that these sanctions play in every existing institutional normative system we characterize as a system of law. The fact that the practices associated with authorizing and imposing sanctions for non-​compliance are utterly central to legal practice in every existing legal system of which we know, rules out the possibility that it is merely happenstance that we converge in characterizing only institutional normative systems backing some mandatory norms with the threat of a coercive sanction as systems of law. The institutional availability of such sanctions is part of what distinguishes systems of law in our world from everything else. Other institutional normative systems punish non-​compliance with something properly characterized as a coercive sanction; even the rules of a chess association are backed with the threat of a coercive sanction in the form of expulsion from the association. But only systems of law punish non-​compliance with mandatory norms prohibiting assaults on persons with the threat of imprisonment or worse. Our linguistic practices are perfectly clear on this point. To fully appreciate the force of the argument here, consider the difference between our ordinary linguistic practices with respect to the concept-​term law and those with respect to the concept-​term bachelor. While the dictionary report of the lexical meaning of bachelor entails that every conceptually possible unmarried adult male counts as a bachelor, our linguistic practices appear indeterminate with respect to whether the Pope counts as a bachelor; insofar as competent speakers split on whether the Pope counts as a bachelor, our linguistic practices

Sanctions as a Paradigmatic Feature of Law  71 are indeterminate with respect to the issue.28 In contrast, the argument of this chapter shows that our linguistic practices are fully determinate with respect to the Coercion Thesis: the dictionary reports of the lexical meaning of law and our universal experience with respect to applying them to things in this world converge on characterizing only institutional normative systems that back some mandatory norms with the threat of a sanction as systems of law. The reason for the indeterminacy with respect to the concept-​term bachelor is ultimately that competent speakers disagree—​though they may not realize it or put the disagreement this way—​on whether it is central to our linguistic practices that only things institutionally eligible for marriage (in a way that the Pope is not) count as bachelors. The deeper issue here has to do with whether the point of our linguistic practices with respect to this term is to identify adult males who are marriageable as a means of promoting the pairing of marriageable persons with one another to encourage the development of stable families. The split on the issue goes to the very heart of why we think it important to distinguish things that count as bachelors from things that do not count as bachelors. It is clear, in contrast, why we might want to single out only institutional normative systems backing some mandatory norms with the threat of a coercive sanction as being properly characterized as systems of law. People have a prudential interest in being able to quickly distinguish norms and systems that might subject them to the prospect of being incarcerated from those that do not. The only reason that I ever care about whether the law has something to say about something I am contemplating doing, as opposed to whether morality has something to say about it, is that I want to avoid trouble with the law because I want to avoid being incarcerated—​and I do not think that I am unusual in this regard: the most prudentially salient feature of law in every existing legal system of which we know is that it punishes non-​compliance with very painful detriment. Either way, it is not enough to establish the Coercion Thesis under MCA that our empirically contingent linguistic practices converge on characterizing only institutional normative systems backing certain mandatory norms with a threat of a coercive sanction; under the modest approach adopted here one must consider both our linguistic practices and the underlying philosophical judgments of ontological salience that ground and condition those linguistic practices. Our conceptual practices with respect to a thing T are defined, under MCA, by our ordinary linguistic practices with respect to using “T” as they are properly understood to incorporate the underlying philosophical assumptions about the metaphysical nature of T. 28 This suggests that the dictionary reports of the lexical meaning of bachelor are not fully accurate, but they could not be improved in any relevant way because it is not clear whether we converge, as a descriptive matter of contingent fact, in characterizing the Pope as a bachelor. These puzzling conceptual issues, which are philosophical in character, call into question the accuracy of such reports without entailing that they require modification.

72  A Prima Facie Case for the Coercion Thesis Nothing has been done in this chapter to identify the philosophical assumptions about the metaphysical nature of law that pick out the salient features that ground and are expressed in our linguistic practices: to say that we regard the availability of such sanctions as central to distinguishing things that count as law from things that do not count as law simply identifies a shared conviction about the centrality of sanctions to law without identifying the underlying philosophical assumptions about the metaphysical nature of law that would explain this shared conviction. The remaining chapters of this volume are concerned with articulating the philosophical assumptions about the metaphysical nature of law that ground our empirically contingent linguistic conventions for using the associated concept-​ term. We adopt the social practices constituting something as a system of law because we assume that these practices are reasonably contrived, by nature, to do all the things we characteristically use legal systems to do; we cannot make sense, as an objective matter of normative practical rationality, of why we adopt legal systems to do what we characteristically deploy them to do unless we assume that they are well-​suited, by nature, to do those things. If we do not assume, for instance, that automobiles are reasonably contrived, by nature, to transport persons and things from one place to another, we cannot make sense of why we use them to do what we use them to do. MCA is concerned ultimately to tell us something about us that is gleaned from our collective efforts to structure the world of our experience through our conceptual practices. While each of the succeeding chapters deals with different dimensions of the relevant conceptual practices, the operative unifying strategy is to show that we cannot make sense of what we do with law unless these practices entail that it is a conceptually necessary condition for something to count as a legal system that it backs some mandatory legal norms governing non-​official behavior with the threat of a coercive sanction.

4

What Law Must Be Able to Do: The Coercion Thesis and the Need to Keep the Peace Chapter 3 presented a prima facie case for the claim that it is a conceptually necessary feature of a legal system that it authorizes courts to impose coercive sanctions for non-​compliance with some mandatory legal norms prohibiting assaults on persons and property. The argument was grounded in the idea that our empirically contingent linguistic practices with respect to using the concept-​term law provide presumptively persuasive support for the idea that it is the metaphysical nature of law that some of these mandatory legal norms governing non-​official behavior are backed by the threat of a coercive sanction. Given that the content of our concept of law, according to the modest approach adopted in this volume, is determined by our empirically contingent linguistic practices as they are conditioned by underlying philosophical assumptions about the metaphysical nature of law, the Coercion Thesis is justified unless it conflicts with these philosophical assumptions. The reason is that these philosophical assumptions explain why we carve up the world conceptually the way we do. It is not happenstance that we distinguish people who count as bachelors from people who do not count as bachelors and normative systems that count as law from normative systems that do not count as law. We make these distinctions because doing so expresses differences among things in the world that we find salient enough to want to incorporate them into the language we adopt to talk about things that matter to us. And these judgments of salience are grounded in philosophical assumptions that are expressed in the metaphysical structure we impose on the world through our conceptual practices. For this reason, a comprehensive case for the Coercion Thesis must, at the very least, show that it is consistent with these underlying assumptions. The remainder of this volume shows something much stronger than that the Coercion Thesis is consistent with these philosophical assumptions about the metaphysical nature of law; it shows that they warrant accepting the Coercion Thesis. This chapter begins that project with an argument grounded in a claim about the function something must be able to perform to be properly characterized as a system of law. It argues that it is a conceptually necessary condition for something to count as a system of law that it is reasonably contrived to keep the peace Coercion and the Nature of Law. Kenneth Einar Himma, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198854937.001.0001

74  What Law Must Be Able to Do among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours by regulating behavior through the governance of norms metaphysically capable of guiding behavior. But the only way that an institutional normative system could be reasonably contrived to perform this function is by backing some mandatory norms prohibiting assaults on persons and property with the threat of a coercive sanction.

1.  The Concept of an Artifact Artifacts are non-​natural objects contrived and used to do things:1 hammers are non-​natural objects contrived and used to do things that require pounding a nail into some surface; automobiles are non-​natural objects contrived and used to do things that require transporting persons or things from one location to another; and computers are non-​natural objects contrived and used to do a variety of things by running other non-​natural objects in the form of computer programs specifically contrived to do these things.2 There are thus two conditions singly necessary and jointly sufficient for an object to be properly characterized as an artifact as the canons of ordinary usage define that term. The first is that the object counts as artificial in virtue of having been created by the efforts of intentional beings; artificial objects, as a conceptual matter, do not exist naturally in nature and are hence brought into existence through the efforts of intentional beings who contrive, manufacture, or otherwise create them and thereby bring them into existence.3 The conceptual claim that artificial objects are brought into existence through the activity of intentional beings does not imply the claim that the point of activity bringing artificial objects into existence is to bring those objects into existence; many such products are simply byproducts of activities that are otherwise 1 Dictionary.com defines artifact to mean “any object made by human beings with a view to subsequent use”; available at: https://​www.dictionary.com/​browse/​artifact. The term is also used to designate artificial objects of particular interest to cultural anthropologists; as Oxford Online Dictionary defines this usage, an artifact is “an object made by a human being, typically one of cultural or historical interest.” Oxford Online Dictionary; available at: https://​en.oxforddictionaries.com/​definition/​ artefact. 2 While conceptual theorizing about the nature of an artifact has largely focused on objects that are manufactured out of material things and are hence themselves material, the application-​conditions of the concept-​term artifact, as they are determined by the canons of ordinary usage, are not limited to material objects. Complex abstract objects like novels and computer programs are properly characterized as artifacts. Although these objects must be communicated in some physical medium, such as a writing or computer code on a flash drive, they are not identical with the physical object in which they are inscribed for human consumption; a novel is not a book, and a computer program is not a flash drive. 3 Oxford Online Dictionary defines “artificial” as “[m]‌ade or produced by human beings rather than occurring naturally”; available at:  http://​www.oxforddictionaries.com/​definition/​english/​artificial. Likewise, Merriam-​Webster defines it as “humanly contrived, often on a natural model”; available at: https://​www.merriam-​webster.com/​dictionary/​artificial.

Uses, Purposes, and Functions of Artifacts  75 intentional. It is conceptually sufficient for something to count as an artificial object that (1) it is not a natural product of biological activity, such as human progeny and bodily waste; and (2) it would not exist if human beings did not exist. The second is that the object is contrived and used by people to do certain things. Not every artificial object counts as an artifact according to our conceptual practices. Air pollution caused by a factory used to manufacture something else counts as artificial but does not count as an artifact insofar as it is not used by someone to do something. But it is not preposterous to think it would count as an artifact if someone figured out how to do something useful with it and deliberately started producing it for that purpose; however, it would be conceptually distinct from the air pollution routinely produced by factories and would likely be referred to with a new term lacking the pejorative connotation of pollution. It is thus a conceptual truth that everything properly characterized as an artifact counts as such in virtue of being an artificial object contrived and used by people to do things. We contrive and use artifacts to do things because doing those things conduces to realizing some end that matters to us; artifacts are thus tools of a sort: we contrive and use automobiles to do things involving transporting people and property from one place to another because doing those things conduces to realizing some end that we value. Understanding the metaphysical nature of an artifact of any kind thus enables us to understand something about ourselves insofar as its contrivance and characteristic use says something about us, what we value, and how we realize those values.

2.  Characteristic Uses, Basic Purposes, and Conceptual Functions of Artifacts Any artifact A can be used to do a multiplicity of things.4 Some of the things A can be used to do are idiosyncratic in the sense that A is not typically used to do these things because there are other artifacts better suited to do them; an automobile can be used as a paperweight but this is an idiosyncratic use insofar as there are other artifacts much better suited for this purpose. But some of these things A can be used to do are non-​idiosyncratic in virtue of its being better suited to do those things than other artifacts; an automobile can be non-​idiosyncratically used to go to the grocery store or take friends to the movies insofar as it is better suited to do these things than other artifacts. There is some non-​idiosyncratic use p to which A can be put that is logically basic in the sense that A can be put to all other non-​idiosyncratic uses only if it can be used to do p. An automobile can be used for going to the grocery store or 4 See Chapter 2, Note 3, for an explanation of the conventions I adopt with respect to using underlines and italics to indicate what kind of thing I am talking about.

76  What Law Must Be Able to Do taking someone to the movies only insofar as it can be used to transport persons and things from one place to another; something that cannot be used to transport persons and things from one place to another cannot, as a conceptual matter, be used for either of these purposes. This logically basic use defines A’s empirically characteristic use in the sense that everyone who uses A non-​idiosyncratically is ultimately using it for this purpose. Someone using an automobile non-​idiosyncratically to go to the grocery store and someone using an automobile non-​idiosyncratically to take someone to the movies are both, of conceptual necessity, using an automobile to transport persons or things from one place to another. Since everyone who uses an automobile non-​idiosyncratically is using it to transport persons or things from one place to another while not everyone who uses an automobile non-​idiosyncratically is using it to go to the grocery store or to take someone to a movie, it follows that neither of these latter uses counts as its characteristic use. Since an automobile cannot be non-​idiosyncratically used for anything it can be used to do without transporting persons or things from one place to another, it follows that transporting persons or things from one place to another is its characteristic use.5 The basic quality of this characteristic use corresponds to and reflects a normative element that is functional in nature. Something counting as an automobile that is transiently, as a descriptive matter of contingent fact, unable to transport persons or things from one place to another can still be put to idiosyncratic uses, such as to kill people by dropping it on them from the top of a building, but none of these uses depends on its having all the functional qualities of an automobile that enable it to perform its basic characteristic function of transporting persons and things from one place to another. Something that counts as an automobile but is transiently unable to perform its basic function is broken or defective qua automobile in virtue of not being utilizable for this purpose. This normativity is purely functional. Something properly characterized as a fork made out of material too flaccid to reliably spear pieces of solid food is functionally deficient but not morally problematic in virtue of this defect. This purely functional normativity of an artifact is not grounded in substantive moral standards governing how people should behave with it; these moral standards dictate, among other things, that one should not use a fork to stab people in the face and that one should not sell a functionally defective fork to people who are unaware of its functional defects; they do not dictate that forks that cannot be used for this

5 The concepts of characteristic, non-​idiosyncratic, and correct as they pertain to uses of artifacts are related roughly as follows. All non-​idiosyncratic uses of an artifact are properly characterized as correct in virtue of being used to do something that can be done only because it can be used to perform its basic function; these uses are not properly characterized as characteristic insofar as they differ in particular respects from other correct uses. Someone who uses a fork to eat a piece of beef liver and someone who uses a fork to eat a Brussels sprout are both using it correctly, but neither of these specific uses is characteristic because most people make it a point to avoid eating beef liver and Brussels sprouts.

Uses, Purposes, and Functions of Artifacts  77 purpose are morally defective. Standards of moral normativity are used to assess the acts of people while standards of functional normativity are used to assess the capacity of something properly characterized as an artifact “A” to do what As are characteristically used to do. This functional normativity presupposes some kind of designation that constitutes an object’s basic characteristic use as what it is supposed to do and thereby defines its conceptual function. The claim that A is characteristically used to do x is a purely descriptive one expressing the most common use of things counting as A; claims about the characteristic use of an artifact A are thus contingent and hence have no immediate or direct logical implications with respect to the concept of A or the conceptual function of A. It is the fact that A’s characteristic use is regarded as functionally normative that tells us something about its nature and conceptual function: what warrants thinking that spearing a piece of solid food for the purpose of placing it in one’s mouth is the fork’s conceptual function is the fact that users of a fork generally think it is “supposed” to be able to spear food for this purpose. There is a second respect in which our practices with respect to A involve a functional normativity that defines A’s conceptual function: it is not just that As are supposed to be used that way; it is also that they can, when neither in a state of disrepair nor improperly designed or constructed, be used in that way. Something is properly characterized as supposed, as a functionally normative matter, to do x only insofar as it is capable, in some sense, of doing x, when neither in a state of disrepair nor improperly designed or constructed; it makes no sense to think an artifact is supposed to do something it cannot do.6 As a general matter, it is a conceptually necessary condition for something to be properly characterized as an artifact “A” that it is capable of being used to do what As are supposed to do when neither damaged nor improperly designed or constructed. The term capable is ambiguous as to the character of the possibility operator it incorporates. The claim that some x is capable of performing A’s conceptual function can be construed to make one of four different assertions: (1) it is logically possible that the correct use of x, when neither damaged nor improperly designed or constructed, succeeds in performing A’s conceptual function; (2) it is conceptually possible that the correct use of x, when neither damaged nor improperly designed or constructed, succeeds in performing A’s conceptual function; (3) it is nomologically possible that the correct use of x, when neither damaged nor improperly designed or constructed, succeeds in performing A’s conceptual function; and (4) it is practically possible that the correct use of x, when neither damaged nor improperly designed or constructed, succeeds in performing A’s conceptual function.

6 The converse is not true: the claim that some artifact is capable of being used to do x does not imply the claim that it is supposed to do x; an automobile is capable of being used to kill people but it misdescribes the situation to say that it is supposed to be used to kill people.

78  What Law Must Be Able to Do Claims (1), (2), and (3) are too weak to express the relevant sense of capability. Consider an artifact contrived to perform the conceptual function of an automobile purely on the strength of energy generated from sunlight; and suppose that the energy it produces from sunlight is sufficient to perform that function only on planets exerting much lower levels of gravitational force than is exerted by the earth; that artifact is logically, conceptually, and nomologically capable of performing the conceptual function of an automobile but is not properly characterized as an automobile because we cannot use it to transport people and things from one location to another. As a conceptual matter, we must be able to use something to perform an artifact’s conceptual function in order to properly characterize it as an artifact of that type.7 The relevant sense in which something must be capable of performing the conceptual function of A to be properly characterized as an “A” is thus practical. Two conceptually related conditions must be satisfied by an object x to be practically capable of performing the function of A in the relevant sense: first, rationally competent subjects like us must be able to use it in the appropriate non-​idiosyncratic way; and, second, it must be reasonably likely that using x in the appropriate non-​ idiosyncratic way, when x is neither damaged nor improperly designed or constructed, will succeed in performing that function. If it is not reasonably likely that our use of x in the appropriate non-​idiosyncratic way succeeds in performing the conceptual function of A and x is neither damaged nor improperly designed or constructed, then x does not count as an A. Something that can be used to tell time only by beings more intelligent than any human being is not properly characterized as a watch; it might count as a prototype of a watch (assuming that it is nomologically possible for human beings to evolve to become beings with such intellectual capacities); but it does not count as a watch under our conceptual practices.

3.  The Problem an Artifact is Needed to Solve as Defining Its Conceptual Function Artifacts are contrived and characteristically used to solve some basic problem people want to solve:  a mousetrap is contrived, characteristically used, and 7 Not surprisingly, there is a parallel between the modest approach to conceptual analysis adopted here and the methodology for ascertaining the conceptual function of an artifact: what counts as the conceptual function of a thing is wholly determined by what we do with it, just as what counts as the content of some concept is wholly determined by what we do with the corresponding concept-​term. The reason is that the language we contrive to enable us to pick out certain concepts is itself an artifact. The semantics of a language consist of a set of symbols, a set of abstract objects expressing content, and a mapping from objects in the former set to objects in the latter set. While the lexical meanings our words pick out might be non-​artifactual abstract objects that exist in logical space regardless of whether there are beings like us in the world, those meanings become ours in virtue of the artifactual mappings that correlate them with the symbols of our language.

The Problem Defining an Artifact’s Function  79 supposed, as a functionally normative matter, to solve the problem of how to remove unwanted mice from some space, while an automobile is contrived, characteristically used, and supposed, as a functionally normative matter, to solve the problem of how to transport persons or things from one place or another. Removing unwanted mice is the basic function of a mousetrap because the problem it is contrived, characteristically used, and supposed, as a functionally normative matter, to solve is how to remove them from some desired space; transporting persons and things from one place to another is the basic function of an automobile because the problem it is contrived, characteristically used, and supposed, as a functionally normative matter, to solve is how to transport persons and things from one place to another. Insofar as it is the basic use of an artifact that solves the problem motivating its characteristic use and contrivance, that problem is the foundational problem that an artifact is needed to solve. The task of identifying the foundational problem an artifact is needed to solve is sometimes straightforward. Identifying the foundational problem that an automobile is needed to solve is easy because we have many obvious reasons to want something that can be used to transport persons or things from one place to another. It is clear that people contrive and use automobiles to perform that basic function and hence that its basic characteristic use solves the foundational problem it is needed and supposed, as a functionally normative matter, to solve—​namely, how to conveniently transport person or things from one place to another. But it can be more challenging to identify the relevant problem. It is not clear, for instance, what foundational problem rings are needed to solve. A ring can be used to express one’s marital status, social class, or educational background, or it can be used for purely decorative purposes. Indeed, it might be that no more can plausibly be said about the foundational problem rings are needed to solve than that the problem is expressive or symbolic in character and that rings are contrived and used to solve that problem by expressing the relevant content through a symbolic object worn on a person’s finger. Either way, the specification of an artifact’s conceptual function must indicate the ultimate end to which use of the artifact conduces, as well as the manner in which the object is supposed to be used to achieve that ultimate end; both the end and the means require specification here. A fork and a pair of chopsticks are both used for the ultimate end of transporting a piece of solid food from plate to mouth without having to touch it with one’s fingers, but they have different conceptual functions: the conceptual function of a fork is to facilitate the sanitary transport of solid food to the mouth by piercing it so that it can be secured on the fork while the conceptual function of a pair of chopsticks is to facilitate its transport by securing it between them; that is part of what distinguishes them conceptually. The specification of this combination of an artifact’s ultimate end and the manner in which it is characteristically used to achieve this end defines the foundational problem that an artifact is needed, contrived, used, and supposed, as a

80  What Law Must Be Able to Do functionally normative matter, to solve. Insofar as an artifact is characteristically used to secure this ultimate end in this manner, it follows that its characteristic use, when neither damaged nor improperly designed or constructed, is reasonably likely to achieve this ultimate end and is hence practically capable of solving the relevant foundational problem. In any case where an artifact has a conceptual function, it can be fully specified only by articulating the ultimate end for which it is contrived and the manner in which it is characteristically used to achieve this ultimate end.

4.  Something Must be Reasonably Contrived to Perform the Conceptual Function of an Artifact to Be Properly Characterized as an Artifact of That Kind The claim that an artifact of some kind is, as a conceptual matter, practically capable of performing the basic function for which artifacts of its kind are characteristically used implies the claim that its functionally correct use, when neither damaged nor improperly designed or constructed, is reasonably likely to succeed in performing that function.8 The claim that a watch is practically capable of informing us of the time of day implies the claim that its functionally correct use, when neither damaged nor improperly designed or constructed, is reasonably likely to succeed in informing us of the time of day. Similarly, the claim that an automobile is practically capable, properly construed, of transporting persons or things from one place to another implies the claim that its functionally correct use, when neither damaged nor improperly designed or constructed, is reasonably likely to succeed in transporting persons or things from one place to another.9 But insofar as it is a conceptually necessary condition for something to be properly characterized as an “A” that it is practically capable, properly construed, of performing the basic function for which it is characteristically used, it is a conceptually necessary condition for something to be properly characterized as an “A” that its functionally correct use is reasonably likely to succeed in performing the conceptual function of A. The basic function an artifact is characteristically used to perform defines the ultimate end of its use and hence its conceptual function. This implies that something must be reasonably contrived to solve the foundational problem an artifact is needed to solve to count as an artifact of that kind.

8 The correct use of an artifact is correlated with what it is supposed to do, as a functionally normative matter. An artifact is supposed to do x, as a purely functionally normative matter, if and only if it is correctly used to do x, also as a purely functional matter. The two notions are extensionally equivalent if not intentionally equivalent (i.e. refer to the same class of objects if not exactly synonymous in lexical meaning). 9 To simplify the discussion, I will omit the locution neither damaged nor improperly designed or constructed below: it should be understood that this requirement applies throughout.

The Conceptual Function of a Legal System  81 Insofar as it is a conceptually necessary condition for something to be properly characterized as an “A” that its correct use is reasonably likely to succeed in performing the conceptual function of A, it follows that it is also a conceptually necessary condition for something to be properly characterized as an “A” that it is reasonably contrived to perform that conceptual function if correctly used. Something that resembles a chair but is made of papier-​mâché might be properly characterized as a replica of a chair, but it does not count as a chair because it is not reasonably contrived to perform the basic function that chairs are characteristically used to perform. Insofar as we characteristically use an artifact A to perform some basic task it is supposed, as a functionally normative matter, to serve as a means of solving some foundational problem it is contrived to solve, it is a conceptually necessary condition for something to count as an A that it is reasonably contrived to perform A’s conceptual function and hence to solve the foundational problem that As are supposed to solve.

5.  The Conceptual Function of a Legal System Legal norms and legal systems are artifacts in virtue of being artificial objects used by human beings to do things like regulate behavior, mediate disputes, facilitate social coordination, etc. They are artificial objects in virtue of being brought into being by the social activities of persons who serve as officials of the system; there are obviously no legal norms or systems in worlds without rationally competent subjects like us. Although legal norms and legal systems are neither material objects nor constructed out of material objects, they are both properly characterized as artifacts, according to the canons of ordinary usage, insofar as they are brought into existence by our social practices and are contrived and used by us to do things.10 Legal systems are artifacts that are manufactured by social practices, understood pre-​theoretically as being legal in character, having to do with regulating our behavior and consist, among other things, of legal norms that are manufactured by related social practices. While legal systems and legal norms depend on each other for their existence, legal systems are artifacts wholly constructed of artifactual institutions defined by artifactual legal norms; both the pieces of the system and the system itself are social artifacts.11 Law is an artifact all the way down—​and clearly so. 10 See Note 2, above. 11 It is not implausible to think that these institutional objects, including a legal system, are ontologically identical with the set of norms that bring them into existence. It is hard to see what a legal system is if it is a set that includes something more than just all the legal norms of the system; it strikes me as inapt to think that it includes, for example, the persons who serve as officials. Whether there is a legal system in force in some society depends on whether there are certain persons doing certain things, but that is a different issue from the ontological issue of what a legal system is.

82  What Law Must Be Able to Do Qua artifact, law has a conceptual function. The idea that legal systems and legal norms have conceptual functions related to regulating behavior through the governance of norms metaphysically capable of guiding it is a truism. Insofar as it is conceptually true that (1) legal norms are contrived to govern behavior by guiding it,12 (2) there cannot be legal systems without legal norms, and (3) something can regulate the behavior of rationally competent subjects only through the governance of norms metaphysically capable of guiding it, it follows that the conceptual function of a legal system involves regulating the behavior of rationally competent subjects through the governance of norms metaphysically capable of guiding behavior. But the idea that a legal system is concerned to regulate behavior in this manner is insufficient to fully specify the conceptual function of a legal system because it specifies only how they are characteristically used—​that is, to regulate the behavior of rationally competent subjects through the governance of norms metaphysically capable of guiding behavior—​and does not specify the ultimate end for which they are characteristically used in this manner. As is conceptually true of all artifacts, there is some functionally normativized ultimate end that legal systems as such (i.e. without regard to the content of any legal norms that distinguish one conceptually possible legal system from another) are characteristically used to achieve by regulating behavior through the governance of norms metaphysically capable of guiding behavior.13 A specification of law’s conceptual function that says no more than that it is to regulate the behavior of rationally competent subjects through the governance of norms metaphysically capable of guiding behavior is incomplete because we do not characteristically contrive or use legal systems simply to regulate behavior; regulating behavior is not, after all, something that we characteristically value for its own sake. We contrive and use legal systems to regulate behavior as a means of achieving other ends that matter to us, whether these ends are explicitly acknowledged or not, and these ends are reflected in the content of the legal norms 12 This is true only of artifactual norms. Moral norms, if objective, govern behavior but are not contrived in the relevant sense because they are not manufactured by our practices. Objective moral norms would be a divine artifact if manufactured by a divine being but not a human artifact. 13 Legal systems differ from material artifacts in that they cannot exist without minimally doing what they are supposed to do, as a functionally normative matter. An institutional normative system that prohibits assaults on persons and property but fails to regulate behavior because it does not succeed in preventing assaults on persons and property is not minimally efficacious and for this reason is not properly characterized, according to our conceptual practices, as a system of law; while such a system might exist in logical space as a system of law, it is not one that regulates the behavior of intended subjects and does not exist as the legal system of some community. In contrast, an object that cannot minimally do what automobiles are supposed to do, as a functionally normative matter, might still be properly characterized, according to our conceptual practices, as an automobile; something that counts as an automobile might be irreparably broken or “totaled” in the sense that it cannot be used to perform the conceptual function of an automobile. While there can be totaled cars, there cannot be totaled legal systems. Material and abstract artifacts differ metaphysically in a number of intriguing ways not relevant here.

The Conceptual Function of a Legal System  83 promulgated as a means to secure them. To fully understand the metaphysical nature of a legal system as it is constituted by its conceptual function, we must specify the ultimate end to which systems of law are characteristically put and hence reasonably contrived to achieve. Theorists have held a variety of views on the ultimate end for which legal systems are characteristically used to regulate the behavior of rationally competent subjects through the governance of norms metaphysically capable of guiding behavior: (1) to morally justify the use of the state’s police power; (2) to tell rationally competent subjects what they should do according to right reason in a manner that makes it more likely than otherwise that they conform their behavior to the requirements of right reason; (3) to settle disputes about what morality requires; (4) to make objective moral requirements more determinate; and so on. All such views presuppose that the conceptual function of law involves regulating behavior through the governance of norms metaphysically capable of guiding behavior but attempt to identify some ultimate end to which such regulation is the means. Although each of these accounts specifies a value that legal systems can be used to serve, none indicates the foundational problem a legal system is contrived or used to solve. It might be true that one purpose of law is to morally justify the state’s use of its coercive police power, but this tells us nothing about why rationally competent self-​interested subjects like us might want to create a state that uses coercive police power. Likewise, it might be true that a second purpose of law is to induce people to behave in a manner that conforms to the requirements of right reason, but this tells us nothing about why rationally competent self-​interested subjects like us might want to construct a system of artifactual norms that accomplishes this purpose when there are already non-​artifactual norms of right reason. Further, it might be true that a third purpose of law is to settle disputes about what morality requires, but this tells us nothing about why rationally competent self-​interested subjects like us might want to construct a system of artifactual norms that accomplishes that purpose. Finally, it might be true that a fourth purpose of law is to make moral requirements more determinate, but this tells us nothing about why rationally competent self-​interested subjects like us might want to make objective moral requirements more determinate—​or, more tellingly, how subjects like us could do that given that no one has privileged access to the content of objective moral requirements. These accounts of law’s conceptual function tell us nothing about law’s ultimate end, which expresses the foundational problem that law is needed and characteristically used to solve. The most helpful place to begin in determining the foundational problem a legal system is needed, characteristically used, and hence supposed to solve, as a functionally normative matter, is to consider what life would be like for rationally competent self-​interested subjects like us in a world of acute material scarcity like ours without an artifactual system that regulates behavior the way law regulates behavior. If the goal is to determine what rationally competent self-​interested

84  What Law Must Be Able to Do subjects like us implicitly regard as the foundational problem that law is needed, characteristically used, and supposed to solve, as a functionally normative matter, the most helpful starting point is to consider what we believe that life would be like for us without law. The so-​called state of nature is a well-​known device in normative political philosophy. The state of nature is not intended to describe some historical condition in which people lived at some point in time; it is intended to provide a useful comparison between life with law and life without it to elicit shared intuitions about the conditions under which law is morally legitimate. The moral issue arises because systems of law coercively constrain what subjects are permitted to do by backing mandatory legal norms governing non-​official behavior with the threat of a coercive sanction and, as such, presumptively infringe the natural moral right to autonomy in a manner creating a need for moral justification.14 Different theorists flesh out the features of the state of nature differently. The most pessimistic depictions conceive it as a condition in which social cooperation and trust are, as a practical matter, impossible. Hobbes describes the state of nature as follows: Hereby it is manifest, that during the time men live without a common Power to keep them all in awe, they are in that condition which is called Warre; and such a warre, as is of every man, against every man. . . . In such condition, there is no place for Industry; because the fruit thereof is uncertain; and consequently no Culture of the Earth; no Navigation, nor use of the commodities that may be imported by Sea; no commodious Building; no Instruments of moving, and removing such things as require much force; no Knowledge of the face of the Earth; no account of Time; no Arts; no Letters; no Society; and which is worst of all, continuall feare, and danger of violent death; And the life of man, solitary, poore, nasty, brutish, and short.15

This all-​encompassing condition of war occurs because material resources in the state of nature are so scarce that rationally competent self-​interested subjects like us must resort to violence to satisfy our needs for them.16 Rationally competent 14 There are no serious issues of moral justification that arise with respect to institutional systems of norms not backed by the threat of a coercive sanction. What makes state regulation presumptively problematic from the standpoint of political morality is that its directives are coercively enforced by a threat of punishment. 15 Thomas Hobbes, Leviathan (first published 1651, Classic Books 2017) 84, 85; available at: . 16 These socially disruptive conflicts are not limited to disagreements about the material goods needed to feed, water, and shelter persons in a fashion they find acceptable. Internecine conflicts can arise over sexual companionship, as people frequently find themselves competing for the romantic affection of the same person. Violence is an all-​too-​common response to a person’s not getting or being able to retain the companionship of someone she finds sexually or otherwise romantically desirable; such violence might be directed at the person whose companionship is desired or at a competitor who

The Conceptual Function of a Legal System  85 self-​interested subjects like us are likely to resort to violence in the state of nature because we should do so; since moral requirements do not apply in the state of nature, on Hobbes’s view, the only objectively relevant standards are prudential norms dictating that beings like us do whatever is necessary to ensure we have what we need to survive.17 But, putting Hobbes aside, even if moral requirements did apply, rationally competent self-​interested subjects like us would be highly likely to ignore them because we are not characteristically willing in extreme conditions of material scarcity to sacrifice our own basic needs to ensure that those of other persons are satisfied. In the state of nature, something like a war of all against all seems all but inevitable. Hobbes utilizes the state-​of-​nature device to show that rationally competent self-​interested subjects would voluntarily contract in such conditions to submit to an absolute sovereign as the only way to end this war, but the argument is problematic for two reasons. The first is that it is not clear how to reconcile the claim that a contract made in the state of nature is morally binding with the claim that the requirements of morality do not apply in the state of nature; it is difficult to see why a promise that is not morally binding because made in the state of nature can morally bind once an absolute sovereign is established. The second, and the more important one for our purposes, is that there is no reason to think we would contract in the state of nature for an absolute sovereign, when all we need is some efficacious system of coercive regulation; we can easily conceive of other forms of governance

succeeded in retaining the desired person’s companionship. Competition for romantic resources, so to speak, can be as destructive as competition for any other materially scarce resource. 17 As Hobbes puts the point, “To this warre of every man against every man, this also is consequent; that nothing can be Unjust. The notions of Right and Wrong, Justice and Injustice have there no place.” Hobbes, Leviathan (n. 15) 85. The idea that objective moral norms do not apply in the state of nature as Hobbes describes it is implausible—​assuming that we are rationally competent in a state of nature and hence capable of both discerning the requirements of morality and freely satisfying them; however, mandatory moral norms do not apply to acts in the state of nature in the same way that they apply under circumstances in which the peace is kept. Consider what typically occurs on the front line of a hot war between country C and country D. C-​soldiers and D-​soldiers are shooting at each other and would be punished severely, usually by death, by their respective countries for desertion. None of the soldiers on the front lines in those circumstances are doing anything morally wrong by shooting at each other—​regardless of which country, if either, is morally justified in waging the war. All of the soldiers seem to be morally justified in shooting as a means of defending their own lives—​given that the option to run is excluded by laws punishing desertion with death or incarceration. Life in the state of nature, on a Hobbesian view, is even more chaotic because there is no possibility of the kind of cooperation needed to create an army. This is not to suggest that it is impossible to do moral wrong in the state of nature, as Hobbes seems to have believed. It is intuitively plausible to think that torture and rape are wrong, as an objective matter of morality, even in a state of nature insofar as they serve no legitimizing defensive or survival purposes. Assuming that a state of nature is less violent than what Hobbes envisions, it should be clear that there will be quite a bit more violence permitted in the state of nature by the norms of an objective morality than is otherwise permitted. Life in the state of nature, under any conception of it, is morally complicated—​even if people in the state of nature do not have the luxury of working out more difficult issues of morality.

86  What Law Must Be Able to Do that can end the war of all against all because we figured out early on in our history that democratic systems can do so.18 Though his argument for the legitimacy of an absolute sovereign fails, it gestures in the direction of the foundational problem a legal system is needed, characteristically used, and supposed to solve, as a functionally normative matter: in an anarchic state, rationally competent self-​interested subjects like us cannot live even a minimally satisfying life because we have to expend too much of our time and energy defending ourselves and our belongings against the violent infringements of others. The foundational problem that a legal system is needed to solve, as Hobbes’s analysis suggests, is to keep the peace among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours because, as a practical matter, the peace cannot be kept in our world without a legal system. Contemporary versions of social contract theories of legitimacy are less pessimistic about life in the state of nature but identify the same foundational problem that systems of law are needed, characteristically used, and supposed to solve, as a functionally normative matter. Robert Nozick rejects the assumption that social cooperation is impossible in the state of nature but grounds the legitimacy of the state in its capacity to keep the peace among rationally competent subjects like us in worlds of acute material scarcity like ours; the minimal state, on his view, arises out of a series of morally permissible “protective associations” voluntarily contrived and used to protect members from wrongful assaults on their natural rights to life, liberty, and property.19 Like Hobbes’s theory, then, Nozick’s presupposes that the basic problem the state is needed, contrived, characteristically used, and supposed to solve, as a functionally normative matter, is that of keeping the peace among rationally competent self-​interested subjects like us who are sometimes disposed to commit assaults against persons and property because material resources are scarce; both theories justify state regulation of behavior on the ground that rationally competent self-​ interested subjects would consent to legal regulation as a means to the ultimate end of minimizing socially disruptive conflicts that potentially threaten everyone’s well-​being. John Rawls has curiously little to say about the coercive mechanisms of law despite their obvious moral salience because his view seems less concerned with identifying the basic problem the state is needed to solve than with identifying the principles 18 We know of legal systems employing democratic mechanisms that go as far back as ancient Greece. Cleisthenes is sometimes thought to have laid the practical groundwork for democracy in 508 bc with reforms that diminished the political power of the aristocracy and increased the influence of ordinary citizens in political decision-​making. Mark Cartwright, “Cleisthenes” Ancient History Encyclopedia; available at: https://​www.ancient.eu/​Cleisthenes/​. 19 Nozick believes that a state of nature is possible in which “people generally satisfy moral constraints and generally act as they should.” Robert Nozick, Anarchy, State, and Utopia (Reprint edn, Basic Books 2013) 4. It might be true that it is possible in some sense, but the relevant issue for our purposes has to do, as discussed below, with whether we believe it is likely and how likely we believe it is.

The Conceptual Function of a Legal System  87 defining the fair terms of social cooperation that the coercive machinery of the law makes possible in the first place in virtue of keeping the peace. His stated concern is to identify the terms that govern “society” conceived as “a fair system of cooperation”—​ and not to justify the coercive practices that make living together in a society possible to begin with.20 Although he is often taken to be presenting a theory that rivals those of Hobbes, Nozick, and other legitimacy theorists, Rawls seems to be presupposing that coercive restriction of behavior can be justified in virtue of its being necessary to make society possible to begin with and hence to be more concerned with how to assure that it is “well-​ordered” in the sense that it creates the conditions for rationally competent self-​interested subjects to cooperate on fair terms.21 That said, the heuristic device of the original position can easily be deployed to identify the basic problem that law is needed to solve. Insofar as persons in the original position lack any information about the social circumstances that distinguish one person’s circumstances from those of others and must decide which principles of cooperation should govern a well-​ordered society wholly on the basis of what they believe conduces to their self-​interest, they would surely converge in insisting that the state protect life, liberty, and property with mandatory norms backed by the threat of a coercive sanction sufficiently severe to deter enough such assaults to enable them to live together in a state of comparative peace; deciding which principles of cooperation should govern a well-​ordered society is a problem that can be solved only once the basic problem of making societal life possible to begin with is solved. Rawls might be concerned with a problem that is either broader or different from the basic problem of keeping the peace (or “problem of order,” as he puts it), but his view clearly presupposes that the foundational legitimizing purpose of the state is to keep the peace. Consequentialist theories agree that the legitimizing purpose of the state is to keep the peace. As John Stuart Mill puts the point: [T]‌he sole end for which mankind are warranted, individually or collectively, in interfering with the liberty of action of any of their number, is self-​protection.

20 John Rawls, Justice as Fairness: A Restatement (Harvard University Press 2001) 5. 21 Rawls seems, oddly, to regard the preoccupation of political philosophy with preventing violence as quaintly anachronistic, as though the basic problem with which other political philosophers have been concerned is easily solved and the real problem is to find a device that resolves political disagreements: “There are long periods in the history of any society during which certain basic questions lead to deep and sharp conflict and it seems difficult if not impossible to find any reasoned common ground for political agreement. To illustrate, one historical origin of liberalism is the Wars of Religion in the sixteenth and seventeenth centuries following the Reformation: these divisions opened a long controversy about the right of resistance and liberty of conscience . . . . Hobbes’s Leviathan . . . is concerned with the problem of order during the turmoil of the English civil war; and so also is Locke’s Second Treatise.” Rawls, Justice as Fairness (n. 20) 1. The problem of order, however, is not specific to any particular moment in time; that is why state-​of-​nature devices are timeless; indeed, Rawls’s own original-​position heuristic is timeless in the same sense since its point is to enable us to identify, regardless of historical context, those principles that distinguish a society that is well-​ordered from one that is not.

88  What Law Must Be Able to Do That the only purpose for which power can be rightfully exercised over any member of a civilised community, against his will, is to prevent harm to others. His own good, either physical or moral, is not a sufficient warrant. He cannot rightfully be compelled to do or forbear because it will be better for him to do so, because it will make him happier, because, in the opinions of others, to do so would be wise, or even right. These are good reasons for remonstrating with him, or reasoning with him, or persuading him, or entreating him, but not for compelling him, or visiting him with any evil in case he do otherwise. To justify that, the conduct from which it is desired to deter him must be calculated to produce evil to someone else. The only part of the conduct of any one, for which he is amenable to society, is that which concerns others. In the part which merely concerns himself, his independence is, of right, absolute.22

Mill’s consequentialist theory justifies legal constraints on individual autonomy wholly on the strength of their capacity to enable a legal system to perform its basic function, as I have described it. Insofar as a legal system can perform no function more basic than keeping the peace by prohibiting acts that harm others, Mill’s view that legal regulation is justified only insofar as it prevents harm to others implies that the foundational problem that law is needed to solve—​and hence its basic legitimizing purpose—​is to keep the peace among rationally competent self-​ interested subjects like us in worlds of acute material scarcity like ours. Law can do none of what it can otherwise be used to do unless it minimally succeeds in deterring rationally competent subjects from harming others and thereby keeps the peace enough to enable us to live together in something plausibly characterized as a community so that we can collectively reap the benefits of social cooperation. It is not happenstance that every moral theory of legitimacy identifies keeping the peace as the basic function of a legal system. The idea that legal systems are needed and can be morally justified in worlds like ours to keep the peace accords with our own common-​sense fears about the prospect of life without law. Few of us would be tempted to do away with the coercive protections of a legal system without substituting other coercive mechanisms able to do the same work with fewer intrusions on natural moral rights. One might reasonably believe that there is much in existing legal practices that should be dramatically reformed; however, it is naïve to think that a legal system could be fruitfully dismantled in any society as complex as those with paradigmatic systems of law in our world of acute material scarcity. Given what we think we know about ourselves, such as we are at this 22 John Stuart Mill, On Liberty (Ticknor and Fields 1863) 23. Lord Patrick Devlin argues that the state may also legitimately enact laws to protect society’s shared morality as a means of keeping the peace. Devlin believes that such enactments are justified to ensure the existence of society because what is essential to the existence of any society is a shared morality. Protecting the shared morality is simply another mechanism for protecting against conflicts that breach the peace and thereby threaten the existence of society. Lord Patrick Devlin, The Enforcement of Morals (Oxford University Press 1970).

The Conceptual Function of a Legal System  89 point in time, a total suspension of law is likely to result in a coast-​to-​coast never-​ ending free-​for-​all that is not in anyone’s self-​interest.23 There is little reason to doubt that the idea that the coercive mechanisms of law are needed to keep the peace is widely shared among the “ordinary” folk whose views condition our social practices, conceptual and otherwise, with respect to law. Popular depictions of life in the state of nature are as darkly pessimistic as Hobbes was with respect to how we would behave under such conditions—​and sometimes even more so. Consider the wildly successful film series entitled The Purge, which depicts a nation that allows citizens to do whatever they want with legal impunity for twelve hours on one day every year; not surprisingly, the nation descends into a terrifying state of violent chaos for that twelve hours.24 But what is particularly disturbing about the story line is that the rationale for this practice is to allow citizens to “blow off steam” in the hope that they will be less likely to commit violent assaults during the year; crime, unemployment, and social unrest have nearly disappeared since the yearly purge program was implemented by the government. In essence, the law allows citizens to slaughter one another for twelve hours once every year to eliminate the incidence of violent assaults during the rest of the year. The idea that rationally competent self-​interest subjects like us need to let off some steam with a once-​a-​year holiday from law is surely overstating things a bit; however, there is clearly some truth in the grim underlying assumptions. What explains why these shows are so successful in drawing viewers is precisely that they are grounded in assumptions about us that appeal to our deepest fears. While they might be properly characterized as fantasy, it is not because they strike us as being utterly implausible; these fantasies can elicit enough interest from us to buy the privilege of consuming them only insofar as we can identify with the characters—​and we can identify with the characters in state-​of-​nature depictions only insofar as they largely conform to our beliefs about what would happen among beings with the psychological characteristics we have if law was not backed with the threat of force in worlds of acute material scarcity like ours. We watch these depictions with lusty fascination precisely because we take ourselves to be watching ourselves.

23 The idea that law’s basic function is to keep the peace is not only true of systems of municipal law; it is also true of systems of international law. International law differs, as a conceptual matter, from municipal law in that its subjects are nation-​states and not natural persons. But the idea that the point of international law is to keep the peace among nations is equally plausible; after all, violent crime is not nearly as dangerous to human well-​being as wars between nations. The foundational problem of international law, as our practices define it, is to minimize the probability of deadly wars among nations. 24 The first installment of The Purge reportedly cost $3 million to make and earned $89 million worldwide. Alex Abad-​Santos, “The purge movie franchise, explained” Vox (July 1, 2016); available at: https://​ www.vox.com/​2016/​7/​1/​12054200/​purge-​movie-​explained.

90  What Law Must Be Able to Do The point here is neither that these widely shared views are true nor that we are epistemically justified in believing they are true; the point is that these are the shared views that ground our social practices, conceptual and otherwise, with respect to law. The claims that matter with respect to explicating our conceptual practices, as discussed in Chapter 2, are views that we share, as a descriptive matter of contingent fact, and not views that are either epistemically justified or objectively true—​although our shared views might, for all we know, be both epistemically justified and objectively true. Insofar as we are trying to understand something about ourselves and why we do what we do with law, we are concerned with identifying and explicating the shared views that lead us to adopt the practices constituting something as a legal system to regulate behavior. But, at the end of the day, it is not what we characteristically believe, as a descriptive matter of contingent fact, that explains why we do what we do with these constitutive practices; what does the work is that we characteristically fear, as a descriptive matter of contingent fact, that a state of nature in our world would be as chaotic as it is depicted in these novels, films, and television shows because we should, as an objective matter of normative practical rationality, fear that it would be that bad. It might or might not be true that rationally competent self-​interested subjects like us could not live together in worlds of acute material scarcity like ours without the coercive strictures of law; but what really matters here is that people characteristically fear, as a descriptive matter of contingent fact, that life without law would be catastrophically dangerous because it is our fears that motivate us to act. We characteristically adopt the practices constituting a legal system to keep the peace not necessarily because we so confidently believe that life without law would be so horrifyingly awful but because we fear it would be. The legal restrictions we want adopted as much reflect our common fears as they do our common beliefs: we want protections against murder because we fear we will be murdered; we want protections against theft because we fear our belongings will be stolen; we want protections against being injured by negligent behavior because we fear we will be injured by negligent behavior; and so on. While those fears are grounded in predictions about what will happen without law, the content of those predictions is what makes the fears grounded in them so persuasive. We adopt the practices constituting something as a legal system because we fear that life without law will approximate something like a Hobbesian state of nature and hope that these constitutive practices will succeed in keeping the peace among us enough to enable us to live together in a community. And, as an objective matter of normative practical rationality, we should fear that we cannot live together in a community without the coercive strictures of law and, equally important here, should adopt the practices constituting something as a legal system to keep the peace. It is worth noting here that contractarian theories of moral legitimacy—​like Hobbes, Locke, Nozick, and Rawls—​are all grounded in

Keeping the Peace and Authorized Coercion  91 the idea that we should, as an objective matter of normative prudential rationality, consent to coercive regulation of our behavior. It is striking that what is utterly uncontentious among normative political philosophers who take themselves to be explicating something implicit in our shared views about legitimacy is so contentious among legal philosophers who take themselves to be explicating our conceptual practices: there is simply no question that our fears about the consequences of an anarchic state of affairs, more than anything else, explain why we adopt the practices constituting something as a system of law. No matter how we come at the problem of identifying its principal function, it is clear that the basic, characteristic, and functionally normative use of a legal system is to keep the peace among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours by regulating behavior through the governance of norms metaphysically capable of guiding behavior. While a system of law can be non-​idiosyncratically used to accomplish many desirable and undesirable purposes, law cannot do any of those things unless it is able to regulate behavior in a manner that minimizes socially disruptive acts to a point where rationally competent subjects can reap the benefits of living together in a community: law cannot mediate disputes, make moral requirements more determinate, enable us to better conform to the dictates of right reason, etc., unless it succeeds in keeping the peace enough for us to live together in a community. Keeping the peace by regulating behavior through the governance of norms metaphysically capable of guiding it is therefore the one thing that law must be able to do in order to do everything else that law can be non-​idiosyncratically used to do. If what an artifact is needed, characteristically used, and supposed to do, as a functionally normative matter, defines its conceptual function, then the conceptual function of a legal system is to keep the peace among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours by regulating behavior through the governance of artifactual norms metaphysically capable of guiding behavior.

6.  Keeping the Peace and Authorized Coercion It is nomologically possible for us to live in peace without the coercive mechanisms of law—​and obviously so. There is nothing in our world that makes it causally impossible for rationally competent self-​interested subjects like us to live together in peace in worlds of even greater material scarcity than we experience in our world without backing some artifactual norms with the threat of a coercive sanction. This is true regardless of whether or not our behavior is causally determined. If, on the one hand, our acts are uncaused by external factors and freely chosen, we are clearly able, as a causal matter, to live in peace without a system of institutional norms to regulate our behavior; all we have to do is simply to choose not to

92  What Law Must Be Able to Do commit violent assaults on persons and property—​which is what we all expect of one another. If, on the other hand, our acts are caused by external factors, the fact that many of us abstain from such acts entails that it is nomologically possible for all of us to do so. There is nothing special about the causal properties of those of us who abstain from violence that distinguishes us from other people in this regard; while it is true under the supposition that our behavior is causally determined that there are causal factors that would explain the bad behavior of those who do violent things, those causal factors could, as a nomological matter, have been otherwise. But the claim that it is nomologically possible for rationally competent self-​ interested subjects like us in worlds of acute material scarcity like ours to live in peace without the coercive mechanisms of law tells us nothing that would help to explain why we use legal systems to perform this foundational task. Regardless of whether we are concerned with logical, conceptual, or nomological possibility, the notion of possibility does not admit of degrees: something is either possible or it is not possible; everything that is possible is as possible as everything else that is possible; and nothing that is impossible is any less or more impossible than anything else that is impossible. Insofar as the claim that it is nomologically possible for a sanctionless system of artifactual norms to keep the peace among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours says nothing about how likely it is to succeed in doing so, that claim lacks the resources to help us understand why we contrive, use, and expect legal systems to keep the peace. What matters with respect to the Coercion Thesis has to do with our fears about the likelihood that the peace would be kept without the coercive mechanisms of law among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours. If an institutional normative system lacking coercive sanctions is not reasonably likely to succeed in minimally keeping the peace among rationally competent subjects like us in worlds like ours, then it is not properly characterized as a system of law because it is not reasonably contrived to succeed in performing law’s conceptual function of keeping the peace among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours; it is, as discussed above, a conceptually necessary condition for something to count as an artifact A that it is reasonably contrived, when used the way As are characteristically used to succeed in performing the conceptual function of A. This much is utterly uncontentious: coercive sanctions deter some assaults in our world on persons and property likely to lead to ever-​escalating cycles of violence that threaten the collective peace. Deterrence-​focused debates about punishment are not concerned with whether the practices associated with norms authorizing the judicial imposition of non-​trivial coercive sanctions for assaults on persons or property deter some of those acts; we all believe that they do. They are instead concerned with the issues of whether increasing the severity of punishment would succeed, as a descriptive matter of contingent fact, in deterring more

Keeping the Peace and Authorized Coercion  93 of those assaults and, if so, whether we should do so, as an objective matter of normative practical rationality. Deterrence might not be the only reason we punish violent crime, but it is one obvious reason we do. Despite the fact that every nation devotes significant resources—​and at great expense25—​to detecting, prosecuting, and punishing violence with a credible threat of life-​changing detriment, none has eliminated all such assaults on persons or property. The incidence of violent assaults remains tragically high in every nation. In fully developed nations, reported annual rates of intentional homicide vary from a low of under 1 per 100,000 persons to significantly higher than the 5.5 per 100,000 persons in the United States.26 In less developed nations, violent crime rates are predictably higher: reported annual rates of intentional homicide in African nations are as high as 41 per 100,000 persons. Worldwide, the intentional homicide rate was 6.2 per 100,000 persons in 2012. Property-​crime rates are higher than intentional homicide rates. There were 7,764,086 property crimes reported in the U.S.  during the year 2017, which amounts to 2,362.2 reported property crimes per 100,000 persons27—​about one every four seconds. While reliable statistics on property crime rates in other fully developed and less developed nations are harder to come by, it is plausible to hypothesize that they are significantly higher than intentional homicide rates in every nation; it would be surprising if there has ever been a nation in our world with an intentional homicide rate higher than its property-​crime rate. Although no society has succeeded in preventing assaults on persons and property, each has reduced their frequency enough to allow people to live together in comparative peace. Without the coercive regulation of a legal system, there would be so many breaches of the peace—​or so we fear—​that we could not live together in any existing society sufficiently complex to have something that counts as a legal system. While rationally competent self-​interested subjects like us might, under the most optimal circumstances, be able to live together in smaller, more intimate, agrarian communities, we would not be able to achieve all the benefits of social cooperation that can be achieved in more complex societies with coercive legal systems. Life in these simple communities might not be as bad as Hobbes would predict, but it would not be optimally meaningful.

25 The cost of incarcerating someone in a federal prison for one year was $36,299.25 ($99.45 per day) in fiscal year 2017. Figures available at: https://​www.federalregister.gov/​documents/​2018/​04/​30/​2018-​ 09062/​annual-​determination-​of-​average-​cost-​of-​incarceration. This figure does not include the costs of investigating or prosecuting defendants. 26 Available at:  https://​en.wikipedia.org/​wiki/​List_​of_​countries_​by_​intentional_​homicide_​ rate#By_​country. The rest of the statistics offered in this paragraph also derive from this source. The statistics for the various countries were taken between the years 2006 and 2015. It bears noting that these statistics do not include unsuccessful homicide attempts, non-​homicidal violent assaults, rapes, or other violent assaults on persons. 27 Available at: https://​www.statista.com/​topics/​1751/​property-​crime-​in-​the-​us/​.

94  What Law Must Be Able to Do This explains why every existing normative system plausibly characterized as one of law has authorized the judicial imposition of coercive sanctions for non-​ compliance with some mandatory legal norms prohibiting assaults on persons and property. Every such system includes norms authorizing the judicial imposition of coercive sanctions intended to deter such crimes because there are no viable alternatives, as a practical matter, that are even remotely likely to keep the peace enough for us to realize the material benefits made possible by social cooperation. While it might be true that we overvalue these material benefits relative to the environmental and psychological costs of producing them, the fact remains that we are very stubbornly attached to them; and that, of course, is why we need law. It should be clear (i) why coercive sanctions are needed to keep the peace among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours; and (ii) why they are reasonably contrived to do so. As to (i), sanctions are needed to keep the peace because the combination of our having psychologically compelling desires for material and non-​material goods and the relative scarcity of such goods to satisfy these desires is highly combustible: the probability that rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours commit assaults on persons and property approximates 1. Rationally competent self-​interested subjects like us who value our own welfare need some sort of effective institutional mechanism to deter these assaults because the costs to our own welfare of being solely responsible for our self-​ protection are prohibitive insofar as they preclude our doing anything else with our lives. As to (ii), backing mandatory legal norms prohibiting violence with periods of incarceration is reasonably contrived to deter assaults on persons and property because the prudential incentives that can lead rationally competent self-​interested subjects like us to commit such acts can be effectively countered by non-​trivial prudential disincentives: we do these things because we expect them to produce something we regard as a prudential benefit, and we can be rationally induced not to do these things by showing that doing them will produce something we regard as a prudential cost that exceeds the prudential benefit of doing them. Backing the relevant legal norms with the threat of a coercive sanction is reasonably contrived to deter non-​compliance insofar as it raises the expected prudential costs of non-​compliance. It is important to understand why it is essential to law’s function of keeping the peace that it prohibits assaults on persons and property—​especially those that are nearly universally thought to be morally wrongful by subjects of the system.28 Rationally competent subjects like us are most likely to respond in socially 28 Insofar as it is conceptually possible to have a legal system in a world of nihilists, it is not a conceptually necessary condition for the existence of a legal system that the assaults it prohibits are generally thought morally wrongful. See Chapter 1, Section 3.

Keeping the Peace and Authorized Coercion  95 undesirable ways to assaults or insults that we believe are morally wrongful and thereby legitimize a retaliatory response. While it is true that people never react favorably to acts that impinge on their perceived vital interests, a violent response that threatens the peace is most likely when people believe they have been morally aggrieved; the sort of anger that fuels retaliatory violence is far more likely to be caused in morally decent people by an attack they believe is morally wrongful.29 In these cases, moral indignation combines with prudential interest to produce a response that is likely to lead to a cycle of escalating retaliatory acts culminating in a larger breach of the peace that adversely affects the psychological well-​being of others in the community, who may themselves respond in socially undesirable ways to this heightened sense of material vulnerability. Coercive sanctions are thus both needed to keep us out of the prudentially catastrophic circumstances that would occur under the most optimistic conditions we are likely to experience in the state of nature and reasonably contrived to do so. There are no other available options likely to do the same work: there is, after all, so much that the most effective socializing processes can do to tame our antisocial tendencies under the best of circumstances; and few of us are raised by parents with the time, money, knowledge, and patience to do everything needed to optimally minimize these tendencies. With sanctions, we can live together under less than optimal circumstances; without them, we cannot live together at all—​or so we sensibly fear. This is not to suggest that rationally competent subjects like us will necessarily commit acts that breach the peace. Most of us rarely experience the sort of intensely painful adversity that might tempt us to do these things; the vast majority of us manage to get through our lives without causing much harm to other people because we are sufficiently decent that we recoil at the prospect of seriously injuring other people, regardless of whether the law punishes such acts with sanctions. But we converge, as a descriptive matter of contingent fact, in believing that every person is psychologically capable under sufficiently adverse circumstances of committing the morally wrongful acts that cause the most serious harms to other people. This, I take it, is ultimately what is meant by religious doctrines asserting we are inherently sinful: we are all capable of doing awful things under the right circumstances; even if we are fortunate enough never to find ourselves in those triggering circumstances, we need spiritual moral redemption because it is luck, and not personal virtue, that explains why we have not done the worst things we are all capable of doing. It is so much a part of our shared conceptions of our nature that we are all

29 We also have a general tendency to think that any negative reaction to us or behavior expressing that reaction is morally undeserved; people almost universally overestimate the extent to which they are morally justified in doing something that has the effect of causing conflict.

96  What Law Must Be Able to Do capable of violent wickedness that this unhappy view is incorporated into the core doctrines of every major religious tradition of which we are aware. The reason we want a system of law with coercive mechanisms is not, of course, that we want to protect other people from whatever violence we are capable of committing; it is rather that we want to be protected from whatever violence other people are capable of committing against us because we fear the suffering those assaults can cause. Even if we trust our own ability to control the anger and frustration we all frequently experience in a world in which we have to compete for nearly everything we need and want, we do not, and should not, trust the ability of other people to control their anger and frustration. Our capacity for socially destructive behavior should not be understated. Moral decency is a luxury that can be purchased in our world only by legal systems with sanctions severe enough that they deter enough violence and theft to allow us to direct our psychic and physical resources to more productive pursuits than defending ourselves against assaults on persons and property. Without the credible threat posed by those sanctions, there would be so much violence that even the meekest among us is likely to aggress against people and property. The coercive strictures of law are what separates our best from our worst selves—​and we are every bit as much our worst selves as we are our best selves. This point is eloquently made by a character in the docudrama Chernobyl explaining how the conditions of war brought out horrifying elements in his character that he never imagined might be part of his personality. Explaining how he felt after the first time he killed a person during the Soviet invasion of Afghanistan, Bacho states: My first time . . . we were moving though a house[,]‌a man was suddenly there, and I shot him in the stomach. . . . I was so scared, I didn’t pull the trigger again for the rest of the day. I thought, well[,] that’s it, Bacho. You put a bullet in someone. You’re not you anymore. You’ll never be you again. But then you wake up the next morning, and . . . you’re still you. And you realize—​that was you all along. You just didn’t know.30

What distinguishes Bacho from those of us harboring silly conceits about our moral purity is simply a matter of opportunity luck: we just have not found ourselves in a situation where we had to do something violent that we thought ourselves constitutionally unable to do. It is the practical ability of law to deter violence that allows us the luxury of such self-​serving fantasies about ourselves. Law, as we conceive and practice it, is contrived to keep the peace among rationally competent self-​interested subjects like us who live as we do in prudentially 30 Craig Mazen, Chernobyl, Episode 4 (2019); available at: https://​johnaugust.com/​wp-​content/​uploads/​2019/​06/​Chernobyl_​Episode-​4The-​Happiness-​Of-​All-​Mankind.pdf,  at  28.

Keeping the Peace and Authorized Coercion  97 challenging circumstances of acute material scarcity, but this cannot be done, as a practical matter, without backing mandatory norms generally prohibiting assaults on persons and property with the threat of a coercive sanction. Insofar as (1) the conceptual function of a legal system is to keep the peace and (2) an institutional normative system lacking sanctions to enforce mandatory norms prohibiting assaults on persons and property is not reasonably contrived to keep the peace, it follows that it is a conceptually necessary feature of a legal system that it authorizes the imposition of coercive sanctions to enforce mandatory legal norms incorporating such content. The Coercion Thesis offers the only plausible explanation for how law is reasonably contrived to do what law is contrived, characteristically used, needed, and supposed to do, as a functionally normative matter, in worlds of acute material scarcity like ours and hence why we do what we characteristically do with law. If we do not have an institutional normative system that succeeds in keeping the peace enough for us to live together, then we cannot do anything we want to do. Our ability to do anything worth doing together in this world depends on the guns of law keeping us off our neighbors’ throats and property.

5

Three Conceptual Problems of Legal Normativity: The Logical Space of Reasons One of the conceptual features of law requiring philosophical explication is its presumed normativity: the practices constituting something as a system of law are presumed normative in the sense that they give rise to something capable of persuading rationally competent subjects to change their minds about doing something prohibited by mandatory legal norms governing non-​official behavior. Systems of law can do this only insofar as the practices constituting something as a system of law give rise to something rationally competent subjects characteristically regard, as a descriptive matter of contingent fact, as normatively relevant in deciding whether to comply because they should, as an objective matter of normative practical rationality, regard it as normatively relevant.1 It makes no sense to adopt the practices constituting something as a system of law to regulate behavior—​regardless of what the ultimate end is for doing so—​unless we presume they can make a difference with respect to whether rationally competent subjects do what law requires. Adopting a legal system would be prudentially unjustified, if law is not conceptually normative, because it is not in anyone’s self-​interest to have her freedom restricted by norms that are not reasonably contrived to make a difference with respect to how other people behave towards her. Adopting a legal system would likewise be morally unjustified, if law is not conceptually normative, because something that is supposed, as a functionally normative matter, to regulate behavior by restricting freedom but lacks the resources to change the way rationally competent subjects behave towards one another is not reasonably contrived to do anything of moral value. To vindicate the rationality of adopting systems of law to regulate behavior, then, it must be shown that the practices constituting something as a system of law 1 It is important to note that the problem of vindicating law’s presumed conceptual normativity arises only in connection with mandatory legal norms governing non-​official behavior. Taking the internal point of view towards the rule of recognition, which governs official behavior, entails regarding it as normatively relevant with respect to deciding whether or not to do what it requires. Given that it is clearly possible for someone to be justified, as an objective matter of normative practical rationality, in taking the internal point of view towards a rule of recognition, there is no mystery as to how officials could be warranted by objective norms of practical rationality in regarding the rule of recognition as normatively relevant. Coercion and the Nature of Law. Kenneth Einar Himma, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198854937.001.0001

100  Three Conceptual Problems of Legal Normativity are reasonably contrived to give rise to something rationally competent subjects characteristically regard, as a descriptive matter of contingent fact, as reasons to comply because they should, as an objective matter of normative practical rationality, regard it as such. Any conceptual theory of law lacking the resources to vindicate law’s presumed conceptual normativity fails, for that reason, as a theory of the metaphysical nature of law—​either because it is incomplete and needs supplementation or because it is false. This part of the book argues that law’s conceptual normativity can be explained and vindicated only with the help of the Coercion Thesis. In the first two sections of this chapter, I explicate the concept of normativity and distinguish several classes of reasons that might be thought to figure into the problems associated with explicating law’s conceptual normativity. In the third section, I  identify the class of reasons that the practices constituting something as a system of law must be presumed equipped to provide. In the final section, I distinguish three conceptual problems of legal normativity that must be solved to vindicate the rationality of adopting legal systems to regulate behavior. The following three chapters are concerned with solving these conceptual problems.

1.  The Concept of Normativity The concept of normativity, as our practices define the corresponding concept-​ term, has to do with the capacity of something to give rise to reasons.2 There are two basic types of normativity: an institution or authority is epistemically normative insofar as it gives rise to considerations rationally competent subjects characteristically regard, as a descriptive matter of contingent fact, as reasons to believe things because rationally competent subjects should, as an objective matter of normative epistemic rationality, regard them as reasons to believe things; an institution or authority is practically normative insofar as it gives rise to considerations rationally competent subjects characteristically regard, as a descriptive matter of contingent fact, as reasons to do things because rationally competent subjects should, as an objective matter of normative practical rationality, regard them as reasons to do things. The next two sections are concerned with explicating and identifying the relevant class of reasons.

2 The notion of normativity is in need of some conceptual clarification; dictionary reports of the meaning of normative are circular and largely uninformative for this reason. Oxford Online Dictionary defines the term normative as “[e]‌stablishing, relating to, or deriving from a standard or norm, especially of behaviour”; available at:  https://​ www.lexico.com/​ en/​ definition/​ normative. Underlined emphasis added.

The Logical Space of Reasons  101

2.  The Logical Space of Reasons The problems associated with law’s normativity cannot be adequately addressed without understanding various distinctions defining different classes of reasons. There are four distinctions needing clarification:  (1) the distinction between reasons to believe (epistemic reasons) and reasons for action (practical reasons); (2) the distinction between reasons that support something but are defeated by other reasons (defeasible reasons) and reasons that support something but are not or cannot be defeated by other reasons (conclusive reasons); (3) the distinction between considerations that provide an incentive for performing an act (motivating reasons) and considerations that justify performing an act (justifying reasons); and (4) the distinction between considerations that ought, as an objective matter of normative epistemic or practical rationality, to be regarded by rationally competent subjects as reasons (objective reasons) and considerations that rationally competent subjects regard, as a descriptive matter of contingent fact, as reasons (subjective reasons). Addressing the various problems of legal normativity requires identifying the character of the reasons that law as such is equipped to provide as it relates to each of these distinctions.

2.1  Epistemic and Practical Reasons Reasons are considerations that favor or oppose something. At the most general level, there are two species of reason having to do with the kind of thing they favor or oppose: reasons for belief and reasons for action. An epistemic reason is a consideration that favors believing some proposition p:  something is an epistemic reason to believe p if and only if it provides some evidence that p is true; that is, r counts as an epistemic reason to believe p if and only if someone who believes p is more likely to be correct in her belief that p if, other things being equal, she believes r than if she does not believe r.3 A proposition expressing the fossil record is an epistemic reason favoring belief that the theory of evolution is true. In contrast, a practical reason is a consideration that favors performing some act a; r counts as a practical reason to do a if and only if r favors doing a. A proposition expressing that I will be punished with incarceration if I rob a bank because doing so is illegal expresses a practical reason for me not to rob a bank.

3 The reason r does not make it more likely that p is true; p is either true or it is not. The truth-​ conducive effect of r with respect to p is as follows: someone who believes that r and p are true is more likely to be correct with respect to her belief that p is true in virtue of believing that r is true than she would be, other things being equal, if she did not believe that r is true. The relevant relationship here has to do with how well supported the belief is by other propositions that are regarded as evidence for the truth of the belief at issue and is hence probabilistic in character.

102  Three Conceptual Problems of Legal Normativity It is important to note that the claim that some piece of propositional content r is a reason for P to do or believe something does not imply that the explanation for why P does or believes that something, as a descriptive matter of contingent fact, is that r has persuaded P to do or believe that something. I believe that the illegality of killing innocent persons is a practical reason not to kill innocent persons yet that belief has nothing to do with why I have never killed an innocent person; I believe that the absence of a plausible account of how John Connally was struck by the “magic” bullet that also struck John F. Kennedy is a reason to think Lee Harvey Oswald was not the lone shooter but nonetheless believe that the available evidence best supports the theory that he was the lone shooter. The class of epistemic reasons and the class of practical reasons are mutually exclusive. Practical reasons to do something might be grounded in, or otherwise appeal to, epistemic reasons to believe something; one cannot have a reason to do something unless one has a reason to believe that it should be done. But epistemic reasons are not practical reasons; an epistemic reason concerning a desire I have to eat a sandwich might give me a reason to believe that I want to eat a sandwich, but it is the proposition expressing the desire to eat the sandwich, and not the proposition expressing the belief that I have that desire, that constitutes the reason to eat the sandwich.4 Believing something is a purely doxastic state that does not usually require moving the body around, while doing something usually involves moving the body around. The class of epistemic reasons and the class of practical reasons are also jointly exhaustive of the class of things counting as reasons in the relevant sense. While we sometimes use the term reason to refer to causal explanations for events, as when we ask for the “reason” the automobile does not start, this usage is not relevant here. This usage of reason is just shorthand for causal explanation and has nothing to do with normativity of any kind. Since there are thus only two classes of reasons and they are mutually exclusive and jointly exhaustive, the class of epistemic reasons and the class of practical reasons partition the logical space of reasons.

2.2  Defeasible and Conclusive Reasons A second distinction is between defeasible and conclusive reasons. A  consideration that favors something is a defeasible reason for that something if and only if it can be defeated by reasons that oppose that something. An epistemic reason r to believe p is defeasible if and only if a rationally competent subject’s belief that r is true makes it more likely, other things being equal, that she is correct in believing 4 Otherwise put, it is the proposition expressed by “I have a desire to eat a sandwich,” and not the proposition expressed by “I believe I have a desire to eat a sandwich,” that constitutes the practical reason to eat a sandwich.

The Logical Space of Reasons  103 p than she would be without the belief that r is true but does not warrant belief because it can be defeated by other epistemic reasons. The proposition that P’s fingerprints were on the gun that killed Q is a defeasible epistemic reason to believe that P killed Q insofar as it can be defeated by evidence that P was framed by someone else who shot Q with P’s gun. A practical reason for performing an act a is defeasible if and only if it expresses a consideration that favors doing a but one that can be defeated by practical reasons opposing doing a. The proposition that I would make Maria happy if I were to pick her up from work is a defeasible practical reason for picking her up from work insofar as it could be defeated by a stronger reason not to pick her up from work—​ such as would be the case if a friend has a medical emergency and needs me to drive her to the hospital. Both kinds of defeasible reason can be defeated by a countervailing reason that is commensurable in the sense that it can be weighed against the other. A defeasible epistemic reason for believing that P shot Q to death in the form of veridical evidence showing that P’s fingerprints were on the gun that killed Q can be outweighed by veridical evidence showing that R was the shooter but used P’s gun while wearing gloves.5 Likewise, a defeasible practical reason to do something because it produces a benefit can be outweighed by a defeasible reason to abstain because it will produce detriment that outweighs the relevant benefit. The proposition that doing heroin will make a person feel euphorically good is a defeasible practical reason for doing heroin that is outweighed by countervailing practical reasons having to do with the dangers associated with overdose and addiction; if experiencing pleasure is all one cares about, one is likely to experience more pleasure over the long term by not getting oneself addicted to heroin. One can have a defeasible practical reason favoring something that, by itself, is outweighed by another practical reason but that nonetheless contributes to a case that, all things considered, warrants that something. Where two countervailing defeasible practical reasons, p and q, express commensurable values, p can be conclusively defeated by q in a situation where they are the only two applicable practical reasons and the weight of q is greater than the weight of p. But where there are more than two commensurable practical reasons, two practical reasons that are each singly defeated by another practical reason might combine to defeat that latter reason; if, for example, there is one practical reason to abstain from doing a that has a weight of 5 and two commensurable practical reasons to do a that each have a weight of 3, then the balance of reasons favors doing a despite the fact that

5 The idea that evidence can be comprehensively assessed by weighing one piece against another is not quite as natural as the idea that values can be weighed against each other; the idea in the former context should be understood to involve assessing certain probabilities. I use the term weight only to simplify the discussion; nothing of substance turns on this.

104  Three Conceptual Problems of Legal Normativity neither of the reasons for doing a, by itself, outweighs the reason for abstaining from doing a. The same is true of defeasible epistemic reasons. Where two countervailing defeasible epistemic reasons, p and q, express considerations that can be weighed against each other, p can be conclusively defeated by q in a situation where they are the only applicable epistemic reasons and the weight of q is greater than the weight of p. Suppose there are two witnesses testifying with respect to whether P shot Q and one testifies that she has known P all her life and knows that P would never shoot a person while another testifies that she saw P shoot Q. One will have to decide, first, how much weight to give the testimony of each witness by assessing, among other things, issues having to do with the witnesses’ credibility and, then, assess the comparative weight of the relevant epistemic reasons after they have been adjusted to reflect the credibility of each witness’s testimony. There can be practical reasons, as far as our shared conceptual judgments are concerned, that trump other conflicting reasons by expressing a winning value in virtue of being qualitatively more important than the others rather than in virtue of being quantitatively heavier. Mill’s hedonistic utilitarian moral theory differs from Bentham’s in the following way: whereas Bentham holds that all pleasures are of the same kind and can always be weighed against one another in assessing which action would produce the most pleasure, Mill holds that the intellectual pleasures are qualitatively higher pleasures than non-​intellectual pleasures and cannot be weighed against those non-​intellectual pleasures because they are of a non-​ commensurable kind. On Mill’s idiosyncratic utilitarian calculus, a higher pleasure necessarily trumps the lower pleasures because of its qualitatively superior character. While Mill’s view might not seem facially plausible, it is because it does not conform to our evaluative practices with respect to assessing practical reasons—​ and not because it is conceptually or metaphysically incoherent. There are more plausible ways in which one reason can be thought to trump another in virtue of expressing a qualitatively more important value. One might think, for example, that there are exclusionary reasons for action that trump other specified reasons in the sense that they preclude acting on those latter reasons; on this view, the moral reasons expressed by a moral obligation to do a include a moral reason to do a and an exclusionary reason not to act on any prudential reasons to abstain from doing a; I have a moral exclusionary reason, on this view, not to kill an innocent person on the strength of any prudential reasons I might have to do so. The first-​order prudential reasons to abstain from a can be thought of as trumped by the exclusionary moral reason that precludes acting on those reasons. Conclusive reasons are reasons that are not, or cannot, be defeated by other reasons. A reason p might be contingently conclusive in the sense that there are other considerations that could outweigh p but do not obtain, as a descriptive matter of contingent fact, in our world. Alternatively, p might be necessarily conclusive in the sense that there could be no considerations that outweigh p. In the

The Logical Space of Reasons  105 case where p is contingently conclusive, there are other possible worlds in which there obtain considerations not obtaining in our world that define reasons that defeat p; in the case where p is necessarily conclusive, there are no possible worlds in which considerations obtain that define reasons that defeat p.6 Epistemic reasons can be either contingently or necessarily conclusive. A reason that supports believing P killed Q might be contingently conclusive in the sense that it justifies belief that P killed Q, all things considered, because there are no facts in this world giving rise to considerations that defeat that reason, although there might be facts obtaining in other possible worlds that do so. In contrast, my belief that there is something to which I refer to as “I,” whatever its nature turns out to be, is arguably necessarily conclusive: if “I believe that p” is true when entertained or uttered by me, then there could not be any evidence to the contrary that would defeat the claim expressed by “I am something that has beliefs” when entertained or uttered by me. The norms of morality are commonly taken to define conclusive practical reasons but the sense in which they are conclusive arguably depends on whether or not we believe those norms are objective in character. If, on the one hand, we believe that morality is objective, then it is not preposterous to think that, as far as our evaluative practices are concerned, moral obligations define practical reasons that are necessarily conclusive; since what determines the truth-​value of a moral statement does not depend on contingent considerations having to do with the beliefs and preferences of subjects, it is reasonable to think that its truth-​value is determined by features instantiated by every possible world—​including worlds in which mandatory moral norms lack application either because there are no moral agents (i.e. things that owe moral obligations) or because there are no moral patients (i.e.

6 There are four interpretations of this claim that correspond to four interpretations of the claim that something is possible. First, a reason p might be contingently conclusive in the sense that there are other considerations that obtain in logically possible worlds that could outweigh p but do not obtain, as a descriptive matter of contingent fact, in the logically possible world that is ours. Second, a reason p might be contingently conclusive in the sense that there are other considerations that obtain in conceptually (or metaphysically) possible worlds that could outweigh p but do not obtain, as a descriptive matter of contingent fact, in our world. Third, a reason p might be contingently conclusive in the sense that there are other considerations that obtain in other nomologically possible worlds (i.e. worlds in which the same propositions describing regular causal relations obtain) that could outweigh p but do not obtain, as a descriptive matter of contingent fact, in our world. Finally, a reason p might be contingently conclusive in the sense that there are other considerations that obtain in practically possible worlds (i.e. nomologically possible worlds that resemble ours in more ways that are relevant with respect to our practical deliberations than other nomologically possible worlds that resemble us only with respect to the material character of our world) but do not obtain, as a descriptive matter of contingent fact, in our world. These practically possible worlds instantiate characteristics reasonably likely to obtain in our world. It might be nomologically possible for a human being to run one hundred meters in under seven seconds, but it does not follow that it is practically possible in our world; if our doing so requires the aid of technologies that we are not likely to develop before some foreseeable extinction event occurs, then it is not practically possible in our world because it is not reasonably likely to happen in worlds that resemble ours in every respect that is salient. The character of the relevant modality, as has been the case throughout this book, is practical and is used to express claims about what is reasonably likely to occur.

106  Three Conceptual Problems of Legal Normativity things that are owed moral obligations).7 If, on the other hand, we believe that morality is conventional or intersubjective, then it is not preposterous to think there could be moral obligations defining practical reasons that, as far as our evaluative practices are concerned, are contingently conclusive insofar as they are outweighed or trumped in other possible worlds in which the relevant community converges on adopting different standards of morality; the idea that insulting P will hurt P’s feelings would thus be a practical moral reason not to insult P only in worlds in which subjects converge on adopting a conventional mandatory moral norm prohibiting acts likely to hurt a person’s feelings. The logical relationship between conclusive and defeasible practical reasons can be expressed as follows. If r is a practical reason that is not necessarily conclusive, then r is a defeasible practical reason to do a in the sense that r can be defeated by other countervailing practical reasons. If r is contingently conclusive in this world, then r is also defeasible, as I define the term here, insofar as there are other possible worlds in which considerations obtain that define practical reasons that would have defeated r in this world had they obtained in this world. The class of necessarily conclusive practical reasons and the class of defeasible practical reasons are thus both mutually exclusive and jointly exhaustive, while the class of contingently conclusive practical reasons and the class of defeasible practical reasons are jointly exhaustive but not mutually exclusive—​since there can be a defeasible practical reason that is defeated in other possible worlds by other practical reasons but is not defeated by other practical reasons in the actual world.

2.3  Subjective and Objective Reasons A third distinction is between subjective and objective reasons.8 The notion of a subjective reason favoring p is concerned with what some rationally competent subject regards, as a descriptive matter of contingent fact, in her thinking as a consideration favoring p. The proposition that P’s fingerprints are on the gun that killed Q is a subjective epistemic reason for R to believe that P killed Q if and only if R both believes that proposition and regards it as evidence for the proposition that P killed Q. R’s experiencing uncomfortable pangs of hunger defines a subjective 7 It is not preposterous to think that moral norms, if objectively true, are necessarily true and hence true even in worlds in which these norms lack application because neither moral agents nor moral patients exist; if it is objectively wrong for human beings to torture non-​human sentient animals, then it is wrong in worlds in which only humans exist, worlds in which only non-​human sentient animals exist, and worlds in which neither exist. On this conception, the principle expresses an objective value protecting beings with certain properties in every possible world but has practical application only in possible worlds in which both humans and non-​human sentient animals exist. 8 I have, of course, presupposed this distinction throughout this volume by my use of the locutions “as a descriptive matter of contingent fact” and “as an objective matter of normative practical rationality” to distinguish what agents believe from what agents should believe; since this distinction is obviously presupposed in our shared conceptual and evaluative practices, its use begs no questions.

The Logical Space of Reasons  107 practical reason for R to get something to eat if and only if R treats the proposition expressing that fact in her deliberations as a consideration that favors eating something. Such reasons are felt, internal reasons that arise from the subject’s experienced mental states and are, for that reason, properly characterized as subjective. Subjective reasons might be good reasons from the standpoint of our shared evaluative practices, or they might be bad reasons. P’s fingerprints being on the gun that killed Q defines a good defeasible epistemic reason, if true, to believe that P killed Q; P’s having written a sympathetic book about a serial killer, if true, is not a good reason to believe that P killed Q. R’s experiencing hunger pangs, other things being equal, defines a good practical reason for R to eat a modest nourishing meal; R’s experiencing a desire to feel the exhilaration that frequently attends committing a crime defines a bad practical reason to rob a bank. The notion of an objective reason arises out of the idea that there can be good and bad reasons for doing and for believing things. Our shared conceptual and evaluative practices presuppose that there are objective standards of normative epistemic and practical rationality that determine whether or not a reason is a good one. On this view, there can be subjective reasons for believing or doing something that are bad in the sense that they do not satisfy the relevant objective standards and hence do not favor doing or believing that something regardless of what anyone thinks. It is the hallmark of objectivity with respect to reasons, as we conceive the notion, that any or all of us could be mistaken with respect to whether something counts as a good reason for doing or believing something.9 The distinction between objective and subjective reasons amounts to this. The claim that some proposition r is a subjective reason for Q to do/​believe something is an empirically descriptive claim about Q’s mental states with respect to r; insofar as Q regards r as a reason to do/​believe something, as a descriptive matter of contingent fact, r is a subjective reason for Q to do/​believe that something. The claim that r is an objective reason for Q to do/​believe something expresses a claim about what we take to be mind-​independent standards of normative rationality; in particular, it asserts that Q should, as an objective matter of normative rationality, regard r as a reason to do/​believe that something. In the case of epistemic reasons, the relevant determinants of correctness are objective norms of epistemic rationality while in the case of practical reasons, the relevant determinants are objective

9 This should not be taken to imply that we have epistemic access to any objective norms governing reasons; the point here is merely that our shared evaluative practices presuppose that there are such standards. The best we can do, as far as the modest approach to conceptual analysis I have adopted is concerned, in evaluating the objective rationality of some belief or behavior is to assess it against our shared conception of what these objective norms require. I speak as though we do have such access only because our conceptual and evaluative practices presuppose that we do; a modest analysis of the concept of a legal system must presuppose the shared views that motivate the associated practices, which I have assumed without endorsing them or presupposing that they are objectively correct. See Chapter 1, Section 7.

108  Three Conceptual Problems of Legal Normativity norms of practical rationality—​or, rather, what we converge on believing these norms require. The claim that an institution is normative in the relevant sense does not entail anything about the strength of the reasons its operation provides, as an objective matter of normative practical or epistemic rationality, or the strength of the reasons that rationally competent subjects are likely to regard it as providing, as a descriptive matter of contingent fact. The strength of the relevant reasons might or might not rise to the level of being conclusive depending on the relevant circumstances. If I know that someone is infallible with respect to what I should do and always tells the truth about what I should do, then the fact that she tells me that I should do a gives rise to a conclusive epistemic reason to believe that I should do a and might also, depending on what justifies someone in asserting practical authority over me, give rise to a conclusive practical reason for me to do a. If I know that someone is more reliable than I am in determining what I should do but is not infallible with respect to what I should do and always tells the truth, then the fact that she tells me that I should do a provides an epistemic reason to believe that I should do a that is weighty but might fall short of being conclusive; it might also, depending on what justifies someone in asserting practical authority over me, provide a practical reason for me to do a that is also weighty but falls short of being conclusive.

2.4  Motivating and Justifying Reasons for Action A final distinction is relevant only with respect to practical reasons. A motivating reason m to do a is a practical reason expressing a consideration that a rationally competent subject either views, as a descriptive matter of contingent fact, or should view, as an objective matter of normative practical rationality, as desirable and that hence inclines, or should incline, her to do a; in the former case m is a subjective motivating reason to do a while in the latter case m is an objective motivating reason to do a. Motivating reasons, as a conceptual matter, function to motivate rationally competent subjects to do what they are motivating reasons to do. In both cases, a motivating reason m to do a is a practical reason for doing a that might, or might not, be defeated by other reasons. In the case of subjective reasons, the subject might, or might not, as a descriptive matter of contingent fact, view m as outweighed or trumped by other considerations she regards as countervailing motivating reasons to abstain from doing a; in the case of objective reasons, the motivating reason to do a might, or might not, be outweighed or trumped by other reasons, as an objective matter of normative practical rationality, to abstain from doing a. A motivating reason m is subjectively conclusive for a rationally competent subject if, as a descriptive matter of contingent fact, she believes that m is not, or could not be, defeated by other practical reasons. Insofar as a rationally competent

The Logical Space of Reasons  109 subject agent believes that m is not but could be defeated by other reasons, m is contingently conclusive for her; insofar as she believes that m could not be defeated by other practical reasons, m is necessarily conclusive for her. Insofar as she regards m as a conclusive motivating reason to do a, she treats m as sufficient, by itself, in her deliberations to rationally induce her to do a and does a for that reason. Subjective conclusive motivating reasons, as a descriptive matter of contingent fact, conclusively motivate the agent for whom they are subjectively conclusive to do what they favor doing.10 A motivating reason m to do a is objectively conclusive for a rationally competent subject if, as an objective matter of normative practical rationality, m is sufficient to induce her to do a; that is, m is objectively conclusive for a rationally competent subject if she should, according to objective standards of normative practical rationality, view it as sufficient to induce her to do a. A motivating reason m is contingently conclusive if, according to objective standards of normative practical rationality, m is not but could be defeated by other reasons; m is necessarily conclusive if it cannot, as an objective matter, be defeated by other reasons. Objective conclusive motivating reasons, as an objective matter of normative practical rationality, should motivate the subject for whom they are objectively conclusive to do what they favor doing.11 A justifying reason j to do a is a practical reason that justifies the performance of a in the sense that j expresses a value that a rationally competent subject either views, as a descriptive matter of contingent fact, or should view, as an objective matter of normative practical rationality, as justifying the doing of a under some set of standards governing her behavior; in the former case j is a subjective justifying reason to do a while in the latter case j is an objective justifying reason to do a. A justifying reason j to do a, as was true of motivating reasons, might be objectively or subjectively conclusive with respect to justifying the act, or it might not be. If j is not necessarily conclusive (as an objective/​subjective matter), then j is a defeasible (objective/​subjective) justifying reason to do a. But, as was true of motivating reasons, if j is contingently but not necessarily conclusive with respect to justifying a under the applicable set of standards, then j is a defeasible justifying reason to do a that is undefeated in the actual world but is defeated in other possible worlds.12 10 To motivate someone to do something is to either rationally incline or, depending on the strength of the motivation, rationally induce her to do that something. 11 A motivating reason m to do a need not be a conclusive motivating reason to do a to contribute to rationally inducing her to do a. It might be that m falls short of being conclusive but contributes enough weight that the subject either regards m, as a descriptive matter of contingent fact, or should regard m, as an objective matter of normative practical rationality, as combining with the weight of other motivating reasons in favor of doing a to tip the balance of the applicable motivating reasons in favor of doing a. 12 One might have a justifying reason to do a in this world under some set of conventional standards, like those of etiquette, that one does not have in some other possible world where the content of the relevant type of conventional standards differs from that of the standards in this world.

110  Three Conceptual Problems of Legal Normativity The logical relationship between motivating and justifying reasons is not clear. One might think that the conceptual distinction between the two amounts to little in practice because motivating reasons also purport to justify an act, but there is no obvious reason to rule out the possibility that someone might be motivated to act on some consideration that she knows would not justify the act. Someone could have, for whatever reason, such a strong desire to do something that she is motivated to act on that desire regardless of whether she believes she is justified in doing so under any set of applicable norms; “I don’t give a fuck” is sometimes used with obvious exasperation on the part of the speaker to express exactly that unfortunate configuration of preferences. While acting on the basis of such preferences might not be rational in the weak sense that it is not justified on the strength of the relevant norms, it is not obviously irrational; the boundary of the distinction between what is irrational and what is merely not rationally justified are not altogether transparent. Motivating and justifying reasons can come apart in other ways. First, a rationally competent subject can act on a subjective motivating reason that does not objectively justify the act; in this case, the subject’s internal motivations do not line up with what objectively justifies the act. Second, a rationally competent subject can act on a subjective motivating reason that she falsely believes does not objectively justify the act; in this case, the subject’s subjective motivating reason is an objectively justifying reason but not a subjective justifying reason. Third, a rationally competent subject can act on a subjective motivating reason without having any beliefs about whether the act is justified. One might plausibly think that there are some acts that one can permissibly perform that need not, as an objective matter of normative practical rationality, be justified under some set of standards that govern the subject; for my part, I do a lot of things without regard for whether they are justified because those acts, being morally permissible, do not seem to trigger a need for prudential, moral, or any other kind of justification. As a conceptual matter, the permissibility of an act under some set of standards seems to logically preclude the need to justify it under that same set of standards. Motivating and justifying reasons can also come apart with respect to the mechanisms by which an institutional normative system provides each type of reason. If it makes sense to think that we have a legally justifying reason not to commit murder, it is the valid legal norm prohibiting murder that is the source of that reason—​and not the coercive sanction that some recognition norm authorizes a court to impose for non-​compliance. But if it makes sense to think that what provides the relevant motivating reason to comply with a law prohibiting murder is the fact that the courts are authorized to impose a coercive sanction for non-​ compliance, then the mechanism by which law creates justifying reasons is distinct from the mechanism by which it creates motivating reasons; the threat of a coercive sanction for doing a does not, and cannot, give rise to a legally justifying

The Logical Space of Reasons  111 reason not to do a. If this is correct, then one can have motivating reasons that are not justifying reasons and conversely. There are as many types of justifying reason as there are types of normative standard. Insofar as a justifying reason to do a makes reference to moral standards and is thereby concerned with what is morally justified, it is a moral justifying reason to do a. Insofar as it makes reference to prudential standards and is thereby concerned with what is justified in virtue of optimally conducing to the agent’s interests, it is a prudential justifying reason to do a. Insofar as a justifying reason to do a makes reference to legal standards and is thereby concerned with what is justified under the law, it is a legal justifying reason. And so on. This feature of justifying reasons also distinguishes them from motivating reasons. It is somewhat counterintuitive, though not obviously incoherent as a conceptual matter, to think of motivating reasons as being relative to particular classes of standards. It is clear that some motivating reasons make reference to considerations that are prudential in character because they have to do with what might or might not conduce to self-​interest and thus count as prudential motivating reasons, but the qualifier functions differently from the way it functions with respect to justifying standards. There is nothing in the notion of a prudential motivating reason that seems clearly to require any reference to some set of behavioral standards of prudential justification; the motivating reason is prudential in the sense that it is concerned with self-​interest but not in the sense that it makes any obvious reference to standards explicitly concerned with prudential justification. The claim that P has a justifying reason to do a, then, neither implies the claim that P has a motivating reason to do a nor is implied by it. The claim that r is an objective/​subjective justifying reason to do a does not, without more, imply anything with respect to whether r is an objective/​subjective motivating reason of any kind to do a; one can have an objective legal justifying reason to do something that one has no objective or subjective motivating reason to do. Conversely, the claim that r is an objective/​subjective motivating reason to do a, by itself, entails nothing about whether it is an objective/​subjective justifying reason to do a; one can have an objective motivating reason to do something that is not subject to the standards of rational justification and hence does not implicate a justifying reason of any kind. Even so, the claim that the relevant considerations are motivating reasons does not preclude that they might also be, or correspond to, justifying reasons. A concern to do the morally right thing can serve both as a motivation for an agent’s act and as part of a justification under the standards of morality. Indeed, it might be the case that the two types of reason always coincide with respect to moral motivations and moral standards, but this is consistent with the claim that the two kinds of reason are conceptually and ontologically distinct. The claim that these two kinds of reason are conceptually distinct does not imply the claim that the considerations that give rise to one type of reason cannot also give rise to the other type; it implies

112  Three Conceptual Problems of Legal Normativity only that, as a conceptual matter, something is not a justifying reason wholly in virtue of being a motivational reason, and conversely.

3.  The Character of the Reasons to Which the Practices Constitutive of Law Are Reasonably Contrived to Give Rise To fully address the problem of legal normativity, one must identify the character of the relevant reasons to which the practices constituting something as a system of law are, as a conceptual matter, reasonably contrived to give rise; there is no way to vindicate law’s presumed conceptual normativity without knowing what kinds of reasons are relevant. This requires determining the character of the applicable reasons with respect to each of the four distinctions discussed above. This section attempts to identify the character of the reasons to which the practices constitutive of law are, as a conceptual matter, reasonably contrived to give rise as a prelude to offering an explication of law’s conceptual normativity that is grounded in the Coercion Thesis.

3.1  The Relevant Reasons Are Practical Reasons and Not Epistemic Reasons There are, as will be discussed in Section 4 of this chapter, a number of conceptual problems of legal normativity but the most important problem is concerned with how the practices constituting something as a system of law can give rise to practical reasons. Insofar as it is a conceptual truth that law as such is reasonably contrived to regulate behavior through the governance of norms metaphysically capable of guiding behavior, it is also a conceptual truth that the practices constituting something as a system of law must be capable, in some sense, of giving rise to reasons to comply with mandatory norms prohibiting acts that might otherwise tempt subjects.13 A system of law can regulate behavior only insofar as it is practically normative in the sense that it does something capable of inducing rationally competent subjects to refrain, when otherwise tempted, from prohibited acts. A system of law can be practically normative in the required sense only insofar as it has resources capable of changing the behavior of rationally competent subjects by providing them with practical reasons to comply with at least some of 13 The argument in this chapter is agnostic, unlike the argument in Chapter 4, with respect to the content of the ultimate end for which legal regulation is the means and with respect to the conceptual function of law. While the last chapter argued the ultimate end and conceptual function of a legal system is to keep the peace, I wish to avoid begging any questions and so will not assume anything more about law’s conceptual function—​or basic task (if one prefers)—​than that it involves regulating behavior.

The Reasons to Which Law Gives Rise  113 its requirements. When a person does something differently because of the law, her behavior has been changed by it; for a legal system to be capable of changing the behavior of subjects, it must be capable of providing reasons for doing something they might not otherwise be inclined to do. At the most basic level, then, the most important conceptual problem of legal normativity is to explain how the practices constituting something as a system of law are capable of changing the behavior of rationally competent subjects inclined to do things prohibited by law by providing something they characteristically regard, as a descriptive matter of contingent fact, as a practical reason not to do those things because they should, as an objective matter of normative practical rationality, regard it as a practical reason not to do those things. It is true that the practices constituting something as a system of law must be capable of providing epistemic reasons with respect to what behaviors are prohibited or otherwise regulated. Law could not make a practical difference with respect to how people behave unless those constitutive practices make it possible for rationally competent subjects to have some reliable way to learn what law requires and hence to be epistemically justified with respect to their beliefs about what it requires.14 One cannot knowingly conform one’s behavior to laws prohibiting copyright infringement unless one knows what acts are treated under those laws as constituting copyright infringement. But there is nothing mysterious about how the practices constituting something as a system of law make it possible for rationally competent subjects to learn what mandatory legal norms require of them. Law is, by nature, a public institution in the sense that the norms treated as valid laws are publicly promulgated, recognized, and applied in something that counts as a common language. If the issue is to explain how law as such is capable of providing reasons to believe, the fact that laws are publicly promulgated, recognized, and applied in a common language is enough to solve that problem; if there is anything left to explain, that is a task for normative and descriptive epistemologists to perform—​and not one for conceptual jurisprudents. A legal system might not always be successful in informing subjects of exactly what the norms require in every case by providing reasons for belief, but the practices constituting something as a system of law are clearly equipped to provide epistemic reasons with respect to what behaviors are legally prohibited or otherwise regulated in the vast majority of cases. It might not always be obvious who counts

14 This seems to be a general truth about norm-​governance. It is implausible to think that a class of beings can be governed by normative content they are causally unable to identify and understand; it is a conceptually necessary condition for norms to govern the behavior of a class of beings that class-​ members can ascertain what, at least, the core content of those norms require of them. This is no truer of law than it is of morality or any other set of norms governing our behavior; these are simply conceptual truths about norm-​governance. See Kenneth Einar Himma, Morality and the Nature of Law (Oxford University Press 2019) Chapter 7.

114  Three Conceptual Problems of Legal Normativity as innocent for purposes of applying a valid legal norm prohibiting the killing of innocent people, but it should be clear to any rationally competent subject that it prohibits bashing, beating, bludgeoning, burning, choking, crushing, drowning, poisoning, shooting, smashing, stabbing, starving, suffocating, and torturing infants to death; insofar as that is not clear to someone who competently speaks the language, that is a reason to think that she is irrational (or clinically insane) and hence ought not, as a moral matter, be held accountable under the law. But, either way, there is no non-​trivial conceptual problem of any kind that concerns how law as such can provide epistemic reasons with respect to beliefs about what behaviors are regulated by law. The practices constituting something as a system of law involve the manufacture, application, and enforcement of legal norms in a manner sufficiently public in character that rationally competent subjects can come to know much, though not everything, about what law requires.15 These practices include publicly recognizing, applying, and enforcing norms that are expressed in language clear enough to communicate to rationally competent subjects the core of what is prohibited. Since it is clear that the practices constitutive of law are sufficiently public to enable rationally competent subjects to determine the core content of those general norms that will be applied and enforced against them, whether by reading the relevant laws or asking an attorney, there is no mystery as to how those practices can provide epistemic reasons; law is arguably better suited, in virtue of these conceptual publicity requirements, than any other set or system of artifactual or non-​artifactual norms to provide epistemic reasons. This should not be construed to suggest that there are no interesting problems having to do with legal epistemology or normativity; it is merely to assert that there are no serious conceptual problems concerning how the practices constitutive of law can give rise to epistemic reasons with respect to legal requirements. Because there is no decision procedure by which any competent subject can determine in any case what law requires of her, there will always be moral and epistemological issues with respect to how judges should decide and apply the law in hard cases. But none of these issues is conceptual. The most important conceptual problem of legal normativity is to explain how law as such is equipped to provide practical reasons because it is not immediately clear how the practices and norms constituting something as a system of law can generate reasons to comply with legal requirements. If, as one might plausibly think, it is conceptually sufficient for the existence of a legal system (1) that the persons serving as its officials practice a social rule of recognition governing the recognition and application of valid norms of 15 Law differs from morality in this respect. Assuming that there are right answers to questions about what is required by morality that do not depend on what subjects believe about the answers to these questions, it is not obvious how we can reliably learn what mind-​independent standards of morality require from us. Unlike the concept of law, there is nothing in the concept of morality that would explain how subjects can come to have epistemically justified beliefs with respect to what norms of morality require of them.

The Reasons to Which Law Gives Rise  115 the system; and (2) that the behavior of subjects generally conforms to the norms valid under the social rule of recognition, it is difficult to see how law as such is equipped to provide practical reasons to comply. There is no obvious reason to think that rationally competent subjects should, or are likely to, treat content with the status of law as giving rise to practical reasons to comply with that content solely in virtue of being recognized and applied as such by officials in a manner that conforms to the social rule of recognition.

3.2  The Relevant Practical Reasons Are Defeasible Reasons and Not Conclusive Reasons The practices constituting something as a system of law, by themselves, do not necessarily result in something that rationally competent subjects characteristically regard, or should regard, as conclusive reasons to comply. It might be true that there are conceptually possible legal systems that provide reasons that subjects characteristically regard, or should regard, as conclusive. A legal system L that is known to always be morally justified in enforcing its valid norms with a coercive sanction might give rise to something subjects should, and characteristically do, regard as conclusive practical reasons; it is hard to see how L could always be morally justified in imposing non-​trivial detriment for violations if it did not provide objectively conclusive practical reasons that are characteristically accepted by subjects as such. But if L provides conclusive practical reasons, it is not the general practices constituting something as a legal system, which are shared by every conceptually possible legal system, that explain how L provides such reasons; it is rather those distinctive practices constituting L as always morally justified in enforcing its norms, which are not shared by every conceptually possible legal system, that explain how L provides those conclusive practical reasons. The most that can be presumed of the capacity of law as such to generate practical reasons is that the practices constituting something as a system of law are equipped to provide considerations likely, as a descriptive matter of contingent fact, to be regarded by rationally competent subjects as practical reasons to comply that can be defeated by practical reasons not to comply because they should, as an objective matter of normative practical rationality, regard them as such. The officials of a morally illegitimate legal system might recognize as law norms governing non-​ official behavior that require acts that are obviously prohibited by objective norms of morality, as is true of laws requiring discriminatory treatment of persons on the basis of race; in such a case, the practical reason favoring compliance to which the practices constituting the system as one of law give rise can and should be regarded as defeated by those practical reasons favoring non-​compliance to which the mandatory moral norms prohibiting discriminatory behaviors give rise.

116  Three Conceptual Problems of Legal Normativity But the practices constitutive of law can give rise, at most, to defeasible practical reasons to comply even with respect to mandatory legal norms prohibiting morally wrongful assaults against persons or nations. Insofar as one has a conclusive practical reason r to refrain from the socially disruptive acts prohibited by that legal content, it is the moral quality of that content that gives rise to r and not the fact that there is a valid mandatory legal norm reproducing that content. While it might be true that a morally legitimate legal system provides novel content-​independent conclusive practical reasons to comply with those norms, this would be so in virtue of the moral properties of the system that give rise to moral reasons to comply—​ and not the practices constitutive of a legal system. The practices constituting something as a system of law, by themselves, are not equipped to provide any kind of conclusive practical reason to comply. No purely artifactual institutional normative system in our world is built that way; rationally competent subjects like us are capable of recognizing, applying, and enforcing morally wicked content that gives rise to, at best, practical reasons that should be regarded by subjects as conclusively defeated by moral reasons because we are both morally flawed and epistemically fallible.

3.3  The Relevant Reasons Are Motivating Reasons and Not Justifying Reasons Insofar as a legal system is conceptually equipped to provide practical reasons capable of inducing compliance with mandatory legal norms governing non-​official behavior in rationally competent self-​interested subjects antecedently disposed to act otherwise, the relevant reasons for action are motivating reasons, and not justifying reasons. To induce someone to refrain from doing something she is antecedently disposed to do, the practices constituting something as a system of law must be equipped to persuasively motivate her to do otherwise. We do what we are most motivated to do. There are a variety of considerations that can motivate us, including a desire to be justified in our behavior, but it makes little sense to think that the practices constituting something as a system of law are equipped to provide reasons that justify, from the standpoint of one’s legal obligations, refraining from acts prohibited by mandatory legal norms governing non-​official behavior. While it is always possible to fetishize the law or some group of officials claiming authority, someone motivated to refrain from killing someone simply to ensure that her behavior is legally justified is confused and possibly in need of serious medical intervention; the idea that valid legal norms prohibiting murder are concerned to legally justify not killing persons who should not be killed is silly enough to be medically worrisome. It is, of course, possible, in any relevant sense of the term, for someone to be motivated by such considerations, but it is hard to see what would warrant thinking

The Reasons to Which Law Gives Rise  117 that the practices constituting something as a system of law are primarily concerned to appeal to such foolishly fetishistic preferences. Given the content of our conceptual practices, which reflect what we commonly believe about the propensities of rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours to behave in socially undesirable ways, it makes far more sense to think that these practices are best understood as providing some sort of motivating reason than it does to think that they should be understood as providing something that justifies compliance under any set of standards—​including standards of moral and prudential justification. This is not to suggest that the concept of a justifying reason is irrelevant with respect to understanding law as such; it is clear that the concept has necessary application with respect to official acts that can be challenged by those putatively bound by the normative output of those acts. The rule of recognition as it applies to adjudication typically requires not only that judges behave in a manner that can be justified by reference to the relevant mandatory legal norms governing judicial behavior but also that they articulate the reasons that would show their decisions properly apply the relevant mandatory legal norms governing non-​official behavior; those decisions can hence be challenged on the ground that they are not legally justified because inconsistent with the balance of applicable legal justifying reasons—​which are determined by the relevant legal norms governing judicial behavior together with the relevant legal norms governing non-​official behavior that give rise to the dispute requiring judicial adjudication. Similarly, the rule of recognition as it applies to the enforcement of law defines standards that legally justify the imposition of coercive sanctions on those who violate mandatory legal norms governing both official and non-​official behavior; judges can be authorized to punish officials who refuse to comply with court orders, as well as persons who violate mandatory legal norms prohibiting assaults on persons or property. Those enforcement decisions can typically be challenged on the ground that they are not legally justified given the balance of applicable legal justifying reasons. Finally, legislative acts can also be challenged on the ground that they do not conform to the procedural or substantive requirements of the rule of recognition and are hence not legally justified by the balance of applicable legal justifying reasons. Constitutions typically define both procedural and substantive requirements with which legislators must comply in order for their enactments to be recognized and enforced by the courts as valid legal norms of the system. The practices constituting something as a system of law are contrived, for these reasons, to utilize different normative mechanisms for regulating official and non-​ official behavior. These practices seem primarily contrived to regulate official behavior through norms that entail justifying reasons that legally justify the acts of legislative, adjudicative, and enforcement agencies; insofar as any official act can be challenged under the relevant legal standards, its rectitude can be established only

118  Three Conceptual Problems of Legal Normativity by showing that the act is legally justified. Even a rule of recognition endowing a sovereign with authority to enact legal norms without restriction defines justifying reasons:  something that is treated as law by officials can be challenged on the ground that it does not have a source in an appropriately configured sovereign act. In contrast, the practices constituting something as a system of law are primarily contrived to regulate non-​official behavior through norms that provide reasons sufficient to motivate subjects to comply with valid mandatory legal norms in circumstances where they are not antecedently motivated to do so. Persons who do what the law requires because it is required by the law regard the balance of motivating reasons as favoring compliance, as do people who do what the law requires because the requirement is backed by the threat of a coercive sanction. It is nonetheless clear that law as such can provide both motivating reasons for official acts and justifying reasons for non-​official acts. To begin, a rule of recognition can include norms providing for impeachment or removal of officials for acts that violate either mandatory legal norms governing non-​official behavior or mandatory legal norms governing official behavior. The U.S. Constitution does both: Article II, Section 4 authorizes the impeachment and removal of presidents for “High Crimes and Misdemeanors”; Section 4 of the 25th Amendment authorizes removal of presidents upon an appropriate showing that “the President is unable to discharge the powers and duties of his office.” Insofar as the required inability to discharge the relevant powers and duties includes a willful refusal to comply with the relevant recognition norms, both constitutional provisions would be the source of motivating reasons to do what is required so as not to be removed from office. Such provisions are not conceptually necessary pieces of a legal system, but they are possible in every sense that matters. It is likewise clear that law can provide justifying reasons for non-​official acts. Valid legal norms empowering persons to alter their legal relations with other people are also contrived to provide justifying reasons.16 A defendant in a lawsuit alleging breach of contract will typically claim that her actions are legally justified either because they satisfied her contractual obligations or because the agreement does not satisfy the formal requirements of contract law and hence does not legally bind her. The legal norms of contract law, like the recognition norms governing official behavior, are reasonably contrived, as a conceptual matter, to provide justifying reasons.17 16 Other civil laws are, in contrast, contrived to provide motivating reasons but these function differently from the civil norms of contract law. Insofar as negligence law authorizes courts to coercively order a defendant to compensate a plaintiff for reasonably foreseeable injuries proximately resulting from negligent acts, negligence law functions as though it prohibits the relevant acts; the prospect of having to pay damages is characteristically likely, as a descriptive matter of contingent fact, to be regarded by subjects as providing motivating reasons to avoid the kind of negligence that triggers such liability because it should, as an objective matter of normative practical rationality, be regarded as such. 17 There are also legal norms having to do with contract law that are equipped to provide motivating reasons to comply with certain directives. Norms authorizing courts to coercively enforce properly formed contracts are equipped, to provide rationally competent subjects who wish to have an

The Reasons to Which Law Gives Rise  119 The norms of contract law thus resemble the recognition norms that prescribe how valid legal norms can be recognized, applied, and enforced by officials. Both types of norm are best conceived as primarily concerned to provide rationally competent subjects with justifying reasons. Insofar as they provide prudential motivating reasons, those reasons depend on the subjects instantiating a constellation of internal motivations that cannot be plausibly attributed to all rationally competent subjects; while all rationally competent subjects can be presumed, other things being equal, to want not to be incarcerated, they cannot be presumed to want to engage in the activities of officials or to formalize their agreements by memorializing them in enforceable instruments. But there is no philosophical mystery with respect to how the relevant legal norms are equipped to provide justifying reasons. It is a conceptual truism that any norm that defines a standard to which some rationally competent subject must conform can provide a justification relative to that norm for conforming behavior where a justification is needed. This is just how norms work:  moral standards are the source of considerations that morally justify conforming behavior; social standards of etiquette are the source of considerations that socially justify conforming behavior from the standpoint of those standards; and legal standards are the source of considerations that legally justify conforming behavior. While it is therefore trivially true that one is legally justified in not committing murder by the mandatory legal norms prohibiting murder, the point of laws prohibiting murder is to provide a decisive motivating reason, where one is needed, for someone to refrain from murder. But since officials do not challenge the legal propriety of acts that do not putatively violate the law, the legally justifying reasons that any legal norm necessarily provides do not play a central role in how law regulates non-​official behavior. There is simply no interesting problem of how the practices constitutive of a system of law can give rise to justifying reasons.

3.4  The Relevant Reasons Are Objective and Not Subjective Reasons It is a conceptual truth that law is an institutional artifact characteristically used, as a descriptive matter of contingent fact, to regulate the behavior of rationally competent subjects sometimes disposed to do things thought to be undesirable. An institutional normative system that does not contribute marginally to preventing these undesirable acts because putative subjects are otherwise conclusively indisposed

agreement enforced by the court with a motivating reason. But such a motivating reason has force only insofar as subjects are antecedently motivated to ensure that their agreements are enforced by the court; it is that pre-​existing desire that leads rationally competent subjects to regard the norms of contract law as motivational reasons to do what is necessary to ensure that their agreements are enforced.

120  Three Conceptual Problems of Legal Normativity to commit such acts has nothing to do with why they abstain from them and is not properly characterized as a system of law; there can be no efficacious legal regulation of the behavior of real angels when it comes to undesirable acts—​because real angels, as opposed to the beings comprising the society of angels discussed in Chapter 10 of this book, are morally infallible, morally impeccable, and hence never disposed to commit such undesirable acts. That is what makes them angels. The point here is not just that real angels hence have no need for a legal system; it is also that they cannot use a legal system to prevent undesirable acts because the practices constituting an institutional normative system as one of law cannot make any difference with respect to whether they commit such acts. One can use an object for a purpose only insofar as that object is needed for that purpose; if subjects are already behaving perfectly well without law, law cannot be used by those subjects to prevent undesirable acts any more than a medication can be used to cure an illness that someone does not have. Law is an institutional artifact conceptually equipped to make a difference with respect to what rationally competent subjects do enough of the time to enable them to live together in a stable community. Where there is a system of law, it is because the practices constituting something as a system of law change the behavior of subjects frequently enough to ensure that they can get along well enough to live peacefully together in a stable community; a group of people can get along well enough to live together only insofar as they refrain from socially disruptive acts that are prohibited because commonly regarded as undesirable. The practices constituting something as a system of law are thus reasonably contrived to make it more likely that rationally competent self-​interested subjects like us can get along together in worlds of acute material scarcity like ours than it would otherwise be—​which presupposes that rationally competent subjects can be, and sometimes are, antecedently disposed to commit acts prohibited by law. The practices constituting something as a system of law are equipped to perform law’s conceptual function of regulating behavior by providing something that rationally competent subjects are likely to treat as additional motivating reasons that persuade them when they need to be persuaded, as is sometimes the case, to refrain from acts that are prohibited by law because they are deemed undesirable. That is, these constitutive practices are equipped to perform law’s conceptual function of regulating behavior, regardless of the ultimate end these practices seek to achieve, only insofar as they are likely to give rise to novel content-​independent subjective motivating reasons that persuade subjects to refrain from prohibited acts, when antecedently disposed to do otherwise, often enough to permit them to live together in a state of comparative peace. Law is contrived to make a practical difference in the behavior of rationally competent subjects whose motivations sometimes dispose them to behave badly by giving them something they are characteristically likely, as a descriptive matter of contingent fact, to regard as

The Reasons to Which Law Gives Rise  121 a persuasive countervailing motivating reason to comply when otherwise antecedently disposed. But this can be done only insofar as the practices constitutive of law are also conceptually equipped to provide new objective motivating reasons to conform to mandatory legal norms governing non-​official behavior. The practices constituting something as a legal system would not be equipped to provide something rationally competent subjects are characteristically likely, as a descriptive matter of contingent fact, to treat as subjective motivating reasons to refrain from prohibited acts if those practices did not provide something that, as an objective matter of normative practical rationality, they should treat as motivating reasons to refrain from such acts. Subjects are plausibly characterized as rationally competent only insofar as they are reason-​responsive in a way that largely conforms to objective norms of practical rationality. The need for law arises among rationally competent self-​interested subjects like us precisely because we are not always predisposed to behave in a manner that conforms to objective norms of practical rationality—​either because we do not always know what these norms require, all things considered, or because we are motivated by considerations of self-​interest to act in ways that we know violate mandatory moral norms we should not violate. The claim that we are rationally competent does not entail the claim that we always satisfy the objective norms of practical rationality dictating how we should assess the applicable reasons; those norms incorporate moral and prudential considerations in a manner that determines what, as an objective matter, we have most reason to do all things considered. The claim that we are rationally competent entails only the claim that we characteristically act for what we take to be reasons and that we are characteristically responsive to what we collectively regard as the requirements of objective norms of practical rationality. What we need law to do is to provide something that rationally competent subjects like us are likely to treat as persuasive motivating reasons to refrain from certain acts when antecedently disposed to do otherwise. The practices constituting something as a legal system are not plausibly thought to be concerned primarily with providing subjective motivating reasons to comply because these practices are concerned to regulate the behavior of subjects who are presumed rationally competent in virtue of being largely reason-​responsive; it would be prohibitively expensive for a system of law to accommodate every subject’s idiosyncratic preferences. The best that a legal system can do, insofar as it is reasonably contrived to prevent acts deemed undesirable through the governance of norms metaphysically capable of guiding behavior, is to provide something rationally competent subjects like us are likely to regard as an additional motivating reason to comply that can tip the balance in favor of complying with mandatory legal norms prohibiting undesirable acts, when we are not sufficiently motivated to comply by considerations that should motivate rationally competent subjects like us, often enough that we can live together. It is thus a conceptually

122  Three Conceptual Problems of Legal Normativity necessary condition for something to count as system of law that it is equipped to provide objective motivating reasons to comply with at least some mandatory legal norms governing non-​official behavior.

3.5  The Relevant Motivating Reasons Must Be Both Novel and Content-​Independent Insofar as a legal system provides objective motivating reasons to comply, those reasons are both novel and content-​independent. They are novel in the sense that rationally competent subjects have those reasons only in virtue of being entailed by the practices constituting a norm as a binding law of the system; while subjects might have other objective motivating reasons for doing what law requires in any given instance, the practices giving rise to a system of law are contrived to give rise to new objective motivating reasons to comply with norms that subjects did not have prior to being treated by the officials as law. They are content-​independent in the sense that the relevant motivating reasons arise from the fact that they are treated by the officials as law and not from the content of the relevant norms. The novelty of these objective motivating reasons is derived from their content-​ independence. Insofar as it is the empirically observable fact that some mandatory norm is recognized, applied, or enforced by officials as law that gives rise to the new motivating reason, the new motivating reason depends on the norm’s being treated as law and not on its content. If it is the case, given any act a required by law, that rationally competent subjects would have had a new motivating reason not to do a had the relevant officials treated a mandatory norm prohibiting a as law, then the legal norms and practices regulating the performance of a provide content-​ independent and hence new objective motivating reasons with respect to doing or not doing a.

3.6  Law Must Be Reasonably Contrived to Provide the Relevant Reasons and Not Merely Capable of Doing So The claim that law is equipped to provide defeasible novel content-​independent objective motivating reasons to comply with mandatory legal norms prohibiting socially undesirable acts should be construed to assert something stronger than the claim that these practices are capable of doing so. It might be possible for something to provide objective motivating reasons to refrain from undesirable acts without its being even minimally likely that it succeeds in doing so among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours. It is possible in every relevant sense for rationally competent subjects like us to regard those practices as providing content-​independent motivating reasons to

The Reasons to Which Law Gives Rise  123 do what the law requires. Since there is nothing in the causal laws obtaining in our universe that precludes us from treating certain institutional practices as providing content-​independent motivating reasons to refrain from prohibited acts, it is nomologically possible—​and hence logically and conceptually possible—​for law to provide such reasons. If the concern is with how the practices constituting something as a system of law are nomologically capable of generating something that rationally competent subjects like us can regard as new content-​independent motivating reasons to comply with such norms, there is nothing to explain. But the claim that it is possible in any of these senses for the practices constitutive of law to provide such motivating reasons does not imply the claim that they are even minimally likely to do so. It might be nomologically possible for something do so without its being even minimally likely to succeed in doing so because the relevant practices implicate a consideration that, while normatively relevant, has insufficiently normative force to do the job; an institutional normative system that rewards desired behavior with candy but does not punish undesired behavior is not minimally likely to succeed in preventing undesired behavior among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours—​though there is nothing in either our nature or that of the material nature of our universe that precludes it from doing so. What is nomologically possible and what is empirically probable are two different matters. The conceptual claim that law is, by nature, equipped to provide such motivating reasons is properly interpreted as expressing the stronger idea that the practices constituting something as a system of law are reasonably contrived to do so in the sense that they give rise to something that rationally competent subjects are characteristically likely, as a descriptive matter of contingent fact, to regard—​because they should do so, as an objective matter of normative practical rationality—​as defeasible novel content-​independent motivating reasons to refrain from prohibited acts. It would make no sense for us to use law as a means of regulating behavior, regardless of the ultimate end of doing so, unless the practices constituting something as a legal system are reasonably contrived in the sense that they are sufficiently likely, as a practical matter, to provide such motivating reasons. This should not be thought surprising. As discussed in Chapter 4, it is a conceptual truth that only things reasonably contrived to do what some artifact A is needed, characteristically used, and supposed to do, as a functionally normative matter, count as A. If (1) it is a conceptually necessary condition for something to count as an artifact-​type A that it is reasonably contrived to do what A is needed, characteristically used, and supposed to do, as a functionally normative matter; (2) it is conceptually true that legal systems are artifactual in character; and (3) something could not be reasonably contrived to do what legal systems are needed, characteristically used, and supposed to do, as a functionally normative matter, without being reasonably contrived to provide new defeasible content-​ independent objective motivating reasons to refrain from prohibited acts, then it is

124  Three Conceptual Problems of Legal Normativity a conceptually necessary condition for something to count as a legal system that it is reasonably contrived to provide novel defeasible content-​independent objective motivating reasons to refrain from prohibited acts.

4.  Three Conceptual Problems of Legal Normativity The normativity of law poses three challenging problems for a conceptual theory of its metaphysical nature.18 The first, and most important, of these problems is to explain how the practices constituting something as a system of law are reasonably contrived to create novel defeasible content-​independent objective motivating reasons to comply with some mandatory legal norms governing non-​official behavior (the How Problem). The second is to identify the order of the objective motivating reasons created by those practices; some theorists believe that those practices are reasonably contrived to create only first-​order motivating reasons to comply while others believe that they are also reasonably contrived to create second-​order exclusionary motivating reasons not to act on some class of first-​ order motivating reasons (the Order Problem). The third is to expose the content of the relevant objective motivating reason to comply that is supposed to motivate compliance among rationally competent subjects antecedently disposed to behave otherwise (the Content Problem); insofar as the practices constituting something as a system of law are reasonably contrived to make a difference in the way rationally competent subjects behave, it is in virtue of being reasonably contrived to create content that can motivate compliance among rationally competent subjects antecedently disposed to behave otherwise. This part of the volume addresses these problems. Chapter 6 addresses the How Problem, arguing that the most plausible account of how law as such creates objective motivating reasons to comply is by backing certain legal directives with sanctions for non-​compliance; the relevant practices give rise to something that rationally competent self-​interested subjects like us are characteristically likely to regard, as a descriptive matter of contingent fact, as defeasible motivating prudential reasons to conform to those norms because they should, as an objective matter of normative practical rationality, regard them as such motivating reasons. Chapter 7 addresses the Order Problem, arguing that it is not a conceptual truth that law provides rationally competent subjects of mandatory legal norms governing non-​official acts with second-​order objective exclusionary reasons to comply; the only reasons that the practices constituting something as a system of law are 18 Another problem related to legal normativity is concerned with explicating how the practices constitutive of a legal system create something properly characterized as a legal obligation, which is the characteristic mechanism through which law purports to provide objective motivating reasons to comply. This problem will be comprehensively addressed in the next volume of this series, entitled Authority and the Nature of Law.

Three Conceptual Problems of Legal Normativity  125 conceptually equipped to provide with respect to non-​official acts are first-​order objective motivating prudential reasons to comply that implicate a subject’s desire to avoid being subject to sanctions. Chapter 8 addresses the Content Problem, rejecting the idea that the content of the objective motivating reason to which the practices constituting something as a system of law are reasonably contrived to give rise is a reason to comply with mandatory legal norms because those norms have the status of law; the content of the objective motivating reason that law as such is equipped to provide, if the Coercion Thesis is true, is to conform to those norms to avoid coercive sanctions. But the arguments of these chapters go further; they comprise a critically important piece of the case for the Coercion Thesis insofar as they also show that the only plausible solutions to the conceptual problems of legal normativity presuppose the Coercion Thesis. This part of the volume argues not only that the Coercion Thesis can figure into a plausible explanation of law’s presumed conceptual normativity; it argues, further, that any conceptual theory of the metaphysical nature of law that does not include the Coercion Thesis cannot explain how the practices constituting something as a system of law are reasonably contrived to regulate behavior through the governance of norms metaphysically capable of guiding it by making a practical difference in the deliberations of rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours.

6

The Coercion Thesis and the How Problem of Legal Normativity It is a conceptual truth that the practices constituting something as a system of law are normative in the sense that they are reasonably contrived to give rise to new content-​independent defeasible objective motivating reasons to comply with at least those mandatory legal norms prohibiting behaviors that must be prohibited to ensure that a legal system can do anything it can be non-​idiosyncratically used to do. Insofar as law as such is reasonably contrived to make a practical difference with respect to what rationally competent subjects do, these motivating reasons must be reasonably contrived to rationally induce—​or persuade—​them to comply when antecedently disposed to do otherwise; this is the only way that the practices constituting something as a system of law can make the kind of difference they must be presumed reasonably contrived to make. Solving the How Problem of Legal Normativity is essential to vindicating law’s presumed conceptual normativity. The How Problem is concerned with explaining how it is that these constitutive practices are reasonably contrived to give rise to something rationally competent subjects characteristically regard, as a descriptive matter of contingent fact, as new content-​independent defeasible motivating reasons to comply with mandatory legal norms governing non-​official behavior because they should, as an objective matter of normative practical rationality, regard it as such. If these constitutive practices are not reasonably contrived to give rise to such reasons, then they are not reasonably contrived to do anything that law is needed, characteristically used, and supposed to do, as a functionally normative matter. The very rationality of adopting these social practices to regulate behavior turns on solving the How Problem. In this chapter, I argue that the only practices plausibly thought to be conceptually necessary features of a legal system that are minimally equipped to explain how the practices constituting something as a system of law are reasonably contrived to give rise to the right kind of motivating reasons to comply are those that back mandatory legal norms governing non-​official behavior with the threat of a coercive sanction. Apart from the existence of the legal norms authorizing courts to impose coercive sanctions for non-​compliance with the relevant mandatory legal norms, there is nothing in the practices constituting something as a system of law with the logical resources to explain law’s conceptual normativity. The Coercion Thesis is essential to vindicating both law’s conceptual normativity and the rationality of Coercion and the Nature of Law. Kenneth Einar Himma, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198854937.001.0001

128  The How Problem of Legal Normativity adopting systems of law to regulate behavior; there is simply nothing else in the conceptually necessary properties of a legal system equipped to explain how it gives rise to objective motivating reasons to comply.

1.  Vindicating the Coercion Thesis (1): Authorized Coercive Sanctions Solve the How Problem of Legal Normativity The Coercion Thesis makes possible a solution to the How Problem that is as intuitive as it is elegant: insofar as it is a conceptually necessary condition for the existence of a legal system that courts are authorized to impose coercive sanctions for non-​compliance with mandatory legal norms governing non-​official behavior, the conceptual capacity of a legal system to provide such motivating reasons is explained by the fact that courts are authorized to impose these sanctions for non-​ compliance; the fact that courts are authorized to impose them for non-​compliance gives rise to something that rationally competent self-​interested subjects like us are characteristically likely, as a descriptive matter of contingent fact, to regard as a new content-​independent defeasible motivating reason to comply because we should, as an objective matter of normative practical rationality, regard it as such.

1.1  The Threat of a Coercive Sanction Gives Rise to Motivating Prudential Reasons The practices authorizing courts to impose coercive sanctions on non-​complying subjects are intended to provide motivating reasons to comply by threatening detriment rationally competent self-​interested subjects are characteristically motivated to avoid. Rationally competent self-​interested subjects characteristically regard, as a descriptive matter of contingent fact, the legal possibility of being punished for doing something as a motivating reason not to do that something because they should, as an objective matter of normative practical rationality, regard it as a motivating reason not to do that something. Coercive sanctions like incarceration and financial penalties are reasonably contrived to give rise to something rationally competent subjects are characteristically likely to regard as motivating reasons to comply because they should regard them as such. The character of these motivating reasons is prudential. While one might have moral reasons for wanting to avoid being on the business end of coercive sanctions (such as might make reference to the interests of persons to whom one owes support obligations), the character of a motivating reason to abstain for self-​regarding considerations from something that triggers the imposition of painful detriment for non-​compliance, such as are derived from the desire to avoid incarceration, is

Vindicating the Coercion Thesis (1)  129 prudential; being incarcerated is not, absent exceptional circumstances, in anyone’s perceived self-​interest. The claim that a rationally competent subject regards something as a motivating reason to comply does not logically imply the claim that, as a descriptive matter of contingent fact, her compliance is explained by that reason. A motivating reason need not tip the agent’s view of what she is most motivated to do; most people would regard the prospect of life imprisonment as a persuasive motivating prudential reason not to commit murder but refrain from doing so for other motivating reasons. The point is not to explain why subjects comply; the point is rather to show that rationally competent self-​interested subjects characteristically regard, as a descriptive matter of contingent fact, the prospect of being on the business end of coercive sanctions as normatively relevant because they should, as an objective matter of normative practical rationality, regard it as normatively relevant—​even if they comply with the law for other reasons and even if they do not comply with the law. The motivating prudential reason to which the desire to avoid coercive sanctions gives rise is equipped to make a difference with respect to whether a rationally competent subject commits a proscribed act in circumstances where her motivations for disobedience are wholly prudential in character. A rationally competent subject motivated to do something for prudential reasons might not be as likely to be moved by the prudential costs of doing that something as she should be; however, she must be motivated to some extent by a desire to avoid the prudential costs insofar as she is moved by prudential considerations of any kind. Backing mandatory legal norms with the threat of a coercive sanction is thus well-​equipped to give rise to considerations that can make a practical difference in the deliberations of rationally competent subjects: rationally competent self-​interested subjects are characteristically likely, as a descriptive matter of contingent fact, to regard those considerations as motivating prudential reasons not to do something that triggers the imposition of coercive sanctions because they should, as an objective matter of normative practical rationality, regard them as such.

1.2  The Motivating Prudential Reasons Are Defeasible The claim that law’s conceptual normativity is explained by the motivating prudential reasons favoring compliance to which the authorization of coercive sanctions for non-​compliance give rise should not be thought to entail either (1) the claim that these motivating prudential reasons to comply are always regarded, as a descriptive matter of contingent fact, by rationally competent subjects as conclusive; or (2) the claim that they should always be regarded, as an objective matter of normative practical rationality, by rationally competent subjects as conclusive.

130  The How Problem of Legal Normativity Claim (1) is false; it is uncontentious that rationally competent subjects sometimes violate mandatory legal norms backed by coercive sanctions—​even when the cost is life imprisonment or execution. Someone who violates the law knowing that violations are punished by a legally authorized coercive sanction does not, as a descriptive matter of contingent fact, view the prospect of being subject to the relevant detriment as a conclusive reason not to violate the law; such a subject, when her violation is deliberative, believes, in effect, that the expected prudential benefits of violating the law outweigh the expected prudential costs of doing so—​ either because she believes that the magnitude of the sanction is mild relative to the magnitude of the expected benefits or because she believes that the non-​zero probability of being caught and sanctioned is sufficiently small to warrant, according to her subjective reckonings, breaking the law to realize the expected prudential value of doing so. But it is not just the obvious fact that rationally competent subjects sometimes violate legal norms that shows they do not always view the relevant motivating reasons as conclusive; it is also that they sometimes, as a descriptive matter of contingent fact, regard some other consideration as a conclusive motivating reason to abstain from doing something prohibited by law. Most of us are morally decent and do not need the criminal law’s threat of punishment to induce us to refrain from intentionally killing innocent people because we live in circumstances safe enough to afford us the psychological space to worry about moral or altruistic considerations in addition to the prudential considerations that would always take deliberative precedence in a Hobbesian state of nature where our interests are constantly under attack. Moral decency is the luxury that law is characteristically used to buy. Claim (2) is also false. To begin, rationally competent subjects sometimes have a motivating moral reason to disobey law that is conclusive, as an objective matter of normative practical rationality, all things considered. Officials had, as far as our shared judgments are concerned, a conclusive objective motivating moral reason to disobey the Fugitive Slave Act of 1850, which required them to capture and return persons who escaped from slavery to the persons legally regarded as their owners and was backed with a compulsory $1,000 fine.1 The motivating moral reasons to disobey the Fugitive Slave Act conclusively defeat, as an objective matter of normative practical rationality, any motivating prudential reasons officials might have had for compliance, including those arising from the threat of the authorized monetary sanction. Sometimes what we take to be the requirements of morality provide motivating moral reasons not to comply with a wicked law that defeat, as an objective matter of normative practical rationality, any countervailing motivating

1 See http://​avalon.law.yale.edu/​19th_​century/​fugitive.asp for the text of the Fugitive Slave Act, which also required non-​officials to cooperate with the efforts of officials to apprehend and return “fugitive slaves” to those who enslaved them.

Vindicating the Coercion Thesis (1)  131 prudential reasons one might have to comply—​including those to which the law gives rise. But subjects might also have a motivating prudential reason to disobey a mandatory legal norm backed by the threat of a coercive sanction that defeats, as an objective matter of normative practical rationality, any countervailing motivating prudential reasons to comply grounded in that threat. The claim that one should necessarily regard the prospect of being subject to coercive sanctions as a motivating prudential reason to comply that defeats any motivating prudential reasons not to comply entails the problematic claim that the expected prudential cost of being caught and subject to coercive sanctions necessarily exceeds the expected prudential benefits of committing the violation. If moral considerations are bracketed and the only relevant consideration is whether complying with the law maximally conduces to one’s self-​interest, one might have a winning reason to commit a violation if the expected benefits of breaking the law exceed the expected costs; if the probability of being caught is low enough and the probability of realizing a substantial benefit is high enough, a purely prudential calculus might dictate doing what must be done to realize that benefit. This is what generally explains the thinking of subjects who decide to violate the law after considering the risks of incurring the authorized coercive sanctions. Rationally competent subjects who deliberatively violate a mandatory legal norm do so, as a descriptive matter of contingent fact, because they believe that the probability of being caught and convicted is sufficiently low to warrant violating the law to achieve even benefits that are comparatively trivial from the standpoint of prudential rationality. A rationally competent subject who, after assessing the expected prudential costs and benefits, risks being incarcerated for shoplifting a $20 bottle of cologne has, absent unusual circumstances having to do with her mental well-​being, done so on the strength of a belief that the risks of stealing the cologne are not enough to offset the expected prudential benefits. There is nothing necessarily problematic about such thinking from an objective standpoint of purely prudential rationality. It might well be true that, given the probabilities of being caught and prosecuted, objective norms of prudential rationality dictate taking the risk to realize the expected prudential benefit of non-​ compliance because people frequently get away with even the most serious crimes. One recent study showed that more than 50% of 54,868 homicides committed in fifty-​five cities over a recent ten-​year period did not result in even an arrest,2 while another showed that the conviction rate for murder in recent years is about 70%3; if those numbers are accurate over the same range of cases, then the probability of

2 “Murder with Impunity,” Washington Post (July 24, 2018); available at: https://​www.washingtonpost. com/​graphics/​2018/​investigations/​unsolved-​homicide-​database/​. 3 Bureau of Justice Statistics, Office of Justice Programs; available at:  https://​www.bjs.gov/​index. cfm?ty=qa&iid=403.

132  The How Problem of Legal Normativity getting away with murder is about .66—​which is, from the standpoint of prudential rationality, not too bad if the expected benefits are thought sufficiently desirable. Insofar as it is less likely that one suffers substantial detriment from being apprehended for shoplifting, as seems plausible, objective norms of purely prudential rationality might dictate shoplifting something if that something is nice enough and there are no cameras or security guards in sight. Sometimes it might be in one’s prudential interests, as an objective matter of normative practical rationality, to commit a crime even when one knows one will be caught and punished. If all that P cares about in her deliberations is doing what maximally conduces to her self-​interest, P might have a conclusive prudential motivating reason to disobey the law even if she knows she will be caught and punished to the full extent of the law. If P knows that (1) she can embezzle $1,000,000,000 from the company for which she works, (2) the maximum penalty for embezzlement is six months of incarceration in the type of country-​club prison to which non-​violent wealthy white financial criminals are too frequently sentenced, and (3) there are no legal mechanisms that can induce her to disgorge or return embezzled funds, then P might have a motivating prudential reason to embezzle the money that objectively defeats, even on the assumption she will be caught and maximally punished, all countervailing motivating prudential reasons. It is thus false both that the coercive sanctions authorized by law are equipped, as a practical matter, to provide motivating reasons that are, as both a descriptive matter of contingent fact and an objective matter of normative practical rationality, either conclusive all things considered or conclusive with respect to just the class of motivating prudential reasons. It might be true that certain sanctions, such as execution, are intended to provide a motivating prudential reason to comply that necessarily defeats countervailing motivating prudential reasons not to comply; however, whether or not a sanction provides such a reason depends on a host of circumstances, including the probability of incurring the sanction and the severity of the sanction. One of the most complex problems any society has to resolve to ensure that law efficaciously regulates non-​official behavior is deciding on coercive sanctions that are sufficiently unpleasant to deter violations of crime but not so severe that they shock the conscience enough to provoke mass resistance—​and that is a problem that varies from society to society because the solution depends on how ordinary people steeped in the cultural norms of the particular society are likely to respond to the various possibilities. The motivating prudential reason to comply to which the authorized threat of coercive sanctions gives rise is defeasible insofar as rationally competent subjects do not, and should not, characteristically regard it as prudentially conclusive—​ even when they are not antecedently motivated to commit the acts triggering their authorized imposition. Insofar as rationally competent subjects take this threat into account in their deliberations about whether to comply, they are likely to regard it, as a descriptive matter of contingent fact, as giving rise only to defeasible

Vindicating the Coercion Thesis (1)  133 motivating prudential reasons to comply because they should regard it, as an objective matter of normative practical rationality, as giving rise only to such reasons. The practices constituting something as a system of law, insofar as they include practices authorizing coercive sanctions for non-​compliance, are therefore equipped to give rise only to motivating reasons for compliance that are defeasible in character; no matter how severe the threatened detriment might be or the probability of being punished with it might be, one might have a prudential motivating reason to run the risk of incurring such sanctions to achieve the expected prudential benefit of non-​compliance.

1.3  The Motivating Prudential Reasons Are Objective Insofar as the practices constituting something as a system of law are equipped, as a practical matter, to regulate behavior though the governance of norms metaphysically capable of guiding it, those practices must be equipped to persuade rationally competent subjects antecedently disposed to violate legal norms to refrain from doing so enough of the time, at the very least, to permit subjects to live together in a stable community. To be properly equipped to do this, those practices must give rise to something that rationally competent subjects are characteristically likely, as a descriptive matter of contingent fact, to regard as motivating reasons not to commit acts that are prohibited because they are deemed undesirable. Backing court orders and valid mandatory legal norms governing non-​official behavior with sufficiently stiff sanctions is well-​equipped to do exactly this. Rationally competent self-​interested subjects, whether good or bad, are likely to regard the legal possibility of being punished by incarceration as normatively relevant in deciding whether to commit a non-​complying act. But even indecent subjects who are not ultimately deterred by the threat of coercive sanctions consider it normatively relevant in their deliberations about what to do; insofar as someone takes precautions to avoid getting caught committing a crime, she is treating the threat of being on the business end of sanctions as normatively relevant. Morally decent agents who do not need to be deterred from committing crimes by the threat of a coercive sanction likewise recognize it as normatively relevant insofar as they would take it into account in deciding whether to commit a crime if they were not antecedently indisposed to violate the law. The reason that rationally competent self-​interested subjects are likely, as a descriptive matter of contingent fact, to regard the legal possibility of being punished for non-​complying behavior as normatively relevant is that they should, as an objective matter of normative practical rationality, regard it as normatively relevant. The threat of coercive sanctions is likely to be regarded by rationally competent self-​interested subjects as giving rise to a motivating prudential reason to

134  The How Problem of Legal Normativity comply precisely because it gives rise to an objective motivating prudential reason to comply. So obviously relevant is the threat of incarceration and punitive fines, as an objective matter of normative practical rationality, that we would consider the fact that someone does not regard the possibility of being subject to such punishments as being even minimally relevant in her deliberations about whether to commit a non-​complying act as a prima facie reason to worry about whether she is suffering from a mental health condition that impairs the operation of her rational faculties. It is one thing to take a chance with the painful consequences of getting caught to realize some expected prudential benefit; it is another thing not to care at all about whether something that is objectively bad will happen to you if you do. The law is concerned to regulate the behavior of subjects who are rationally competent in virtue of being characteristically responsive to objective norms of practical rationality by providing something they are likely, as a descriptive matter of contingent fact, to regard as normatively relevant in deliberating about whether to do something that violates the law precisely because they are rationally competent and hence should, as an objective matter of normative practical rationality, regard it as normatively relevant.4 Since someone is rationally competent only insofar as she characteristically deliberates about what to do in a manner that largely conforms to objective norms of practical rationality, the law is equipped to efficaciously regulate the behavior of rationally competent subjects only insofar as the practices constitutive of a legal system are equipped to give rise to objective motivating reasons to comply.

1.4  The Objective Motivating Prudential Reasons Are Novel and Content-​Independent The relevant objective motivating reasons are both novel and content-​independent. They are novel in virtue of being created by the legal practices authorizing the imposition of coercive sanctions for non-​compliance. Without those practices, a rationally competent subject of a legal norm does not have this particular objective motivating prudential reason to comply—​though she might have others; the point 4 Rational competence also implicates objective norms of epistemic rationality. Someone who suffers from delusions and characteristically forms false beliefs on the basis of such delusions is not rationally competent—​regardless of whether she remains capable of deliberating about what to do in a manner that conforms to objective norms of practical rationality. If it is possible to be practically competent in some sense while being epistemically incompetent for reasons of mental illness, then there are two ways that the behavior of a subject cannot be norm-​governed: either because she is chronically unable to accurately discern what she should do or because she is chronically unable to do what she believes she should do. My guess is that a serious mental health condition that implicates one’s competence involves both disabilities in combinations that unpredictably affect behavior.

Vindicating the Coercion Thesis (1)  135 of backing the relevant directives with the threat of a coercive sanction is precisely to provide a new objective motivating prudential reason to comply in the form of a painful deterrent to non-​complying acts. While one might want morally problematic criminal acts punished for retributive reasons, a legal system can succeed in regulating the behavior of rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours, regardless of the ultimate end of doing so, only insofar as it deters enough acts deemed undesirable to enable us to live together. Our most basic reason for wanting a legal system is prudential: we know that rationally competent self-​interested subjects like us cannot live together and benefit from social cooperation in worlds of acute material scarcity like ours without an institutional normative system that efficaciously prohibits certain undesirable acts, and it is in everyone’s prudential interest that enough of these acts are prevented to enable us to live together in a community. From the standpoint of our need for an institutional normative system to regulate behavior, deterrence is the point of backing mandatory norms prohibiting assaults on persons and property. Retributive motives are secondary insofar as one cannot effectively deter people from committing prohibited acts if punishment is routinely imposed on people who have not been shown to have committed one of the proscribed acts. Deterrence is the basic point of any institutional normative system that is supposed to, as a functionally normative matter, enable us to live in comparative peace with one another. One might think that mandatory legal norms prohibiting undesirable acts that are independently prohibited by the minimum content of natural law are metaphysically incapable of providing new content-​independent objective reasons because there could not be a legal system that does not include such norms.5 It is a conceptual truth that a legal system can provide a new content-​independent objective motivating reason to do something only insofar as it can provide a new content-​independent objective motivating reason not to do that something. Since, on this line of reasoning, it is not conceptually possible for officials to adopt mandatory norms properly characterized as law that require rationally competent subjects to do what is prohibited by the minimum content of natural law, it is not conceptually possible for there to be a legal system that provides objective motivating reasons to do things violating the minimum content of natural law. But if it is not conceptually possible for there to be a legal system that provides content-​ independent objective motivating reasons to do things violating the minimum content of natural law, then it is not conceptually possible for there to be a legal system that provides content-​independent objective motivating reasons not to do things violating the minimum content of natural law. 5 The minimum content of natural law consists of mandatory moral norms prohibiting acts that adversely impact interests of subjects that are vital to well-​being—​like assaults on persons and property.

136  The How Problem of Legal Normativity The problem with this line of objection is that it is not a conceptually necessary condition for the existence of a legal system that it prohibits every undesirable act prohibited by the minimum content of natural law; the claim that it is a conceptually necessary condition for something to count as a system of law that it prevents enough undesirable acts to allow us to live together does not entail the claim that it is a conceptually necessary condition for something to count as a system of law that it prohibits all such undesirable acts. While it is plausible to think that we could not live together if law did not prohibit some of the acts prohibited by the minimum content of natural law, the claim that we could not live together if law does not prohibit all of these acts is demonstrably false: there have obviously been many legal systems that coercively require what is prohibited by the minimum content of natural law—​such as the Fugitive Slave Act discussed above. An institutional normative system that seeks to regulate behavior as broadly as law does could not be efficacious in doing so if it generally required or allowed too many acts of violence. There is only so much violence compatible with the existence of a stable legal system that succeeds in regulating behavior through the governance of norms metaphysically capable of guiding it. The more violence a system of law permits or requires, the less likely it is that the system will succeed in enabling rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours to live together in something that counts as a community. Such a system would not be one of law because it would likely generate so much “legal” violence and illegal violence in the form of resistance and possibly outright rebellion that it could not prevent enough undesirable acts to enable beings like us in worlds like ours to live together in a community—​and not because it is impossible for artifactual norms to give rise to defeasible objective motivating reasons to perform undesirable acts breaching the peace. But insofar as it is conceptually possible for an institutional normative system to provide defeasible objective motivating prudential reasons to perform undesirable acts prohibited by the minimum content of natural law, such as was true of the Fugitive Slave Act, it is conceptually possible for a legal system to provide defeasible content-​independent objective motivating prudential reasons to abstain from such acts. Insofar as the reason for performing those acts derives from the fact that some artifactual body, which is properly characterized as a lawmaking or adjudicative body in virtue of being part of a legal system, has recognized, applied, and enforced a norm requiring such acts, the fact that some artifactual body has recognized, applied, and enforced a norm prohibiting such acts would also give rise to a new objective motivating prudential reason to abstain from such acts that has nothing to do with the content-​based considerations giving rise to conclusive objective motivating moral reasons to abstain from them. The source of the new objective motivating reason to comply with such norms is the fact that courts are authorized to impose a coercive sanction for non-​ compliance. The relevant motivating reasons are content-​independent because

Vindicating the Coercion Thesis (1)  137 they ultimately depend on whether the content of the particular norm is enforced by a coercive sanction—​and not on whether that content is recognized as law. The only objective motivating reasons to which the practices constituting something as a system of law can, by themselves, give rise are content-​independent reasons; a legal system cannot, as a conceptual matter, create any new content-​based reasons to do something that one does not already have. Law is just not that kind of thing.

1.5  The Relevant Practices Are Reasonably Contrived to Provide Objective Motivating Prudential Reasons to Comply On the assumption that law is conceptually normative, it is a conceptually necessary condition for something to count as a system of law that it is, in some sense, capable of giving rise to objective motivating reasons to comply with mandatory norms governing non-​official behavior. An institutional system of regulation is not properly characterized as normative if it is not capable in the relevant sense of giving rise to considerations that rationally competent subjects should, as an objective matter of normative practical rationality, regard as a motivating reason to comply; that is just an obvious fact about the concept of normativity as it straightforwardly applies to institutional systems of behavioral regulation. The Coercion Thesis explains how the practices constitutive of law are equipped to regulate non-​official behavior in virtue of being capable of giving rise to such reasons. The operative claim is not that it is logically, metaphysically, conceptually, and nomologically possible for these practices to give rise to something that rationally competent subjects characteristically regard, as a descriptive matter of contingent fact, as a new content-​independent motivating prudential reason to comply because they should, as an objective matter of normative practical rationality, regard it as such. It is trivially true that it is possible in all of these respects for these practices to give rise to such reasons—​regardless of whether the Coercion Thesis is true. The operative claim is rather that these practices—​construed to include legal norms governing official behavior that authorize courts to impose coercive sanctions for non-​compliance with mandatory legal norms governing non-​official behavior—​are reasonably contrived to produce something that rationally competent subjects characteristically regard, as a descriptive matter of contingent fact, as a new content-​independent motivating prudential reason to comply because they should, as an objective matter of normative practical rationality, regard it as such a reason. It is the fact that courts are authorized to impose painful detriment in response to non-​compliance that constitutes these practices as reasonably contrived to give rise to such motivating reasons. Further, it is precisely this feature of these practices that constitutes an institutional normative system as something that can be used, as a practical matter, by

138  The How Problem of Legal Normativity rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours to enable us to live together in comparative peace. Rationally competent subjects who are sometimes moved to commit undesirable acts for prudential motivating reasons can be dissuaded from doing so by countervailing prudential motivating reasons. Backing mandatory norms with the threat of a coercive sanction is exactly the kind of thing that must be done by an institutional normative system to give rise to the kind of motivating prudential reasons to comply that rationally competent self-​interested subjects are characteristically likely, as a descriptive matter of contingent fact, to regard as outweighing countervailing motivating prudential reasons because they should, as an objective matter of normative practical rationality, regard them as such. These enforcement mechanisms are not, then, just capable of providing new defeasible objective motivating reasons for compliance that can deter violations of mandatory legal norms governing non-​official behavior prohibiting undesirable acts; they are reasonably contrived to do so in the sense that they are reasonably likely to succeed in minimally enabling us to live together in comparative peace by reducing the incidence of such acts. The practices constituting something as a system of law, construed to include legal norms governing official behavior that authorize courts to impose non-​trivial detriment to deter non-​ compliance with the relevant mandatory legal norms governing non-​official behavior, are better tailored to provide such reasons for rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours than any other available institutional mechanism of which we know. The Coercion Thesis makes it possible to explain what, as will be argued below, cannot be explained by any remotely plausible theory of the metaphysical nature of law that denies that thesis.

1.6  Conclusion: The Coercion Thesis Explains Law’s Presumed Conceptual Normativity The How Problem of Legal Normativity is trivially easy to solve with the help of the Coercion Thesis. If the Coercion Thesis is true, then it is the practices authorizing courts to punish non-​compliance with something rationally competent self-​ interested subjects should regard as normatively relevant detriment that explain how the practices constituting something as a system of law are reasonably contrived to give rise to novel content-​independent defeasible objective motivating reasons to comply. Insofar as any institutional normative system backs its mandatory norms with the threat of something likely to be experienced by rationally competent self-​interested subjects as detriment, it creates circumstances that give rise to something they are likely, as a descriptive matter of contingent fact, to regard as a new defeasible content-​independent motivating prudential reason to comply

Vindicating the Coercion Thesis (2)  139 because they should, as an objective matter of normative practical rationality, regard it as such.

2.  Vindicating the Coercion Thesis (2): There Is No Other Plausible Explanation for Law’s Conceptual Normativity Any plausible theory of the metaphysical nature of law must be able to explain law’s conceptual normativity. It is impossible to make sense of why we characteristically adopt the practices constituting something as a system of law to regulate non-​official behavior unless those practices give rise to something that rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours characteristically regard, as a descriptive matter of contingent fact, as a new content-​independent defeasible motivating reason to comply with mandatory legal norms governing that behavior because we should, as an objective matter of normative practical rationality, regard it as such. Any theory of the metaphysical nature of law, then, that lacks the resources to explain how law is reasonably contrived to give rise to such objective motivating reasons is problematic for that reason—​either because it is false or because it is incomplete. This section argues that no other conceptually necessary property of a legal system can explain how the practices constituting something as a system of law are reasonably contrived to give rise to new objective motivating reasons to comply with mandatory legal norms governing non-​official behavior. The only other properties plausibly thought conceptually necessary for an institutional normative system to count as one of law are as follows: (1) the system must satisfy Hart’s minimum conditions for the existence of a legal system; (2) it must contain norms governing non-​official behavior that prohibit acts violating the minimum content of natural law; (3) it must be metaphysically capable of morally legitimate authority; and (4) it must regulate behavior by the recognition, application, and enforcement of norms metaphysically capable of motivationally and epistemically guiding behavior. I argue that none of those properties is reasonably contrived to give rise to new objective motivating reasons to comply.

2.1  The System Satisfies Hart’s Minimum Conditions for the Existence of a Legal System H.L.A. Hart argues that there are two “minimum” conditions that an institutional normative system S must satisfy to count as one of law: (1) officials of S must converge in accepting and complying with the requirements of a conventional rule of recognition governing official behavior that defines the recipes for making, changing, and adjudicating what counts as valid law in S; and (2) the behavior of

140  The How Problem of Legal Normativity rationally competent subjects of S generally satisfies the valid mandatory norms of S governing non-​official behavior. Neither of these conditions constitutes a legal system as reasonably contrived to give rise to considerations rationally competent subjects should regard as new content-​independent motivating reasons in virtue of being characteristically responsive to objective norms of practical rationality. As to condition (1), there is nothing in the fact that some group of officials converge in accepting and complying with a conventional rule of recognition governing official behavior that, by itself, gives rise to something that rationally competent subjects should regard as an objective motivating reason of any kind to comply with valid norms governing non-​official behavior. While it might be true that a morally legitimate legal system gives rise to objective motivating moral reasons to comply with such mandatory norms in virtue of creating official procedures for making, changing, and adjudicating mandatory legal norms that are morally just, a legal system is legitimate only in virtue of instantiating moral properties that distinguish it from conceptually possible legal systems that are not legitimate. As to condition (2), the claim that the behavior of subjects generally satisfies the valid mandatory norms of the system explains nothing that needs to be explained to vindicate law’s conceptual normativity. The claim that the behavior of subjects generally satisfies the valid norms of the system tells us nothing about why subjects behave the way they do and hence cannot tell us anything about how the practices constituting something as a system of law are reasonably contrived to induce rationally competent subjects to comply when antecedently disposed to do otherwise by providing objective motivating reasons to comply. It might be, for all we can tell, that the conforming behavior of subjects has nothing at all to do with the requirements of the system; they just happen to abstain from acts they regard as morally problematic because they are simply decent people. While such a state of affairs is not reasonably likely to obtain among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours, the probability of such a state of affairs is nonetheless non-​zero. Neither condition (1) nor condition (2), then, is logically equipped to explicate law’s conceptual normativity.

2.2  The System Must Contain Norms Governing Non-​Official Behavior that Reproduce the Minimum Content of Natural Law Hart argues that it is a naturally necessary condition for the existence of a system of municipal law, given the psychological propensities of rationally competent self-​ interested subjects like us in worlds of acute material scarcity like ours, that it includes mandatory legal norms incorporating the minimum content of the natural

Vindicating the Coercion Thesis (2)  141 law—​in essence, because the peace cannot be kept among rationally competent subjects like us in worlds like ours without such norms: In considering the simple truisms which we set forth here, and their connection with law and morals, it is important to observe that in each case the facts mentioned afford a reason why, given survival as an aim, law and morals should include a specific content. The general form of the argument is simply that without such a content laws and morals could not forward the minimum purpose of survival which men have in associating with each other. In the absence of this content men, as they are, would have no reason for obeying voluntarily any rules.6

Hart realized that the only way that the practices constituting something as a system of law can be reasonably contrived, as a conceptual matter, to keep the peace among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours is by including mandatory norms prohibiting acts likely to create or escalate conflicts that can threaten the community, such as are prohibited by the minimum content of natural law. It makes no sense for us to adopt the practices constituting something as a system of law as a means of enabling us to live together in comparative peace if the norms recognized and applied as law do not include mandatory norms prohibiting behaviors likely to lead to breaches of the very peace that these constitutive practices must be able to prevent in order to do anything else that law can be non-​idiosyncratically used to do. But there is nothing in the fact that officials of the system recognize and apply valid mandatory norms prohibiting such behaviors that, by itself, would explain how the practices constituting something as a system of law are reasonably contrived to give rise to something that rationally competent subjects are characteristically likely, as a descriptive matter of contingent fact, to regard as a new motivating reason to abstain from such behaviors because they should, as an objective matter of normative practical rationality, regard it as such. Legal systems are needed to make it more likely that subjects abstain from such acts because rationally competent subjects do not always abstain from doing what they antecedently have conclusive objective motivating moral reason to abstain from doing: any rationally competent subject can be presumed to know that killing an innocent person to take her money is wrong; yet rationally competent subjects commit a theoretically significant number of those murders every year.

6 H.L.A. Hart, The Concept of Law (Clarendon Press 1994) 1993. Hart also believed, for similar reasons, that the authorization of coercive sanctions was a naturally necessary feature of a legal system. Hart’s opposition to the Coercion Thesis was ultimately grounded in two mistakes: (1) thinking that the U.N. Charter norms of international regulation are not backed up by something that counts as a coercive sanction; and (2) thinking that the Coercion Thesis implies the claim that every mandatory legal norm must be backed by the threat of a coercive sanction, as Austin thought.

142  The How Problem of Legal Normativity But the fact that officials recognize and apply mandatory legal norms incorporating the minimum content of natural law does nothing that would give rise to new objective motivating reasons not to commit such acts. A rationally competent subject who is not antecedently motivated by the underlying moral considerations to abstain from such acts is not, as a descriptive matter of contingent fact, likely to be persuaded to do so simply because some arbitrary group of persons serving as officials happen to recognize and apply certain norms prohibiting those acts as law. Nor should rationally competent subjects be persuaded to comply with these norms, as an objective matter of normative practical rationality, simply because the persons who serve as officials recognize and apply those norms as law. While it might be true that officials in morally legitimate legal systems objectively deserve such deference, there is no motivating reason, as an objective matter of normative practical rationality, to defer to officials without any regard for the moral quality of those practices. There is nothing in objective norms of practical rationality that would require that rationally competent subjects defer to the directives of someone simply because she serves as an official in some legal system because those directives might require doing something that should not be done, as an objective matter of normative practical rationality.

2.3  The System Must Be Metaphysically Capable of Instantiating Morally Legitimate Authority The truism that every conceptually possible legal system is metaphysically capable of instantiating legitimate authority entails the claim that every conceptually possible legal system has properties that constitute it as metaphysically capable of instantiating such authority. If the so-​called service conception of authority is true, then it is a conceptual truth that the valid norms of anything counting as a system of law are (1) metaphysically capable of expressing the authority’s view about what subjects ought to do according to the requirements of right reason; (2) metaphysically capable of replacing the judgment of subjects in their practical deliberations about what they should do according to the requirements of right reason; and (3) framed in a manner metaphysically capable of making it more likely that subjects comply with the requirements of right reason if they follow the judgment of officials with respect to what right reason requires by doing what the norms require than if they follow their own judgments with respect to what it requires.7

7 See Joseph Raz, “Authority, Law, and Morality” reprinted in Raz, Ethics in the Public Domain (Clarendon Press 1996). For a critique of the service conception of authority, see K.E. Himma, Morality and the Nature of Law (Oxford University Press, 2019), Chapter 6.

Vindicating the Coercion Thesis (2)  143 At the outset, it is worth noting that the claim that something must be metaphysically capable of being morally legitimate to count as a legal system asserts nothing stronger than that it must be the right kind of thing to count as a legal system. As Joseph Raz explains the relevant sense of capability here: [W]‌hat cannot communicate with people cannot have authority over them. Trees cannot have authority over people. . . . If I say that trees have authority over people, you will know that either my grasp of the concepts of authority or of trees is deficient or that I am trying to deceive.8

Trees cannot count as legal systems because they are metaphysically incapable of authority in virtue of lacking the conceptual prerequisites for being the kind of thing that can have authority and hence count as a legal system; something that cannot communicate with rationally competent subjects is metaphysically incapable of having authority over them.9 The claim that a system of law is the kind of thing that is metaphysically capable of instantiating morally legitimate authority tells us nothing that would even begin to explain how the practices constituting something as a system of law are reasonably contrived to give rise to objective motivating reasons to comply with mandatory legal norms governing non-​official behavior. While it is might be true that only things metaphysically capable of moral legitimacy are metaphysically capable of giving rise to new objective motivating reasons to comply, the latter claim likewise tells us nothing that matters here. What matters is how likely it is that something gives rise to objective motivating reasons to comply—​and not whether it is simply metaphysically possible for something to do so. It should be clear, moreover, that none of the three conditions described above that flesh out the service conception of authority contributes anything that could help to explain how the practices constitutive of a legal system are reasonably contrived to give rise to new objective motivating reasons to comply. An institutional system of norms satisfying Hart’s minimum conditions for the existence of a legal system that consists of norms satisfying the three conditions above simply articulates its requirements with sufficient clarity that rationally competent subjects can understand what those norms require without having to decide for themselves what the norms should require according to right reason.

8 Raz, “Authority, Law, and Morality” (n. 7) at 217. 9 There are hence both non-​moral (i.e. conceptual) and moral requirements for something to be morally legitimate. A tree is the wrong kind of thing, as a conceptual matter, to instantiate morally legitimate authority; it necessarily lacks the right properties to be a practical authority of any kind. In contrast, a totalitarian legal system does the wrong things, as a moral matter, to instantiate morally legitimate authority; it contingently lacks the necessary properties insofar as it could be legitimate if it were to do the right things. The former disability is metaphysical and necessary while the latter disability is moral and contingent.

144  The How Problem of Legal Normativity But the claim that an institutional normative system with these properties articulates its requirements with such clarity, by itself, entails nothing that would enable us to explain how the practices constituting a system of law are reasonably contrived to give rise to new objective motivating reasons to comply; the officials who manufacture and apply legal norms, after all, might allow what right reason prohibits or prohibit what right reason requires. There is simply nothing in the conceptual claim that the norms of a legal system are sufficiently clear to constitute the system as metaphysically capable of instantiating morally legitimacy that, by itself, has the resources to explain how the practices constituting something as a system of law are reasonably contrived to give rise to new objective motivating reasons to comply.10

2.4  The System Must Include Norms Metaphysically Capable of Motivationally and Epistemically Guiding Behavior It is a conceptual truth that every possible legal system includes some mandatory norms metaphysically capable of motivationally and epistemically guiding the behavior of rationally competent subjects.11 To be motivationally guided by a norm n is to conform to n because n is a norm that governs one’s conduct; that is, a mandatory norm n motivationally guides a person P if and only if P’s conformity to n is motivated by the fact that n is a valid mandatory norm of the system. To be epistemically guided by a mandatory norm n is to conform one’s behavior to the requirements of n after one learns what n requires by consulting the norm; that is, a mandatory norm n epistemically guides P if and only if P learns of her obligations under n by consulting n and then satisfies those obligations under n. But the claim that a legal system must include mandatory norms metaphysically capable of motivationally and epistemically guiding behavior does not entail anything about what a system including them is reasonably contrived to do. To begin, the claim that an artifact must be metaphysically capable of doing something does not imply that it is reasonably contrived to do that something: something must be metaphysically capable of being used as a murder weapon to count as a chair because chairs must be solid and movable; however, that does not imply 10 It might be true that a legal system known by its subjects to be morally legitimate gives rise to new objective motivating moral reasons to comply with its requirements but that is of no help. It is clearly possible to have legal systems that are morally illegitimate in virtue of requiring subjects to do what is morally odious, such as is true of totalitarian legal systems that systematically violate human rights. We simply cannot learn anything about the nature of law or legal authority, if understood to allow for the conceptual possibility of morally illegitimate legal authority, by looking at either the moral or non-​ moral properties a legal system must instantiate to be morally legitimate. 11 For a discussion of the notions of motivational and epistemic guidance, see Scott Shapiro, “On Hart’s Way Out” Legal Theory, vol. 4, no. 4 (1998) 469–​507. For a critique of the claim that an inclusive legal system cannot motivationally guide the behavior of subjects, see Himma, Morality and the Nature of Law (n. 7) Chapter 7.

Vindicating the Coercion Thesis (2)  145 that it must be reasonably contrived to be used as a murder weapon to count as a chair. The claim that some piece of an artifact (in this case, the norms of a system) must be metaphysically capable of doing something likewise does not imply that the artifact itself (in this case, the system) is reasonably contrived to do that something: some pieces of a chair must be metaphysically capable of being used as a murder weapon in order for a chair to be reasonably contrived to support a person’s weight; however, that does not imply that something must be reasonably contrived to be used as a murder weapon to count as a chair. Neither claim about a mandatory norm’s capacity to guide the behavior of a rationally competent subject tells us anything about whether it gives rise to an objective motivating reason to comply. To begin, the claim that it is metaphysically possible for a rationally competent subject to deliberatively conform her behavior to the requirements of a mandatory norm because she accepts the norm as governing her behavior entails nothing with respect to whether she should, as an objective matter of normative practical rationality, regard it as giving rise to a new motivating reason to do what is required by the norm; at most, it shows that it is a conceptually necessary condition for something to count as a mandatory norm governing the behavior of a class of subjects that it is metaphysically possible for rationally competent subjects to treat it as giving rise to a new motivating reason to do what it requires—​which seems to express a trivial conceptual truth about norm-​governance. Further, the claim that it is metaphysically possible for a rationally competent subject to conform her behavior to a mandatory norm after learning of its requirements by consulting the norm likewise entails nothing with respect to whether she should regard it, as an objective matter of normative practical rationality, as giving rise to a new motivating reason to do what is required by the norm; at most, this claim shows that it is a conceptually necessary condition for something to count as a mandatory norm governing the behavior of a class of rationally competent subjects that it is metaphysically possible for them to learn what it requires by consulting it.

2.5  Conclusion: Only the Coercion Thesis Can Explain Law’s Presumed Conceptual Normativity There is no other property plausibly thought conceptually necessary for the existence of a legal system that is even remotely equipped to help explain how the practices constituting something as a system of law are reasonably contrived to give rise to new objective motivating reasons of any kind. As far our empirically contingent linguistic practices with respect to the descriptive concept of law are concerned, the properties discussed in this section, assuming that the Coercion Thesis is false, exhaust the properties that something must have to count as a system of

146  The How Problem of Legal Normativity law; and none of them has the resources to explain how the practices constituting something as a system of law are reasonably contrived to give rise to new objective motivating reasons to comply with mandatory legal norms governing non-​official behavior. This assumes that there are no conceptually necessary moral constraints on the content of law; however, even if there were such conceptual constraints on what counts as law, they add nothing that would help us to explain how the practices constituting something as a system of law are reasonably contrived to give rise to new objective motivating reasons to comply—​for the same reason that the claim that a legal system must incorporate the minimum content of natural law does not. The fact that courts treat a mandatory moral norm as law does nothing that would equip the system to give rise to a new objective motivating reason to comply; the fact that the law prohibits killing innocent persons, by itself, does nothing to augment the objective motivating moral reasons that rationally competent subjects antecedently have not to kill innocent persons.12 The Coercion Thesis, thus, exclusively vindicates law’s presumed conceptual normativity. There is nothing else in any plausible candidate for a conceptually necessary property of a legal system, singly or together, that is reasonably contrived to give rise to new objective motivating reasons to comply with mandatory legal norms governing non-​official behavior. Only the Coercion Thesis has the resources to explain how the practices constituting something as a system of law are reasonably contrived to give rise to such motivating reasons.

3.  Are Objective Motivating Prudential Reasons the Wrong Kind of Reason to Solve the How Problem of Legal Normativity? One might object to the analysis above that the kind of objective motivating prudential reasons to which the threat of coercive sanctions gives rise cannot ground a plausible solution to the How Problem. Raz argues that prudential reasons are the wrong kind of objective motivating reasons to explain law’s presumed conceptual normativity because they are “partial” reasons that depend for their normative force on the subject’s having an antecedent desire to avoid sanctions:

12 Nor does it help to aggregate all of these other properties. Insofar as each of them is normatively inert, the aggregation of them is also normatively inert. Aggregating a number of properties that can contribute nothing to an explanation as to how the practices constitutive of a legal system are reasonably contrived to give rise to new objective motivating reasons to comply can do nothing, taken together, to explain this conceptual feature of law. It is analogous to repetitively adding zero to zero: no matter how many times one adds another zero, the sum will always be zero.

Are Prudential Reasons the Wrong Kind?  147 The fact that a law of a system in force provides for a sanction is no doubt a reason for action, but it is a reason of the wrong kind. A sanction-​backed norm is at best only an auxiliary partial reason. The complete reason must include the agent’s desire to avoid the sanction or the fact that it is against his interests for it to be applied to him. This is the operative reason: it is because the agent wants to avoid the sanction that he has reason to take notice of law.13

On Raz’s view, everything that counts as a mandatory norm of any kind, including those of law, is a complete operative motivating reason to comply in the sense that it is a reason to comply that contains all of the relevant factors needed to properly motivate a subject to comply. Raz’s concern here is with objective motivating reasons, and not subjective motivating reasons, since he believes that it is the metaphysical nature of a mandatory norm that it is a complete operative motivating reason to comply. Since it is clearly not true that rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours characteristically treat, as a descriptive matter of contingent fact, anything that counts as a mandatory norm as either being or giving rise to a complete operative motivating reason to comply, he is properly construed as making a point only about how mandatory norms should, as an objective matter of normative practical rationality, function in deliberations about what to do: rationally competent subjects should, on this view, treat mandatory norms as complete operative motivating reasons to comply. Raz’s claim here should not be construed as asserting that everything that has the logical form of a mandatory norm should, as an objective matter of normative practical rationality, be treated by rationally competent subjects as complete operative motivating reasons; the claim must be construed as the weaker and more plausible one that all and only mandatory norms that are objectively valid in some respect should be treated by rationally competent subjects as complete operative motivating reasons. Regardless of what Raz might intend here, it is straightforwardly false that something with the form of a mandatory norm that purports to require someone to torture babies gives rise—​simply in virtue of having the form of a mandatory norm—​to something that rationally competent subjects should, as an objective matter of normative practical rationality, treat as a complete operative motivating reason to torture them. It is not preposterous to think that valid mandatory moral norms give rise to objective complete operative motivating reasons to comply. The claim that something having the form of a mandatory norm that requires rationally competent subjects to torture babies does not give rise to such reasons relies on the intuition that someone could not have an objective complete operative motivating reason to



13

Joseph Raz, Practical Reason and Norms (Princeton University Press 1990) 161.

148  The How Problem of Legal Normativity do something clearly prohibited by mandatory moral norms simply because doing that something is required by something else with the form of a mandatory norm. While one might have a defeasible objective motivating prudential reason to torture even the cutest baby if, say, one is threatened with torture for non-​compliance, that is a reason grounded in objective norms of purely prudential rationality; new defeasible objective motivating reasons to do a of any kind cannot be derived, as an objective matter of normative practical rationality, from just the fact that something having the form of a mandatory norm requires doing a. This much is certainly right about Raz’s view:  rationally competent subjects should not, as an objective matter of normative practical rationality, treat the threat of a coercive sanction for non-​compliance as a complete operative motivating reason to comply; however, there is no plausible candidate for a conceptually necessary condition for the existence of a legal system that is remotely equipped even to contribute to an explanation of how the practices constituting something as a system of law are reasonably contrived to provide objective complete operative motivating reason to comply. As was shown in Section 2, there is nothing in any plausible candidate for a conceptually necessary property of a legal system, assuming the Coercion Thesis is false, that is reasonably contrived to give rise to objective motivating reasons of any kind.14 Beyond this, it is quite puzzling as to why Raz would think, apart from his idiosyncratic analysis of a mandatory norm, that the objective motivating prudential reasons to which backing a norm with the threat of a coercive sanction gives rise is the wrong kind of reason to vindicate law’s presumed conceptual normativity. If the problem is to explain how the practices constituting something as a system of law are reasonably contrived to give rise to normative considerations that make it more likely than otherwise that the system succeeds in doing anything that a legal system can be non-​idiosyncratically used to do, prudential reasons are obviously up to the job. As Raz himself concedes, “[t]‌he fact that a law of a system in force provides for a sanction is no doubt a reason for action” because “the agent wants to avoid the sanction.” This is all that is needed to explain everything that must be explained about law’s normative capacity to do anything a legal system can be non-​ idiosyncratically used to do. There is simply nothing here that needs to be done to explain how the practices constituting something as a system of law are reasonably 14 Apart from the claim that a legal system must authorize coercive sanctions for non-​compliance, the only other properties plausibly thought conceptually necessary for an institutional normative system to count as one of law, as discussed above, are: (1) the system must satisfy Hart’s minimum conditions for the existence of a legal system; (2) it must contain norms governing non-​official behavior that prohibit acts violating the minimum content of natural law; (3) it must be metaphysically capable of morally legitimate authority; and (4) it must regulate behavior by the recognition, application, and enforcement of norms metaphysically capable of motivationally and epistemically guiding behavior. There is simply nothing in any of this that is so objectively valuable that subjects should treat the normative output of any system with these properties as giving rise to objective motivating reasons, complete or otherwise, of any kind.

Valid Norms Do Not Provide the Reason  149 contrived to regulate non-​official behavior—​regardless of what one might think the ultimate end of these practices are. But if there is a conceptual problem that requires explaining how the practices constituting something as a system of law are reasonably contrived to give rise to objective complete operative motivating reasons, the claim that the Coercion Thesis cannot solve this problem tells us nothing about whether the Coercion Thesis is true. If this is a conceptual problem that theories of the metaphysical nature of law must solve, as Raz seems to believe, it is precisely because there is nothing in any other plausible candidate for a conceptually necessary condition for the existence of a legal system equipped to explain how these constitutive practices are reasonably contrived to give rise to complete reasons. It should be clear that there is nothing in any of the properties discussed above in Section 2 that is reasonably contrived to give rise to something plausibly characterized as any kind of objective motivating reason, complete or otherwise; the reason that law’s conceptual normativity needs an explication is precisely that there is nothing in these other features reasonably contrived to give rise to anything that rationally competent subjects characteristically regard, as a descriptive matter of contingent fact, as normatively relevant because they should, as an objective matter of normative practical rationality, regard it as normatively relevant.

4.  Does It Matter that It Is Not the Valid Norms Governing the Subject’s Behavior That Provide the Objective Motivating Reason to Comply? A second line of objection argues that this solution to the How Problem is problematic because it is the recognition norms authorizing judges to impose coercive sanctions for violations of mandatory legal norms governing non-​official behavior—​and not those mandatory legal norms themselves—​that are the source of the new content-​independent defeasible objective motivating prudential reasons to comply. The How Problem, on this line of reasoning, can be solved only by showing how the mandatory legal norms governing non-​official behavior can themselves be normatively relevant for rationally competent subjects of those norms. There is no reason to think that the normativity of some empirical phenomenon involving directives can be explained only by showing how those directives themselves are normative. The demand of an armed robber P to “hand over your money or else,” construed to include all the relevant facts, gives rise to a new objective motivating prudential reason for the subject of the demand Q to give P her money, but it is not the “hand over your money” piece of the sentence that gives rise to the new objective motivating reason. It is rather the “or else” piece—​which is, strictly speaking, not part of what expresses the directive with which P expects Q to comply—​that is reasonably contrived to give rise to an objective motivating reason

150  The How Problem of Legal Normativity for Q to comply. The expressed threat asserts the purely descriptive claim that the probable consequence of Q’s failing to comply with the directive requiring Q to surrender her money is a gunshot wound—​something P and Q both know Q has a strong objective motivating prudential reason to avoid. Without that threat, there is nothing in P’s behavior reasonably contrived to give rise to an objective motivating reason of any kind for Q to surrender her money. The expressed threat, purely descriptive in character, that P will shoot Q is intended and reasonably contrived to activate in Q a latent normative conviction that Q should perform any morally permissible act to avoid being shot. Purely descriptive claims can express states of affairs that activate latent normative convictions that are rarely, if ever, relevant in a person’s practical deliberations about what to do. I have never been robbed at gunpoint and hence have never acted on my latent normative conviction that I should do whatever is morally permissible to avoid being shot; but if I am ever robbed, I will not hesitate to act on it, barring circumstances too exceptional for me to imagine, much less to anticipate. The point here is not just that P’s behavior is reasonably contrived to serve its purpose only insofar as she backs her demand with the threat that Q will be shot for non-​compliance; it is also that without the threat P’s behavior is, as a conceptual matter, not properly characterized as even an attempted robbery. Although the sentence “give me your money” is expressed in the imperative mood and implicitly incorporates the deontic operator expressed by the term “must,” it is, by itself, neither objectively normative nor likely to be regarded by rationally competent self-​ interested subjects of the demand as normative. The reason a threat is needed in this case to accomplish P’s objective of separating Q from her cash is precisely that P knows that Q is not likely, as a descriptive matter of contingent fact, to regard the directive expressed by “give me all your money,” by itself, as providing a motivating reason of any kind because Q should not, as an objective matter of normative practical rationality, regard it as such. It is precisely because propositions expressed by such unreasonable sentences purporting to be normative in virtue of incorporating a subject, deontic operator, and act are not objectively normative that a threat is needed to provide an objective motivating prudential reason to comply. Threats are needed to induce compliance with a directive among rationally competent self-​interested subjects who are not characteristically likely, as a descriptive matter of contingent fact, to regard the directive as motivationally normative because they should not, as an objective matter of normative practical rationality, regard it as motivationally normative. I have no motivating reason of any kind, subjective or objective, to comply with a stranger’s demand for my cash until she points a loaded gun at my head. The content of a directive does not change depending on whether it is backed by the threat of a sanction. The content of the directive expressed by the sentence “give me your money” is exactly the same regardless of whether it is enforced by the threat expressed by “or else”; it is the threat of being shot that gives Q a reason

Valid Norms Do Not Provide the Reason  151 to do what she has no reason to do otherwise—​namely, to give P her money. The content of the Fifth Commandment expressed by the sentence “thou shalt not kill” is likewise exactly the same regardless of whether it is enforced by a divine threat of eternal torment; it is the threat of experiencing eternal torment for killing someone that provides an additional reason for rationally competent self-​interested subjects to abstain from doing something they already have a conclusive objective motivating moral reason not to do—​i.e. kill people. Regardless of whether one antecedently has sufficient motivating reason to comply with the relevant directive, it cannot be plausibly denied that the threats of being shot and of being punished with eternal torment change the prudential calculus, both objectively and subjectively, with respect to what one should do. The same threat can provide a prudential motivating reason, both objectively and subjectively, to comply with a mandatory norm regardless of that content. The threat of being killed by someone pointing a loaded gun at me, by itself, provides exactly the same objective motivating prudential reason to comply with the accompanying directive regardless of its content: the objective motivating prudential reason provided by the threat is exactly the same whether it is backing a demand for my cash or a demand that I kill an innocent person. A rationally competent self-​interested subject might, without objective practical irrationality, treat that prudentially motivating reason as conclusive in the latter case, but the relevant prudentially motivating reason will be defeated, as an objective matter of normative practical rationality, by moral considerations in the latter case that are not usually applicable in the former case. Similarly, the threat of being punished with eternal torment would provide exactly the same objective motivating prudential reason to comply with a divine commandment to kill innocent persons as it does to comply with a divine commandment not to kill innocent persons. A  rationally competent self-​interested subject might sensibly regard the threat of eternal torment as a conclusive reason to do what is commanded in both cases, but whether one has a conclusive reason to do so all things considered, as an objective matter of normative practical rationality, depends on whether one correctly believes that God’s commandments are necessarily morally binding: if so, then one will have both a defeasible objective motivating prudential reason and a conclusive objective motivating moral reason for doing whatever God commands—​including killing innocent persons, as the Old Testament figure Abraham can attest; if not, then one might have a conclusive objective motivating moral reason that defeats the objective motivating prudential reason to do what God commands—​such as in a case where God’s commands are not morally binding and God commands that one impermissibly kill one’s son. A demand backed by such a threat is reasonably contrived to be both subjectively and objectively normative. It is reasonably contrived to be subjectively normative in the sense that rationally competent self-​interested subjects of the demand are characteristically likely, as a descriptive matter of contingent fact, to regard the

152  The How Problem of Legal Normativity threat as relevant in deciding whether to comply with the demand with which the threat is bundled; that is, the threat is reasonably contrived to be subjectively normative in the sense that rationally competent self-​interested subjects are characteristically likely, as a descriptive matter of contingent fact, to regard it as giving rise to a defeasible motivating prudential reason to comply in virtue of activating otherwise latent normative convictions. It is also reasonably contrived to be objectively normative in the sense that rationally competent self-​interested subjects of the threat should regard it, as an objective matter of normative practical rationality, as relevant in deciding whether to comply with the demand with which the threat is bundled; that is, the threat is reasonably contrived to be objectively normative in the sense that rationally competent subjects should, as an objective matter of normative practical rationality, regard it as giving rise to a defeasible motivating prudential reason to comply. It is not a problem, then, for the solution to the How Problem proposed here that it does not explain the conceptual normativity of law in terms of the normativity of its valid norms because the very point of backing those norms with the threat of a sanction is exactly that rationally competent subjects do not—​and need not—​regard those norms as providing persuasive motivating reasons to comply, even when they should, as an objective matter of normative practical rationality, and there is nothing irrational about this. Not doing what one has most objective motivating reason to do might not be the “most rational thing to do,” whatever that might mean, but it is not thereby practically irrational in virtue of not being what is best to do as an objective matter of normative practical rationality. Acting on the basis of even a decision one believes might be bad is not the same as acting irrationally. Coercive sanctions are needed to ensure that law is reasonably contrived to keep the peace among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours precisely because the other practices constituting something as a system of law, by themselves, are not reasonably contrived to give rise to new content-​independent objective motivating reasons of any kind. If the practices constituting something as a legal system lack those authorizing the imposition of coercive sanctions for non-​compliance, then those practices are simply not reasonably contrived to do anything that law can be non-​idiosyncratically used to do among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours; there is no existing society in our world that would not be destroyed if the sanctions backing the valid mandatory legal norms in its legal system were suddenly de-​authorized. The fact that it is not the valid legal norms governing non-​official behavior that are objectively normative is not only not a problem for the solution defended in this chapter; it also illustrates why the solution defended in this chapter is the only plausible solution to the How Problem: there is simply no other way to endow a valid legal norm that does not itself give rise to objective motivating reasons to

Sanctions and the How Problem  153 comply with normative relevance except by bundling it with something else that does give rise to objective motivating reasons to comply. The practices constituting something as a system of law are reasonably contrived to regulate behavior through the governance of norms metaphysically capable of guiding it only insofar as they publicly back some of those norms with the threat of detriment sufficiently painful that it can rationally induce compliance among rationally competent self-​ interested subjects antecedently disposed to do otherwise often enough to succeed in efficaciously regulating behavior.

5.  Coercive Sanctions and the How Problem of Legal Normativity On the account defended in this chapter, the only elements of legal practice reasonably contrived to provide a new defeasible content-​independent objective motivating reason to comply have to do with the legal norms governing official behavior that authorize courts to impose coercive sanctions on subjects for violating mandatory legal norms governing non-​official behavior. It is, thus, the fact that there is a legal recognition norm that authorizes courts to impose coercive sanctions on subjects of other legal norms for violations of those norms that constitutes a legal system as reasonably contrived to give rise to a new objective motivating reason to comply with those other norms. There is nothing else in the practices constituting something as a legal system reasonably contrived to give rise to new content-​independent defeasible objective motivating reasons to comply with mandatory legal norms governing non-​official behavior. The contingent fact that some collection of people who happen to serve as officials in an institutional normative system satisfying Hart’s minimum conditions for the existence of a legal system that includes the minimum content of natural law and is metaphysically capable of epistemically and motivationally guiding behavior and hence of instantiating moral legitimacy tells us nothing that would help to explain how the practices constitutive of a legal system are reasonably contrived to give rise to such motivating reasons. The Coercion Thesis is uniquely equipped to provide a plausible solution to the How Problem of Legal Normativity.

7

The Coercion Thesis and the Order Problem of Legal Normativity One of the foundational problems in conceptual jurisprudence is to vindicate the presumed normativity of law as such. The practices constituting something as a system of law are reasonably contrived to efficaciously regulate non-​official behavior through the governance of norms metaphysically capable of guiding it only if those practices are reasonably contrived to give rise to normative considerations that can make a practical difference with respect to whether rationally competent subjects comply with mandatory legal norms governing non-​official behavior. Vindicating law’s conceptual normativity hence requires explaining how these constitutive practices are reasonably contrived to give rise to objective motivating reasons to comply with mandatory legal norms governing non-​official behavior. It is sometimes argued that law’s conceptual normativity cannot be fully vindicated simply by showing that these constitutive practices are reasonably contrived to give rise to the objective motivating prudential reasons favoring compliance to which the threat of a coercive sanction gives rise. Since mandatory legal norms are necessarily applied and enforced as excluding certain excusing or justifying motivating reasons for non-​compliance, the practices constitutive of law must also be reasonably contrived to give rise to objective exclusionary motivating reasons that bar acting on countervailing motivating prudential reasons favoring non-​ compliance. The Order Problem of Legal Normativity is concerned with determining whether and how these constitutive practices are properly equipped to give rise to objective exclusionary motivating reasons to comply with mandatory legal norms governing non-​official behavior.1 This chapter argues that there is no reason to think that the practices constituting something as a system of law must be reasonably contrived to give rise to objective exclusionary motivating reasons to comply with mandatory legal norms governing non-​official behavior. I argue, to begin, that there are three basic sources of objective motivating reasons and that there is nothing in these constitutive 1 Strictly speaking, exclusionary motivating reasons are not motivating reasons to comply; since they only bar acting on motivating reasons favoring non-​compliance, they provide nothing that would count as a rational inducement to comply among rationally competent subjects antecedently disposed to do otherwise. For the sake of simplicity, I will refer to them as motivating reasons to comply insofar as they are supposed to operate by decreasing the probability of non-​compliance by barring subjects from acting on first-​order motivating reasons favoring non-​compliance. Coercion and the Nature of Law. Kenneth Einar Himma, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198854937.001.0001

156  The Order Problem of Legal Normativity practices reasonably contrived to give rise to objective exclusionary motivating reasons from any of these sources. I argue, further, that these constitutive practices are equipped to do everything that a legal system can be non-​idiosyncratically used to do simply in virtue of providing objective motivating prudential reasons to comply with mandatory legal norms governing non-​official behavior; since there is no mystery as to how these constitutive practices give rise to objective motivating reasons for officials to comply with mandatory legal norms governing their behavior, the claim that these constitutive practices can produce objective exclusionary motivating reasons explains nothing about legal normativity that needs to be explained.

1.  First-​Order, Second-​Order, and Exclusionary Reasons Certain kinds of normative consideration can be distinguished, as a conceptual matter, from one another in terms of what they are about. A normative consideration can be about an external state-​of-​affairs valued by the agent or it can be about some other normative consideration valued by the agent: the normative consideration is first-​order in the former case and second-​order in the latter case. Both types of consideration can properly figure into a rationally competent subject’s deliberations about what ought to be done. Desires are normative considerations capable of playing a role in the deliberations of a rationally competent subject about what to do and can be distinguished from one another in terms of what they are about and hence in terms of their order.2 Suppose P loves Q but regrets loving Q—​a common configuration of preferences in people unhappy about how a romantic relationship is going (or not going, as the case may be). In this case, P has a first-​order desire that has Q’s romantic companionship as its object but also has a second-​order desire that has P’s first-​order desire as its object: P wants Q’s romantic companionship but wishes that she did not want it. Similarly, if P is unhappily addicted to some substance, she will have first-​and second-​order desires that, taken together, express ambivalence about that substance. In this case, P has a compelling first-​order desire for the substance but a second-​order desire that has P’s first-​order desire as its object: P wants the substance but wishes she did not want it. Although these first-​order and second-​order desires are in tension with one another, there is no logical inconsistency between them. Such tension expresses an ambivalence towards the relevant phenomena that is not only not irrational in any applicable sense but can also be psychologically healthy. If P is addicted to some unhealthy substance or attracted to an abusive person, it is healthy from a 2 Strictly speaking, only propositions expressing desires can play a role in practical deliberations since one can reason only with things having propositional content.

First-Order, Second-Order, and Exclusionary Reasons  157 psychological standpoint that P has a second-​order desire not to have a first-​order desire for that something or someone. Given that we frequently desire things we should not desire, it is indicative of good psychological health that we recognize the morally or prudentially problematic character of these desires and respond with a second-​order desire that we not have those desires. One can define an nth-​order desire pertaining to some basic thing or phenomena p, for any natural number n, recursively as follows. For the base case n = 1, an nth-​order desire pertaining to p is just a first-​order desire for p or not-​p. For any n > 1, an nth-​order desire pertaining to p is a desire to have, or not to have, an (n  –​ 1)th-​order desire pertaining to p.3 First-​order desires are, as a conceptual matter, desires about things other than desires while every higher-​order desire is a desire having to do with a lower-​order desire. It is not clear how far our desires pertaining to some thing or phenomena can go, but it is clear, as a descriptive matter of contingent fact, that we can have a complex of desires that include higher-​order desires than second-​order desires. A person P might have a desire for the romantic companionship of someone Q about whom she is conflicted and have a desire that she not be conflicted about Q’s companionship because it causes her unnecessary suffering. In such a case, P would have a first-​order desire for Q’s companionship; a second-​order desire that P not have a first-​order desire for Q’s companionship; and a third-​order desire that P not have a second-​order desire not to have a first-​order desire for Q’s companionship. If P is somewhat conflicted about this desire not to question her second-​order desire, P might have a fourth-​order desire not to have this third-​order desire—​and so on. The same is true of motivating reasons to do something. One might have a first-​ order motivating reason pertaining to the doing of a and higher-​order motivating reasons with respect to lower-​order motivating reasons pertaining to the doing of a. A first-​order motivating reason to do a has to do with something that functions, or ought to function, in practical deliberations directly as a rational inducement of some kind to do a. P’s first-​order desire to smoke a cigarette might give rise to a first-​order motivating reason to smoke a cigarette either as a purely descriptive matter of contingent fact having to do with what is going on in P’s thinking or as an objective matter of normative practical rationality having to do with what should be going on in P’s thinking. In the former case P treats her first-​order desire to smoke in her deliberations as a first-​order motivating reason to smoke, while in the latter case P should treat her first-​order desire as giving rise to a first-​order motivating reason to smoke.

3 Notice that the relevant lower-​order desire is just one order lower than the higher-​order desire: one cannot, as a conceptual matter, have a fourth-​order desire about a second-​order desire. The same is true of motivating reasons: the relevant lower-​order motivating reason is just one order lower than the higher-​order motivating reason; one cannot, as a conceptual matter, have a fourth-​order motivating reason about a second-​order motivating reason.

158  The Order Problem of Legal Normativity Similarly, a second-​order motivating reason pertaining to the doing of a has to do with something that functions, or ought to function, as a rational inducement in the form of a motivating reason to act or refrain from acting, as the case may be, on any first-​order motivating reason to do a. If P has a first-​order desire to smoke a cigarette and a second-​order desire not to want to smoke a cigarette, then P’s second-​order desire not to want to smoke might give rise to a second-​order motivating reason not to act on any first-​order motivating reasons to smoke either as a purely descriptive matter having to do with what is going on in P’s thinking or as an objective matter of normative practical rationality having to do with what should be going on in P’s thinking. In the former case P treats her second-​order desire not to want to smoke cigarettes as giving rise to a second-​order motivating reason not to act on any first-​order motivating reasons to smoke a cigarette, while in the latter case P should treat her second-​order desire as a second-​order motivating reason not to act on any first-​order motivating reasons to smoke a cigarette. One can define an nth-​order motivating reason pertaining to the performance of some act a, for any natural number n, recursively as follows. For the base case n = 1, an nth-​order motivating reason to do a is just a first-​order motivating reason to do a. For any n > 1, an nth-​order reason to do a is a motivating reason to have, or not to have, an (n –​1)th-​order reason to do a. First-​order motivating reasons are, as a conceptual matter, motivating reasons pertaining to things other than motivating reasons while every higher-​order motivating reason is a motivating reason having to do with a lower-​order motivating reason. Higher-​order motivating reasons, like first-​order motivating reasons, can be either subjective or objective, but the ones that matter for our purposes are second-​ order motivating reasons. A subjective second-​order motivating reason is one that functions, as a descriptive matter of contingent fact, as a second-​order motivating reason in the practical deliberations of some rationally competent subject, while an objective second-​order motivating reason is one that should function, as an objective matter of normative practical rationality, as a second-​order motivating reason in the deliberations of a rationally competent subject. Given that what we want and what we should want can come apart, as a conceptual matter, so can what we treat as a second-​order motivating reason and what we should treat as a second-​ order motivating reason. Two classes of second-​order motivating reasons can be distinguished according to whether the reasons favor a performance or an abstention. An inclusionary motivating reason is a second-​order motivating reason to act on some class of first-​ order reasons. There might be objective inclusionary motivating moral reasons that dictate that an agent should act on certain first-​order motivating reasons: P might have a second-​order motivating moral reason to act on a first-​order motivating reason to marry Q as a means of legitimizing P’s relationship with Q in a manner that conforms to P’s religious beliefs. An exclusionary motivating reason is a second-​order motivating reason not to act on some class of first-​order motivating

First-Order, Second-Order, and Exclusionary Reasons  159 reasons.4 There seem to be objective moral exclusionary reasons that dictate that an agent not act on morally undesirable first-​order prudential motivating reasons: P might have a second-​order motivating moral reason not to marry Q for first-​order motivating prudential reasons having to do with financial considerations or other considerations not having to do with P’s romantic feelings as they are expressed in first-​order motivating reasons. The logical relationship between inclusionary motivating reasons and exclusionary motivating reasons is not as straightforward as might initially appear. The claim that one has an inclusionary motivating reason to do a does not imply the claim that one has an exclusionary motivating reason that bars acting on other first-​ order motivating reasons not to do a. The claim that one has a motivating reason to get married for first-​order motivating reasons having to do with romantic desires does not imply the claim that one should not act on certain first-​order motivating reasons not to get married; there might be other first-​order motivating reasons not to get married that defeat any first-​order motivating reasons to get married. Conversely, the claim that one has an exclusionary motivating reason that bars acting on first-​order motivating reasons not to do a does not imply that one has an inclusionary motivating reason to do a. The claim that one should not abstain from getting married because one’s parents disapprove of one’s partner does not imply that one should act on other first-​order motivating reasons to get married; there might not be any other first-​order motivating reasons to get married. Despite initial appearances to the contrary, the two classes of reasons are ontologically distinct, logically independent, and partition the class of second-​order motivating reasons into two mutually exclusive and jointly exhaustive subclasses. The notion of an exclusionary motivating reason should be distinguished from that of a peremptory motivating reason. A peremptory motivating reason for action is a second-​order motivating reason that bars an agent from weighing the various first-​order motivating reasons for herself.5 There is nothing in the notion of an exclusionary reason that bars the agent from weighing such reasons. An exclusionary reason precludes only the agent’s acting on the excluded first-​order motivating reasons; unlike a peremptory motivating reason, it does not bar weighing those motivating reasons. An agent with an exclusionary motivating reason to comply with a norm requiring her to do a but not a peremptory motivating reason to comply with that norm may—​consistent with that reason—​assess the first-​order motivating reasons pertaining to the doing of a and may act on her assessment of the balance if it favors doing a but not if it favors not-​doing a.

4 See, e.g., Joseph Raz, Practical Reason and Norms (Princeton University Press 1994) 39. Hereinafter PRN. 5 H.L.A. Hart, “Commands and Authoritative Legal Reasons” in Hart, Essays in Bentham (Clarendon Press 1982) 253.

160  The Order Problem of Legal Normativity The notion of an exclusionary motivating reason is inspired, as the marriage examples above might suggest, by considerations having to do with morality. Morality is commonly thought to trump considerations of self-​interest in the sense that a moral obligation to do a, as an objective matter of normative practical rationality, cannot be defeated by first-​order prudential reasons not to do a. The mechanism by which a moral obligation trumps countervailing prudential considerations, one might think, is by entailing a second-​order objective exclusionary motivating moral reason that bars acting on first-​order motivating prudential reasons favoring non-​compliance. The norms of an objective morality, on this plausible line of reasoning, thus give rise not only to first-​order objective motivating moral reasons to comply but also to objective second-​order exclusionary motivating moral reasons not to act on countervailing first-​order motivating prudential reasons. These norms give rise to objective second-​order exclusionary motivating reasons insofar as it is true that, as an objective matter of normative practical rationality, rationally competent subjects should not act on first-​order motivating prudential reasons favoring non-​compliance. They are also likely to give rise to subjective second-​ order exclusionary motivating reasons in rationally competent subjects insofar as subjects are characteristically likely in virtue of being properly reason-​ responsive, as a descriptive matter of contingent fact, to treat these norms as giving rise to second-​order motivating reasons not to act on countervailing first-​order motivating reasons.

2.  Sanctions, Exclusionary Reasons, and the Problems of Legal Normativity One might think that the objective first-​order motivating prudential reasons to which the threat of coercive sanctions is reasonably contrived to give rise do not wholly explicate law’s conceptual normativity because they cannot explain how legal norms prohibiting acts are mandatory in the conceptually requisite sense. As Joseph Raz makes the point: The fact that a law of a system in force provides for a sanction is no doubt a reason for action, but it is a reason of the wrong kind. A sanction-​backed norm . . . cannot explain how it is that some laws are mandatory norms. . . . [I]‌f some laws are mandatory norms then they are exclusionary reasons as well as first-​order reasons. But the fact that a law is backed by a sanction is never an exclusionary reason. . . . The inevitable conclusion is that, despite the undoubted importance of sanctions and the use of force to enforce them in all human legal systems, the sanction-​ directed attempt to explain the normativity of the law leads to a dead end (PRN 161–​62).

Sanctions and Exclusionary Reasons  161 On this line of reasoning, there is nothing in the practices authorizing the imposition of sanctions for non-​compliance with mandatory legal norms governing non-​ official behavior equipped to figure into an explanation of how these norms give rise to objective second-​order motivating reasons to comply; the claim that these practices are reasonably contrived to give rise to objective first-​order motivating prudential reasons tells us nothing about how they are reasonably contrived to give rise to objective second-​order exclusionary motivating reasons. Raz’s view is ambiguous in one crucial respect. It is clear, on the one hand, that he explains law’s capacity to give rise to objective exclusionary motivating reasons in terms of the fact that judges apply mandatory legal norms governing non-​official behavior as excluding certain excuses or justifications for non-​compliance: “[The courts] must judge individuals as if they should take the legal requirements of exclusionary reasons” (PRN 144). But it is not clear, on the other hand, to whose behavior the relevant objective exclusionary motivating reasons apply: The courts are in effect bound to regard individuals as acting in accordance with legal standards to the exclusion of all other reasons. Since. . . the rules of the system which apply to ordinary individuals are identical with the rules by which the primary organs ought to judge individuals it follows that all the legal rules are both first-​order and exclusionary reasons (PRN 144).

The issue is whether the relevant exclusionary motivating reasons apply to the judge’s decision-​making in deciding disputes under mandatory legal norms governing non-​official behavior or whether, alternatively, they apply directly to the subjects of those mandatory legal norms: in the former case it would be the recognition norms governing the judge’s official behavior that give rise to the objective exclusionary reason to apply the relevant mandatory legal norms governing non-​ official behavior in an exclusionary fashion in legal disputes implicating those norms; in the latter case it would be the fact that judges apply those norms in an exclusionary manner that gives rise to an objective motivating exclusionary reason on the part of subjects of mandatory legal norms governing non-​official behavior not to act on first-​order motivating reasons favoring non-​compliance with those mandatory legal norms. Both alternatives are problematic. If, on the one hand, the claim is that the relevant objective exclusionary motivating reasons apply to the judge’s decision-​ making, then there is nothing here that would explain law’s conceptual normativity as it pertains to subjects of mandatory legal norms governing non-​official behavior. The only legal norms governing the judge’s decision-​making are recognition norms governing official behavior; if it is true, as seems plausible, that only a norm that governs an act can give rise to an objective motivating reason to perform or abstain from that act, the only norms that could give rise to objective motivating reasons that apply to the judge are legal norms governing the judge’s behavior. But since,

162  The Order Problem of Legal Normativity as a conceptual matter, norms that govern only official behavior cannot, by themselves, give rise to objective motivating reasons to comply with norms governing non-​official behavior, there is nothing in the claim that the relevant objective exclusionary motivating reasons apply to the judge’s decision-​making that would tell us anything about law’s conceptual normativity as it pertains to subjects of legal norms governing non-​official behavior.6 But if, on the other hand, the claim is that it is the fact that judges apply legal norms governing non-​official behavior in an exclusionary fashion that gives rise to an objective motivating exclusionary reasons on the part of subjects of valid mandatory legal norms governing non-​official behavior not to act on first-​order motivating reasons favoring non-​compliance, the claim is utterly lacking plausible support. A vindication of law’s presumed conceptual normativity is needed precisely because there is nothing in the practices constituting something as a system of law, apart from enforcement practices, reasonably contrived to give rise to something that rationally competent subjects are characteristically likely, as a descriptive matter of contingent fact, to regard as motivating reasons to comply with mandatory legal norms governing non-​official behavior because they should, as an objective matter of normative practical rationality, regard them as such reasons. The idea that the practices constituting something as a system of law are reasonably contrived to give rise to objective exclusionary motivating reasons to comply with mandatory legal norms governing non-​official behavior has, despite these problems, become something of a dogma in certain circles of theorists and, for this reason, will be evaluated in this chapter. Although I will discuss the issue of whether it is plausible to think that mandatory legal norms governing official behavior give rise to second-​order objective exclusionary motivating reasons for judges to apply mandatory legal norms governing non-​official behavior in an exclusionary manner, this volume is concerned to vindicate only law’s conceptual capacity to give rise to objective motivating reasons to comply with mandatory legal norms governing non-​official behavior; the Coercion Thesis says nothing about mandatory legal norms governing official behavior.

3.  The Basic Sources of Objective Motivating Reasons for Action One initial concern with the claim that the practices constituting something as a system of law are reasonably contrived to give rise to objective exclusionary 6 The presumed normativity of norms governing non-​official behavior is what needs to be explained by a conceptual theory of law. The only serious conceptual problems of legal normativity concern how the practices constituting something as a system of law are reasonably contrived to regulate non-​official behavior; it is a trivial matter to explain the conceptual normativity of the rule of recognition: insofar as officials take the internal point of view toward rule of recognition, they have accepted it as normative.

Basic Sources of Motivating Reasons  163 motivating reasons to comply with mandatory legal norms governing non-​official behavior has to do with identifying a plausible source of those reasons. Rationally competent subjects like us are characteristically motivated by three kinds of consideration that cannot be reduced to, or explained in terms of, simpler considerations we view as desirable: these basic considerations have to do with whether an act (1) satisfies standards of morality; (2) conduces to self-​interest; or (3) results in beauty.7 There are thus only three types of ontologically irreducible objective motivating reasons: (1) moral reasons; (2) prudential reasons; and (3) aesthetic reasons. Each of these basic considerations gives rise not only to subjective motivating reasons but also to objective motivating reasons: we do not just happen to come to value them; we characteristically value these basic considerations, as a descriptive matter of contingent fact, because we should value them, as an objective matter of normative practical rationality. Rationally competent self-​interested subjects like us who must compete for all we want and need in worlds of acute material scarcity like ours characteristically care about these considerations because things will simply not go well for us if we do not; the mental health of someone not motivated in the least by one or more of these classes of considerations is suspect for precisely that reason.8 The severity and character of the mental health issue differ according to which of these basic considerations is not regarded as normatively relevant. Someone not responsive to either moral or prudential considerations is dangerous but for different reasons: someone not motivated by moral considerations is a sociopath who poses a threat to other people in virtue of not having scruples that would prevent her from doing something harmful to other people; someone not motivated by prudential considerations likely suffers from the kind of depression that makes her a threat to do something harmful to herself. Someone unresponsive to aesthetic considerations is not necessarily dangerous but is not likely to live an optimally fulfilling life; it is prohibitively difficult to make deep friendships when one takes no interest in aesthetic phenomena. There are a couple of related observations that should be made about the logical relationships among the three classes of objective motivating reasons. First, these three basic sources of reasons are ontologically distinct and define three mutually exclusive subclasses of objective motivating reasons that jointly exhaust the class 7 I assume here that altruistic concerns are plausibly explained in terms of moral and prudential concerns: if there are standards dictating how other people’s interests should objectively be promoted, they are moral in character; if there are standards having to do with how to promote the interests of other people insofar as one values them for reasons having to do with what makes one happy, they are prudential in character. Any other considerations properly characterized as altruistic are clearly irrelevant with respect to whether the practices constituting something as a system of law are reasonably contrived to give rise to objective exclusionary motivating reasons. 8 Aesthetic considerations are considerably less important than prudential and moral considerations in this regard but are nonetheless crucial to our being able to thrive.

164  The Order Problem of Legal Normativity of objective motivating reasons. It should be clear that one cannot validly derive (1)  an objective motivating prudential reason from a set consisting only of objective motivating moral reasons and objective motivating aesthetic reasons; (2) an objective motivating moral reason from a set consisting only of objective motivating aesthetic reasons and objective motivating prudential reasons; or (3) an objective motivating aesthetic reason from a set consisting only of objective motivating prudential reasons and objective motivating moral reasons. Second, there are no objective motivating reasons that are properly characterized as hybrid in the sense that they count as objective motivating reasons of more than one kind. The proof is straightforward: if one cannot derive an objective motivating reason of one class from objective motivating reasons from other classes, then an objective motivating reason of one kind can be validly derived from only objective motivating reasons of the same kind; however, something is properly characterized as a reason of one kind insofar as it derives only from considerations belonging to the basic source defining that kind: a compound objective motivating aesthetic/​moral/​prudential reason is properly characterized as hybrid insofar as it can be logically derived only from more than just one class of basic objective motivating aesthetic/​moral/​prudential reasons. It is true that one can have objective motivating reasons of different kinds to do the same thing, but this is because certain acts implicate values from more than one basic source—​and not because there are objective motivating reasons that can be validly derived only from sets including objective motivating reasons of more than one kind. There is no reason to rule out the possibility that one can have objective motivating reasons that are prudential, moral, and aesthetic to do nice things for other people—​assuming that there can be aesthetic motivating reasons for doing things unconnected with the production of artworks; however, that does not imply that any of those objective motivating reasons can be validly derived only from a set containing objective motivating reasons of more than one basic kind. Law’s conceptual normativity must, then, be explained in terms of its capacity to give rise to objective motivating reasons from one of these basic sources. The practices constituting something as a system of law are reasonably contrived to efficaciously regulate behavior only insofar as they are capable of making a practical difference with respect to what rationally competent subjects do. But since law can make a practical difference through the governance of norms metaphysically capable of guiding behavior only by providing something rationally competent subjects should, as an objective matter of normative practical rationality, regard as a motivating reason to comply, these constitutive practices must be presumed to be reasonably contrived to give rise to objective motivating reasons from one of these sources because only reasons from these sources are characteristically likely to motivate rationally competent subjects to comply. This should not be taken to preclude the possibility that the practices constituting something as a system of law give rise to more than one kind of objective

Only Sanctions Explain Exclusionary Reasons  165 motivating reasons to comply with mandatory legal norms governing non-​official behavior. There is nothing in what has been said up until now that is inconsistent with the conceptual possibility that these constitutive practices might give rise to, for instance, objective motivating prudential reasons to comply and objective motivating moral reasons to comply in virtue of implicating values from more than one basic source. But it turns out, as will be seen below, that the only kind of objective motivating reasons to which these constitutive practices are reasonably contrived to give rise are prudential reasons.

4.  Only the Coercion Thesis Can Explain How Mandatory Legal Norms Governing Non-​Official Behavior Give Rise to Objective Motivating Exclusionary Reasons There is nothing in the practices constituting something as a system of law, apart from those authorizing the judicial imposition of coercive sanctions for non-​ compliance, reasonably contrived to give rise to objective exclusionary motivating reasons from any of these basic sources. First, the practices constituting something as a system of law are not reasonably contrived to give rise to objective exclusionary motivating moral reasons because only morally legitimate legal systems give rise to such reasons; insofar as there are clearly conceptually possible legal systems that are morally illegitimate in virtue of applying and enforcing morally wicked norms as norms of law, it is the moral properties of a system of law, and not its conceptual properties, that would explain why that system gives rise to such reasons—​ assuming that legitimate systems give rise to them. Second, the practices constituting something as a system of law are not reasonably contrived to give rise to objective exclusionary motivating aesthetic reasons to comply with mandatory legal norms governing non-​official behavior.9 To begin, the idea that there is something aesthetically valuable about a rationally competent subject taking an authority’s judgment as a second-​order motivating reason to do something that bars her from acting on her own assessment of the applicable first-​order motivating reasons is implausible: there is nothing cute or lovely about someone completely deferring to the judgment of an authority; it might be true that someone should defer to someone else’s judgment but not because doing so is aesthetically pleasing. Further, it is unclear how objective aesthetic exclusionary motivating reasons could bar acting on some other class of basic motivating reasons. The idea that there are aesthetic norms that trump competing mandatory moral norms is a silly 9 This is not to deny that law might require doing something that is beautiful or that promotes aesthetic value; in such a case, the law would implicate objective first-​order motivating aesthetic reasons for doing that something but not in virtue of giving rise to new aesthetic reasons to do it.

166  The Order Problem of Legal Normativity conceit of people who overvalue the production of aesthetic experiences: if an example is needed, the aesthetic value of doing unconsented-​to harm to an innocent person simply cannot, as an objective matter of normative practical rationality, trump the moral motivating reasons to which the moral obligation not to harm innocent persons gives rise. Likewise, the idea that there are aesthetic norms that trump competing prudential norms is problematic because they intersect only in connection with the production of artworks; the correct aesthetic choice might not optimally conduce to the artist’s self-​interest in these rare cases of conflict insofar as they reduce the perceived monetary value of the piece. Either way, there is nothing about the practices constituting something as a system of law, regardless of whether the Coercion Thesis is true, plausibly thought to be intended or reasonably contrived to give rise to either objective first-​or second-​order motivating aesthetic reasons. Third, the practices constituting something as a system of law, on the assumption that the Coercion Thesis is false, are not reasonably contrived to give rise to objective exclusionary motivating prudential reasons to comply with mandatory legal norms governing non-​official behavior. There is simply nothing in the fact that officials recognize and apply certain norms as binding requirements of law, if not backed by the threat of a coercive sanction, reasonably contrived to give rise to something that rationally competent subjects are characteristically likely, as a descriptive matter of contingent fact, to regard as a prudential motivating reason to comply with legal requirements because they should, as an objective matter of normative practical rationality, regard it as such. The only constitutive practices of law reasonably contrived to give rise to first-​or second-​ order objective motivating reasons to comply with mandatory legal norms governing non-​official behavior are those that back them with coercive sanctions for non-​compliance. But if these constitutive practices are reasonably contrived to give rise to objective second-​order exclusionary motivating reasons, it would have to be in virtue of being reasonably contrived to give rise to objective first-​order motivating reasons that have their basic source in prudential norms. If a theory of the metaphysical nature of law must explain how the practices constituting something as a system of law are reasonably contrived to give rise to objective exclusionary motivating reasons, then it can do so only insofar as it includes the Coercion Thesis because only the Coercion Thesis can explain how those practices can give rise to first-​order objective motivating reasons of any kind. Since (1) there can be no second-​order objective exclusionary reasons to do what there are no first-​order objective motivating reasons to do and (2) the only objective first-​order motivating reasons to comply to which the practices constituting something as a system of law are reasonably contrived to give rise are motivating prudential reasons grounded in those practices backing the relevant mandatory norms with the threat of a coercive sanction, the claim that these constitutive practices give rise to objective

Sanctions Can Explain Exclusionary Reasons  167 exclusionary motivating reasons to comply with mandatory legal norms governing non-​official behavior presupposes, rather than refutes, the Coercion Thesis. The Coercion Thesis thus plays a logically indispensable role here in determining whether the practices constituting something as a system of law give rise to second-​order objective exclusionary motivating reasons to comply. If the Coercion Thesis is false, then there is nothing else in these constitutive practices reasonably contrived to give rise to objective first-​order motivating reasons to comply with mandatory legal norms governing non-​official behavior. If what work the Coercion Thesis can do by way of showing how law is reasonably contrived to give rise to such reasons is not sufficient to vindicate the claim that these constitutive practices are reasonably contrived to give rise to objective exclusionary motivating reasons, then that claim is false. If the Coercion Thesis cannot explain how the practices constituting something as a system of law are reasonably contrived to do this, then that fact is a refutation of the claim that these constitutive practices are reasonably contrived to give rise to exclusionary motivating reasons.

5.  Can the Coercion Thesis Explain How Law Creates Objective Exclusionary Motivating Reasons? Raz denies that law’s conceptual normativity can be comprehensively explained by the Coercion Thesis: if it is true that “the fact that a law is backed by a sanction is never an exclusionary reason (PRN 161–​62),” it follows that prudential considerations cannot give rise to objective exclusionary motivating reasons; prudential considerations are necessarily irrelevant, then, with respect to explaining how the practices constituting something as a legal system are reasonably contrived to give rise to objective second-​order exclusionary motivating reasons favoring compliance with mandatory legal norms governing non-​official behavior. This is a metaphysical claim grounded in what Raz takes to be objective norms of practical rationality. Since (1) there are no conceptually possible worlds like ours in which beings like us are justified, as an objective matter of normative practical rationality, in treating prudential considerations as giving rise to exclusionary motivating reasons to comply and (2) the only objective motivating reasons to which the threat of coercive sanctions can give rise are prudential in character, the threat of a coercive sanction cannot, as a conceptual matter, give rise to an objective exclusionary motivating reason to comply with norms backed by such a threat. Claims (1) and (2) entail that it is practically irrational, and not just practically unjustified, to treat mandatory legal norms governing non-​official behavior as giving rise to an exclusionary motivating reason to comply in virtue of being backed by the threat of a coercive sanction. Insofar as there are no conceptually possible worlds in which rationally competent subjects like us are justified, as an objective matter of normative practical rationality, in treating the threat of a

168  The Order Problem of Legal Normativity coercive sanction as giving rise to an exclusionary motivating reason not to act on first-​order motivating reasons favoring non-​compliance, it would be practically irrational to treat the threat of a coercive sanction as such: while it can be practically rational to do what is contingently unjustified as an objective matter of normative practical rationality, it is practically irrational to do what is necessarily unjustified as an objective matter of normative practical rationality. Doing what is practically irrational in virtue of being necessarily unjustified as an objective matter of normative practical rationality need not call the agent’s general competence into question but it does raise questions about what might explain the agent’s decision in circumstances where it is nonetheless reasonable to presume her general competence. When someone generally competent in virtue of being characteristically responsive to objective norms of practical rationality departs in such unpredictably spectacular fashion from what those norms permit or require, it raises questions about whether her decision was subject to illicit influences, such as unjustified coercive pressure, or about whether it was made under circumstances in which the operation of her practical faculties was impaired, such as occurs when under the influence of an intoxicant. The claim that prudential considerations are metaphysically incapable of giving rise to objective exclusionary motivating reasons is not wholly implausible. To begin, the idea that prudential standards can give rise to objective exclusionary motivating prudential reasons that bar acting on first-​order motivating aesthetic reasons is problematic if prudential and aesthetic standards govern classes of acts with little overlap. More to the point, it is not preposterous to think that objective motivating prudential reasons to do a can never trump competing objective motivating moral reasons to which moral obligations give rise; the claim that one has a moral obligation not to do a seems to imply that one is morally barred from doing a regardless of the extent to which it conduces to one’s self-​interest. Even so, the idea that prudential considerations cannot give rise to second-​order exclusionary motivating reasons that objectively bar acting on some class of first-​ order motivating reasons is less plausible than might initially appear. Suppose, as a first step towards seeing why, that one believes correctly, is epistemically justified in believing, and knows one is epistemically justified in believing the following claims:  (1) there exists an all-​perfect being G who knows everything about the world that can be known; (2) G enforces a norm prohibiting a with eternal torment unmatched by any suffering that can be experienced in this world; (3) such torment is imposed as punishment on an agent immediately after the agent does a; (4) G is morally justified in immediately enforcing divine norms with eternal torment because one deserves immediate eternal torment for violating any objective mandatory moral norm that G deems important enough to enforce with the threat of eternal torment; and (5)  doing a typically results in transiently tremendous pleasure without any other prudential costs than eternal torment.

Sanctions Can Explain Exclusionary Reasons  169 Under these far-​fetched but logically coherent assumptions, it is plausible to think that one has an objective exclusionary motivating moral reason not to act on first-​order prudential motivating reasons that favor doing a. If mandatory moral norms give rise to objective exclusionary motivating reasons in virtue of being justified under the relevant objective norms of practical rationality, then G’s command that one not do a expresses a mandatory moral norm that gives rise to an objective first-​order motivating moral reason not to do a and an objective second-​ order exclusionary motivating moral reason not to act on first-​order motivating reasons that favor doing a; if it is morally wrong to do a, then it is morally wrong to act on any first-​order motivating reasons to do a. Taken together, these moral reasons objectively trump all first-​order prudential motivating reasons to do a.10 But one arguably also has a prudential motivating reason not to do a that, as an objective matter of normative practical rationality, trumps all first-​order prudential motivating reasons to do a. If prudential considerations are the only relevant considerations, then one has a motivating prudential reason to abstain from doing a that, as an objective matter of normative practical rationality, is necessarily conclusive with respect to every other motivating prudential reason one might have to do a; there could not be an objective motivating prudential reason to do a, consistent with the world being as much like ours as is compatible with (1) through (5), that defeats the objective motivating prudential reason to abstain from doing a. There are a couple of ways, as discussed in Chapter 5, that one motivating reason might be objectively trumped by another. One motivating reason p might trump another q in virtue of expressing a value that is non-​commensurable with, but qualitatively higher than, the value expressed by q; Mill believes, as we saw, that intellectual pleasures are qualitatively higher pleasures that necessarily trump any lower pleasures—​no matter how large the quantity of lower pleasure or how small the quantity of higher pleasure. Or p might trump q, as is the case here, in virtue of expressing a commensurate value so great that it cannot be defeated by any countervailing values expressed by q.

10 That said, it is not clear what work an objective second-​order exclusionary motivating moral reason is needed to do in moral deliberations about what should be done. An objective exclusionary motivating moral reason would be needed to trump a set of first-​order motivating prudential reasons only if it is true that the first-​order motivating prudential reasons to perform some morally prohibited act a objectively outweigh the first-​order motivating moral reason to abstain from a. That might not be obviously preposterous, but it is odd: the idea that whatever first-​order moral reason we have not to kill an innocent person might be outweighed, as an objective matter of normative practical rationality, by the prudential benefits of doing so (e.g. someone is really rich and about to write me out of her will) is difficult to make sense of if our ordinary evaluative practices are any indication. But if it is assumed that the first-​order motivating moral reasons provided by a mandatory moral norm objectively outweigh all prudential reasons favoring non-​compliance, then there is simply no work that an objective exclusionary motivating moral reason could do in guiding one’s practical deliberations; the first-​order motivating moral reasons necessarily decide the issue. See K.E. Himma, Authority and the Nature of Law (forthcoming 2021) for more discussion of these issues.

170  The Order Problem of Legal Normativity Even so, it should nonetheless be clear that, as an objective matter of normative practical rationality, there is nothing problematic about treating the relevant trumping prudential considerations as giving rise to an exclusionary motivating reason not to act on any first-​order prudential motivating reasons to do a. Even if p objectively trumps q because it expresses a value that is commensurate but infinitely weightier than the value expressed by q, it is objectively unproblematic to treat p as giving rise to an exclusionary motivating reason not to do what q favors doing. The prudential stakes are so high that if that is what an agent needs to do to ensure that she not do a, there is nothing objectively irrational about doing so—​even if doing so is not warranted by objective norms of practical rationality. But a stronger claim is plausible here: insofar as we are easily tempted to do what we should not do for the sake of experiencing a lower pleasure, we should, as an objective matter of normative practical rationality, treat the threat of eternal torment as a peremptory motivating prudential reason not to think about any first-​order motivating reasons favoring doing a. Since there is nothing that could be worse, on any remotely plausible objective prudential calculus, than eternal torment, there are no first-​order prudential motivating reasons to do a that would defeat, as an objective matter of purely prudential rationality, the first-​order prudential motivating reasons not to do a; if the only way beings like us can avoid succumbing to a temptation that has eternal torment as its consequence is to steer clear of even thinking about doing so, then objective norms of practical rationality would require treating the threat of eternal torment as giving rise to a peremptory motivating reason not to think about any first-​order motivating prudential reasons to do a. One need not resort to such fantastic thought-​experiments to see that one might have an objective peremptory motivating prudential reason to comply with a directive. An armed robber’s admonition “Don’t even think about it” expresses an obvious objective requirement of normative practical rationality; if just thinking about “it” makes it even slightly more likely that one does it, then one should not, as an objective matter of normative practical rationality, even think about it. Prudential considerations can ground peremptory reasons that objectively bar thinking about first-​order motivating prudential reasons favoring non-​compliance. But insofar as a competent agent would be justified under objective norms of practical rationality in treating the threat of eternal torment as a second-​order peremptory motivating prudential reason not to think about any first-​order motivating prudential reasons to do a, she would also be justified in treating that threat as giving rise to a second-​order exclusionary motivating prudential reason not to act on any first-​order motivating prudential reasons to do a. Any considerations grounding a second-​order objective motivating prudential reason that bars thinking about first-​order motivating prudential reasons to do a also ground a second-​order objective motivating prudential reason that bars acting on first-​ order motivating prudential reasons to do a; one can, as a logical matter, have an

Sanctions Can Explain Exclusionary Reasons  171 objective motivating reason not to think about some class of reasons only if one also has an objective motivating reason not to act on that class of reasons. It is true that one would have two second-​order objective motivating exclusionary reasons not to act on any first-​order prudential reasons that favor doing a—​one that has its source in moral considerations and one that has its source in prudential considerations, but this is not a problem. There is nothing in our conceptual or evaluative practices that precludes the possibility of two objective second-​order motivating reasons grounded in different basic sources of normativity that exclude the same considerations. While it is true that objective norms of practical rationality would overdetermine the result, one can have multiple objectively compelling motivating reasons not to do something immoral that would result in prudentially catastrophic consequences: if suicide is morally impermissible and punished by hell, we have compelling prudential and moral reasons not to do something intended and reasonably contrived to bring about our own premature deaths. More to the point, it is not clear that there could not be a second-​order exclusionary motivating prudential reason that objectively bars acting on all first-​order moral motivating reasons to do something. Consider this modification of the thought-​experiment above: (1*) there exists an omnipotent being G* who knows everything about the world that can be known; (2*) G* enforces a norm prohibiting a with eternal torment unmatched by any suffering that can be experienced in this world; (3*) such torment is imposed as punishment on an agent immediately after she does a; and (4*) G* is not morally justified in enforcing the norm prohibiting a with eternal torment because doing a is required by objective norms of substantive morality. The world described by (1*) through (4*) differs from the one described earlier by (1) through (5) in that G* punishes what is morally required while G punishes only what is morally prohibited; if G is a morally benevolent sovereign in the world described by (1) through (5) above, G* is a morally flawed sovereign in the world described by (1*) through (4*). There is nothing in objective norms of practical rationality that obviously precludes one’s being justified in not doing a in this modified example—​despite the fact that doing a is required by objective norms of substantive morality. Suppose that the act a that G* has prohibited with the threat of eternal torment is merely calling an ambulance; and suppose, further, that P’s life is in danger and that Q can save P’s life only by doing a. While it might be morally wrong for Q to refrain from doing a to save P’s life, it is not obvious that Q would not be justified, as an objective matter of normative practical rationality, in letting P die to avoid eternal torment; there are moral limits, after all, to what prudential sacrifices can be legitimately demanded of someone. But if Q can be objectively justified in not doing a, then Q can also be objectively justified in treating the threat of eternal torment as an exclusionary motivating prudential reason not to do a that bars acting on first-​order motivating moral reasons to do a.

172  The Order Problem of Legal Normativity It is plausible to hypothesize that rationally competent self-​interested subjects like us are likely, in effect if not intent, to treat G*’s threat of eternal torment as giving rise to an exclusionary motivating prudential reason not to do a that bars acting on any first-​order motivating reason—​moral or prudential—​to do a. Even if we were to regard someone who does the morally right thing a here as a hero, and not as someone with a psychologically problematic desire to martyr herself to the worst suffering imaginable, the fact that we regard someone as a hero for doing so suggests, if not entails, that objective norms of practical rationality do not require doing the morally right thing here. As far as the shared views about normative practical rationality underwriting our practices with respect to law are concerned, we should regard the threat of eternal torment as an exclusionary motivating reason that bars acting on the objective motivating moral reason not to do a. Further, it is not preposterous to think that rationally competent self-​interested subjects like us are more likely, as a descriptive matter of contingent fact, to treat G*’s threat of eternal torment for doing a as an exclusionary prudential motivating reason not to act on any first-​order reasons to do a than we are to regard valid or justified mandatory moral norms as giving rise to exclusionary motivating moral reasons not to act on first-​order motivating prudential reasons to do a. Perhaps Abraham was willing to kill his son at God’s command because he believed he had an absolute moral obligation to do what God commands; or perhaps he was willing to subordinate his moral beliefs to placate an angry sovereign God who demands fear and trembling rather than love and devotion. But, either way, anyone familiar with that story would surely sympathize with Abraham had he killed Isaac out of a desire to save himself from eternal torment. It might be, all things considered, morally wrong to do so; but insofar as it is it would tempt even the morally best of us, it must be because doing so is consistent with, if not dictated by, objective norms of practical rationality. One might think that what most of us would do with respect to G*’s threat, as a descriptive matter of contingent fact, and what we should do, as an objective matter of normative practical rationality, come apart here, but this is not obvious. Insofar as it is true that rationally competent self-​interested subjects like us would characteristically refrain from calling an ambulance to save P’s life because we believe doing so is objectively justified all things considered, it is plausible to think that objective norms of practical rationality allow, or even require, us to refrain from doing something morally good when doing so is punished with eternal torment. It would be quite odd if objective norms of practical rationality require doing something that few of us would be remotely willing to do because doing so would incur such prudential detriment. But it does not matter for my purposes whether we are justified, as an objective matter of normative practical rationality, in treating the threat of eternal torment as a second-​order exclusionary motivating prudential reason not to act on first-​order motivating reasons to do a; what matters is that nearly all of us, as a descriptive

Sanctions Can Explain Exclusionary Reasons  173 matter of contingent fact, would treat it as such in our deliberations about whether to do a. If rationally competent self-​interested subjects like us are characteristically likely, as a descriptive matter of contingent fact, to treat the threat of eternal torment as an exclusionary motivating prudential reason not to do a, then there is no reason to think that the practices constituting something as a system of law must be reasonably contrived to give rise to objective exclusionary motivating reasons to succeed in regulating non-​official behavior. It is enough, then, that these practices are reasonably contrived to give rise to something rationally competent self-​interested subjects like us are characteristically likely, as a descriptive matter of contingent fact, to treat it as an exclusionary motivating prudential reason to comply with mandatory legal norms governing non-​official behavior that are backed by the threat of a non-​trivial coercive sanction. At the end of the day, law is concerned to regulate the non-​official behavior of rationally competent self-​interested subjects like us who are not always optimally responsive to objective norms of practical rationality. While law typically does this by predicting we will conform our behavior to these objective norms of practical rationality, there is no reason to think that law cannot anticipate these occasional departures from those norms. The threat of eternal torment unmatched in suffering by anything that can be experienced in this world is one thing and incarceration is another. But the difference is a matter of degree: suffering unmatched by anything that can be experienced in this world is still pain—​and pain is pain, regardless of whether its duration is finite or eternal, and regardless of whether such pain is of a better or worse kind. There is simply no reason to think it is objectively irrational for rationally competent subjects like us to regard the threat of coercive sanctions for non-​compliance with the law as giving rise to exclusionary motivating reasons not to act on first-​ order motivating reasons favoring non-​compliance; rationally competent subjects like us frequently do things that are not the best things to do, as an objective matter of normative practical rationality, and this need not involve objective irrationality. The practices constituting something as a system of law are, assuming the Coercion Thesis is true, nonetheless reasonably contrived to give rise to something rationally competent subjects like us are characteristically likely, as a descriptive matter of contingent fact, to regard as exclusionary motivating prudential reasons not to act on first-​order motivating moral reasons favoring non-​compliance, even if they often fail to do so in our world because the sanctions authorized for non-​ compliance are not severe enough; the practices constituting something as a system of law are as reasonably contrived to require morally problematic acts as they are to prohibit them. If there is any element in the practices constituting something as a system of law, construed to include the authorization of coercive sanctions for non-​compliance, likely to give rise to something rationally competent subjects characteristically regard as exclusionary motivating reasons of the right sort, it is the practices that consistently apply sanctions for non-​compliance.

174  The Order Problem of Legal Normativity

6.  The Claim That Law Provides Objective Exclusionary Motivating Reasons to Comply Explains Nothing That Needs to Be Explained Law’s conceptual normativity needs to be vindicated in order for us to make sense of why we adopt the practices constituting something as a system of law to do anything that law can be non-​idiosyncratically used to do. There is no plausible rationale for adopting the practices constituting something as a system of law to keep the peace, mediate disputes, make moral requirements more determinate, legitimately govern behavior, make it more likely that subjects comply with the requirements of right reason than they would otherwise be, or anything else that law can be non-​idiosyncratically used to do if these practices are not reasonably contrived to give rise to objective motivating reasons to comply with mandatory legal norms governing non-​official behavior. There is nothing about law’s conceptual normativity that cannot be explained without showing that these constitutive practices are reasonably contrived to give rise to objective exclusionary motivating reasons to comply with mandatory legal norms governing non-​official behavior. It suffices to explain how these constitutive practices are reasonably contrived to do everything that they are needed, characteristically used, and supposed to do, as a functionally normative matter, that they back mandatory legal norms governing non-​ official behavior with detriment sufficiently severe to deter enough violations to enable us to live together in something that counts as a community; if a rationally competent subject is not moved by the objective first-​order motivating prudential reasons favoring compliance to which the threat of a sanction gives rise, it is hard to see why she would be moved by an objective second-​order exclusionary motivating prudential reason to which that threat might also be thought to give rise. Nor is there any reason, for what it is worth, to think that these constitutive practices are ill-​equipped to efficaciously regulate non-​official behavior unless rationally competent judges characteristically regard themselves as having, as a descriptive matter of contingent fact, a second-​order exclusionary motivating reason to comply with mandatory legal norms governing official behavior. Judges characteristically regard the mandatory legal norms governing official behavior as giving rise to strong first-​order motivating reasons to apply the relevant norms governing non-​official behavior in the appropriate way when deciding disputes that arise under these norms in virtue of taking the internal point of view towards the norms governing their behavior; insofar as judges characteristically regard the norms governing their behavior as giving rise to such first-​order motivating reasons, they can resolve disputes and enforce these norms governing non-​official behavior in a manner that enables the system to do anything that a system of law can be used non-​idiosyncratically to do.

No Need to Explain Exclusionary Reasons  175 One might think that, even so, judges regard, as a matter of conceptual necessity, mandatory legal norms governing their behavior as also giving rise to second-​ order exclusionary motivating reasons to apply mandatory legal norms governing non-​official behavior in an exclusionary manner. On this line of reasoning, it is a conceptual truth that officials take the internal point of view towards the mandatory legal norms governing their behavior and hence regard it as providing second-​ order exclusionary motivating reasons to comply with those norms, which require judges to apply mandatory legal norms governing non-​official behavior in an exclusionary manner. This line of reasoning is problematic. Even if we assume that taking the internal point of view towards a norm for moral reasons involves, as a conceptual matter, regarding the norm as giving rise to second-​order exclusionary motivating moral reasons to comply, there is no reason to think that judges take the internal point of view towards the norms governing their behavior for moral reasons. Rationally competent self-​ interested judges, like any other rationally competent self-​ interested subject, make decisions about what to do with their lives for prudential reasons. Insofar as a rationally competent subject decides to become a judge for purely prudential reasons, she is thereby committed to regarding the norms governing her behavior as giving rise to first-​order motivating prudential reasons to comply. While she might also regard them as giving rise to second-​order exclusionary motivating moral reasons, it would be in virtue of additional moral commitments she might have—​and not in virtue of her taking the internal point of view towards those norms insofar as she is motivated to do so for purely prudential reasons. It is hence not a conceptual truth that judges view mandatory legal norms governing their behavior as giving rise to second-​order exclusionary motivating moral reasons to comply in virtue of taking the internal point of view—​insofar as it is conceptually possible to take the internal point of view towards a norm for purely prudential reasons. Whether it is a conceptual truth that judges view mandatory legal norms governing their behavior as giving rise to second-​order exclusionary motivating reasons of some kind depends on whether prudential considerations can give rise to such reasons and on whether the only possible motivations for taking the internal point of view towards these norms are moral and prudential; if prudential considerations can give rise to such reasons and there are no other possible motivations than moral and prudential motivations, then it follows that judges who take the internal point of view towards these norms, as a conceptual matter, view them as giving rise to second-​order exclusionary motivating reasons that are either moral or prudential in character. One might worry that judges whose commitment to conform to mandatory legal norms governing official behavior is motivated exclusively by prudential considerations are vulnerable to corrupting influence and need something that can offset any prudential motivating reasons they might take themselves to have that

176  The Order Problem of Legal Normativity favor non-​compliance in certain cases. Judges are no less rationally self-​interested than anyone else and should not be presumed to be less responsive to the normative prudential considerations that motivate rationally competent subjects to violate mandatory legal norms governing non-​official behavior. But this is a problem that we have to worry about in our world not because it is conceptually possible for rationally competent self-​interested judges to take the internal point of view towards mandatory legal norms governing their behavior for purely prudential reasons but because it is sufficiently likely that they do that we need to enact measures to protect against the dangers of corruption. There is no need, after all, for us to protect ourselves against risks that have no appreciable probability of being realized. Not surprisingly, systems of law provide enforcement mechanisms to deter such departures that are grounded in the same objective norms of practical rationality as those that systems of law authorize to ensure that rationally competent subjects comply with mandatory legal norms governing non-​official behavior—​in the form of coerced removal sanctions. Legal norms authorizing the coerced removal of judges from office for misconduct or incompetence are enough to provide objective first-​order motivating prudential reasons to comply with mandatory legal norms governing their official behavior—​in cases where an additional inducement might be necessary to augment their own commitments expressed by the internal point of view that they take towards these norms. Statutes authorizing other officials of the system to remove judges for misconduct or incompetence are reasonably contrived to provide something that rationally competent self-​interested judges characteristically regard, as a descriptive matter of contingent fact, as a strong first-​order motivating prudential reason to comply with mandatory legal norms governing their behavior because they should regard, as an objective matter of normative practical rationality, it as such. Those statutes constitute a legal system as reasonably contrived to give rise to objective first-​order motivating reasons to comply with such norms that are strong enough to ensure that the system can do everything that it is needed, used, and supposed to do, as a functionally normative matter. This is not to suggest that judges characteristically regard themselves as having only first-​order motivating prudential reasons to comply with mandatory legal norms governing official behavior that are defined by the threat of coerced removal—​any more than the arguments in this volume suggest that ordinary people characteristically regard themselves as having only first-​order motivating reasons to comply with mandatory legal norms governing non-​official behavior that are defined by the threat of incarceration. Judges are as plausibly assumed, absent exceptional circumstances, to be as morally decent as anyone else. The point of authorizing coercive sanctions to enforce mandatory legal norms of any kind is to provide an additional objective first-​order motivating reason to abstain from undesirable acts in cases where someone’s prudential interests in doing something

No Need to Explain Exclusionary Reasons  177 wrong are sufficiently seductive, without the threat of a coercive sanction, to tempt her to do that something. Either way, it is false that an institutional normative system is not reasonably contrived to efficaciously regulate non-​official behavior unless judges characteristically regard themselves as having exclusionary motivating reasons to do anything. Even if we assume implausibly that rationally competent judges in legal systems like ours are characteristically familiar with the technical notion of an objective exclusionary motivating reason, there is little reason to think they believe that mandatory legal norms governing their official behavior give rise to an objective second-​order exclusionary motivating reason that bars applying mandatory legal norms governing non-​official behavior in a non-​ exclusionary manner. Insofar as judges in legal systems like ours know they are legally authorized to exercise a quasi-​lawmaking discretion to modify or overrule some pre-​existing legal directive or holding, they are not likely to believe that they have such objective second-​order exclusionary motivating reasons. Given that legal systems like ours are clearly sufficiently efficacious in regulating non-​official behavior in communities like ours, it follows that a legal system in which judges do not view mandatory legal norms governing their behavior as giving rise to such exclusionary motivating reasons is nonetheless reasonably contrived to regulate such behavior. The claim that these constitutive practices are reasonably contrived to give rise to objective exclusionary reasons to comply with mandatory legal norms governing either official or non-​official behavior, then, is simply not needed to explain anything that needs explanation in the conceptual theory of law. While it is plausible to think that our conceptual and evaluative practices entail that mandatory moral norms give rise to objective exclusionary motivating moral reasons that bar acting on first-​order motivating prudential reasons favoring non-​compliance, they do not entail that the practices constituting something as a system of law are reasonably contrived to give rise to objective exclusionary motivating reasons to comply with mandatory legal norms governing either official or non-​official behavior. A legal system is reasonably contrived to do anything it is needed, characteristically used, and supposed to do, as a functionally normative matter, only insofar as the practices constituting something as a system of law are reasonably contrived to give rise to objective first-​order motivating reasons for officials to comply with mandatory legal norms governing their behavior. Systems of law are needed and contrived to efficaciously regulate the non-​official behavior of rationally competent self-​interested subjects, but they can do this only by efficaciously regulating the official behavior of subjects who are rationally competent in virtue of being largely responsive to objective norms of practical rationality; if judges do not do what they are supposed to in adjudicating disputes arising under mandatory legal norms governing non-​official behavior, then subjects of these norms are not plausibly presumed to be likely to do what they are supposed to under these norms.

178  The Order Problem of Legal Normativity But, as we have seen, there is no mystery as to how the mandatory legal norms governing official behavior are reasonably contrived to give rise to objective first-​ order motivating reasons to comply. Insofar as it is a conceptual truth that officials take the internal point of view towards mandatory legal norms governing official behavior, it is also a conceptual truth that they regard these norms as giving rise to first-​order motivating reasons to comply. Insofar as one doubts that these subjective commitments can be optimally relied upon to ensure judges do what they are supposed to do, they can be supplemented, as is done in many existing legal systems, by authorizing the imposition of coercive removal sanctions for chronic or flagrant non-​compliance. Everything that needs to be explained about law’s capacity to do everything it can be non-​idiosyncratically used to do can be explained by the Coercion Thesis without assuming the practices constituting something as a system of law are reasonably contrived to give rise to objective exclusionary motivating reasons favoring compliance with mandatory legal norms governing non-​official behavior. Given that there is nothing in these constitutive practices other than those authorizing courts to impose sanctions for non-​compliance that is reasonably contrived to give rise to objective exclusionary motivating reasons to comply with mandatory legal norms governing non-​official behavior, law’s conceptual normativity cannot be vindicated without the Coercion Thesis. Given that vindicating law’s conceptual normativity does not require showing it is reasonably contrived to give rise to objective exclusionary motivating reasons, there is nothing needed to fully vindicate it that is not entailed by the Coercion Thesis. The Coercion Thesis has the logical resources to do all that is needed to enable a theory of law’s metaphysical nature to fully vindicate law’s presumed conceptual normativity.

8

The Coercion Thesis and the Content Problem of Legal Normativity A conceptual theory of law must explain how the practices constituting something as a system of law are reasonably contrived to efficaciously regulate non-​official behavior through the governance of norms metaphysically capable of guiding it. This requires showing how these practices are reasonably contrived (1) to inform subjects of what acts are prohibited by mandatory legal norms governing non-​official behavior and (2) to provide something rationally competent subjects characteristically regard, as a descriptive matter of contingent fact, as a rational inducement to comply because they should, as an objective matter of practical normativity, regard it as such. If an artifactual normative system is not reasonably contrived to do both of those things, it does not count as a system of law.1 The first problem is epistemic and easy to solve. Since these constitutive practices consist, as a conceptual matter, in the public application of publicly promulgated mandatory norms to subjects, these practices are well-​suited to ensure that subjects can determine what is prohibited enough of the time to enable law to do anything a system of law can be non-​idiosyncratically used to do, including minimally keeping the peace. It might not always be clear what is prohibited by some mandatory legal norm governing non-​official behavior because the language used to express the norm is vague or ambiguous; but the core content of anything consistently applied by courts as a mandatory legal norm can be identified, except in unusual cases, by observing judicial practices, reading the text of the relevant norm, or consulting an attorney. The second problem is practical and more challenging. To address this problem, one must (1) identify the source, order, and content of the objective motivating reasons to which systems of law are intended, characteristically used, and supposed, as a functionally normative matter, to give rise and (2) explicate how the practices constituting something as a system of law are reasonably contrived to give rise to such objective motivating reasons. Plausible answers to (1) and (2) are essential to explaining why beings like us adopt these constitutive practices to do everything law can be non-​idiosyncratically used to do; if we cannot explain how 1 If an artifact-​type A has a conceptual function, then something must be reasonably contrived to perform each of the tasks essential to performing that function to be properly characterized as an A. See Chapter 4. Coercion and the Nature of Law. Kenneth Einar Himma, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198854937.001.0001

180  The Content Problem of Legal Normativity these constitutive practices are reasonably contrived to give rise to objective motivating reasons of some plausible source, order, and content, then it is utterly opaque as to why we think law can be used to do anything that we characteristically use it to do. The Content Problem is concerned with explicating the content of the objective first-​order motivating reason to which the practices constituting something as a system of law are reasonably contrived to give rise. First-​order motivating reasons function to guide deliberations by favoring doings or abstentions; but since we can deliberate only by considering the content of various propositions, an objective first-​order motivating reason can guide deliberation only insofar as it has content that, as an objective matter of normative practical rationality, favors some doing or abstention. Law’s presumed conceptual normativity cannot be fully vindicated without identifying the content of the objective first-​order motivating reason to which these constitutive practices are reasonably contrived to give rise. This chapter rejects one possible solution to the Content Problem, arguing that there is nothing in objective norms of practical rationality that would encourage us, even presumptively, to obey a norm simply because it has the status of law.2 Since there is nothing else in these practices reasonably contrived to give rise to an objective motivating reason with different content, neither the How Problem nor the Content Problem can be solved without assuming that it is a conceptual truth that the practices constituting something as a system of law back some mandatory legal norms governing non-​official behavior with the threat of a coercive sanction that is reasonably contrived to give rise to a first-​order objective motivating reason to comply as a means of avoiding that sanction. We simply cannot make sense of why we do what we do with law unless the Coercion Thesis is true.

1.  Intrinsic and Instrumental Motivating Reasons Reasons for action express values. If motivating and justifying reasons make reference to standards having to do with what people should do, it is because those reasons express values the relevant standards are concerned to protect. These values are expressed in familiar substantive standards that govern publicly observable behaviors: prudential values are expressed in prudential standards; moral values are expressed in moral standards; and so on. But they are also expressed in

2 Obedience is conceptually distinct from compliance; one can comply with a norm without obeying it, but one cannot obey a norm without complying with it. The difference has to do with the motivation for satisfying the requirements of the norm: one can comply with a norm without being motivated to satisfy its requirements, but one cannot obey a norm without being motivated to satisfy its requirements; to do what a norm requires because the norm requires it is to obey the norm. While an intrinsic motivating reason is more accurately characterized as a reason to obey because of the motivational element it connotes, I will use both comply and obey here. Nothing of import turns on this.

Intrinsic and Instrumental Motivating Reasons  181 standards of practical rationality governing private deliberations about behavior. Both our public behaviors and private deliberations are norm-​governed.3 There are two kinds of value expressed by standards governing public behaviors: instrumental and intrinsic value. Something is instrumentally valuable if and only if it has value as a means to some antecedently valuable end. Money is an example of something with instrumental value; while money clearly has value as a means to desirable ends like health, knowledge, and pleasure, it is not typically regarded, and should not be regarded, as having value as an end-​in-​itself. Something is, in contrast, intrinsically valuable if and only if it has value for its own sake as an end-​in-​itself. One’s own happiness is something with intrinsic value; while it might make sense to value some other person’s happiness as a means to an end, it is not typically regarded, and should not be regarded, as having value as a means to some other end. Corresponding to the two kinds of value are two kinds of practical reasons. A practical reason is an instrumental motivating reason if the value it expresses is, or should be, regarded by a rationally competent subject P as instrumentally valuable as a means to some desirable end: it is a subjective instrumental motivating reason for P if, as a descriptive matter of contingent fact, P treats it in her practical deliberations as expressing something she ought to pursue as a means to realizing some end; it is an objective instrumental motivating reason for P if she should, as an objective matter of normative practical rationality, treat it in her practical deliberations as expressing something she ought to pursue as a means to realizing some end. The content of an instrumental motivating reason expresses the instrumental value promoted by the relevant class of standards by favoring that something be done as a means of realizing the value promoted by those standards: a prudential reason to go to a doctor when one has an easily treatable but otherwise fatal illness is an instrumental motivating reason to go to the doctor expressing the instrumental value of having it treated. An instrumental motivating reason functions to encourage acts and abstentions necessary to actualize some state of affairs that is valuable as a means to actualizing a state of affairs that is intrinsically valuable: the prudential reason to go to the doctor favors going to the doctor as a means of actualizing the intrinsically valuable state of affairs in which one survives the illness. A practical reason is an intrinsic motivating reason if the value it expresses is, or should be, regarded by the subject as intrinsically valuable in virtue of being valuable as an end-​in-​itself: it is a subjective intrinsic motivating reason for P if, as a descriptive matter of contingent fact, P treats it as expressing intrinsic value her behavior should realize; it is an objective intrinsic motivating reason for P if 3 By public, I mean acts that are outward in the sense that they can, in principle, be directly observed by other persons; by private, I mean acts that are mental and hence cannot, in principle, be directly observed by other persons.

182  The Content Problem of Legal Normativity she should, as an objective matter of normative practical rationality, treat it as expressing intrinsic value her behavior should realize. The content of an intrinsic motivating reason expresses the intrinsic value promoted by the relevant class of standards by favoring that something be done because those standards assert that it should be done: an intrinsic motivating moral reason not to kill expresses the content that one should not kill because abstaining from killing is required by morality. An intrinsic motivating reason functions to encourage acts or abstentions necessary to actualize some state of affairs that is intrinsically valuable because its value is independent of the value of any further states of affairs to which its actualization might conduce: the moral reason not to kill favors refraining from killing because the state of affairs in which one refrains from killing is intrinsically valuable.4

2.  Order and Value: The Relationships Among the Various Kinds of Reason A number of distinctions have been made in Chapters 5, 6, and 7 among various kinds of reason: (1) the distinction between practical and epistemic reasons; (2) the distinction between motivating practical reasons and justifying practical reasons; (3) the distinction between first-​order motivating reasons and second-​order exclusionary motivating reasons; (4) the distinction between subjective and objective motivating reasons; and (5)  the distinction between intrinsic and instrumental motivating reasons. Each helps us to understand the various respects in which the practices constitutive of law are reasonably contrived to do everything a system of law can be non-​idiosyncratically used to do. The distinction between intrinsic and instrumental motivating reasons is related to the other distinctions in some ways that are clear. The relationship between the distinction between intrinsic and instrumental motivating reasons and the distinction between subjective and objective motivating reasons is straightforward. One can, as discussed above, have intrinsic motivating reasons that are either subjective or objective (or both) in character, and one can have instrumental motivating reasons that are either subjective or objective (or both) in character; intrinsic and instrumental motivating reasons can hence be either subjective or objective (or both) in character.

4 More aptly put, this moral reason favors not-​killing because the state of affairs in which one kills is significantly dis-​valuable as an end-​in-​itself; a state of affairs in which one kills should be avoided for its own sake—​and not just because actualizing it might cause suffering or otherwise diminish utility. Killing is inherently wrong, though it might also be instrumentally wrong in virtue of causing suffering to others. Not-​killing is not like consensual hugging because not-​killing does not add positive value to the world; unlike consensual hugging, it simply avoids adding negative value.

Order and Value: The Various Kinds of Reason  183 The relationship between the distinction between intrinsic and instrumental motivating reasons and the distinction between first-​order and second-​order exclusionary motivating reasons is not as clear. One might think that first-​order motivating reasons are instrumental motivating reasons and second-​order exclusionary motivating reasons are intrinsic motivating reasons. First-​order motivating reasons, on this line of reasoning, can defeat other first-​order motivating reasons only by outweighing them because they express commensurable instrumental values that can be directly assessed against each other. Second-​order motivating reasons can defeat first-​order motivating reasons only by excluding or neutralizing their normative significance insofar as they express intrinsic value incommensurable with the instrumental value expressed by first-​order motivating reasons. Though plausible, this line of reasoning overlooks the complexity of the relevant values. Intrinsic motivating reasons can be treated as exclusionary motivating reasons that bar acting on countervailing first-​order motivating instrumental reasons, but they can also be treated as having the dimension of weight like first-​ order motivating reasons. Two intrinsic values might come into conflict in a manner that can be resolved only by comparing them to determine which wins in virtue of being weightier. Assuming that non-​human and human lives both have intrinsic value, the value of some human life might outweigh the value of some non-​human life in situations where both lives are in jeopardy and only one can be saved; it is commonly thought that the life of a human being should be saved in such circumstances because human lives are intrinsically more valuable than non-​human  lives. Further, an intrinsic value can come into conflict with an instrumental value in a manner that can be resolved only by directly weighing them against each other: if I regard rare stamps as valuable for their own sake and I regard baseball games as valuable as a means to the end of making new friends, I might have to decide whether to spend my money on one or the other by weighing the two values against each other. It is not implausible to think that the intrinsic value of a rare stamp might be outweighed by the instrumental value of a baseball game—​depending on how much I value the stamp relative to how much I value the baseball game given my past luck with making friends at games. Whatever the relationship between the two distinctions turns out to be, they do not directly track each other. It is not preposterous to think that only intrinsic motivating reasons function, both as an objective matter of normative practical rationality and as a descriptive matter of human psychology, as exclusionary motivating reasons not to act on first-​order motivating reasons. If moral obligations always objectively trump prudential considerations in practical deliberations, it is arguably because they express intrinsic values that, as an objective matter of normative practical rationality, trump prudential considerations not by outweighing them but by excluding them in a manner that bars agents from acting on them. While it is not obviously absurd to think there could be prudential considerations so weighty that they can

184  The Content Problem of Legal Normativity rationally be treated as exclusionary reasons expressing only instrumental values not to act on conflicting first-​order reasons, it is more natural to think they are properly treated only as outweighing all other prudential considerations.

3.  Morality, Conventions, and Intrinsic Motivating Reasons to Comply The distinction between intrinsic and instrumental reasons seems most conspicuously relevant in explicating our shared moral practices having to do with making judgments about a person’s motivations or character. As far as these practices go, it is plausible to think that one should be motivated to do the morally right thing, at least in part, because it is the morally right thing—​and not only because it conduces to self-​interest. Refraining from killing an innocent person is always the right thing to do but doing so only because one wants to avoid a lengthy period of incarceration expresses a configuration of motivational preferences that is problematic from a moral point of view. Doing the right thing reflects positively on one’s moral character in our ordinary judgments only insofar as one has the right kinds of motivation for doing it—​and these motivations include doing the right thing because it is the right thing to do. Doing the morally right thing because it is morally right is not the only legitimate motivation for doing it. It is morally appropriate, as an objective matter of normative practical rationality, to refrain from killing people because killing people is wrong; but it is also appropriate to refrain from killing because one empathizes with other people. Someone who refrains from killing because she empathizes with other people has done nothing that should elicit any moral concerns about her character, unlike someone who refrains from killing only for purely prudential concerns. But doing the morally right thing because it is morally right is not necessarily indicative of good moral character. It is not absurd to think that one should, as an objective matter of normative practical rationality, have other motivating reasons not to kill than the intrinsic motivating reasons associated with the mandatory moral norm prohibiting killing. Someone who lacks empathy and refrains from killing only because it is the morally right thing to do is certainly less scary than someone who refrains only to avoid incarceration, but she is still pretty damn scary: “if morality did not prohibit murder, I might just kill you” affords less reassurance than one might hope for in cases of doubt about someone’s intentions. A truly good person refrains from killing not just for intrinsic motivating moral reasons but also for the right kind of instrumental motivating prudential reason in virtue of connecting her sense of well-​being to that of people she does not know. The issue might seem more complicated if morality is consequentialist in character, but it is no less plausible to think that moral norms provide objective

Morality, Conventions, and Intrinsic Reasons  185 intrinsic motivating reasons even if substantive moral norms are conceived as concerned only with promoting certain consequences. It is perfectly appropriate, on a consequentialist view, to do something for the instrumental motivating reason of promoting the favored consequences in the sense that doing it for this reason does not raise any moral concerns about the quality of an agent’s character. But it is also appropriate, on a consequentialist view, to do something that maximally promotes well-​being because morality requires it. Every objectivist moral theory, deontological or consequentialist, is ultimately concerned with protecting some sort of objective intrinsic value. The ultimate justification, on any objectivist theory of morality, for doing something morally required is to protect something objectively valuable for its own sake. A mandatory moral norm requires doing a, on a deontological view, because doing a is valuable in virtue of its inherent properties and not because it is a means to achieve some favored inherently valuable state of affairs; it is wrong to lie, on Kant’s deontological view, even if doing so would save the life of an innocent person, because deception is inherently wrong. A mandatory moral norm requires doing a, on a consequentialist view, because doing a is a means to achieve some inherently valuable state of affairs and not because doing a is valuable in virtue of its inherent properties; whether it is wrong to lie about something, on a consequentialist view, depends on whether doing so conduces to a state of affairs that is intrinsically valuable. Morality is concerned with protecting something intrinsically valuable even if it is intersubjective in the sense that the content of any moral norm is determined by the conventional practices of the group with respect to evaluating the behavior of subjects governed by the norm. The relevant practices give rise to a system of norms that counts as a morality only insofar as those practices are concerned with protecting what is intrinsically valuable. In this case, what is intrinsically valuable is determined, within limits compatible with a plausible account of the nature of morality, by the practices of the relevant group.5 Insofar as morality is concerned to protect what is intrinsically valuable through the governance of norms that duly promote that value, those norms give rise to objective intrinsic motivating reasons to comply. Since mandatory moral norms necessarily give rise to objective motivating reasons to do a only insofar as doing a is necessary to evince due respect for an intrinsically valuable quality or state of affairs by actualizing it, those motivating reasons express a value that should be regarded by the subject as intrinsically valuable. Since morality protects what is

5 A configuration of practices that allow indiscriminate cheating, hitting, hurting, killing, lying, and stealing cannot give rise to a system of norms properly characterized as a morality. Something counts as a morality only insofar as it functions to trump the prudential interests of one agent in order to protect those of another, and a set of norms that generally allow people to infringe the prudential interests of others does not function to constrain agents in the conceptually required respects. There are thus conceptual content-​based conditions that the norms of a system must satisfy to be properly characterized as a morality.

186  The Content Problem of Legal Normativity objectively valuable for its own sake, mandatory moral norms give rise to what should be regarded, as an objective matter of normative practical rationality, by subjects as intrinsic motivating reasons. Not every class of norms is plausibly thought to give rise to something that is, or should be, regarded by subjects as intrinsic motivating reasons. The membership norms of an institutional organization, for instance, are framed to require behaviors valuable only in virtue of promoting ends deemed valuable by appropriate representatives of the organization. A membership norm governing people employed by an institution might require employees to keep time sheets not because doing so is intrinsically valuable but because doing so conduces to the employer’s goal of paying only for work done. While there might be something morally desirable about paying employees only for work done, this is not usually the motivation for requiring them to keep time sheets; the motivation is rooted in a prudential desire to avoid unnecessarily transferring instrumentally valuable resources from employers to employees. One might think that mandatory norms of etiquette require acts that should be done for their own sake and not simply as a means to some other end, but this is incorrect. The requirements of etiquette are concerned to ensure that people avoid acts likely to make public social interactions unpleasant: killing an innocent person might be a breach of etiquette but only if done at some public social gathering in a manner that makes social interactions awkward and difficult; killing someone in such circumstances would be a breach of etiquette in virtue of being rude. Although some mandatory moral norms prohibit acts also prohibited by norms of etiquette, that is because acts causing gratuitous unpleasantness are also morally problematic; there can be instrumental and intrinsic motivating reasons for doing the same thing. Artifactual norms of etiquette are reasonably contrived to give rise only to objective instrumental motivating reasons. The foregoing examples call attention to a problem with the idea that an artifactual normative system might be reasonably contrived to give rise to objective intrinsic motivating reasons to comply. Artifacts are, as a conceptual matter, contrived and used to do things; as such, they are used instrumentally to secure ends characteristically regarded by potential users as valuable—​whether as ultimate ends-​in-​themselves or as intermediate means to some ultimate end valued for its own sake. The problem is to explain how something used, by nature, for instrumental purposes could give rise to objective intrinsic motivating reasons that presuppose paradoxically that its use is intrinsically valuable. Artifactual membership norms differ from norms of law and morality in a number of ways relevant with respect to understanding law’s conceptual normativity. While one can opt out of being accountable under institutional membership norms or those of etiquette by removing oneself from the relevant social group, one cannot opt out of being accountable under moral or legal norms. One can avoid being judged under these moral or legal norms simply by avoiding people because

Morality, Conventions, and Intrinsic Reasons  187 judging is something that must be done by other people, but one cannot opt out of being accountable under those norms; there is no habitable place in this world where one can escape being accountable under either moral or legal requirements. Even if there is no one around to judge you morally or legally, you are accountable under moral or legal norms as long as there is someone with standing to judge you because these norms govern your behavior regardless of whether or not you have consented to being judged. Though norms of morality and law resemble each other in this respect, they differ in another: unlike norms of morality, norms of law do not have such reach in virtue of their content or in virtue of their being universally applicable across cultures.6 The rules of law have this kind of reach in our world only in virtue of the occurrence of certain contingent social facts—​namely those constituting something as a system of law that efficaciously regulates behavior within some geographic area; since every habitable space hosts a community governed by some existing legal system,7 one can voluntarily leave a legal system in our world only by entering into another. In contrast, there is no possibility of escaping being subject to moral norms if morality is objective in character because those norms govern rationally competent subjects wholly in virtue of their content—​and not even partly in virtue of the instantiation of any social facts like those giving rise to law.8 Norms of law and norms of objective morality resemble each other in another respect relevant to the problems of legal normativity. Insofar as it is a conceptually necessary condition for something to count as a system of law that it includes norms prohibiting morally wrongful assaults on persons and property, it follows that law and morality share certain normative content. Norms of morality and law resemble each other with respect to this particular but differ from institutional membership norms, norms of etiquette, and other artifactual norms, such as the rules of a game, insofar as these latter norms do not prohibit these wrongful assaults.

6 This is true even if morality is wholly conventional in the sense that the only valid moral norms are manufactured by “culture.” The relevant norms might be valid in virtue of cultural convergence; however, if they are not framed to apply to people in other cultures, they are not properly characterized as norms of morality. They are more appropriately characterized as a hybrid (and somewhat bizarre) species of social membership norms that typically apply to the kind of behaviors with which a set of norms must be concerned, as a conceptual matter, to count as comprising a morality. A cultural norm that prohibits intentional killing of innocent persons among members of that culture but is agnostic with respect to whether it is morally permissible for members of another culture to hack one another to pieces without provocation, whatever else it might be, is not properly characterized, according to our conceptual practices, as a norm of morality. 7 The existence of a community of persons in any geographic space is part of what makes it habitable. 8 The assumption that the norms of morality are objective, rather than artifactual, in character is crucial to this point. If the norms of morality are conventional and hence manufactured by some social group, then one’s behavior is not properly characterized as subject to moral evaluation regardless of whether there are any people in the world who could do the evaluating under the right circumstances; if there are no such people, then there is no community of people who could manufacture the relevant rules.

188  The Content Problem of Legal Normativity The problems associated with vindicating law’s conceptual normativity are complicated by the fact that norms of law, unlike objective norms of morality, resemble other artifactual norms in virtue of being contingently cultural in character but, unlike other artifactual norms, resemble objective norms of morality in virtue of their reach and content. Whether law’s culturally manufactured normativity can plausibly be explained in terms of objective intrinsic motivating reasons depends on whether law resembles morality with respect to qualities that explain why subjects should, as an objective matter of normative practical rationality, treat mandatory moral norms as giving rise to intrinsic motivating reasons.

4.  Law as a Source of Subjective Intrinsic Motivating Reasons to Comply One might argue that the practices constituting something as a system of law are reasonably contrived to give rise to objective intrinsic motivating reasons to comply because subjects characteristically believe these practices give rise to objective intrinsic motivating reasons to comply and treat them subjectively as such in deciding what to do. Since what competent agents should treat, as an objective matter of normative practical rationality, as motivating reasons depends in part on what competent agents, as a descriptive matter of contingent fact, characteristically treat as motivating reasons, the claim that the practices constitutive of law are reasonably contrived to give rise to objective intrinsic motivating reasons can, on this line of reasoning, be inferred from the claim that these practices characteristically give rise to subjective intrinsic motivating reasons. It is not uncommon to hear ordinary folks make assertions suggesting they believe something like this. Assertions that “it is wrong to disobey the law” and “people shouldn’t break the law” express the claim that we have some sort of objectively unqualified motivating reason to obey a mandatory norm with the status of law. The idea that we have such an objective motivating reason to obey anything with the status of law is plausibly interpreted as expressing the claim that we have an objective motivating reason to obey mandatory legal norms governing non-​ official behavior because these norms have the status of law—​i.e. we have an objective intrinsic motivating reason to obey law. The claim that we have an objective motivating reason to obey law because it is law is ambiguous between two interpretations. On the first interpretation, it expresses something about the laws of the particular legal system in which one is a subject: thus construed, the claim is that if l is a mandatory norm recognized according to the requirements of the particular legal system L in which one is a subject, then one has an objective motivating reason to obey l because it is a law in L. On the second interpretation, it expresses a more general claim about the laws of anything that counts as a legal system: thus construed, the claim is that if l is

Law as Source of Subjective Intrinsic Reasons  189 a mandatory norm in something properly characterized as a legal system, then l gives rise to an objective motivating reason to comply in subjects of that system because l was enacted or recognized in a system that instantiates the practices constituting something as a legal system. Only the second interpretation is relevant for our purposes. On the first interpretation, the claim makes a statement about certain distinguishing properties of the particular legal system in which the speaker is a subject: properly fleshed out to expose what is implicit in this claim, the idea is that rationally competent subjects of L have an objective motivating moral reason to obey the laws of L because L satisfies the morally necessary conditions for being justified in governing behavior—​and not because L satisfies the conceptually necessary conditions for something to count as a system of law. On the second interpretation, that claim makes a statement about properties shared by every conceptually possible legal system: properly fleshed out to expose what is implicit in this claim, the idea is that the properties common to every conceptually possible legal system give rise to objective intrinsic motivating reasons for rationally competent subjects of that system to obey the law. Since the first interpretation does not purport to express a conceptual truth about law while the second does, only the second interpretation is equipped to do anything that would contribute to explicating law’s conceptual normativity. The more plausible characterization of what people can be presumed to believe on the basis of assertions like the ones described above is expressed by the first interpretation. Since, on the one hand, people do not usually have theoretical beliefs about the concept of law, there is no reason to interpret these common assertions as expressing a claim about all conceptually possible legal systems. Since, on the other, they usually do have beliefs about the moral legitimacy of the legal system in which they are subjects, these assertions are more plausibly interpreted as expressing a belief about what they take to be morally favorable features of the system in which they are subjects. Given that the more plausible interpretation of these assertions entails nothing about what objective motivating reasons rationally competent subjects of other legal systems might have to obey law, they neither assert nor entail that the practices constituting something as a system of law are reasonably contrived to give rise to objective intrinsic motivating reasons to comply. The inference of objective intrinsic motivating reasons to comply from these oft-​heard assertions is logically fallacious, then, because the only belief that can be plausibly attributed to people on the basis of sentences like “it is wrong to break the law” is that they believe it is morally wrong to break the laws of the system of which they are subjects. It is not just that there is nothing in that interpretation of the claim that would imply the second interpretation; it is also that it is eminently plausible to hypothesize that, as a descriptive matter of contingent fact, competent agents would reject the attribution expressed by the second interpretation—​ either because they have not thought about the matter or because they believe it is

190  The Content Problem of Legal Normativity necessarily wrong to obey laws requiring wicked acts.9 The argument fails, then, because it falsely presupposes that rationally competent agents characteristically believe that the practices constituting something as a legal system give rise to an objective intrinsic motivating reason to comply.

5.  Law as a Source of Objective Intrinsic Motivating Reasons to Comply The most reliable way to ascertain whether there is anything in the practices constituting something as a system of law reasonably contrived to give rise to objective intrinsic motivating reasons is to consider the matter directly from the standpoint of objective norms of practical rationality. I argue below that (1) there is no plausible intuitive source of an objective intrinsic motivating reason to obey mandatory legal norms governing non-​official behavior; and (2) even if there were, there is nothing in the content of objective norms of practical rationality, as we understand them, that requires that we treat these constitutive practices as giving rise to an objective intrinsic motivating reason to comply.10

5.1  There Is No Plausible Source of an Objective Intrinsic Motivating Reason to Comply There are only three basic sources of objective motivating reasons. Insofar as practically rational agents characteristically value things from only aesthetic, moral, and prudential points of view, an objective intrinsic motivating reason to comply would have to be either aesthetic, moral, or prudential in character. If the practices constituting something as a system of law are not reasonably contrived to give rise to objective intrinsic motivating reasons that are either aesthetic, moral, or prudential in character but are nonetheless reasonably contrived to give rise to such 9 There could be, on this latter view, no objective motivating reason to kill an innocent person because it is required by something that counts as a system of law. 10 A  methodological reminder might be helpful here. The defense of claims (1)  and (2)  will be grounded in what I take to be shared views about what is required by objective norms of practical rationality—​and not in claims about what is actually required by these mind-​independent norms. The reason is not that I am assuming that we cannot be epistemically justified in beliefs about what is required by objective norms to which we lack direct access; that is for normative epistemologists to decide, though I am admittedly skeptical about the idea that we have epistemically reliable access to the content of such standards. The reason is that what matters from the standpoint of a modest approach to conceptual theory is not what is objectively correct about the content of these norms; what matters are the ordinary views that help to explain why we do what we do with the thing we call law. We converge in doing what we do with law because we converge in believing things having to do with why we need law.

Law as Source of Objective Intrinsic Reasons  191 reasons, they would have to be grounded in a source that is autonomous in the sense that reasons grounded in this source cannot be explained in terms of aesthetic, moral, or prudential motivating reasons. None of these basic sources give rise to an objective intrinsic motivating reason to obey law. Aesthetic considerations are clearly irrelevant. Even if we assume that the practices constituting something as a system of law should be valued from an aesthetic point of view as ends-​in-​themselves in virtue of their aesthetic qualities, such value could not give rise to an objective intrinsic motivating reason to obey law. Aesthetic norms might entail that these constitutive practices are beautiful and hence should be pursued, but that would not tell us anything about whether the normative output of those practices is beautiful; a beautiful set of social practices might produce norms, like those of the Fugitive Slave Act, that are neither beautiful nor require what is beautiful. Moral considerations cannot give rise to an objective intrinsic motivating reason to obey law because such considerations can give rise only to objective motivating moral reasons to obey the laws of a morally legitimate legal system. Since there can be legal systems so wicked they do not give rise to objective motivating moral reasons to comply, moral considerations cannot give rise to something subjects should, as an objective matter of normative practical rationality, regard as a motivating reason to obey law because it is law. There is no plausible theory of anything that entails that law-​ subjects in Nazi Germany had even a defeasible objective motivating moral reason to obey a law of that system simply because it was the law in that system. Prudential considerations cannot, as a conceptual matter, give rise to an objective intrinsic motivating reason to obey law because the value expressed by an intrinsic motivating reason is independent of any prudential benefits that compliance might produce. An objective intrinsic motivating reason to comply favors compliance regardless of whether it conduces to self-​interest and would hence favor compliance even if it causes harm to the agent; such a reason cannot, for this reason, be logically derived from the objective prudential standards defining objective motivating prudential reasons. While it is always possible for a rationally competent subject to mistreat prudential considerations as giving rise to such reasons, the existence of objective intrinsic motivating prudential reasons is conceptually impossible—​and straightforwardly so. The only possible source of objective intrinsic motivating reasons to obey law would have to be logically independent of the other autonomous sources of motivating reasons. If we have an objective intrinsic motivating reason to obey the law, it is in virtue of normative considerations independent of aesthetic, moral, and prudential considerations that would have to arise from some practice with objective normative significance common to all conceptually possible legal systems. The practices constituting something as a legal system would thus have to be, if there

192  The Content Problem of Legal Normativity are such reasons, an autonomous source of objective intrinsic motivating reasons that are neither aesthetic, moral, nor prudential in character.11 The claim that these constitutive practices are an autonomous source of objective intrinsic motivating reasons is implausible. With the exceptions of those practices recognizing, applying, and enforcing norms prohibiting acts likely to breach the peace among subjects and those asserting law’s normative supremacy, the practices constitutive of law are also partly constitutive of every other artifactual system of norms; artifactual normative systems are created by a proper subset of the social practices that create legal systems. It is, on the one hand, clear that there is nothing in this proper subset of practices reasonably contrived to give rise to objective motivating reasons of any kind; it is silly to think that every artifactual normative system is reasonably contrived to autonomously give rise to objective intrinsic motivating reasons to comply. But there is, on the other hand, nothing in the practices distinguishing systems of law from other classes of artifactual normative systems that jumps out as reasonably contrived to give rise to an ontologically novel class of objective motivating reasons. More importantly, there is nothing in our behavior, verbal or non-​verbal, that suggests we are even remotely aware of the existence of such autonomous objective motivating reasons. The idea that we could have something that counts as an objective motivating reason grounded in an autonomous source of normativity utterly unknown to us borders on incoherence. It is true that we are sometimes unaware that we have some specific objective motivating reason to do something; otherwise we could not be mistaken about what our objective motivating reasons are. But the idea that we have a class of objective motivating reasons of which we are characteristically unaware, as a descriptive matter of contingent fact, implies that we should believe, as an objective matter of normative practical rationality, something we have no epistemic reason to believe—​a claim that is strikingly counterintuitive if not outright nonsense.

5.2  Objective Norms of Practical Rationality Do Not Require Treating Law as Giving Rise to an Intrinsic Motivating Reason Even if one is convinced by the argument of the last subsection, it would be helpful to consider the content of these objective standards more directly. Whether the practices constituting something as a legal system, according to these standards, are reasonably contrived to give rise to intrinsic motivating reasons depends on whether there is some element of these constitutive practices that rationally 11 These autonomous motivating reasons would be logically basic in the sense that they cannot be derived from any simpler sources of normativity but not intuitively basic in the sense that aesthetic, moral, and prudential motivating reasons are.

Law as Source of Objective Intrinsic Reasons  193 competent subjects should regard as having enough intrinsic value to give rise to objective intrinsic motivating reasons to comply. If there is nothing in those practices, which define the conceptually necessary properties for the existence of a legal system, with such value, then it is false that they are reasonably contrived to give rise to objective motivating reasons to obey the law because it is law. On any plausible account of these constitutive practices that excludes the Coercion Thesis, the existence conditions for a legal system include at least the following properties: to count as a system of law in the purely descriptive sense of the term, an institutional normative system must (1) include a social rule of recognition constructed by the cooperative convergence of officials on satisfying its requirements with respect to what they may recognize and apply as law; (2) contain mandatory legal norms governing non-​official behavior that prohibit some violent assaults on persons and property; (3) be metaphysically capable of moral legitimacy; (4) regulate behavior through the governance of norms metaphysically capable of guiding behavior; and (5) keep the peace enough to enable subjects to live together in something plausibly characterized as a community. There is nothing here that should, as an objective matter of normative practical rationality, be regarded by competent agents as intrinsically valuable—​much less as so valuable for its own sake that it gives rise to objective intrinsic motivating reasons to comply. To begin, there is nothing in the idea that officials in the system cooperatively converge in following certain recipes for recognizing and applying law that should, as an objective matter of normative practical rationality, be regarded by rationally competent subjects of the system as intrinsically valuable. The fact that members of a criminal gang cooperate with one another to achieve its objectives does not, as far as ordinary views are concerned, give rise to something that anyone should regard as even prima facie valuable for its own sake.12 But even if we assume that the cooperation of officials in sustaining a legal system has objective intrinsic value, there is no reason to think it gives rise to an objective motivating reason to obey mandatory legal norms governing non-​official behavior. While the cooperation of officials in manufacturing mandatory legal norms governing non-​official behavior might give them an objective motivating reason to obey those mandatory legal norms,13 it would not give non-​officials an objective motivating reason to obey them. A cooperative endeavor on the part of members 12 This is not surprising. Cooperation is not something we pursue for its own sake; we pursue it as a means to some end that we are more likely to efficiently realize by working with other people than we are by working by ourselves. One can, of course, judge the value of the end to which a cooperative endeavor is deployed to achieve; however, the ultimate value of the end is the same regardless of whether or not someone needs to enlist the cooperation of other people to achieve it. Doing something good has positive value and doing something bad has negative value—​regardless of whether that end is achieved by one person or whether it is achieved by a group of people who cooperate with one another to achieve that end. 13 A commitment to cooperate on the part of officials to comply with mandatory norms governing official behavior surely gives them a defeasible objective motivating reason to honor that commitment by complying with those norms.

194  The Content Problem of Legal Normativity of some proper subset of S to manufacture mandatory norms governing the acts of all members of S, by itself, cannot give rise to an objective motivating reason to comply with those mandatory norms in members of S who are non-​participants of the endeavor. This is not to deny that such cooperation might have objective normative significance for non-​participants. The cooperative efforts of officials might be objectively deserving of some kind of moral or non-​moral respect in virtue of being deployed to realize some intrinsically valuable end, but there is no reason to think that due respect requires non-​participants to obey the directives of officials with respect to non-​official behavior. One can surely pay due respect to the cooperative efforts of officials that figure into explaining the continuing existence of a legal system without regarding the normative output of the resulting official practices as giving rise to any kind of motivating reason to comply; if there is an objective intrinsic motivating reason to obey law, it is not because obedience is necessary to express due respect for the cooperative activities of officials that constitute something as a system of law. Nor does the claim that the system includes norms prohibiting some morally wrongful assaults on persons and property entail the claim that subjects should, as an objective matter of normative practical rationality, regard the system as having intrinsic value. It might be true that not doing what is prohibited by those norms is intrinsically valuable because doing otherwise would be morally wrong, but this would not imply that a system incorporating these norms should be valued for its own sake. Insofar as a system of law applies and enforces such moral content against subjects of the system, it conduces to a state of affairs in which people commit fewer of these morally proscribed acts. But that property of a legal system, which explains why we characteristically adopt systems of law to regulate behavior, is valuable only instrumentally insofar as it conduces to a state of affairs in which the peace is minimally kept. Insofar as the claim that law necessarily includes norms prohibiting some morally wrongful assaults on persons and properly can give rise only to instrumental value, there is nothing in that conceptual feature of law that gives rise to an objective intrinsic motivating reason to obey law. The idea that an institutional normative system S is metaphysically capable of legitimate authority is logically related to the idea that its norms are metaphysically capable of guiding behavior. An institutional normative system S is metaphysically capable of legitimate authority only if it is metaphysically possible for subjects to follow the norms of S. But it is metaphysically possible for a subject to follow a norm only if it is metaphysically possible for her to identify what it requires without having to decide for herself what she ought to do; if she can identify what it requires only by deciding for herself what she ought to do, then she is doing what she thinks she ought to do—​and not following the norm. For S to be metaphysically capable

Law as Source of Objective Intrinsic Reasons  195 of legitimate authority, then, it must be metaphysically possible for subjects to determine what S’s norms require without having to decide for themselves what they ought to do. No more than this is required for the norms of S to be metaphysically capable of guiding the behavior they govern. For a norm to be metaphysically capable of guiding behavior in the relevant sense, n must be metaphysically capable of making a practical difference with respect to what a subject does. But a norm can do this only if its content can replace the subject’s own judgment about what she should do: if the subject cannot identify what a norm requires without deciding what she thinks she ought to do, then the norm cannot make a practical difference with respect to what the subject does because she cannot obey it without doing what she thinks she ought to do. For the norms of S to be metaphysically capable of guiding behavior, then, it must be metaphysically possible for subjects to determine what S’s norms require without having to decide for themselves what they ought to do. The two conceptual requirements are logically equivalent. Either way, there is nothing in any of this that rationally competent subjects should regard, as an objective matter of normative practical rationality, as valuable for its own sake. These properties, regardless of whether they amount to the same thing, simply express what is conceptually necessary for something to be metaphysically capable of governing a subject’s behavior, as that notion is ordinarily understood; being metaphysically capable of guiding behavior in virtue of being capable of replacing the subject’s own assessment of what she ought to do is a trivial conceptual prerequisite of norm-​governance. But there is nothing about the fact that something is metaphysically capable of norm-​governance that should be regarded, as an objective matter of normative practical rationality, as valuable for its own sake; otherwise, every conceivable coherent system of norms, no matter what it requires or is used to do, would have objective intrinsic value in virtue of being conceptually capable of governing their behavior—​regardless of whether it actually governs their behavior. Finally, the fact that the practices constituting something as a system of law succeed in keeping the peace enough for us to live together in a community does not give rise to an objective motivating reason to obey law because it is law. Even if we assume that a state of affairs in which the peace is kept enough to permit subjects to live in a community is, as an objective matter of normative practical rationality, intrinsically valuable and not merely instrumentally valuable, the fact that an artifactual normative system, like law, is reasonably contrived to keep the peace endows it only with value as a means to that intrinsically valuable end. Artifacts do not become intrinsically valuable simply in virtue of being characteristically used to produce some state of affairs that should be regarded as intrinsically valuable. Adding the Coercion Thesis to properties (1)  through (5)  above adds nothing that would endow the practices constituting something as a system

196  The Content Problem of Legal Normativity of law with sufficient objective intrinsic value to give rise to objective intrinsic motivating reasons to comply. It should be clear that the only objective motivating reasons to which the enforcement practices expressed by the Coercion Thesis give rise are objective instrumental prudential motivating reasons to comply with mandatory legal norms governing non-​official behavior as a means of avoiding being on the receiving end of the coercive sanctions authorized for non-​compliance.14 Aggregating these properties is not plausibly thought to give rise to such value because the practices constituting something as a legal system can obviously be deployed for wicked purposes. The idea that an artifactual normative system with all these conceptually necessary properties is reasonably contrived to give rise to objective intrinsic motivating reasons is problematic insofar as it is conceptually possible for such a system to require acts so objectively wicked that there could not be an objective intrinsic motivating reason—​of any kind—​to perform them; there could be no objective intrinsic motivating reason to treat persons of some disfavored minority as property. One can, of course, have objective motivating reasons of some kind to do something wicked: one might have an objective defeasible motivating prudential reason to do something wicked if required by a norm backed by the threat of a sufficiently severe sanction, but that reason would be instrumental in virtue of being concerned to realize the prudential benefit of avoiding the sanction. But a wicked norm simply cannot give rise to an objective intrinsic motivating reason—​of any kind—​to comply with it. The practices constituting something as a system of law are as reasonably contrived to require wicked acts as they are to prohibit them. While it is a conceptually necessary condition for the existence of a legal system that it prohibits enough wrongful violence to ensure that the peace is minimally kept, this does not preclude there being norms requiring wrongful acts, including morally egregious acts of violence against members of a disfavored minority, because a state of affairs in which the peace is minimally kept is compatible with a state of affairs in which law institutionalizes racism, sexism, xenophobia, homophobia, transphobia, and more. Given that every legal system in our world includes norms like this, it is plausible to conclude that the practices constitutive of law are as reasonably contrived to

14 Insofar as one regards the practices authorizing the judicial imposition of coercive sanctions as a necessary moral evil, one might plausibly think that those practices give rise to negative intrinsic value that is defeated in some manner by the instrumental value they provide in virtue of making it far more likely that the peace is minimally kept among subjects; they increase the probability that the peace is kept because rationally competent subjects are characteristically likely to treat those practices as giving rise to an instrumental motivating reason to comply as a means of avoiding having sanctions imposed against them and refrain from some non-​complying acts they would have otherwise committed. If such views are correct, they illustrate the complexities involved in reasoning with claims involving intrinsic and instrumental value.

Law as Source of Objective Intrinsic Reasons  197 validate norms that cannot give rise to objective intrinsic motivating reasons as they are to validate norms that can give rise to such reasons. But this implies that these constitutive practices are not reasonably contrived to give rise to objective intrinsic motivating reasons. An artifact that is as reasonably contrived to do not-​X as it is to do X is not properly thought of as reasonably contrived to do either of those things. An artifact that is as reasonably contrived to fail to transport persons or things from one destination to another as it is to succeed in transporting persons or things from one destination is not properly characterized as an automobile because it is simply not reasonably contrived to do what automobiles are contrived to do; something counts as an automobile only if its non-​idiosyncratic use is significantly more likely than not to do what automobiles are intended, characteristically used, and supposed to do, as a functionally normative matter. Likewise, since the practices constitutive of law are as reasonably contrived to produce norms that do not give rise to objective intrinsic motivating reasons as they are to produce norms that do give rise to such reasons, those practices are not reasonably contrived to give rise to objective intrinsic motivating reasons. One might argue that law’s characteristic assertion of normative supremacy favorably distinguishes it from other artifactual normative systems in this respect, but there is nothing here equipped to explain how the practices constitutive of law are reasonably contrived to give rise to objective intrinsic motivating reasons. Since practices treating mandatory legal norms as trumping mandatory moral norms can obviously be objectively illegitimate, they are not reasonably contrived to give rise to objective intrinsic motivating reasons to obey law. Some legal systems are so wicked that it is simply implausible to think that mandatory legal norms enjoy the kind of objective normative supremacy needed to support the claim that these constitutive practices are a source of objective intrinsic motivating reasons to obey law. There is nothing in any plausible candidate for an existence condition for law reasonably contrived to give rise to objective intrinsic motivating reasons to comply. Insofar as conditions (1) through (5) together with the Coercion Thesis and the idea that law claims normative supremacy exhaust the plausible existence conditions for a legal system, there is nothing in the existence conditions for a legal system that should, as an objective matter of normative practical rationality, be regarded as having the kind of intrinsic value needed to give rise to objective intrinsic motivating reasons to comply. If those properties do not give rise to something that should, as an objective matter of normative practical rationality, be regarded as having intrinsic value, then they are not reasonably contrived to give rise to something that should, also as an objective matter of normative practical rationality, be regarded as an intrinsic motivating reason to comply.

198  The Content Problem of Legal Normativity

5.3  Does It Make a Difference If the System Contains Only Legal Norms Reproducing the Content of Mandatory Moral Norms? One might think that the arguments above succeed in establishing that law is not reasonably contrived to give rise to objective intrinsic motivating reasons depend on its being true that there can be morally unjust legal norms and hence presuppose a positivist view. On this line of reasoning, an institutional normative system that, of conceptual necessity, includes only mandatory norms reproducing the content of mandatory moral norms gives rise to objective intrinsic motivating moral reasons to obey law; if there are objective intrinsic motivating moral reasons to comply with mandatory moral norms, then there are also objective intrinsic motivating moral reasons to comply with any mandatory legal norm reproducing the content of these mandatory moral norms. There are two problems with this line of reasoning. First, it would not imply that the practices constituting something as a system of law give rise to objective intrinsic motivating reasons on any plausible conceptual theory of the metaphysical nature of law insofar as it is clearly not a conceptually necessary condition, according to the canons of ordinary usage, for the existence of a legal system that it includes only mandatory legal norms reproducing the content of mandatory moral norms. Even on a strong natural law view, there can be mandatory legal norms that do not reproduce the content of mandatory moral norms; there is no plausible theory of the metaphysical nature of law consistent with our conceptual practices that entails that it is a conceptually necessary condition for something to count as a system of law that it prohibits all and only non-​official behaviors prohibited by mandatory moral norms. Second, even if we assume that there cannot be a morally unjust law, there is nothing in this idea that would explain how the practices constituting something as a system of law are reasonably contrived to give rise to objective intrinsic motivating reasons to comply. While it might be true that any legal norm reproducing the content of a mandatory moral norm implicates or triggers an objective intrinsic motivating moral reason to comply, the fact that the content of the mandatory moral norm has the status of law has nothing to do with why it gives rise to an objective intrinsic motivating moral reason to comply: it is the moral content of the legal norm that gives rise to that reason; the fact that it is also recognized as law contributes nothing to its moral normativity.15

15 Anti-​positivism and positivism, then, live or die together on this issue in the following sense: either both theories can explain how these constitutive practices are reasonably contrived to give rise to new objective intrinsic motivating reasons to comply with mandatory legal norms governing non-​ official behavior or neither theory can.

No Need to Explain Intrinsic Reasons  199

6.  The Capacity to Give Rise to Objective Intrinsic Motivating Reasons Explains Nothing That Needs to Be Explained about Law The idea that law is an autonomous source of objective intrinsic motivating reasons to comply has this, I suppose, going for it: if one believes that law’s conceptual normativity is distinctive in some sense, the claim that law is an autonomous source of objective motivating reasons would explain what otherwise cannot be explained. An artifactual normative system conceptually equipped to give rise to autonomous objective motivating reasons is clearly normatively distinctive in virtue of being able to create a novel basic source of motivating reasons. The idea that law’s conceptual normativity is distinctive in some sense needed to explain why we do what we do with law is facially plausible. It is not happenstance that every reasonably complex society of which we know has adopted the practices constitutive of law to do everything a legal system can be non-​idiosyncratically used to do. Societies characteristically adopt these constitutive practices to regulate behavior instead of other artifactual regulative devices because there is something distinctive about law that makes it uniquely well-​suited to regulate behavior. What makes it especially suitable to do this regulatory work, the argument concludes, is that these constitutive practices are equipped to create something normative that no other artifactual normative system is equipped to create. This line of reasoning can be deployed to explain a number of normative properties sometimes attributed to law that are not plausibly attributed to other artifactual normative systems. The practices constituting something as a system of law are sometimes thought, as discussed in Chapters 6 and 7, to give rise to objective first-​order complete motivating reasons and objective second-​order exclusionary motivating reasons. Law’s conceptual capacity to give rise to both types of objective motivating reasons would, if it has this capacity, distinguish law as a kind from every other kind of artifactual normative system and would constitute its normativity as distinctive among the broader class of artifactual normative systems to which it belongs. It is true that the practices constituting something as a system of law are distinctive in a stronger sense than the trivial one that our conceptual practices distinguish law from every other type of artifactual normative system of which we know: among other features, systems of law (i) regulate a distinctively broad range of ordinary behaviors; (ii) create the police power they deploy to back mandatory legal norms with the threat of a coercive sanction; and (iii) apply mandatory legal norms governing these ordinary behaviors in an exclusionary manner as though they trump other artifactual and non-​artifactual norms, such as those of morality. Each of these obviously salient features helps to explain why the practices constituting something as a system of law are uniquely well-​suited among artifactual normative systems to do everything that law can be non-​idiosyncratically used to do.

200  The Content Problem of Legal Normativity But the issue is whether we need the idea that the practices constituting something as a system of law are reasonably contrived to give rise to objective intrinsic motivating reasons to explain anything about law needing explanation to make sense of why we characteristically adopt systems of law to regulate behavior. Once the Coercion Thesis is adopted, it is trivially easy to explain why law is uniquely well-​suited to efficaciously regulate behavior in a manner that enables us to live together in a community—​namely that certain mandatory norms are backed by a threat of detriment that beings like us have strong objective motivating prudential reasons to avoid. The only plausible explanation for the fact that every existing legal system of which we know backs mandatory legal norms prohibiting certain breaches of the peace is that doing so uniquely renders the system able to keep the peace among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours and hence to do everything that law can be non-​ idiosyncratically used to do. The idea that the practices constituting something as a system of law are reasonably contrived to give rise to objective intrinsic motivating reasons to comply explains nothing about the metaphysical nature of law that needs explaining. Since beings like us neither characteristically regard these constitutive practices as giving rise to objective intrinsic motivating reasons nor ought to regard them as such, the idea that law gives rise to such reasons cannot explain how it is reasonably contrived to make a practical difference with respect to what we do; something that is neither objectively relevant nor characteristically regarded as such in deliberations about what to do lacks the resources to explain anything that needs to be explained about law’s conceptual capacity to make such a difference. If the claim that law is reasonably contrived to give rise to objective intrinsic motivating reasons cannot contribute to vindicating law’s conceptual normativity, there is nothing of importance it can explain about the nature of law. This is not to deny that such reasons exist in logical space as a purely metaphysical matter; however, there is nothing in the conceptual theory of law that needs to account for them because, for all practical purposes, they do not exist. Positing that law is reasonably contrived to give rise to such reasons does nothing more helpful in enabling us to understand law’s conceptual normativity than positing the existence of phlogiston does to enable us to understand combustion. Once the Coercion Thesis is adopted, it becomes clear why no one has been able to persuasively argue that these constitutive practices are reasonably contrived to give rise to objective motivating reasons that are complete, exclusionary, or intrinsic in character: there is simply nothing about law’s conceptual normativity that any of these types of motivating reason is needed to explain. What constitutes law’s conceptual normativity as distinctive among artifactual and non-​artifactual systems of norms, according to the Coercion Thesis, is that the practices constituting something as a system of law apply and enforce mandatory legal norms in a manner that trumps all other artifactual and non-​artifactual

The Problems of Legal Normativity  201 norms with the threat of severe detriment like incarceration. Law is distinguished from other artifactual normative systems, like those regulating the behavior of players in the National Football League, not in virtue of backing its norms with the threat of detriment; professional football players can be fined or suspended for non-​compliance with mandatory norms governing league players. Law is distinguished from these other artifactual normative systems in virtue of the severity of the threatened detriment and in virtue of its asserted normative supremacy; there is no other artifactual normative system that characteristically punishes non-​ complying behaviors with detriment as severe as incarceration. Law is distinctively normative among artifactual normative systems, then, in virtue of its conceptual capacity to give rise to motivating prudential reasons to comply that subjects are minimally likely, as a descriptive matter of fact, to regard as trumping any motivating prudential reasons for non-​compliance to which other artifactual normative systems are reasonably contrived to give rise because they should, as an objective matter of normative practical rationality, regard them as such. There are no other social practices contrived to do this work, in essence, because law has the longest arms and the biggest guns. That is what distinguishes law’s conceptual capacity to rationally induce compliance with its norms through the operation of objective motivating reasons from that of any other artifactual normative system.

7.  The Coercion Thesis and the Problems of Legal Normativity Law’s conceptual normativity must be presumed to make practical sense of why we use systems of law to regulate ordinary behavior. It makes practical sense to devote so many scarce material resources into deploying and maintaining systems of law only if they are reasonably contrived, as a conceptual matter, to give rise to objective first-​order motivating reasons to comply with mandatory legal norms governing ordinary behavior.16 If we cannot explain how the practices constituting

16 Operating a legal system in any reasonably large society costs trillions of dollars per year. The 2019 federal budget for the U.S. legal system is approximately $4.407 trillion. See https://​www.thebalance. com/​u-​s-​federal-​budget-​breakdown-​3305789. It is important to note that this figure includes costs that are not, strictly speaking, needed to maintain the existence of a legal system; those functions that are necessary to sustain the existence of a legal system are probably limited to tasks directly connected with minimally keeping the peace. It is also worth noting that this figure does not include the cost of maintaining the state systems that comprise part of the legal system in the U.S. Governor Jay Inslee of Washington proposed a state budget for 2019 of $54 billion; this figure does not include the costs of sustaining local country and municipal systems. See https://​www.seattletimes.com/​seattle-​news/​politics/​ gov-​inslee-​proposes-​54-​4b-​state-​budget-​that-​includes-​new-​taxes-​on-​capital-​gains-​and-​business-​ services/​. Assuming this is the average size of a state budget, the states expend a total of about $2.7 billion to sustain. While this figure covers more than the basic services needed to sustain the state systems, it should be clear, when all the costs are taken into account, that legal systems are very expensive artifacts.

202  The Content Problem of Legal Normativity something as a system of law are reasonably contrived to give rise to objective first-​ order motivating reasons to comply, then we cannot explain something that should be easily explained—​namely, why we use these systems to regulate the behaviors we use them to regulate. The very rationality of our doing what we do with law rests on our being able to vindicate the presumption that it is conceptually normative. Everything that needs to be explained to vindicate law’s conceptual normativity is explained by the claim that backing certain mandatory legal norms with the threat of a non-​trivial sanction gives rise to an objective first-​order motivating prudential reason to do what those mandatory norms require as a means of avoiding sanctions we characteristically regard, as a descriptive matter of contingent fact, as something to be avoided because we should regard them, as an objective matter of normative practical rationality, as something to be avoided. The How, Order, and Content Problems of Legal Normativity can be fully resolved if—​and only if—​it is assumed that the practices constituting something as a legal system include practices authorizing courts to impose coercive sanctions as a formal response to non-​compliance with mandatory legal norms prohibiting acts likely to breach the peace. There is simply nothing else in the existence conditions for law equipped to explain how the practices constituting something as a system of law are reasonably contrived to give rise to objective motivating reasons to comply with mandatory legal norms governing non-​official behavior. The Coercion Thesis fully and exclusively vindicates law’s conceptual normativity.

9

Coercive Sanctions and International Law* It is sometimes thought that the institutional normative system to which we refer as our system of international law is a counterexample to the Coercion Thesis. On this line of reasoning, the system of international regulation defined by the U.N. Charter wholly lacks authorized coercive sanctions to enforce its norms. Insofar as this system is properly characterized as one of law despite lacking legal norms governing official behavior authorizing the imposition of coercive sanctions for non-​compliance with mandatory legal norms of the system governing non-​official behavior, it is a counterexample to the Coercion Thesis. H.L.A. Hart was the first major figure in conceptual jurisprudence to reject the Coercion Thesis and he did so for this reason. As he puts the matter, “[t]‌here are no settled principles forbidding the use of the word ‘law’ of systems where there are no centrally organized sanctions, and there is good reason (though no compulsion) for using the expression ‘international law’ of a system, which has none.”1 If, as seems plausible, there is good reason to characterize the system defined by the Charter as a legal system despite lacking sanctions, then the Coercion Thesis is false. Hart’s argument is weaker than might initially appear because it is agnostic with respect to whether the institutional system defined by the Charter is properly characterized as one of law according to the canons of ordinary usage. The crucial premise is not that our conceptual practices clearly warrant characterizing this system as one of law despite its lacking sanctions; the premise is rather that there is nothing in these practices that obviously entails that it is not a system of law in virtue of lacking the right kind of coercive sanctions. Given that our conceptual practices—​as these are defined by the canons of ordinary usage and the philosophical assumptions informing these canons—​are epistemically indeterminate with * I am indebted to Luka Burazin and Corrado Roversi for helpful comments on this chapter. 1 H.L.A. Hart, The Concept of Law, 2nd edn (Oxford University Press 1961) 199. Hereinafter CL. John Austin also rejects the idea that international law is backed by something properly characterized as sanctions, but he rejects the claim that international law is really law for precisely this reason: “[T]‌he law obtaining between nations is not positive law: for every positive law is set by a given sovereign to a person or persons in a state of subjection to its author. . . . [T]he law obtaining between nations is law (improperly so called) set by general opinion. The duties which it imposes are enforced by moral sanctions: by fear on the part of nations, or by fear on the part of sovereigns, of provoking general hostility, and incurring its probable evils, in case they shall violate maxims generally received and respected.” John Austin, The Province of Jurisprudence Determined (Cambridge University Press 1995) 171. Austin wrote before the first supranational organizations were chartered by multilateral acceptance of a formal covenant or charter purporting to have the status of law; this is, in part, why he believed that the norms of international law are purely customary. Coercion and the Nature of Law. Kenneth Einar Himma, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198854937.001.0001

204  Coercive Sanctions and International Law respect to whether this system is one of law, there is nothing in those practices that entails the Coercion Thesis. On Hart’s view, this entails that the Coercion Thesis should be rejected because, as far as we can tell, our practices allow for the conceptual possibility of a legal system without sanctions. This chapter adopts Hart’s agnosticism with respect to the Charter system and argues that, regardless of whether it is properly characterized as law, the Charter system authorizes the imposition of economic and diplomatic measures that count as coercive sanctions, as I have explicated the notion, insofar as they are reasonably contrived to deter and punish non-​compliance. While these measures differ in salient ways from the coercive sanctions authorized by systems of municipal law, they nonetheless satisfy the application-​conditions for counting as coercive sanctions; all that is required for something to count as a coercive sanction, as I have explicated the notion, is that it imposes non-​trivial detriment that rationally competent self-​interested subjects are characteristically likely to experience as something that deters and punishes non-​compliance because they should experience it as such. Given that the relevant economic and diplomatic measures are reasonably contrived to do just that, the Charter system, regardless of whether it counts as a system of law, is consistent with the Coercion Thesis and is hence not a counterexample to it.

1.  Ordinary Usages of the Term International Law International law is comprised, as a conceptual matter, by a system of institutional norms governing relations among nations. Oxford Online Dictionary defines the term as “a body of rules established by custom or treaty and recognized by nations as binding in their relations with one another.”2 Merriam-​Webster defines it as “a body of rules that control or affect the rights of nations in their relations with each other.”3 Insofar as a system of international law purports to regulate only those behaviors bearing directly on the relations among nations, it is contrived to respect the regulatory sovereignty of those nations over persons within their boundaries insofar as its exercise does not directly bear on international relations. It is worth noting that the lexical meanings reported by Oxford and Merriam-​ Webster are, like Hart’s argument, agnostic with respect to whether the U.N. Charter system of international regulation is really law. This suggests that our empirically contingent linguistic practices allow for the existence of institutional systems of international regulation that are not properly characterized as law. One

2 Oxford Online Dictionary, available at: https://​www.lexico.com/​en/​definition/​international_​ law. 3 Merriam-​Webster Online Dictionary, available at:  https://​www.merriam-​webster.com/​dictionary/​international%20law.

Ordinary Usages of the Term International Law  205 reason for this might be that the prevailing view prior to the implementation of systems of international regulation grounded in instruments like the Covenant of the League of Nations and the U.N. Charter, which are analogous to the constitutions of municipal legal systems, was that the existing system of international law lacked certain properties (such as a formal instrument creating the system) that anything must have to count as a system of law. The compound concept-​term international law is used, as a descriptive matter of contingent fact, to refer to distinct classes of institutional norms governing legal relations among subjects in different nations, but only one is relevant for our purposes. While the term private international law is used to pick out a body of institutional norms governing disputes between ordinary parties in different nations, those norms are applied by courts in systems of municipal law to resolve those disputes and thus count as laws of a municipal legal system; indeed, the term private international law is typically defined as “primarily concerned with the resolution of conflict of national laws, determining the law of which country is applicable to specific situations.”4 The norms of private international law are properly thought of as conflict-​of-​laws municipal legal norms that can, as a conceptual matter, differ from one municipal legal system to the next. This use of the term international law to refer to these municipal laws might be thought facially misleading in virtue of suggesting that the content of private international law is manufactured by some international agency that is not a part of any municipal legal system; however, this is unremarkable. Analogous confusions abound throughout our conceptual practices: a void contract is a contract, as Lon L. Fuller put the point, in only a Pickwickian sense; it is, strictly speaking, no more plausibly characterized as a real contract than private international law is as being international law.5 These confusions arise because our empirically contingent linguistic conventions are sometimes imprecise in ways that seem to call into question the coherence of the corresponding conceptual practices. The usage of the term that is relevant for our purposes is what is commonly called public international law, which is typically defined along the following lines: Public International Law is composed of the laws, rules, and principles of general application that deal with the conduct of nation states and international organisations among themselves as well as the relationships between nation states and international organisations with persons, whether natural or juridical. Public International Law is sometimes called the “law of nations” or just simply International Law. It should not be confused with Private International Law,

4 “Public International Law: Introduction to Public International Law Research,” University of Melbourne Library Guides; available at: https://​unimelb.libguides.com/​internationallaw. 5 Lon L. Fuller, The Morality of Law (Yale University Press 1960) 39.

206  Coercive Sanctions and International Law which is primarily concerned with the resolution of conflict of national laws, determining the law of which country is applicable to specific situations.6

Insofar as the norms of international law thought to define a counterexample to the Coercion Thesis govern, as the matter is commonly put, the behavior of nations, the relevant norms for my purposes are those of public international law—​and not those of private international law. The term international law is used to refer to two classes of norm that have no more in common than that they happen to involve disputes among parties in different nations; it is misleadingly used in private international law because the relevant norms are laws of municipal systems that govern disputes between persons rather than laws of a transnational system that govern disputes between nations. It is important not to make too much of such noise because the seemingly discrepant practices express different usages that do not really conflict with each other. It is convenient to refer to a small plastic object contrived to resemble chairs as a toy chair—​despite the fact that it is not reasonably contrived to do what chairs are needed, characteristically used, and supposed to do, as a functionally normative matter, and is hence not really a chair—​to indicate that it is a replica of a chair contrived for novelty purposes that involve depicting chairs in the settings in which they are typically used.7 It is likewise convenient to refer to municipal laws governing the choice of law in disputes involving subjects of different nations as international laws because they are applied to resolve legal disputes between subjects in two different nations. While conceptual analysis must be grounded in empirical claims describing our contingent linguistic practices with respect to the relevant terms, a sound conceptual methodology should not require the rejection of common sense. The traditional description of public international law, as it turns out, evinces an obvious conceptual confusion. While these norms are commonly characterized as governing the behavior of nations, they are more aptly characterized as regulating acts of national officials that implicate, in some potentially problematic way, the interests of people in other nations through the governance of norms metaphysically capable of guiding the behavior of those officials with respect to such acts.

6 “Public International Law: Introduction to Public International Law Research,” Note 4, available at: https://​unimelb.libguides.com/​internationallaw. 7 Toys present an interesting class of seemingly pathological usages from the standpoint of the claims defended in Chapter 4 about the character of artifacts. One might think that so-​called magic wands are a counterexample to the claim that it is a conceptually necessary condition for something to be properly characterized as an artifact A that it is reasonably contrived to perform the conceptual function of A; since what we call magic wands are not reasonably contrived to perform the conceptual function of a magic wand, it is a counterexample to that claim. The problem with this reasoning is that these are toys contrived to resemble what is portrayed in films and novels as wands that do things that are magical; they are thus novelty items reasonably contrived to perform the role-​playing function that is the conceptual function of such things. I am indebted to Kenneth Ehrenberg for this example.

The U.N. Charter System of International Law  207 The problem with characterizing these norms as governing the behavior of nations is that a nation is an abstract object metaphysically incapable of doing anything that can be governed or guided by norms: a nation is no more capable, as a conceptual matter, of doing things that can be governed or guided by norms than is the abstract object referred to by the symbol 2 because neither is metaphysically capable of anything properly characterized as a doing. The relevant norms are hence properly characterized as governing the behavior of officials of those nations with respect to certain acts implicating the well-​being of persons in other nations—​and not the behavior of the nations of which they are officials because nations are metaphysically incapable of behavior. The relevant norms of public international law are those of what is sometimes called supranational law in virtue of their having their origin in the acts of some agency distinct from all the agencies of municipal law. These norms govern those acts of officials of national municipal legal systems implicating the interests of persons in other nations—​especially insofar as those acts are likely to lead to conflicts breaching the peace because potentially regarded as morally wrongful by persons in adversely affected nations. Supranational law is manufactured and applied by bodies of officials properly characterized as international in virtue of their having authority to bind officials of various nations with norms restricting those acts. These norms are of particular interest here insofar as they potentially define a unified system of legal norms ontologically distinct from systems of municipal law and give rise to something properly characterized as a system of international law. There are no conceptually possible legal norms that are not norms of some conceptually possible legal system; while the existence of some fundamental set of legal norms might come into existence simultaneously with the existence of the legal system it defines, those norms are not properly characterized as legal until they are properly characterized as members of a system that is also properly characterized as legal. Just as there is no law without laws, there are no laws without law: the norms of international law are laws only insofar as the norms of international law comprise a system of international law.

2.  The System of International Regulation Defined by the U.N. Charter Following Hart, I consider only the system of international norms defined by the U.N. Charter. The Charter defines a body of institutional norms comprising something plausibly, if not correctly, characterized as a system of international law created for the purpose of keeping the peace among nations: 1. To maintain international peace and security, and to that end: to take effective collective measures for the prevention and removal of threats to the peace, and

208  Coercive Sanctions and International Law for the suppression of acts of aggression or other breaches of the peace, and to bring about by peaceful means, and in conformity with the principles of justice and international law, adjustment or settlement of international disputes or situations which might lead to a breach of the peace; 2. To develop friendly relations among nations based on respect for the principle of equal rights and self-​determination of peoples, and to take other appropriate measures to strengthen universal peace; 3. To achieve international co-​operation in solving international problems of an economic, social, cultural, or humanitarian character, and in promoting and encouraging respect for human rights and for fundamental freedoms for all without distinction as to race, sex, language, or religion; and 4. To be a centre for harmonizing the actions of nations in the attainment of these common ends.8

Insofar as this system (1) minimally succeeds in achieving its basic function of minimizing breaches of the peace that threaten the welfare of all members in the community (2) through the governance of norms metaphysically capable of guiding the behavior of its subjects and (3) satisfies the minimum conditions for the existence of a legal system, it is properly characterized as a system of law. But insofar as it lacks the right kind of sanctions but nonetheless counts as a system of law, it is a counterexample to the Coercion Thesis.

3.  Enforcement Mechanisms Authorized by the U.N. Charter The U.N. Charter, regardless of whether it creates something that is really law, authorizes a number of institutional enforcement mechanisms seemingly intended, as is conceptually true of anything that counts as a coercive sanction, to deter and punish non-​compliance. Chapter VII authorizes the Security Council to impose “preventive [and] enforcement measures” for the purpose of “maintain[ing] or restor[ing] international peace and security.”9 These measures include military,

8 Charter of the United Nations, Article 1; available at: https://​www.un.org/​en/​sections/​un-​charter/​ un-​charter-​full-​text/​. 9 Article 50 provides that “[i]‌f preventive or enforcement measures against any state are taken by the Security Council, any other state, whether a Member of the United Nations or not, which finds itself confronted with special economic problems arising from the carrying out of those measures shall have the right to consult the Security Council with regard to a solution of those problems.” Article 39 provides that “[t]he Security Council shall determine the existence of any threat to the peace, breach of the peace, or act of aggression and shall make recommendations, or decide what measures shall be taken in accordance with Articles 41 and 42, to maintain or restore international peace and security.” Charter of the United Nations; available at: https://​www.un.org/​en/​sections/​un-​charter/​chapter-​vii/​index.html. Emphasis added.

Enforcement Mechanisms Authorized by the U.N. Charter  209 economic, and diplomatic actions intended to minimize “threat[s]‌to the peace, breach[es] of the peace, [and] act[s] of aggression.”10 It is important to distinguish the notion of an enforcement mechanism from the notion of a coercive sanction. Everything properly characterized as a coercive sanction according to the canons of ordinary usage is also properly characterized as an enforcement mechanism: as Oxford Online Dictionary defines the term, enforcement is an attempt to “[c]‌ompel observance of or compliance with . . . a law, rule, or obligation”;11 insofar as an enforcement mechanism, as a conceptual matter, seeks to compel compliance by threatening the forced imposition of a penalty, it is also a sanction.12 But it is not obviously true that everything properly characterized as an enforcement mechanism is also properly characterized as a sanction: a threat not to enforce an improperly formed agreement provides the right kind of prudential inducement to count as a negative enforcement mechanism—​though that inducement is conditional upon someone’s desiring to have an agreement enforced; however, it is not properly characterized as a coercive sanction, as I have explicated the notion.13 The question for purposes of evaluating the Coercion Thesis is whether the enforcement mechanisms authorized by the Charter count as coercive sanctions, as that notion is explicated in Chapter 1.The issue of whether these mechanisms have the necessary properties to warrant characterizing them as coercive sanctions is considered below. Since the relevant military, economic, and diplomatic enforcement mechanisms are distinct forms of institutional intervention differing with respect to their conceptually essential properties, they are considered separately.

3.1  Military Enforcement Mechanisms Article 42 of the Charter authorizes the Security Council to undertake a number of military interventions deemed necessary to accomplish the U.N.’s basic organizing purpose of ensuring that the peace is kept among nations: Should the Security Council consider that measures provided for in Article 41 [i.e. those not involving the use of armed force] would be inadequate or have proved to be inadequate, it may take such action by air, sea, or land forces as may 10 Article 39, Charter of the United Nations; available at:  https://​www.un.org/​en/​sections/​un-​ charter/​chapter-​vii/​index.html. 11 Oxford Online Dictionary; available at: https://​www.lexico.com/​en/​definition/​enforce. 12 Oxford Online Dictionary defines the term sanction as “a threatened penalty for disobeying a law or rule.” See: https://​www.lexico.com/​en/​definition/​sanction. 13 See Chapter 3, Section 2.3. As will be seen below, the military measures authorized by the Charter count as enforcement mechanisms but do not also count as coercive sanctions; this, if correct, confirms the point that not every conceptually possible enforcement mechanism counts as a coercive sanction, according to our conceptual practices.

210  Coercive Sanctions and International Law be necessary to maintain or restore international peace and security. Such action may include demonstrations, blockade, and other operations by air, sea, or land forces of Members of the United Nations.

Among this somewhat underspecified range of interventions authorized by the Charter are the creation and deployment of a force for the purpose of ending armed hostilities between two nations. This force is authorized by the Charter to conduct operations analogous to those performed by national military forces: these forces, like those of a national armed force, may conduct such operations through the “air, sea, or land” and are no less plausibly characterized as a military force, as far as our conceptual practices are concerned, than is any national air force, navy, or army. The implied threat of military intervention is not an empty one; these interventions have been deployed at least twice by the Security Council, albeit in different ways. The first deployment was for purely pacific purposes in that it was intended simply to oversee a cessation of hostilities, rather than to engage militarily with warring forces. The First U.N. Emergency Force was created in response to the invasion of Egypt by France, Israel, and the United Kingdom after Gamal Abdel Nasser, then-​president of Egypt, nationalized the Suez Canal; it was deployed under the auspices of the Charter to do no more than supervise the withdrawal of French, Israeli, and United Kingdom forces from Egypt: The mandate of the Force was to secure and supervise the cessation of hostilities, including the withdrawal of the armed forces of France, Israel and the United Kingdom from Egyptian territory and, after the withdrawal, to serve as a buffer between the Egyptian and Israeli forces and to provide impartial supervision of the ceasefire. UNEF was withdrawn in May-​June 1967, at Egypt’s request.14

A military peacekeeping intervention limited to ending armed hostilities between nations with as little violence as possible to combatants is no more contrived to deter or punish non-​compliance than is a deployment of police to peacefully end a fistfight between two people. An expressed continuing commitment on the part of some official agency to restore the peace with minimal violence by separating combatants is properly characterized as enforcing norms that prohibit fighting but is neither intended to deter or punish rationally competent subjects from fighting nor reasonably contrived to do so: “if you two don’t stop fighting, you will be forcibly separated and held against your will until you calm down” is not plausibly characterized as expressing a threat to impose a coercive sanction that deters or punishes fighting.



14

See: https://​peacekeeping.un.org/​en/​mission/​past/​unef1mandate.html.

Enforcement Mechanisms Authorized by the U.N. Charter  211 This is not to deny that there is a coercive element to pacific measures deployed by either a domestic police force or an international peacekeeping military force. Such measures, whether undertaken by an armed international peacekeeping force or an armed domestic police force, come with an obvious threat of violence: violent resistance to peacekeeping efforts will be met with violence. While this might seem paradoxical, it is straightforwardly coherent: sometimes the only way to stop a school shooter or a soldier from killing as many innocent people as she can is to shoot her before she can shoot them. U.N. peacekeeping troops have sometimes engaged aggressively with warring forces in response to an attack. Despite the fact that various norms require treating peacekeeping forces as non-​combatants, Norwegian members of the First U.N. Emergency Force were attacked by Egyptian gunman and returned fire.15 Although that force was deployed for purely pacific purposes, it could not have accomplished those purposes without being authorized to defend itself against armed attacks. These missions, though pacific in character, are carried out by military personnel authorized to respond to force with force in the hope that the threat of armed retaliation will dissuade parties from attacking U.N. peacekeeping forces. But the fact that these purely pacific military measures authorize the imposition of detriment in the form of an armed response to attacks does not, by itself, constitute them as coercive sanctions.16 Consider a legal system that criminalizes the distribution of alcohol by adults to minors but does not criminalize the possession of alcohol by minors. This configuration of prohibitions involves the forced imposition of something likely to be experienced as detriment by both adults and minors; adults experience detriment in the form of a penalty for distributing alcohol to minors while minors experience detriment in the form of being denied certain means of obtaining it. The only coercive sanctions these prohibitions create are the penalties for distributing alcohol to minors: it is clear that these penalties are sanctions in virtue of enforcing the distribution-​prohibition by deterring and punishing violations by adults; it is likewise clear, in contrast, that there is nothing in this configuration of prohibitions that deters or punishes efforts on the part of minors to obtain alcohol. Detriment must be reasonably contrived to deter and punish to constitute a coercive sanction, as I have explicated a notion; and a purely defensive armed response to armed attacks is not reasonably contrived to punish them. Sometimes the peace can be kept or restored only by more aggressive measures, such as those deployed by the U.N. in response to North Korea’s invasion 15 Peter Nadin (ed.), The Use of Force in UN Peacekeeping (Routledge 2018) Introduction. 16 The term pacific is ambiguous between two lexical meanings that can conflict. Merriam-​Webster offers two definitions corresponding to two usages: “tending to lessen conflict: conciliatory” and “rejecting the use of force as an instrument of policy.” See: https://​www.merriam-​webster.com/​dictionary/​ pacific. A purely defensive peacekeeping effort on the part of an armed military authorized to use force might tend to lessen conflict but it does not reject the use of force. I use the term here to refer to military actions that are initiated for the ultimate purpose of keeping the peace among parties in violent conflict and do not involve other potentially related objectives such as regime change.

212  Coercive Sanctions and International Law of South Korea in 1950; this second deployment was intended, unlike the first, to actively engage with an attacking military force. The U.N. Security Council authorized member-​nations to assist South Korean military forces in repelling the invasion; the U.S.—​ostensibly acting as an agent of the U.N.—​deployed approximately 150,000 U.S. soldiers to drive the invading North Korean forces out of South Korea.17 Although the U.N. forces succeeded in repelling the invasion and in provisionally restoring the peace between North and South Korea, the war took the lives of approximately five million people, the vast majority of whom were civilians.18 It would be helpful to distinguish the two types of peacekeeping missions described above in terms of their ability to deter. Purely peacekeeping missions intended to enforce cessations of hostilities, like the Suez intervention, are not reasonably contrived to deter invasions that might trigger them—​though the threat of an armed response to a wrongful attack on a peacekeeping force is reasonably contrived to deter those attacks. In contrast, the threat of an armed response to a wrongful invasion that is contrived to repel the attackers, such as deployed by the U.S. during the Korean war, is as reasonably contrived to induce compliance with international norms prohibiting such invasions as the threat of the contempt sanction in municipal law is to induce compliance with court orders because both threaten an assertive deployment of violence—​or force—​and are hence properly characterized as enforcement mechanisms. But that, again, does not make them coercive sanctions. We distinguish forceful acts intended to sanction others from forceful acts intended to defend against threats regarded as wrongful. Merely defending against a violent attack with force sufficient to stop or repel the attack does not count as the imposition of a coercive sanction: hitting me back for no other reason than to stop me from hitting you, by itself, is self-​defense—​and not the imposition of a coercive sanction; however, if you continue to hit me after I surrender as a means of punishing me for fighting with you and of deterring me from ever starting a fight with you again, your behavior is not implausibly characterized as imposing a sanction—​provided that it meets the other conceptually necessary conditions for counting as a coercive sanction.19 But, as far as the Coercion Thesis is concerned, none of the military measures authorized by the Charter for peacekeeping or defensive purposes count as coercive sanctions, as I have explicated the term.

17 See, “Korean War,” Wikipedia.com; available at: https://​en.wikipedia.org/​wiki/​Korean_​War#Factors_​ in_​US_​intervention. 18 See, e.g., “Korean War,” History.com; available at: https://​www.history.com/​topics/​korea/​korean-​war. 19 As far as the canons of ordinary usage are concerned, sanctions are penalties that are threatened or imposed in response to violations of some sort of norm. As Oxford Online Dictionary defines it, a sanction is “[a]‌threatened penalty for disobeying a law or rule.” See: https://​www.lexico.com/​en/​definition/​sanction. The issue here is whether the notion of a rule is broad enough to encompass a personal policy on your part to repel an attack with more force than is needed to stop it as a means of punishing it and deterring future attacks.

Enforcement Mechanisms Authorized by the U.N. Charter  213 This is not to suggest that the term sanction is never properly applied to military interventions; it is merely to point out that purely pacific and defensive interventions are not properly characterized as sanctions. An armed military operation counts as the imposition of a sanction, as we use the term, if deployed to punish and deter non-​compliance with some norm; the idea of a military sanction is not, after all, an obvious oxymoron according to the canons of ordinary usage.20 Just as your continuing to hit me after you end my assault to punish me for attacking you and to deter future attacks on my part is plausibly characterized as a coercive sanction, an aggressive military action intended to punish a nation for some act it wrongfully committed under some set of international norms and to deter future violations (by the offending nation and by other nations) is plausibly characterized as a coercive sanction. Coercive sanctions are, as a conceptual matter, contrived to deter non-​ compliance by doing something plausibly characterized as punishing it. An institutional system of international norms authorizing the destruction of some military facility or national institution as retribution for an aggressive act not rising to the level of an act of war counts as a sanction, as a matter of ordinary usage. In this case, the intent and the effect are both punitive: the destruction of the facility is intended to punish, given its retributive intent, and would be experienced as punishment by the recipient regime. The intent need not be explicitly punitive in character.21 Had the U.S.-​led forces invaded North Korea for the purpose of toppling and replacing the regime of Kim Il-​sung to minimize the threat North Korea posed to the global peace, it would be experienced by Kim as punishment despite not being intended as punishment. For my purposes, a threat of coerced regime-​change motivated by purely pacific considerations counts as a coercive sanction because it is reasonably contrived to deter and punish non-​compliance with norms prohibiting invasion of one nation by another—​despite the fact that it is intended as neither retribution nor punitive retaliation. Either way, neither the armed deployment of the First U.N. Emergency Force in response to the Suez Crisis nor the deployment of U.S.-​led forces to repel the North Korean attack on South Korea counts as a coercive sanction because neither was intended nor contrived to punish non-​compliance. The deployment of each was motivated solely by a concern to restore and keep the peace and was contrived—​in intent if not in execution—​to do no more than restore the peace with as little violence as possible. Insofar as the military measures authorized by the Charter are 20 While some dictionary reports are facially agnostic with respect to the possibility of military sanctions, others are clear that there can be military sanctions. Merriam-​Webster defines one usage of sanction as “an economic or military coercive measure adopted usually by several nations in concert for forcing a nation violating international law to desist or yield to adjudication.” See: https://​www. merriam-​webster.com/​dictionary/​sanctions. 21 See Chapter 1, Section 2 for more discussion on this point.

214  Coercive Sanctions and International Law intended only to keep or restore the peace in response to an act of military aggression violating the Charter, they do not count as coercive sanctions, as that notion was explicated in Chapter 1.

3.2  Economic Enforcement Mechanisms Although Hart has little to say about the military measures discussed in the last subsection, he clearly rejects the idea that they count as coercive sanctions insofar as he believes that neither the Covenant of the League of Nations nor the U.N. Charter, which authorizes the use of such military measures, authorizes any enforcement mechanisms properly characterized as a sanction, as that notion is properly understood: We shall take it that neither Article 16 of the Covenant of the League of Nations nor Chapter VII of the United Nations Charter introduced anything which can be equated with the sanctions of municipal law. In spite of the Korean war and of whatever moral may be drawn from the Suez incident, we shall suppose that whenever their use is of importance, the law enforcement provisions of the Charter are likely to be paralysed by the veto and must be said to exist only on paper (CL 217).

Given that he rejects the idea that these instruments authorize something counting as a sanction on the ground that the military measures deployed in the Korean war and Suez incident do not count as sanctions, Hart assumes that there are no non-​ military measures authorized by the Charter plausibly characterized, even at first blush, as coercive sanctions. Hart does not say much by way of explanation here, but it is reasonable to hypothesize that he believes that the non-​military measures can be ruled out because they do not involve a show of armed force analogous to that which backs the paradigmatic municipal sanctions of criminal law. The idea seems to be that there is no need for further philosophical explication because non-​military enforcement mechanisms not involving some direct show of force are indisputably not properly characterized as sanctions, according to the canons of ordinary usage. Since they are conclusively ruled out by the unambiguous lexical meanings of the terms, there is no need to explicitly consider non-​military measures. But if this is Hart’s view, it is problematic. Merriam-​Webster explicitly defines the term as including such measures: a sanction, according to this definition, is a “detriment, loss of reward, or coercive intervention annexed to a violation of a law as a means of enforcing the law.”22 Likewise, Oxford Online Dictionary 22 Merriam-​Webster Online Dictionary; available at: https://​www.merriam-​webster.com/​dictionary/​sanction. Emphasis added.

Enforcement Mechanisms Authorized by the U.N. Charter  215 characterizes a “measure[] taken by a state to coerce another to conform to an international agreement or norms of conduct” as an example of something that counts as a sanction.23 Insofar as one can coerce without threatening or applying physical force,24 there is nothing in the lexical meaning of the term sanction that obviously precludes its correct application to non-​military measures. Further, the Charter authorizes a number of economic “enforcement measures” that are intended and characteristically used to do what coercive sanctions are supposed to do, as a functionally normative matter. Article 41 provides as follows: The Security Council may decide what measures not involving the use of armed force are to be employed to give effect to its decisions, and it may call upon the Members of the United Nations to apply such measures. These may include complete or partial interruption of economic relations and of rail, sea, air, postal, telegraphic, radio, and other means of communication, and the severance of diplomatic relations.25

These economic enforcement mechanisms seem, at first blush, to operate exactly the way coercive sanctions are contrived to operate, as a conceptual matter—​ namely, enforce the norms they back with a threat of detriment that functions to deter and punish non-​compliance. To begin, the “interruption of economic relations and of rail, sea, air, postal, telegraphic, radio, and other means of communication, and the severance of diplomatic relations” constitutes detriment likely to deter acts of aggression against other nations because it can be imposed as a punitive measure. Further, it is plausible to think that these non-​military measures have contributed, as a descriptive matter of contingent fact, to keeping us out of another world war; but these non-​military measures can have contributed to keeping the peace among member-​nations only insofar as they have functioned to deter prohibited acts of aggression by member-​nations.

23 This example falls under the definition of sanction cited above in Note 21. 24 “Blackmailing” someone is an example; threatening to disclose sensitive information about an individual if she does not do something demanded is coercive, on any plausible analysis of the notion. 25 Charter of the United Nations, Article 41, Chapter VII; available at:  https://​www.un.org/​en/​ sections/​un-​charter/​chapter-​vii/​. It is worth noting that the Covenant of the League of Nations provides for similar enforcement measures; in particular, Article 16 provides that “Should any member of the League resort to war in disregard of its covenants under Articles 12, 13, or 15, it shall ipso facto be deemed to have committed an act of war against all other Members of the League, which hereby undertake immediately to subject it to the severance of all trade or financial relations, the prohibition of all intercourse between their nationals and the nationals of the covenant-​breaking State, and the prevention of all financial, commercial or personal intercourse between the nationals of the covenant-​ breaking State and the nationals of any other State, whether a Member of the League of Nations or not.” Covenant of the League of Nations, Article 16; available at: https://​avalon.law.yale.edu/​20th_​century/​ leagcov.asp#art16.

216  Coercive Sanctions and International Law In this connection, it is worth noting that the analogical distance between these economic enforcement mechanisms and the punitive sanctions of municipal law is not as great as Hart seems to believe; many paradigmatic sanctions of municipal law, after all, involve purely financial penalties. The judicial imposition of a municipal fine is reasonably contrived to deter and punish a range of minor civil infractions, such as exceeding posted speed limits or parking illegally; while these authorized fines are not always efficacious in deterring such violations, as is evident immediately upon merging onto an ordinary highway, they are reasonably contrived to do both those things and would seem, contra Hart, to count as coercive sanctions on any remotely plausible analysis of the concept. Municipal financial penalties like fines are backed by the court’s authority to impose the contempt sanction; however, that does not obviously distinguish these municipal penalties from the economic measures authorized by the Charter. As long as the Charter allows the Security Council to impose further measures reasonably contrived to induce compliance, such as a military blockade and other aggressive measures, there is no theoretically significant difference between the financial penalties of municipal law and the economic enforcement mechanisms of the Charter. That said, it is worth considering the matter in more depth. Each of the following is a conceptually necessary condition for something to count a coercive sanction as I have explicated the concept: a non-​military measure counts as a coercive sanction only if (1) its deployment is authorized by a recognition norm governing officials of the system that permits them to impose it as an institutional response to violations of valid mandatory norms of the system governing non-​official behavior; (2) it is intended and reasonably contrived to deter and punish non-​compliance with the applicable norms of the system governing non-​official behavior; and (3) it is plausibly thought to succeed to some non-​trivial extent in deterring rationally competent subjects from violating the applicable norms.26 As discussed below, the economic enforcement mechanisms authorized by the Charter satisfy each of these conditions.

26 I  say “plausibly thought” because it is not clear how we could determine conclusively that the reason we have not had a third world war is that the practices associated with the Charter have prevented it; one might sensibly think that the costs of war are so prohibitively high, given the destructive power of nuclear weapons, that the desire to avoid a nuclear apocalypse has been sufficient to prevent an outbreak. While one might think there is a similar epistemic issue with respect to municipal law, it seems clear that the result of doing away with the sanctions in any municipal legal system would be a coast-​to-​coast free-​for-​all; whereas many people believe (falsely, on my view) that the U.N. and its practices could safely be done away with, far fewer people believe that the domestic law of a nation could safely be done away with. These numbers do not tell us much about the ultimate merits of such views; truth is not something that can be settled, on any intuitively plausible view of the matter, by a show of hands. But they do tell us something about the comparative first-​blush plausibility of the two views.

Enforcement Mechanisms Authorized by the U.N. Charter  217

3.2.1 The Economic Measures Are Authorized as a Response to Non-​Compliance There are two conceptually necessary conditions for something to count as a coercive sanction according to condition (1) above. First, the deployment of the relevant measure must, as a conceptual matter, be authorized by some recognition norm of the system governing official behavior. Second, the norm must authorize its deployment as a response to violations of valid mandatory norms of the system governing non-​official behavior by rationally competent subjects who fall within the jurisdictional limits of the system; the deployment of measures intended to inflict detriment on someone not within the jurisdictional limits of the system is, as a general matter, more accurately characterized as either an act of unprovoked aggression or an act of defense, whether of oneself or others. Both conditions are satisfied by these Charter mechanisms. Article 16 authorizes officials of the international system to impose the relevant non-​military measures in response to non-​compliance by member-​nations with norms governing non-​official acts within the jurisdictional limits of the system: Should any Member of the League resort to war in disregard of its covenants under Articles 12, 13 or 15, it shall ipso facto be deemed to have committed an act of war against all other Members of the League, which hereby undertake immediately to subject it to the severance of all trade or financial relations, the prohibition of all intercourse between their nationals and the nationals of the covenant-​breaking State, and the prevention of all financial, commercial or personal intercourse between the nationals of the covenant-​breaking State and the nationals of any other State, whether a Member of the League or not.27

As the italicized text makes clear, Article 16 authorizes the imposition of these measures as a response to violations of Articles 12, 13, or 15 by any “Member of the League.” Article 16 thus authorizes the deployment of the relevant measures for acts committed by nations within the jurisdictional limits of the system in virtue of the right kind of act (consent by ratification in this case) as an institutional response to non-​compliance with valid mandatory norms of the system governing the acts of member-​nations.28 Insofar as the relevant measures are explicitly authorized to enforce the applicable valid norms of the Charter, they satisfy Condition (1).

27 Covenant of the League of Nations; available at: http://​avalon.law.yale.edu/​20th_​century/​leagcov. asp#art16. Emphasis added. 28 These latter acts are official acts within the system of municipal law constituting the member-​ nation as a nation but non-​official acts within the institutional system defined by the U.N. Charter; the officials authorized to recognize, apply, and enforce norms of the Charter are officials of that system but not of the municipal systems constituting member-​nations as something plausibly characterized as nations.

218  Coercive Sanctions and International Law

3.2.2 The Economic Measures Are Intended and Reasonably Contrived to Deter and Punish Misbehavior One possible explanation for why Hart might have thought the lexical meaning of sanction precludes applying the term to these economic measures is that they function quite differently from the coercive sanctions of municipal law. The enforcement mechanisms of municipal law are contrived to inflict painful detriment on only the person whose conduct violates a criminal law or court order; while the sanction of incarceration inevitably causes suffering to friends and family of the offender, it is the offender who will experience, by far, the worst suffering. In contrast, the economic measures authorized by the Charter are not contrived to inflict the most painful detriment on the authority responsible for violating Charter norms; rather they are contrived to inflict detriment on other people by causing harm to the economy of the offending nation. The effects of such measures might seem mild compared to the municipal sanction of incarceration, but the material hardship they inflict on ordinary people should not be understated. Diminished access to material resources can cause life-​ altering detriment even to someone of comparative affluence; losing a job causes significant disruption to the life of anyone who is not independently wealthy. Diminished access to material resources to less affluent persons can mean the difference between getting enough to eat and not getting enough to eat.29 If I were forced to choose between doing a short bid in prison where I know I will be properly cared for and protected in an ethically humane manner—​which is utterly atypical of prison-​conditions in even the most enlightened nations—​and experiencing life-​threatening poverty in a nation targeted by these economic measures, I would choose doing the bid with no hesitation, although both strike me as terrifyingly awful.30 29 Institutional norms authorizing the forced imposition of painful detriment on anyone for violating other institutional norms always raise issues of moral justification, but part of what makes these economic enforcement measures particularly worrisome is that they inflict detriment on ordinary persons who have nothing to do with the offending acts. The sanctions imposed on Iraq after the First Gulf War in 1990 are thought to have killed more than 500,000 Iraqi children without accomplishing what they were deployed to do, as Saddam Hussein remained in power until after the Iraq War in 2002. See, e.g., Barbara Crossette, “Iraq Sanctions Kill Children, U.N. Reports” NY Times (December 1, 1995), available at:  https://​www.nytimes.com/​1995/​12/​01/​world/​iraq-​sanctions-​kill-​children-​un-​reports.html; and “Sanctions Against Iraq, Global Policy Forum, available at: https://​www.globalpolicy.org/​previous-​ issues-​and-​debate-​on-​iraq/​sanctions-​against-​iraq.html. It is one thing, morally speaking, to impose life-​altering detriment on someone who arguably deserves it in virtue of having culpably caused harm to another person; it is another thing entirely to knowingly impose life-​altering detriment on innocent persons for the sole purpose of inducing someone else to refrain from certain acts. 30 Elderly persons sometimes find living in prison preferable to living on their own because they cannot afford to properly care for themselves. While older persons lacking an appropriate support system who have spent most of their adulthood in prison are likely to find the prospect of release terrifying, someone need not become elderly while serving a prison sentence to understand the comparative utility of prison life: the crime rate among elderly persons in Japan has increased dramatically for such reasons. See Mark Abadi, “Elderly people in Japan are getting arrested on purpose because they want to go to prison,” Business Insider (March 19, 2018); available at: https://​www.businessinsider.com/​ japan-​aging-​prison-​2018-​3. At the end of the day, it would be morally and economically preferable to provide more safety-​net protections for the elderly; imprisoning a person is expensive.

Enforcement Mechanisms Authorized by the U.N. Charter  219 These economic enforcement mechanisms are intended and reasonably contrived to impose collateral detriment on policy-​ making officials by directly imposing detriment on the subjects on whose acquiescence those officials ultimately depend for their continuing authority.31 Policy-​makers can be presumed to care about the well-​being of residents for prudential, rather than altruistic, reasons: there is only so much active resistance, after all, to a regime compatible with its continuing power. If people get angry enough because they believe policy-​ making officials are not doing what they are supposed to do, they might just rebel; suppressing armed resistance among civilians is not something even the most callously corrupt ruler ever wants to do—​even if it is clear to all that civilians cannot muster enough force to effect regime change.32 The point of these non-​military measures is to inflict sufficient collateral detriment on policy-​making officials to deter non-​complying behavior and to punish it if it occurs. The infliction of detriment on the subjects of a municipal (i.e. national) authority is merely a means to the end of deterring the authority from doing something that violates Charter norms; indeed, there is no reason to think that the ultimate objective of these economic measures is ever to inflict suffering on the subjects of the authority for its own sake. The ordinary persons who experience the most painful effects of these international economic measures are the immediate but not ultimate target of those measures; the ultimate target is the policy-​making officials. In contrast, both the immediate and ultimate targets of the detriment inflicted by the municipal sanction of incarnation are the offending party. These enforcement mechanisms are nonetheless reasonably contrived to deter and punish non-​compliance with mandatory norms prohibiting non-​violent acts that can lead to war. The acts that Charter norms attempt to deter and punish do not all necessarily lead to war, but most can destabilize relations among nations because they are likely to benefit citizens of the acting nations at the expense of citizens of affected nations and are hence likely to be perceived as culpably aggressive in the targeted nations. Given that the dangers of war are so grave to the international community, it makes sense to regulate not only acts of war but also non-​violent acts that run an appreciable risk of beginning a cycle of events that can escalate into armed hostilities.

31 By collateral, I do not mean to suggest such detriment is accidental. While the compound term collateral damage is used to refer to unintended damage to third parties, the term collateral connotes only that the effect is “concomitant,” “indirect,” and “ancillary.” See, e.g., Merriam Webster Online Dictionary; available at: https://​www.merriam-​webster.com/​dictionary/​collateral. 32 More than any other consideration, this probably explains why authoritarian regimes limit speech rights; while no one likes to be criticized, the real danger to a repressive regime is that public criticism might amplify existing dissatisfaction to the point where people begin to more openly protest and resist those practices because they start to believe they are enough of them to induce the regime to make the desired changes. State suppression of speech also runs the risk of exacerbating unrest, but allowing critical speech is usually regarded as creating far more dangerous risks to the regime.

220  Coercive Sanctions and International Law The very point of the Charter is “to save succeeding generations from the scourge of war, which twice in our lifetime has brought untold sorrow to mankind.”33 The Charter authorizes the U.N. to “employ international machinery for the promotion of the economic and social advancement of all peoples” as a means of diminishing the likelihood of economic and social problems that might lead policy-​making officials to commit acts that detrimentally affect people of other nations and create tensions that can escalate into armed hostilities threatening the global peace. While the economic measures authorized by the Charter add a deterrent of less normative force than the natural deterrent posed by the likely effects of war, such economic detriment is reasonably contrived to add an additional deterrent to non-​ compliance: insofar as the imposition of economic detriment on the subjects of an authority is likely to weaken their support for municipal authorities, they are reasonably contrived to deter and punish non-​conforming behavior on the part of policy-​making officials enough to decrease to some non-​trivial extent the likelihood of a war that potentially threatens the well-​being of people in every nation. Condition (2) is satisfied.

3.2.3 The Economic Measures Are Plausibly Thought to Succeed in Helping to Keep the Peace One might argue that the dangers posed by a third world war, which the Charter is explicitly concerned to prevent, are so great that it is possible, as a practical matter, that the global peace might be kept without a system of international law.34 Since policy-​making officials are sensibly presumed motivated by prudential considerations, if not necessarily by moral considerations, it is plausible to think that the natural prudential disincentives to war are so great that a system of international law is not needed to keep the peace. But, on this line of reasoning, insofar as the global peace can be kept without a system of international law, the economic enforcement mechanisms of the Charter are not reasonably contrived to keep the peace. There are a number of problems with this argument. To begin, the claim that something can be done without using an artifact A that is, as a conceptual matter, created, characteristically used, and supposed to do as a functionally normative matter does not imply the claim that A is not reasonably contrived to do that something. I might be able to get from point a to point b without using an automobile because they are not that far apart, but that does not imply that an automobile is not 33 Preamble to the Charter of the United Nations; available at: https://​www.un.org/​en/​sections/​un-​ charter/​preamble/​index.html. 34 The relevant notion of possibility here is not nomological, conceptual, or logical possibility; it is obvious that it is possible in each of those respects. The idea expressed by this practical sense of possibility is that there is a sufficiently high probability that we would be able to avoid a global war without the mediation of Charter practices that it warrants abolishing the Charter practices as a means of avoiding costs to national sovereignty that one might believe (falsely) are not worth bearing.

Enforcement Mechanisms Authorized by the U.N. Charter  221 reasonably contrived to transport me from a to b. Artifacts can be reasonably contrived to do things that can be done without them. Further, the fact that policy-​making officials are rationally competent might make it less likely they behave in aggressive ways that lead to war, but it certainly does not preclude their doing so. National leaders frequently act more aggressively than seems justified under objective norms of practical rationality because they not uncommonly underestimate both the likelihood that their acts will lead to war and the amount of damage that will be done to their nations: if leaders always assessed these risks accurately, the Second World War would never have happened; it is obviously a really bad idea in a world of rising tension to start invading one’s neighbors. Although the Second World War predated the system defined by Charter, a similar system of international regulation for the purpose of keeping the peace was in place—​the Covenant of the League of Nations, which the Charter replaced after the Second World War. The idea, then, that additional mechanisms reasonably contrived to discourage war are not needed because the natural deterrents to behaving aggressively are sufficient to prevent such behaviors is patently false. The economic measures of the Charter are reasonably contrived, as shown in the last subsection, to reduce the risk that such miscalculations culminate in aggressive acts that increase the likelihood of war insofar as they are well-​suited to alter the prudential calculus by inflicting economic hardship on those people whose acquiescence policy-​makers need for their continuing authority. The claim that A is reasonably contrived to do p does not imply the claim that the use of A will succeed in doing p: chemotherapy is properly characterized as a cancer treatment because it is intended, characteristically used, reasonably contrived, and supposed to induce remission, as a functionally normative matter; however, it falls far short of always being successful in doing so.35 It is enough, given our conceptual practices, to warrant characterizing chemotherapy as a cancer treatment that it is successful in inducing remission often enough that, despite its toxicity to non-​ cancerous cells and dangerous side effects, oncologists routinely administer it to lengthen the life-​span of patients with cancer that, left untreated, would kill them. No more than this is needed, as far as our conceptual practices are concerned, to constitute something as reasonably contrived to perform the function for which it is characteristically used. This does not imply that we are epistemically justified in believing that these Charter practices have helped to keep the peace, but it does imply that they are plausibly thought to have done so, which is all that Condition (3) demands. The facts that (i) these economic enforcement mechanisms are authorized by an instrument signed by fifty nations in 1945 after the Second World War and (ii) there has not been a third world war since the Charter went into effect shows that these 35 It is also routinely administered, where there is no reasonable likelihood of remission, to delay the progression of the disease in order to lengthen what is left of the patient’s life.

222  Coercive Sanctions and International Law measures are plausibly thought to have helped to prevent a third world war. To claim otherwise is to challenge the competence of the national officials who signed and ratified the Charter; the claim that we cannot make practical sense of why so many nations would continue to participate in Charter practices is a refutation of any claim that implies it. Condition (3) is satisfied.

3.3  Diplomatic Enforcement Mechanisms It is enough to vindicate the Coercion Thesis against the challenge that Charter practices present to show that the economic measures considered above count as coercive sanctions, as I have explicated the notion. The Coercion Thesis asserts only that it is a conceptually necessary condition for the existence of a legal system that some of its mandatory norms governing non-​official behavior are backed by the threat of something properly characterized, according to the canons of ordinary usage, as a coercive sanction; it says nothing about the character of the conceptually requisite measures beyond the idea that they must impose detriment reasonably likely to function as a deterrent to non-​compliance in some instances and to punish it when it occurs. That said, it is worth considering the Charter’s diplomatic measures, which are characterized as “enforcement measures” by Article 50, not because doing so is needed to defend the Coercion Thesis, but because doing so would help to further clarify the operative notion of a coercive sanction, as it has been explicated in this volume. Article 41 authorizes the Security Council to enforce its norms by calling upon member-​nations to sever diplomatic relationships with violating nations. Article 16 authorizes “the prohibition of all intercourse between their nationals and the nationals of the covenant-​breaking State, and the prevention of all . . . personal intercourse between the nationals of the covenant-​breaking State and the nationals of any other State, whether a Member of the League or not.”36 There is nothing in our ordinary social practices, conceptual or otherwise, that precludes characterizing severance measures like those described above as coercive sanctions. A threat to end a romantic relationship or ordinary friendship with someone for committing some specified act is plausibly characterized as a sanction. It is true that some of these acts, such as those involving sexual infidelity and other forms of betrayal, threaten a relationship because they damage the mutual trust any relationship needs to be successful; ending a relationship because one has lost confidence in the other person’s trustworthiness does not necessarily involve an intent to punish. But insofar as the point of an expressed threat to leave a relationship if certain acts are performed is to deter and punish those acts, it counts as a



36

Italicized emphasis added.

Enforcement Mechanisms Authorized by the U.N. Charter  223 coercive sanction—​though if things have reached this point, it is just a matter of time before the relationship ends. The “severance of diplomatic relations” authorized by Article 41 also counts as a coercive sanction, as I  have explicated the notion. As will be recalled, a non-​military measure counts as a coercive sanction only if (1) its deployment is authorized by a recognition norm governing officials of the system that permits them to impose it as a means of enforcing valid mandatory norms of the system governing non-​official behavior; (2) it is intended and reasonably contrived to deter and punish non-​compliance with the applicable norms of the system governing non-​official behavior; and (3) it is plausibly thought to succeed to some non-​trivial extent in deterring competent agents from violating the applicable norms. Like the non-​military economic measures considered above, the severance of diplomatic relations is explicitly authorized by the Charter as a means of enforcing its norms; imposes something likely to be experienced by the subject as detriment that is intended and reasonably contrived to function as a prudential disincentive to non-​compliance; and is plausibly thought to play some non-​ trivial role in deterring non-​compliance, as the threat of complete diplomatic isolation clearly poses a non-​trivial deterrent to engaging in acts likely to trigger such diplomatic sanctions. There are conspicuous differences between the threat of incarceration that backs municipal law and such non-​military measures, but those differences are not as great as one might think. The severance of all diplomatic relations, when this includes prohibition of all interaction between subjects in the sanctioned state and other subjects, does much of what incarceration is contrived to do but without the walls, armed guards, and razor wire: it limits intercourse between the offending subject and other rationally competent subjects by restricting the former’s freedom of movement. While it is true that incarceration enacts much narrower limits on with whom one may communicate and interact, that is a difference in degree and not in kind. But even if it is plausible to think that none of these non-​military measures counts as a coercive sanction according to our conceptual practices, it should be clear that combining them to sever all intercourse—​economic, personal, and diplomatic—​is properly characterized as a coercive sanction. Given that nothing in the Charter precludes the imposition of such a combination of non-​military measures for violations, it should be clear that the Charter authorizes the imposition of something with non-​trivial deterrent force that can be used to punish violations and hence counts as a coercive sanction for purposes of vindicating the Coercion Thesis. Insofar as such a combination of non-​military measures can be backed up by military measures involving blockades and more aggressive measures, it is all the more reason to think that these Charter enforcement mechanisms count as coercive sanctions on any plausible, non-​question-​begging explication of that notion.

224  Coercive Sanctions and International Law

4.  Objections and Replies The considerations above provisionally vindicate the Coercion Thesis from the challenge posed by the institutional system defined by the U.N. Charter. Regardless of whether it is properly characterized as a system of law according to our conceptual practices, we are presumptively justified, as an epistemic matter, in thinking that the Charter system authorizes the imposition of enforcement measures that count as coercive sanctions, as I have explicated the concept. To complete the argument of this chapter, this section responds to the two most plausible objections to the claim that these measures are coercive sanctions.

4.1  Are Acts That Merely Withhold a Benefit Not Properly Characterized as Sanctions? One might think that the economic and diplomatic enforcement mechanisms authorized by the Charter function as a deterrent by withholding various benefits but are not properly characterized as coercive sanctions because they merely withhold benefits. On this line of reasoning, it is a conceptually necessary condition for something to count as a coercive sanction that it forcibly inflicts detriment in a direct manner. Incarceration counts as a sanction because one is forcibly imprisoned against one’s will; likewise, a financial penalty counts as a sanction because the fine is backed by a threat of incarceration. But insofar as a trade restriction or a severance of diplomatic relations merely withdraws a benefit, it is not properly characterized as a sanction. Although facially plausible, there is nothing in the canons of ordinary usage that precludes characterizing a withdrawal of benefits to punish or deter non-​ compliance as a sanction. Merriam-​Webster explicitly defines the term as including a “loss of reward . . . annexed to a violation of a law as a means of enforcing the law.”37 Likewise, Oxford Online Dictionary explicitly includes as sanctions “state . . . restrictions on trade or official sporting participation”;38 while this does not expressly assert, unlike the Merriam-​Webster definition above, that a measure imposing a loss of benefits counts as a sanction, it is clear that a restriction on trade and sporting activities amounts to no more than a loss of a benefit insofar as it merely deprives a nation of beneficial exchange relations. Such restrictions do exactly the work that coercive sanctions are contrived to do in ordinary relationships. Parents commonly threaten to withhold some desired

37 Merriam-​Webster Online Dictionary; available at: https://​www.merriam-​webster.com/​dictionary/​sanction. Emphasis added. 38 Oxford Online Dictionary; available at:  https://​www.lexico.com/​en/​definition/​sanction. Emphasis added.

Objections and Replies  225 benefit from their children as a means of punishing and deterring misbehavior. A parent might revoke a privilege, such as the use of the family automobile, or withhold a financial benefit, such as payment of a weekly allowance, as a means of deterring and punishing misbehavior on the part of a child; and it should be clear that, depending on how much the subject values the benefit, such measures can, as a practical matter, succeed in deterring behavior. Likewise, an employer might revoke the benefit of employment as a threatened response to various instances of misconduct, such as sexual harassment and other harmful behaviors, as a means of deterring and punishing such acts. The line between acts that indirectly impose detriment by withholding benefits and acts that directly impose detriment by inflicting harm is not always clear. While it is clearly absurd to characterize the death penalty as merely withholding the benefit of life, the revocation of a driver’s license for driving under the influence is as plausibly characterized as punishment—​and hence as a coercive sanction—​as the death penalty is; the former punishes by indirectly imposing detriment in the form of a withdrawn benefit while the latter punishes by directly inflicting harm. Both measures function to deter and punish non-​compliance in virtue of inflicting the kind of harm that one has a strong prudential reason to avoid. The essence of the capacity of a coercive sanction to deter and punish, as a conceptual matter, consists in it having significant prudential normative force in virtue of being contrived to significantly diminish perceived well-​being. To get a sense for how much prudential normative force a threat to withhold a financial benefit can have, consider again the facts giving rise to the Riggs v. Palmer case.39 Elmer so valued what he expected to take under his grandfather’s will that he murdered his grandfather to prevent him from changing the will after he remarried. Given that Elmer was convicted of the murder and imprisoned, this was probably not the best thing to do as an objective matter of purely prudential rationality; however, it was not prudentially irrational to do so: if the probability of being caught and convicted had been sufficiently low and the amount of the gift sufficiently large, and there had been no other prudentially relevant considerations, then Elmer might have been prudentially justified in acting as he did. People are willing to do horrible things to each other precisely because they can conduce significantly, other things being equal, to self-​interest. Doing something horrible to someone else is usually morally wrong, but that does not make it practically irrational—​at least as far as our conceptual and

39 Riggs v. Palmer, 115 NY 506 (1889). Alternatively, consider the case of Thomas Gilbert, Jr., a 30-​ year-​old Princeton graduate who allegedly killed his father for cutting his weekly allowance from $1000 to $300. Sonia Moghe, “Princeton grad killed father over allowance, prosecutors say” CNN (June 27, 2019); available at:  https://​www.cnn.com/​2019/​06/​27/​us/​nyc-​princeton-​grad-​murder-​trial/​index. html. Unlike Riggs who killed his grandfather to prevent him from changing his will, Gilbert seems to have killed his father to punish him for cutting his allowance. In both cases, the murders were committed as a response to something properly characterized as merely a loss of benefit.

226  Coercive Sanctions and International Law evaluative practices are concerned. As long as we can make sense of why someone did something morally wrong in terms of its being reasonably contrived to produce some substantial prudential benefit given the facts of the world as they appear to someone without a serious mental illness, that is enough to warrant characterizing them as rationally competent: if something is prudentially rational, then it is also practically rational—​though it may not be practically justified, all things considered. As far as our conceptual practices are concerned, which are the ones that matter for purposes of explicating the content of our concepts on the modest approach adopted here, the claim that p is immoral does not imply the claim that p is irrational.40 This tells us something about objective norms of practical rationality (or shared convictions about what those norms require) that helps to explain why threats to withhold benefits can be reasonably contrived to do the work that coercive sanctions are deployed to do in law. If it can be prudentially and hence practically rational (though not justified) for Elmer to ensure that he is not cut out of the will by murdering his grandfather, then it can be prudentially and hence practically rational for the officials of a nation to conform their behavior to the Charter to avoid having an important economic or diplomatic benefit withheld by member-​nations. Such measures are intended and conceptually equipped to do exactly the work that sanctions are deployed to do and are properly characterized as sanctions—​ especially insofar as they contribute, as is plausible to think, to preventing violations of the Charter. That said, it is somewhat misleading, if not conceptually inapt, to characterize the non-​military enforcement mechanisms authorized by the Charter as merely withholding a benefit.41 The effects of sanctions that sever trade and diplomatic relationships can be devastating on the people who live in the sanctioned nation and go well beyond anything that counts as merely withholding a benefit in the sense that revoking a non-​vital privilege amounts to just withholding a benefit. As one commentator describes the effects of the economic sanctions imposed by the U.N. on Iran as a means of deterring its uranium enrichment program:

40 One might think that the claim that doing p is morally wrong implies that doing p is rational, but this is not obvious. While it is true that only competent rational agents are subject to the requirements of morality, the claim that someone generally satisfies the standards of rational competence does not imply that everything she does is rational. It is arguable that a rationally competent subject can do something wrongful that is practically irrational but that she is nonetheless accountable under moral standards. It is surely true that being irrational entails that one does something irrational, but it is not true that doing something irrational entails being irrational. 41 In contrast, it is not wholly implausible to characterize the so-​called “sanction of nullity” as it operates in contract law as merely withholding benefits. While it is true that one might enter into a contract as a means of avoiding catastrophic detriment, such as personal bankruptcy, the characteristic motivation for entering into a contract is to obtain some desired set of prudential benefits. In such cases, a judicial refusal to enforce the terms of the contract merely denies to one of the parties the prudential benefit she sought to obtain by entering into the agreement.

Objections and Replies  227 The sanctions on Iran caused a fall of [the] country’s revenues, devaluation of national currency, and increase of inflation and unemployment. These all resulted in deterioration of people’s overall welfare and lowering their ability to access the necessities of a standard life such as nutritious food, healthcare and medicine. . . . The impacts of sanctions were more immense on the lives of the poor, patients, women and children. Humanitarian exemptions did not protect Iranians from the adverse effects of sanctions.42

Characterizing such effects as merely a matter of withholding benefits not only suppresses the punitive intent and effect of such measures; it also understates—​and significantly—​the prudential undesirability of such effects. While there is a trivial sense in which incarceration can be correctly characterized as withholding the benefit of freedom, it is clear that doing so suppresses the punitive intent and effect of incarceration as well as understates the prudential undesirability of incarceration and its ability to deter violations. The same is true of the economic sanctions of international law; it is as intuitively plausible to think that the diplomatic and economic enforcement mechanisms authorized by the Charter deter some acts of non-​compliance that threaten the global peace as it is to think that the sanction of life imprisonment deters some murders.43

4.2  Does the Veto Power Imply That the Authorized Enforcement Mechanisms Are Not Sanctions? Hart ultimately rejects the idea that the Charter authorizes anything plausibly characterized as coercive sanctions on the ground that every member of the Security Council can veto any resolution applying these mechanisms. The underlying idea seems to be that it is not enough for an institutional normative system

42 Fatemeh Kokabisaghi, “Assessment of the Effects of Economic Sanctions on Iranians’ Right to Health by Using Human Rights Impact Assessment Tool: A Systematic Review” International Journal of Health Policy and Management vol. 7. no. 5 (May 2018) 374–​93. 43 This is not to suggest that economic sanctions are, by themselves, necessarily or even generally successful in effecting the desired changes in the behavior of the targeted nations. When they are unsuccessful, it is because their direct deleterious effects are not felt by either the persons making the objectionable decisions or the economic elites whose support is perhaps most vital to the continuing authority of those persons; the worst effects are felt by the most impoverished citizens who are least able to defend their interests because, being poor, they also lack political power. Even so, it is nonetheless plausible to think that economic sanctions help to deter some aggressive acts that could lead to conflicts that escalate into armed hostilities precisely because policy-​makers can underestimate the probable risks where the ostensible probable benefits seem deceptively desirable. But in cases where they are imposed to deter behaviors that violate human rights, such as in the case of the examples listed above, they are not as likely to be successful in changing the relevant behavior. In many, if not most, of those humanitarian cases, economic sanctions accomplish no more than to gratify the need of those politicians who vote to impose them to appear as though they are doing something significant to alleviate the human rights abuses that trigger the perceived need for them.

228  Coercive Sanctions and International Law to authorize the deployment of detriment as a means of punishing and deterring non-​compliance; it must also ensure that such detriment is sufficiently likely to be imposed against offending parties to be reasonably contrived to do what sanctions are contrived to do. Insofar as any member of the Security Council can veto a resolution deploying such authorized detriment, the likelihood of application does not rise to the conceptually requisite level; the veto power effectively negates the practical capability of these measures to do what coercive sanctions are reasonably contrived to do. There are two problems with this argument. First, it is false that the provisions authorizing any member of the Security Council to veto a resolution imposing sanctions negates, in any practical sense that matters, the operation and desired effects of those provisions. These enforcement mechanisms have frequently been deployed by member-​nations of the international community, individually or collectively, against other nations for what is deemed to be misbehavior under the Charter. The list of nations subjected to U.N. economic sanctions and travel bans includes Afghanistan, Central African Republic, Democratic People’s Republic of Korea, Democratic Republic of the Congo, Eritrea, Guinea-​Bissau, Iran, Libya, Mali, Somalia, South Sudan, Sudan, and Yemen.44 Each of these nations has been conspicuously targeted by sanctions for violations of norms accepted by member-​ nations of the U.N. It is plausible to hypothesize that what explains the application of such sanctions in these cases is that the targeted nations have committed violations deemed so potentially disruptive to the peace among nations and hence to the well-​being of people in every member-​nation of the Security Council that no member-​nation was willing to veto their application because doing so would adversely affect their respective domestic interests as well.45 The idea that members of the Security Council are authorized to veto any resolution requiring the application of sanctions to an offending nation no more disqualifies the relevant measure from counting as a coercive sanction than either the fact that judges can suspend a sentence for non-​ violent offenses involving marijuana or the fact that juries can nullify the imposition of a prison sentence they deem unjust by voting for acquittal disqualifies the authorized prison sentence from counting as a coercive sanction. The powers of a judge to suspend a sentence and of a jury to nullify sanctions by acquitting the defendant amount to de facto, if not de jure, veto powers. 44 “Policy Responsibilities:  United Nations Security Counsel Sanctions,” Commerce & Economic Development Bureau, The Government of the Hong Kong Special Administrative Region; available at: https://​www.cedb.gov.hk/​citb/​txt_​en/​Policy_​Responsibilities/​united_​nations_​sanctions.html. 45 Of course, this also has something to do with the fact that the sanctioned nations are all quite small and impoverished—​which is part of what makes the application of these sanctions in such cases morally worrisome. They not only disproportionately impact people who are not responsible for the non-​ complying acts but also disproportionately impact impoverished nations. While such sanctions might not be morally objectionable as a matter of principle, the way in which they have been applied raises ethical concerns.

Objections and Replies  229 Second, and more importantly, the argument misses the mark. The Coercion Thesis asserts no more than that it is a conceptually necessary condition for something to be a legal system that it authorizes the imposition of coercive sanctions in response to non-​compliance. As discussed in Chapter  1, the claim that such sanctions are authorized neither asserts nor entails that they must sometimes be applied; it is nomologically possible for there to be a legal system in which sanctions are authorized but never applied because people never violate valid mandatory legal norms governing non-​official behavior that are backed by the threat of a sanction.46 While such a state of affairs is astronomically unlikely for rational self-​interested beings like us who live in worlds of material scarcity like ours, it is nomologically (and hence conceptually) possible. But a state of affairs in which sanctions are authorized by something that counts as a legal system but never applied is not merely nomologically possible; it is practically possible in the sense that the authorization of sanctions—​under the right social circumstances—​is, by itself, reasonably contrived to produce such a state of affairs among rationally competent subjects like us—​even in worlds of acute material scarcity like ours; this is, after all, the very state of affairs that totalitarian police-​states try to achieve by conspicuously deploying ever more sophisticated surveillance technologies and Draconian sanctions to deter non-​compliance. Rationally competent self-​interested subjects like us are not likely to violate a mandatory norm if we know that the probability we will be caught and severely punished is very high: if it were known that an omnipotent and omniscient God immediately punished a sin by dipping the offender in hell for periods that increase in duration with each successive sin, it would not take long for competent agents to largely, if not completely, cease committing sins that trigger the expected period of torture. Notice, however, that such a state of affairs is practically possible among rationally competent subjects like us in worlds of acute material scarcity like ours only on the assumption that the necessary mandatory norms governing non-​official behavior are backed by severe sanctions. Rationally competent self-​interested subjects like us sometimes need a prudential disincentive to commit harmful acts that the practices constitutive of a legal system can provide only insofar as the relevant norms are backed by coercive sanctions. But insofar as such a state of affairs could, as a practical matter, be achieved by a sufficiently violent and intrusive legal system, it should not be thought that sanctions have to be applied to succeed in producing such a state. It is, in part, for this reason, that there is nothing in any of the relevant concepts as they have been explicated here that would entail that authorized sanctions must be applied in a legal system. The claim that sanctions can



46

See Chapter 1, Section 2.2.

230  Coercive Sanctions and International Law be vetoed by any member of the Security Council cannot, for this reason, bear the weight that Hart intends it to bear in evaluating the Coercion Thesis.

5.  Conclusion: The Normative System Defined by the U.N. Charter Is Not a Counterexample to the Coercion Thesis The non-​military enforcement mechanisms satisfy all the conceptually necessary conditions to count as coercive sanctions, as I have explicated the notion for purposes of clarifying the Coercion Thesis. According to the canons of ordinary usage as they have been philosophically explicated here, a measure counts as a coercive sanction if and only if (1)  its deployment is authorized by a recognition norm governing officials of the system that permits or requires them to impose it as an institutional response to violations of valid mandatory norms of the system governing non-​official behavior; (2) it is intended and reasonably contrived to deter and punish non-​compliance with the applicable mandatory norms of the system governing non-​official behavior; and (3) it is plausibly thought to succeed to some non-​trivial extent in deterring competent agents from violating these mandatory norms. The non-​military, economic, and diplomatic enforcement mechanisms authorized by the U.N. Charter satisfy each of these conditions. No claims have been made here as to whether the system of international regulation grounded in the Charter counts as a system of law; it might or might not be as far as the claims of this chapter are concerned. What matters for my purposes is not whether those norms are really law; what matters is whether our system of international regulation, if properly characterized as a system of law, is a counterexample to the Coercion Thesis. It is true that if international law is not really law, then it is not a counterexample to the Coercion Thesis—​no matter what properties it might otherwise have. But, regardless of whether this system is really law, the U.N. Charter system cannot ground a counterexample to the Coercion Thesis because it authorizes enforcement mechanisms that are properly characterized as coercive sanctions, as that notion has been explicated throughout this volume.

10

Can There Be Law in a Society of Angels? Until Hart published The Concept of Law in 1961, it was taken for granted that everything that counts as a legal system backs some mandatory legal norms governing non-​official behavior with the threat of a coercive sanction. The first conceptual theorist of significance, St. Thomas Aquinas, argued that “human” law must include sanctions to ensure that we behave in accordance with the eternal moral law. On his view, fallen self-​interested subjects like us are prone to behave in dangerous ways because we can be presumed to understand the general principles of the eternal law, but not their specific applications: “[since] man has a natural participation of the eternal law, according to certain general principles, but not as regards the particular determinations of individual cases[,]‌. . . .[there is a] need for human reason to sanction them by law.”1 Subsequent theorists took for granted that a system of law cannot succeed in doing what it is needed to do unless some mandatory legal norms prohibiting violence and theft are backed by the threat of a coercive sanction. Blackstone, Bentham, Austin, and Kelsen went further, holding that it is a conceptual truth that every mandatory legal norm is backed by the threat of a coercive sanction, which is what constitutes the norm as mandatory and hence as giving rise to something that counts as an obligation. While the Coercion Thesis, as I have explicated it, does not entail that every mandatory legal norm is backed by a coercive sanction, the latter view clearly entails the Coercion Thesis. Every major philosopher who has theorized about the metaphysical nature of law since the 1600s holds the view that law consists, in part, of mandatory norms backed by coercive sanctions. Hobbes and Hume take the view, later more fully explicated by Bentham and Austin, that law consists in the commands of a sovereign willing and able to impose coercive sanctions for non-​compliance. Locke argues that the government’s power to coercively punish violations of natural rights, which is transferred from people in the state of nature to the government, is constitutive of its political power.2 Kant holds that every mandatory norm defining a duty, other than those of morality, is supported by an external constraint in the form of a sanction.3 And so on. 1 St. Thomas Aquinas, Summa Theologica, Question 91, Article 3; available at:  http://​www. newadvent.org/​summa/​2091.htm. 2 See Alex Tuckness, “Locke’s Political Philosophy” Stanford Encyclopedia of Philosophy (Edward Zalta, ed.); available at: https://​plato.stanford.edu/​entries/​locke-​political/​. 3 Mandatory moral norms, on Kant’s view, are distinct from mandatory social norms in virtue of expressing categorical imperatives fully derivable from reason and are hence, unlike these Coercion and the Nature of Law. Kenneth Einar Himma, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198854937.001.0001

232 

Can There Be Law in a Society of Angels?

Joseph Raz follows Hart in challenging the Coercion Thesis but rejects Hart’s view that our conceptual practices are indeterminate with respect to the Coercion Thesis in virtue of being indeterminate with respect to whether the U.N. Charter system counts as a system of international law.4 Raz argues, more forcefully, that our conceptual practices are consistent with the possibility of a legal system lacking authorized coercive sanctions: one can conceive of a system of law among beings who are angels in the sense of being always conclusively motivated, without regard for whether the system authorizes the imposition of coercive sanctions for non-​ compliance, to do what mandatory norms governing non-​official behavior require. Our conceptual practices, for Raz, are determinately clear on this point. This chapter rejects the society-​of-​angels argument on the ground that the psychological features of the “angels” are too far removed from what is remotely probable for rationally competent self-​interested subjects like us who live in worlds of acute material scarcity like ours to tell us anything of theoretical significance about the content of our conceptual practices with respect to the term law. There is nothing that counts as a legal system in the society of angels because our conceptual practices presuppose that the practices constituting something as a system of law are intended and reasonably contrived to regulate the behavior of rationally competent self-​interested subjects like us who would never, as a descriptive matter of contingent fact, conclusively defer to the dictates of a purely institutional authority the way the angels do because we should not do so, as an objective matter of normative practical rationality.

1.  Law and Real Angels The question of whether it is conceptually possible to have a system of law regulating the behavior of beings plausibly characterized as angels is a topic of perennial interest among analytic philosophers of law; however, this much is quite clear. There cannot be a system of law regulating the behavior of subjects who have no disposition to behave in socially undesirable ways because law is needed, as a conceptual matter, to regulate only the behavior of rationally competent subjects sometimes inclined to perform wrongfully harmful acts: as Lon Fuller puts this uncontentious point, “the need for rules . . . arises wholly out of man’s selfish, quarrelsome, and disputatious nature.”5 artifactual norms, supported by the internal constraint of human rationality. See, e.g., Leslie Green, “Legal Positivism” Stanford Encyclopedia of Philosophy (Edward Zalta, ed.); available at: https://​plato. stanford.edu/​entries/​legal-​positivism/​. 4 See Chapter 9 for a discussion of Hart’s views on the Coercion Thesis and the U.N. Charter system of international regulation. 5 Lon L. Fuller, The Morality of Law (Yale University Press 1964). As James Madison puts the point in The Federalist #51, “if men were angels, no government would be necessary.”

Law and Real Angels  233 Since, on this line of reasoning, angels are morally impeccable in the sense that they always do what they correctly believe morality requires, they have no need for law because they are constitutionally indisposed to behave in ways likely to create conflicts that might breach the peace among subjects. Insofar as the angels have no need for law to minimize such conflicts, they could not have a system of law because it could not make any difference with respect to how they behave; if it is a conceptually necessary condition for something to count as a legal system that it is needed to deter socially disruptive behavior and is hence needed to make a difference with respect to how rationally competent subjects behave, then there could not be something that counts as a legal system in a society of such beings. This is not to suggest that there cannot be law in a society in which rationally competent subjects always conform their behavior to the law. A totalitarian legal system employing highly intrusive surveillance technology to detect violations in conjunction with a system of draconianly disproportionate sanctions might succeed in eradicating socially disruptive behavior. The more likely it is that a rationally competent self-​interested subject believes she will be quickly caught and punished severely for doing something prohibited by the law, the less likely she is to do that something; rationally competent self-​interested subjects like us can be induced to abstain from all undesirable behavior if we can be persuaded that we will be rapidly punished with egregiously disproportionate suffering for such behavior. The Old Testament story about the Fall, if construed as a literal description of our history accurately illustrating facts about our psychology, is implausible for this reason. The narrative of the Fall begins with God’s warning to Adam and Eve not to eat from the tree of the knowledge of good and evil: You are free to eat from any tree in the garden; but you must not eat from the tree of the knowledge of good and evil, for when you eat from it you will certainly die (Genesis 2:17–​18; NIV).

Yet they were easily persuaded by the serpent to ignore God’s unambiguously pointed warning: When the woman saw that the fruit of the tree was good for food and pleasing to the eye, and also desirable for gaining wisdom, she took some and ate it. She also gave some to her husband, who was with her, and he ate it. Then the eyes of both of them were opened, and they realized they were naked; so they sewed fig leaves together and made coverings for themselves. Then the man and his wife heard the sound of the Lord God as he was walking in the garden in the cool of the day, and they hid from the Lord God among the trees of the garden (Genesis 3:6–​8).

There is much in this Biblical story that, construed literally, defies credulity, but harder to understand than anything else is that Adam and Eve were so

234 

Can There Be Law in a Society of Angels?

easily persuaded to disobey the command of their omnipotent creator, knowing that doing so would result in their immediate expulsion from Eden and death.6 While they lacked knowledge of good and evil up to that point, Adam and Eve were otherwise competent subjects who could be presumed to behave in a way that was prudentially and hence practically rational; otherwise, it makes no sense for God to threaten them with prudential detriment.7 They had to have understood that a being who created the universe from nothing has the causal ability to punish them with as much or as little pain as He wishes and that there could be nothing worse than what God had threatened them with. On the assumption that Adam and Eve were minimally competent, their decision to eat the apple was sheer lunacy. I can be far more easily tempted to do wrong than I would like to be, but there is not a chance in hell that I would make the decision Adam and Eve made given the information that they had: very few people are imprudent enough to disobey the command of an armed robber, but at least one has a non-​zero probability of escaping unharmed; given all that Adam and Even had witnessed, they had to know that the probability of getting away with disobedience was nil. The error from the standpoint of objective norms of prudential rationality is bad enough to raise questions about whether they were sufficiently competent to begin with and about whether an omnibenevolent being would punish them as harshly as God did. That they chose the way they did under those circumstances suggests that the more appropriate response on God’s part was to recognize that their brains needed to be either rewired or reprogrammed because they were wildly off in solving an utterly trivial problem in prudential arithmetic: refrain from doing the one thing that God demanded they not do, which they had no compelling need to do, or face immediate expulsion from paradise. Either way, if the thinking of Adam and Eve is indicative of what we can expect from rationally competent self-​interested subjects like us, then we are so driven by problematic desires that the most repressive totalitarian legal system imaginable could not succeed in even minimally keeping the peace. Those two were incorrigibly out-​of-​control: if having someone you know to be omnipotent threatening you with certain death is not enough to deter you from eating a piece of fruit, there is simply nothing the rest of us can do that would keep you from doing anything you might want to do, no matter how trivial or harmful.

6 While this punishment is most plausibly construed to be spiritual in the form of perpetual alienation from God, it is thought to create a condition of profound suffering. On this somewhat more humane conception of divine punishment, the torment of hell results from being deprived of divine love and not from being plunged into an everlasting fire where one is perpetually burned alive without being destroyed. 7 In a world where morality is irrelevant, such as was the case for Adam and Eve, the norms of prudential rationality exhaust the norms of practical rationality.

Law and Razian Angels  235 It is not a conceptually necessary condition for the existence of a legal system that it succeeds in deterring all violence, but it is a conceptually necessary condition for the existence of a legal system that it deters enough violence for subjects to live together in something that counts as a community. The more we resemble Adam and Eve, the less plausible it is to think that there could be a legal system that even minimally keeps the peace among rationally competent self-​interested subjects like us in a world, unlike theirs, of acutely painful material scarcity. Their sin is certainly original in more than one sense of the word, but it is not ours. We are not like Adam and Eve, whose behavior cannot be efficaciously regulated by a legal system, but we are also not like morally impeccable angels, whose behavior does not need efficacious regulation of any kind because they are never disposed to do what is morally wrong. As long as there is some institutional normative system that keeps us out of a Hobbesian state of nature, we are largely a population of rationally self-​interested but morally decent subjects who can nonetheless be tempted to compromise our decency if the prudential benefits are sufficiently seductive. The idea that it is a conceptually necessary condition for something to count as a legal system that it is reasonably contrived to keep the peace presupposes that rationally competent subjects of the system are psychologically capable of performing socially undesirable acts that breach the peace. If angels are morally impeccable in virtue of being psychologically incapable of ever performing an act that violates mandatory moral norms, an institutional normative system, no matter what content and features it might have, is not reasonably contrived to keep the peace among them. An artifact of any kind cannot, as a conceptual matter, be used to do what it is not needed to do; law can no more be used to keep the peace in a society in which the peace is perfectly kept without it than a medication can be used to cure an illness one does not have.

2.  Law and Razian Angels One might concede that law is not possible in a society of real angels but is possible in a society of beings conclusively motivated always to do what law requires regardless of what it requires. Such beings are rationally competent self-​interested subjects, like us, capable of acting in socially disruptive ways but can nonetheless be induced always to abstain from these acts by someone they accept as an institutional authority because they regard the authority’s directives as always giving rise to conclusive motivating reasons to conform. The beings of this imaginary society are commonly referred to as “angels,” but they are not real angels insofar as they are psychologically capable of acting in socially disruptive ways. They are angels only in the lexically contorted sense that, despite these latent antisocial propensities, they are always conclusively motivated

236 

Can There Be Law in a Society of Angels?

to conform their behavior to the mandatory norms of an institutional system without regard to whether they are backed by coercive sanctions; they need no sanctions backing the norms of the system because they regard the system as conclusively authoritative and always do what the norms require. There is nothing in this description of the angels that implicates the idea, discussed in Chapter 7, that the practices constituting something as a system of law are reasonably contrived to provide objective second-​order exclusionary motivating reasons for action. While it is clear that the angels are treating the authority’s directives as giving rise to conclusive motivating reasons to comply, they could be treating them as conclusive in virtue of providing prohibitively weighty first-​order motivating reasons to comply—​and not in virtue of also giving rise to second-​ order exclusionary motivating reasons to comply.8 But even if the angels are properly construed as treating these directives as second-​order exclusionary motivating reasons to comply, the argument shows, at most, that there can be something counting as a legal system in which subjects treat mandatory legal norms governing their non-​official behavior as second-​order exclusionary motivating reasons to comply; it does not show the stronger claim that these constitutive practices are reasonably contrived to give rise to something rationally competent subjects characteristically regard, as a descriptive matter of contingent fact, as second-​order exclusionary motivating reasons to comply with mandatory legal norms governing non-​official behavior because they should, as an objective matter of normative practical rationality, regard it as such. Either way, if the system of the angels satisfies Hart’s minimum conditions for the existence of a legal system and is metaphysically capable of authoritative regulation of behavior, the angels have, on this line of reasoning, something properly characterized, according to the canons of ordinary usage, as a system of law without coercive sanctions: insofar as their institutional normative system has all the paradigmatic properties of law and does everything for the angels that law must be able to do without authorizing coercive enforcement mechanisms, it can perform every task that law is needed, characteristically used, and supposed to perform, as a functionally normative matter—​including that of keeping the peace—​and hence counts as a system of law. Proponents seem to think that the claim that the system of the angels counts as a system of law is sufficiently uncontentious as to require no substantive defense. The entirety of the society-​of-​angels argument against the Coercion Thesis, as Raz articulates it, is given in the following passage:

8 See Chapter 7, Note 11, for some worries about the idea that exclusionary motivating reasons are equipped to play a useful role even in moral deliberations; the angels’ deference to authority seems to be moral in character—​though, as discussed below, it is problematic as an objective matter of normative practical rationality.

Law and Razian Angels  237 Is it possible for there to be a legal system in force which does not provide for sanctions or which does not authorize their enforcement by force? The answer seems to be that it is humanly impossible but logically possible. It is humanly impossible because for human beings as they are the support of sanctions, to be enforced by law if necessary, is required to assure a reasonable degree of conformity to law and prevent its complete breakdown. And yet we can imagine other rational beings who may be subject to law, who have, and who would acknowledge that they have, more than enough reasons to obey the law regardless of sanctions. Perhaps even human beings may be transformed to become such creatures. It is reasonable to suppose that in such a society the legislator would not bother to enact sanctions since they would be unnecessary and superfluous. If such a normative system has all the features of a legal system described above, then it would be recognized as one by all despite its lack of sanctions.9

This argument is surprisingly brief, given its theoretical import. Raz describes an institutional normative system without sanctions and simply asserts that ordinary speakers would characterize it as a legal system if it has all the other conceptually necessary properties of a legal system: “If such a normative system has all the features of a legal system . . . then it would be recognized as one by all despite its lack of sanctions” (PRN 160).10 In essence, the argument assumes, without reference to the lexical meanings of the relevant terms or anything that might do the work of a sociological survey of ordinary speakers, that competent speakers would converge on characterizing the system of the angels as one of law. While it is epistemically legitimate—​and argumentatively necessary on a modest approach to conceptual analysis—​to appeal to ordinary intuitions, something must be said in defense of the claim that the relevant intuitions are “ordinary” in the sense of being commonly shared among competent speakers of the language; some sort of empirical evidence is needed to show that these intuitions are widespread among competent speakers because conceptual claims like this can be properly grounded only in claims about the lexical meanings of the terms, which are contingently determined by our linguistic conventions for 9 Joseph Raz, Practical Reason and Norms (Oxford University Press 2002) 159–​60; emphasis added. Henceforth PRN. 10 Emphasis added. As is evident from the italicized emphasis, Raz’s argument assumes a modest approach to conceptual analysis insofar as it is concerned with how competent speakers would characterize the system. See also Raz, Ethics in the Public Domain (Oxford University Press 1994) 217. (“[G]‌iven the centrality of legal institutions in our structures of authority, their claims and conceptions are formed by and contribute to our concept of authority. It is what it is in part as a result of the claims and conceptions of legal institutions. This answer applies where the legal institutions themselves employ the concept of authority. But there may be law in societies which do not have our concept of authority. We say of their legal institutions that they claim authority because they claim to impose duties, confer rights, etc. Not having the concept, they cannot be confused about it, though we can be confused in attributing the claim of authority to them.”)

238 

Can There Be Law in a Society of Angels?

using them. The society-​of-​angels argument is prima facie problematic because it offers no defense of the claim that the relevant system is properly characterized as law according to our conceptual practices and hence begs the question against the Coercion Thesis.

3.  The Society of Angels and the Canons of Ordinary Usage At the outset, it is worth noting that the conclusion of the society-​of-​angels argument is inconsistent with the canons of ordinary usage as our linguistic practices define them. As discussed in Chapter 3, the empirically contingent linguistic conventions defining the application-​conditions for the term law as it picks out systems, rather than norms, of law clearly entail the Coercion Thesis: the dictionary reports of the term as it refers to systems of law all assert that systems of law consist of enforceable norms—​and norms are enforceable in the relevant sense only if enforcement is authorized by the system; every norm, after all, is enforceable if the character of the relevant possibility operator is logical, conceptual, or nomological, rather than practical. The modest approach adopted in this volume does not entail that every conceptual issue can be decided simply by consulting the canons of ordinary usage as they are reported in dictionaries. The content of our concepts, on this approach, is also determined by shared philosophical assumptions about the metaphysical nature of things that condition and are conditioned by our conventional linguistic practices. Insofar as the canons of ordinary usage are inconsistent with these assumptions, they do not accurately express the content of our conceptual practices with respect to the relevant concept-​term. Given that the canons of ordinary usage clearly entail the Coercion Thesis, the society-​of-​angels argument cannot warrant rejecting the Coercion Thesis without an epistemically justified showing that the relevant assumptions about the metaphysical nature of law are inconsistent on this point with the canons of ordinary usage. The proponent of the society-​of-​angels argument must do more than anyone has done with the argument up till now to warrant rejecting the Coercion Thesis; one must show that the denial of the claim that the angels have a system of law is logically inconsistent with the shared philosophical assumptions about the metaphysical nature of law that ground these canons of usage and define the ultimate touchstone for evaluating conceptual claims about law—​something that has been done nowhere in the literature. That said, this chapter cannot rest on the claim, though true, that the society-​ of-​angels argument has not received an epistemically adequate defense; that would leave open a door that must be closed decisively if this book is to succeed in defending the Coercion Thesis. This chapter is concerned, for this reason, to show

The Society of Angels and the Function of Law  239 the stronger claim that characterizing the system of the angels as one of law is inconsistent with these shared philosophical assumptions and hence with the social practices that determine the content of our concepts, on a modest approach.

4.  The Society of Angels and the Conceptual Function of Law The claim that the society-​of-​angels has a legal system despite lacking coercive sanctions is inconsistent with a number of philosophical assumptions about the metaphysical nature of law plausibly thought to ground our empirically contingent linguistic practices with respect to the term law as it picks out systems, rather than norms, of law. To begin, this claim is inconsistent with the claim that the conceptual function of a legal system is to keep the peace. If it is true that anything that counts as an artifact-​type A must be reasonably contrived to perform the conceptual function of A, then it is false, assuming the angels have a legal system, that anything counting as a system of law must be reasonably contrived to keep the peace among beings like us in worlds like ours. Insofar as the angels have something that counts as a system of law despite lacking coercive sanctions, it is conceptually possible for there to be a legal system without coercive sanctions. But a legal system without sanctions is not reasonably contrived to keep the peace among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours; remove all sanctions from any legal system in our world without replacing them with some other norms backed by sanctions, and the eventual, if not immediate, result will be a war-​of-​all-​against all. There are certainly other things that an institutional normative system without sanctions might be reasonably contrived to do in regulating the non-​official behavior of rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours. Such a system might be reasonably contrived to resolve disputes about what morality requires; guide behavior through the governance of norms; inform subjects of what right reason requires in a manner that does not require them to weigh the balance of reasons for themselves; make more determinate the requirements of morality; and so on. There are, after all, many artifactual systems efficaciously regulating non-​official behavior in our world that lack the kind of coercive sanctions characteristically authorized by legal systems for non-​compliance. But such a system is reasonably contrived to do that only insofar as there is some other artifactual mechanism that succeeds in keeping the peace. An institutional normative system lacking coercive sanctions simply could not do anything normative systems are needed, characteristically used, and supposed to do, as a functionally normative matter, if the peace is not minimally kept. One of the reasons, as discussed in Chapter 4, for thinking that the conceptual function of a legal system

240 

Can There Be Law in a Society of Angels?

is to keep the peace among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours is that a legal system cannot be non-​ idiosyncratically used, as a practical matter, to do anything that legal systems can be non-​idiosyncratically used to do unless the peace is minimally kept. The problem is that there are no other artifactual mechanisms reasonably contrived to minimally keep the peace among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours. All that stands between civilization and the state of nature is an institutional normative system that backs mandatory norms prohibiting certain assaults on persons or property with the threat of severe detriment; this is why every existing legal system in our world, as discussed in Chapter 3, backs mandatory norms prohibiting acts likely to lead to breaches of the peace with the threat of incarceration—​or worse. We know of nothing else reasonably contrived to do what legal systems that back certain mandatory norms governing non-​official behavior with the threat of incarceration are needed, characteristically used, and supposed to do, as a functionally normative matter, in our world.

5.  The Society of Angels and the How Problem of Legal Normativity The possibility of a sanctionless legal system is also inconsistent with the claim that the practices constituting something as a system of law are reasonably contrived to give rise to objective motivating reasons to comply. As argued in Chapter 6, there is no plausible solution to the How Problem of Legal Normativity on the assumption that the Coercion Thesis is false. Insofar as it is true that every conceptually possible legal system backs some mandatory legal norms with the threat of a coercive sanction, it is also true that the practices constitutive of a system of law are reasonably contrived to give rise to something rationally competent subjects characteristically regard as a motivating reason to comply, as a descriptive matter of contingent fact, because they should regard it as a motivating reason to comply, as an objective matter of normative practical normativity. If the Coercion Thesis is false, we cannot make sense of how the practices constitutive of law are reasonably contrived to provide either subjective or objective motivating reasons to comply. The cost, then, of characterizing the institutional normative system of the angels as a system of law is that we can no longer make sense of why we characteristically adopt the practices constituting something as a system of law to regulate the non-​official behavior of rationally competent subjects. It is a conceptual truism that something is reasonably contrived to regulate behavior only insofar as it is reasonably contrived to give rise to something that rationally competent subjects are characteristically likely, as a descriptive matter of contingent fact, to regard as normatively relevant because they should, as an objective matter of normative

The Society of Angels and the Content Problem  241 practical rationality, regard it as normatively relevant. The claim that the practices constitutive of law are not reasonably contrived to give rise to subjective and objective motivating reasons to comply is a straightforward refutation of any claim that implies it.

6.  The Society of Angels and the Content Problem of Legal Normativity Most immediately relevant with respect to assessing the coherence of the society-​ of-​angels argument is the analysis of Chapter  8, which addresses the Content Problem of Legal Normativity. The Content Problem is concerned with identifying the content of the objective first-​order motivating reasons to which the practices constituting something as a system of law are reasonably contrived to give rise. On the assumption that the Coercion Thesis is true, these constitutive practices are reasonably contrived to give rise to an objective first-​order instrumental motivating prudential reason to comply with the law as a means of avoiding the imposition of coercive sanctions that rationally competent self-​interested subjects prefer to avoid, as a descriptive matter of contingent fact, because they should prefer to avoid them, as an objective matter of normative practical rationality. The only other reason to which one might think these constitutive practices are reasonably contrived to give rise is an objective first-​order intrinsic motivating reason to obey the law because it is the law. But, as discussed in Chapter 8, the idea that the practices constituting something as a legal system are reasonably contrived to give rise to objective intrinsic motivating reasons is problematic insofar as there is no plausible source for such reasons: neither aesthetic, prudential. nor moral considerations give rise to such reasons; and the artifactual mechanisms of law are not equipped to give rise to an autonomous source of basic irreducible normativity, such as would be needed to give rise to a novel kind of motivating reason. This disability stands in stark contrast to the objective instrumental motivating reasons to which these artifactual mechanisms give rise if the Coercion Thesis is true. There is no mystery, after all, how backing a directive, legitimate or otherwise, with the threat of detriment that rationally competent subjects characteristically prefer to avoid, as a descriptive matter of contingent fact, because they should prefer to avoid it, as an objective matter of normative practical rationality, gives rise to an objective instrumental motivating prudential reason to comply as a means of avoiding the detriment. Insofar as the angels are always conclusively motivated to do what their authoritative system requires because it requires it and hence always treat the authority’s directives as conclusive intrinsic motivating reasons to comply, some of the arguments adduced in Chapter 8 on the Content Problem are potentially relevant here. Even so, the operative claim of the society-​of-​angels argument is not that these

242 

Can There Be Law in a Society of Angels?

constitutive practices are reasonably contrived to give rise to objective intrinsic motivating reasons to obey law because it is law, which was refuted in Chapter 8; it is, rather, that there can be an institutional normative system that counts as a system of law among rationally competent self-​interested subjects who regard the fact that some group of persons serving as officials recognize a norm as, by itself, sufficient to give rise to conclusive motivating reasons to comply with it. This claim will be evaluated in the next section.

7.  Evaluating the Normative Commitments of the Angels One might sensibly demand more by way of critical response to the society-​of-​angels argument than the responses of the last four sections because they presuppose the arguments made in the preceding chapters of this book in defense of the Coercion Thesis. While these arguments strike me as sufficient to warrant rejecting the society-​ of-​angels argument, it is not unreasonable to demand more since, given the abiding influence of the argument, the burden of persuasion is mine. The subsections below articulate independent grounds for rejecting the argument that augment the arguments of those chapters and put the finishing touch on the case for the Coercion Thesis.

7.1  The Improbable Oddness of the Angels’ Normative Commitments The claim that the angels are always conclusively motivated to obey the mandatory norms of the system without regard to whether they are enforced by coercive sanctions presupposes two claims about their behavior and mental states. Insofar as they are presumed to be rationally competent and self-​interested, the claim that the angels are always conclusively motivated to obey the norms of the system regardless of sanctions presupposes not only that (1) the behavior of the angels always satisfies those norms but also that (2) each angel justifiably believes that every other angel will always obey those norms. If either of these claims are false, then it would be straightforwardly problematic, as an objective matter of normative practical rationality, for the angels always to defer to the requirements of law without demanding they be backed by coercive sanctions to make it more likely that other angels also comply.11 11 The problem arises because the law, like morality, characteristically demands that subjects refrain from acts that, though harmful to others, conduce to their own self-​interest. Insofar as criminal norms prohibit theft, they ultimately require that self-​interested agents refrain from expropriating property that can substantially benefit them; insofar as tax laws punish non-​payment of taxes, they require that self-​interested agents pay funds to the government that they could spend in more prudentially

Evaluating the Normative Commitments of the Angels  243 Both (1) and (2) are implausible if intended to express properties that rationally competent subjects like us are likely to instantiate. As to (1), it is vanishingly improbable that rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours would converge—​given the obvious limitations of our shared criminal justice practices in identifying and punishing persons who violate these norms—​in always satisfying the norms of an institutional normative system that comprehensively regulates behavior the way modern municipal legal systems do. It might be true that many of us manage to live our lives without violating legal prohibitions on violence and theft. But the probability that all of us converge, at any given moment, on always satisfying all mandatory legal norms in a world like ours in which there are not enough material resources to satisfy all of our wants and needs is negligible—​though it is admittedly non-​zero. As to (2), there is little reason to think that rationally competent self-​interested subjects like us inhabiting a world of acute material scarcity like ours could ever come to trust, and be epistemically justified in doing so, that other rationally competent self-​interested subjects will never violently infringe against other persons or their property—​much less that they will never violate even mandatory legal norms defining civil infractions, like speeding and illegal parking. Even those who believe that human nature is inherently good or morally perfectible recognize that this is a difficult world in which the best of us will sometimes be tempted to do bad things that, as far as shared judgments of morality are concerned, impermissibly benefit ourselves at the expense of others; we all experience our desires with a painful intensity that we sometimes allow to trump the less urgently felt demands of conscience. This is true even in the highly improbable circumstance in which people happen, for whatever reason, always to converge on satisfying legal norms that prohibit assaults on persons or property. If, as seems plausible, it is epistemically sensible to believe that otherwise domesticated agents might nonetheless sometimes be tempted to disobey these laws, we are not even remotely likely to believe that other people will never commit such assaults; the prospect that some people might commit socially disruptive acts will always be a live possibility in our minds—​no matter how well-​behaved everyone else might otherwise appear to be. The line between what we take to be a justified belief and what we take to be wishful thinking can get blurry at times; however, the angels’ shared belief that other angels will never commit such acts would clearly be pathologically wishful thinking in the mind of anyone who lives in this world. But even if we could somehow make ourselves adopt more optimistic assumptions about the benevolence of others, it is doubtful that we could, given the facts desirable ways. It is prudentially irrational, absent exceptional circumstances, for a self-​interested agent to commit conclusively to always obeying legal norms unless she believes correctly that other subjects will always obey at least those norms that prohibit violent assaults.

244 

Can There Be Law in a Society of Angels?

of our experience, ever be comfortable replacing our legal system with an institutional normative system that does not authorize the deployment of coercive sanctions for violations of mandatory legal norms governing non-​official behavior that prohibit assaults on persons and property. The probability that any of us could so completely trust every other person that we would be willing to do without institutional regulation by norms backed by sanctions is theoretically insignificant; the probability that all of us would converge in developing such an attitude is even smaller. We all lock our doors to keep ourselves safe. This is worrisome because the “angels” are conceived to resemble us in ways that ensure that they are sometimes tempted to commit acts that breach the peace even if they are successful most of the time in resisting temptation: [The angels] may pursue many different and conflicting goals and they may share our difficulties in settling disputes and resolving conflicts of interests by mutual agreement. They differ from us only in having universal and deep-​rooted respect towards their legal institutions and in lacking all desire to disobey their rulings (PRN 159).12

Insofar as the angels share these difficulties, it is because they share those characteristics of ours that sometimes tempt rationally competent self-​interested subjects in worlds of acute material scarcity like ours to behave in socially disruptive ways. Despite their odd willingness always to obey the law, these angels know, to the extent that they resemble us psychologically, that every other angel is capable of violating mandatory legal norms prohibiting assaults on persons and property. It is simply implausible to think that rationally competent subjects like us in worlds like ours would be willing, as is presupposed of the angels by claim (2), to trust that all the others will always abstain from undesirable assaults without backing the norms of their system with coercive sanctions. That said, it would be helpful to consider two possibilities having to do with the character of the angels’ system: either the institutional normative system of the angels includes mandatory legal norms governing non-​official behavior that prohibit undesirable assaults on persons and property or it does not. Insofar as the angels always refrain from committing assaults likely to lead to escalating cycles of retaliatory assaults among increasingly wider circles of subjects, there are thus two possibilities corresponding to whether the system includes the relevant prohibitions:  either (i)  the angels abstain from committing such assaults because 12 For similar reasons, Raz makes it a point to assert that “[p]‌erhaps even human beings may be transformed to become such creatures” PRN 160, n. 9. It is clear that Raz understands that the success of the argument critically depends on defining the characteristics of the angels so that they resemble us in conceptually crucial respects. Assertions about the possibility of a group of lions having something resembling a system of norms simply cannot tell us anything about our conceptual practices with respect to systems of what we characterize as law.

Evaluating the Normative Commitments of the Angels  245 mandatory norms of the system prohibit them or (ii) the angels abstain from committing such assaults even though the system does not include mandatory norms prohibiting them. Both possibilities are problematic. If, to begin, the institutional normative system of the angels does not include mandatory norms governing non-​official behavior that prohibit such assaults, it is not properly characterized as a system of municipal law. Such a system would include only norms defining civil wrongs or infractions properly remedied by the payment of compensatory damages or fines—​and there is simply nothing in our ordinary experience that would dictate characterizing such an institutional normative system as a system of municipal law. It is neither a conceptually necessary nor conceptually sufficient condition for the existence of a legal system that it includes norms defining civil wrongs of the sort associated with contract and tort law. If, as seems plausible, it is conceptually possible to have a municipal legal system that includes only mandatory criminal norms prohibiting undesirable assaults on persons or property, it is not a conceptually necessary condition for the existence of a legal system that it includes civil norms like those of contract and tort law. If, as also seems plausible, it is a conceptually necessary condition for the existence of a legal system that it includes mandatory legal norms governing non-​official behavior that prohibit some of these assaults, it is also not a sufficient condition for the existence of a municipal legal system that it includes civil norms like those of contract and tort law. It might be true that every existing municipal legal system includes legal norms defining these civil wrongs; but insofar as it is neither a necessary nor sufficient condition for the existence of a legal system that it includes norms defining civil wrongs like those of contract and tort law, the inclusion of such norms plays no essential role in explicating the content of our concept of law. What is conceptually essential to the existence of a municipal legal system is that it includes mandatory norms governing non-​official behavior that prohibit enough assaults on persons and property to enable rationally competent self-​interested subjects to live together in comparative peace; while it might be true that such prohibitions need not be characterized as criminal in character, there could not be a municipal legal system that does not prohibit some such behaviors. The society-​of-​angels argument, then, fails if it is assumed that the institutional normative municipal system of the angels does not include mandatory legal norms governing non-​official behavior that prohibit some undesirable assaults on persons and property. The argument likewise fails if it is assumed that the institutional normative system of the angels includes mandatory norms governing non-​official behavior that prohibit undesirable assaults on persons and property. If this is assumed, then the fact that these behaviors are prohibited by morality or by norms of prudential rationality plays no role at all in explaining their compliance. What wholly explains why the angels always abstain from these morally problematic assaults, then, is that some persons serving as officials have recognized and

246 

Can There Be Law in a Society of Angels?

applied mandatory norms that prohibit violent assaults. Since the angels are assumed to resemble us with respect to the psychological characteristics that explain why we sometimes commit these undesirable assaults, they would behave the way Hobbes believes we would behave in the state of nature if not for just the fact that these officials have prohibited them. Such norms do not need to be enforced, on this line of argument, because there is no need to provide angels with a prudential disincentive to commit murders. The angels, in essence, simply do what the officials require because it has been required by the officials. Rationally competent self-​interested subjects like us who inhabit worlds of acute material scarcity like ours are not even remotely likely to converge in having the motivational preferences of the angels. It is surely true that there are people in our world who fetishize persons in positions of institutional authority to the extent that the angels do, affording every statement or directive with some special normative force; however, the devotion of even those persons is conditioned upon their belief that the authority is uniquely inclined to favor only laws that its devotees believe should be recognized and enforced. If Donald Trump were to change his position with respect to immigration, security at the U.S.-​Mexico border, gun rights, the rights of LGBTQA+ persons, or all things having to do with former President Barack Obama, his seemingly immovable base support would begin to erode; and the more changes he made in these positions, the further it would erode. The angels, in contrast, are committed to doing whatever the authority requires regardless of the content of what the authority requires and regardless of whether the directives are backed by the threat of coercive sanctions. While nomologically possible, the probability that rationally competent self-​interested subjects like us who inhabit worlds of acute material scarcity like ours would ever converge in regarding some non-​coercive institutional authority with such conclusive content-​ independent deference is theoretically negligible; as Raz himself concedes, the idea that beings like us could live peacefully in a society regulated by an institutional system of norms lacking sanctions for non-​compliance is logically but not “humanly” possible (PRN 158). One might argue that what explains this conclusive commitment on the part of the angels to doing what the authority requires because the authority requires it is that they converge in regarding the authority as morally legitimate—​and maximally so; however, rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours are not minimally likely to converge in adopting an attitude of such wholly conclusive deference towards legal requirements even where we converge in believing that the system’s authority is morally legitimate. Insofar as the view that an authority is morally legitimate depends on the content of what is recognized and applied as law, it is unlikely, as a descriptive matter of contingent fact, that even rationally competent self-​interested subjects who regard the authority as legitimate would treat the authority’s directives as giving rise to conclusive motivating reasons to comply.

Evaluating the Normative Commitments of the Angels  247 There is no inconsistency between the claim that S is a morally legitimate legal system and the claim that some laws of S are morally problematic; while there is clearly a limit on how much immoral content can be incorporated into a legal system plausibly characterized as legitimate, it is surely compatible with the legitimacy of a legal system that it contains some morally problematic laws. A tax law that affords affluent people a morally unjustified reduction of taxes is not, by itself, sufficient to show that the legal system in which that norm is valid is illegitimate. It is not preposterous, moreover, to think that a morally legitimate system might include mandatory legal norms that require what is morally wicked. The Fugitive Slave Act of 1850 clearly required something morally wicked, but it is not obvious that the inclusion of this deeply problematic legal norm vitiated the legitimacy of the U.S. legal system; whether or not it did depends on one’s intuitions about the outer limits of legitimacy.13 While there is clearly a limit to how much wickedness a legal system can allow or require compatible with its being morally legitimate, it seems that the legitimacy of a legal system is compatible with the inclusion of a small number of wicked laws. Given that morally legitimate legal systems can include some morally problematic content, it is reasonable to think that rationally competent self-​interested subjects like us who accept the legitimacy of a system are nonetheless prohibitively unlikely to converge on adopting the angels’ attitude of conclusive deference. It is plausible to hypothesize that a theoretically significant number of us would refuse to do something we believe is wicked simply because it is required by law—​ especially if we assume that the expected prudential costs are low because we are not likely to be caught and punished. But insofar as it is true that we would be disinclined to comply with unjust laws of a system we nonetheless continue to regard as morally legitimate, it follows that we do not regard the system as giving rise to conclusive motivating reasons to comply; we characteristically regard, as a descriptive matter of contingent fact, even a morally legitimate legal system as giving rise to only defeasible motivating reasons. Our practices with respect to law are grounded in shared views about what is reasonably likely for rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours. It might be nomologically possible for us to live in peace in a community without an institutional system of norms backed by coercive sanctions; however, this tells us nothing about any of our normative practices. What matters with respect to our conceptual and normative practices has to do with what features the practices constitutive of a legal system must have to be 13 The issue here is complicated by the fact that norms like this can be valid in a system only against the backdrop of a cluster of other norms that require what is wicked. The Fugitive Slave Act would have had no practical consequences if not for the cluster of laws that permitted and protected the institution of slavery in the Antebellum South; and one might plausibly regard the system as wholly illegitimate in virtue of including these norms. It is difficult to think of an example of a possible legal norm requiring something wicked that does not presuppose a cluster of wicked legal norms.

248 

Can There Be Law in a Society of Angels?

reasonably contrived to do what legal systems are needed, characteristically used, and supposed to do, as a functionally normative matter, in worlds of acute material scarcity like ours. The society-​of-​angels argument tells us nothing about the content of our concept of law as it is determined by our practices.

7.2  The Practical Rationality of the Angels’ Deference to the Authority The society-​of-​angels argument presupposes that there could be circumstances in which conclusive deference to law is justified under objective norms of practical rationality. If there are no such circumstances, then conclusive deference to legal authority is practically unjustified in every conceptually possible world and is hence practically irrational; while it might be practically rational to do what is contingently unjustified under the circumstances, it is practically irrational to do what is necessarily unjustified. We can learn absolutely nothing about law from the behavior of beings whose attitudes towards the law are so thoroughgoingly irrational; the practices constituting something as a legal system are needed, characteristically used, and supposed, as a functionally normative matter, to regulate the behavior of subjects who are rationally competent in virtue of being largely responsive to objective norms of practical rationality. If such conclusive deference to authority is irrational, then the society of angels can tell us nothing about our concept of law. The idea that conclusive deference to authority can be objectively justified is problematic. The only circumstance in which conclusive deference would be justified by the norms of theoretical and practical reason is one in which subjects are epistemically justified in believing the following claims: (1) the behavior of every other rationally competent subject will always satisfy the mandatory norms of the authority prohibiting undesirable assaults on persons and property; (2)  each rationally competent subject believes that every other rationally competent subject will always obey those mandatory norms; and (3) the lawmaking officials are morally impeccable—​i.e. that the officials always know what is objectively right, as a moral matter, with respect to exercising their authority and will always do what they know to be objectively right. Claim (3) is utterly crucial in objectively justifying conclusive deference on the part of rationally competent subjects in these circumstances. As long as subjects are epistemically justified in believing that rationally competent self-​interested officials might make a mistake with respect to what morality requires and might sometimes do wrong—​either because they are mistaken about what morality requires or because they are motivated to do what they correctly believe is wrong by prudentially rational considerations—​they would be unjustified, as an objective matter of normative practical rationality, in committing to always obeying the law regardless of content.

Evaluating the Normative Commitments of the Angels  249 It would be both prudentially and morally unjustified to conclusively defer to the mandatory norms of an authority who is not known to be morally impeccable with respect to her dictates. It would be unjustified, as an objective matter of purely prudential rationality, because rationally competent subjects like us are epistemically justified, given the facts that obtain in our world, in believing that morally imperfect officials might, as a practical matter, require that they do something sufficiently contrary to their interests that they are better off not doing it—​especially in an institutional normative system lacking coercive sanctions. It would be unjustified, as an objective matter of moral rationality, because rationally competent subjects like us are epistemically justified, given the facts that obtain in our world, in believing that morally imperfect officials, even in a system that satisfies the standards of moral legitimacy up to that moment, might sometimes require that they act in ways that cause harms they have objectively conclusive motivating moral reasons to abstain from causing. The claim that the angels have irrationally committed themselves to regarding the mandatory norms of the authority as giving rise to conclusive motivating reasons to comply does not imply the claim that they are, as a general matter, practically irrational in a sense that would insulate them from moral accountability on our shared meta-​ethical views. Rationally competent subjects can, and often do, act in ways that are practically irrational; rational competence does not require always doing what is practically justified and never doing what is practically unjustified. Subjects like us are, after all, also highly emotional beings whose passionate responses to the world are not always appropriately countered by some constraining Cartesian light of reason; this is why we sometimes do things that are objectively irrational. But the idea that we can learn something worth learning about the concept of law as our social practices determine it from a thought-​experiment in which otherwise competent subjects characteristically behave with respect to an authority in a manner that is objectively irrational is epistemically illegitimate—​and obviously so. Our conceptual and normative practices with respect to law are grounded in shared views about how rationally competent self-​interested beings like us are characteristically likely, as a descriptive matter of contingent fact, to behave in worlds of material scarcity like ours. A society of beings that behave like the angels do simply cannot tell us anything that would explain why we characteristically use law to do the things we use it to do. We could not put together a coherent set of institutional practices to regulate behavior on the assumptions that rationally competent subjects will behave in unpredictably irrational ways in response to those practices. The angels’ conclusive deference to authority would be regarded in our world as cult-​like and hence indicative of psychological pathology; the proper response when we are confronted with persons who have come under the spell of a cult-​leader is to try to remove

250 

Can There Be Law in a Society of Angels?

them from the cult and place them in the care of physicians and psychologists who can “deprogram” them. To ground a putative counterexample to the Coercion Thesis in a society of beings needing to be deprogrammed is straightforwardly problematic; the normative practices of the society-​of-​angels are no more plausibly characterized as law than the normative practices of your favorite cult—​regardless of whether the relevant norms are backed with coercive sanctions. With or without sanctions, there is nothing in that society properly characterized as a system of law, as far as our conceptual practices are concerned. While it is epistemically legitimate to ground counterexamples to claims that purport to be conceptually or metaphysically necessary in nomological possibilities that are somewhat far-​fetched, there are limits to how far-​fetched those possibilities can be. The society-​of-​angels counterargument to the Coercion Thesis exceeds those limits and fails for that reason.

8.  Revisiting Existing Legal Systems It is difficult to see a plausible motivation for denying the Coercion Thesis that is even remotely grounded in our empirical experience with law. If we look at all existing legal systems, which centrally figure into the construction of our concepts, each includes a number of institutional features that jump out as paradigmatic of law as our conceptual practices define the notion: (1) every existing legal system contains legal norms that constrain or empower ordinary persons; (2) every existing legal system has an official body that performs a legislative function; (3) every existing legal system includes a set of norms authorizing certain persons to perform the legislative function, as well as a set of norms prescribing the manner in which that function is to be performed by those authorized to do so; (4) every existing legal system has an official body that performs a judicial function; (5) every existing legal system includes a set of norms authorizing certain persons to perform the judicial function, as well as a set of norms prescribing the manner in which that function is to be performed by those authorized to do so; (6) every existing legal system has an official body that performs an executive function; (7) every existing legal system includes a set of norms authorizing certain persons to perform the executive function, as well as a set of norms prescribing the manner in which that function is to be performed by those authorized to do so; and (8) every existing legal system includes a set of norms authorizing the application of coercive enforcement mechanisms for some violations of legal norms. There are a number of conceptual relationships between the various features worth noting. Feature (1) is basic in some logical sense and related to each of the other features in theoretically significant ways, although the relationships between (1) and the others are surely dialectical in character. The other features are related in the following respects. Features (2) and (3) specify the existence of a legislative

Revisiting Existing Legal Systems  251 body and function, which are concerned with the promulgation of laws, as well as the procedures by which certain persons are authorized to perform those functions and the manner in which those functions are to be performed. Features (4) and (5) specify the existence of a judicial body and function, which are concerned with the recognition and application of laws, as well as the procedures by which certain persons are authorized to perform those functions and the manner in which those functions are to be performed. Features (6) and (7) specify the existence of an executive body and function, as well as the procedures by which certain persons are authorized to perform those functions and the manner in which those functions are to be performed. But, as a conceptual matter, the executive function is concerned with the enforcement of existing laws, and laws cannot be enforced unless there are norms governing official behavior authorizing coercive enforcement mechanisms and norms governing official behavior specifying who is authorized to perform the executive function and how those functions should be performed. This means that features (6) and (7) presuppose the existence of authorized coercive enforcement mechanisms and hence feature (8). Given any sensible methodology for conceptual jurisprudence, a conceptual theory of law would have to include the features picked out by all of (1) through (5) as conceptually necessary features of a legal system. No matter how one conceives of the function of a legal system, a legal system could not do what it is needed to do without containing those features. If the function of a legal system is to resolve disputes, then there must be norms, as well as institutions authorized to make, recognize, and apply those norms. If the function of a legal system is to guide behavior, then there must be norms, as well as institutions authorized to make, recognize, and apply those norms, that are reasonably contrived to perform that function. If the function of a legal system is to keep the peace by regulating behavior through the governance of norms metaphysically capable of guiding behavior, then there must be norms, as well as institutions authorized to make, recognize, and apply those norms. This is why it is a commonplace that the existence of legislative and judicial institutions, along with the recognition norms that create and regulate those institutions, are conceptually necessary features of law.14 The angels have a need for an institutional normative system that has only the features picked out by (1) through (5). Insofar as they need an institutional normative system to coordinate activities and to mediate disagreements that could lead to breaches of the peace, they need a set of institutional norms, as well as a legislative

14 It might be true, as a contingent empirical matter, that every existing legal system defines separate legislative, judicial, and executive bodies, but it should be noted that there is nothing in our conceptual practices that entails that distinct bodies must perform these necessary functions. While it is surely not possible for, say, one person or agency to perform all of these functions in any legal system as complex as the municipal legal systems that define the paradigms for conceptual theorizing about law, there is nothing in our conceptual practices that precludes the possibility of a legal system in which one person or agency performs all three of these functions.

252 

Can There Be Law in a Society of Angels?

institution to make them and a judicial institution to apply them in resolving disagreements. Insofar as they are always conclusively motivated to obey these norms without regard for whether they are backed by sanctions, they have no need for those features specified in (6), (7), and (8). But insofar as a conceptual theory must express all the features of law needed to keep the peace among rationally competent self-​interested beings like us in worlds of material scarcity like ours, it would have to include as conceptually necessary features of a legal system those expressed by (6), (7), and (8). As discussed above, we differ from the angels in that we are not characteristically motivated, as a descriptive matter of contingent fact, conclusively always to follow the law no matter what it requires and are all-​too-​frequently inclined to act in ways that can create conflicts threatening the stability of the community we need in order to thrive in a world of limited material and social resources. There will always be Holmesian “bad men” among us, of course, but there will also always be people of conscience who must sometimes be induced with the threat of a coercive sanction to comply with some mandatory legal norm they might otherwise disobey out of a conviction that it is unjust or imprudent. Unlike the angels, we need sanctions to ensure that the peace is kept enough for us to live and work together in comparative peace and safety.

9.  Conclusions This chapter completes what I take to be a comprehensive case for the Coercion Thesis. The first part of this book showed that the canons of ordinary usage, empirical experience, and plausible assumptions about what we characteristically use systems of law to do converge in supporting the claim that it is a conceptually necessary condition for the existence of a legal system that it backs some mandatory norms prohibiting socially disruptive acts with the threat of a coercive sanction. The second part distinguished three conceptual problems of legal normativity and showed that the only account that can succeed in vindicating law’s conceptual normativity presupposes the Coercion Thesis. The last part rebutted two putative counterexamples to the Coercion Thesis; neither the institutional normative system of international regulation defined by the U.N. Charter nor the institutional system of the angels, such as it is, refutes the Coercion Thesis. The arguments in this volume thus vindicate what was largely regarded as obvious from core features of ubiquitous legal practice prior to the late twentieth century. There is simply no plausible ground to reject the historically predominant view that law is, by nature, coercive; what is needed to counter the prudential incentives to commit acts that threaten the peace is a countervailing prudential disincentive in the form of authorized coercive sanctions that are contrived to punish and deter non-​compliance. Keeping the peace is exactly what law is contrived,

Conclusions  253 characteristically used, and supposed, as a functionally normative matter, to do in worlds of acute material scarcity like ours; unless the peace is kept, law cannot be used to do anything else that law is non-​idiosyncratically used to do. It is law’s essentially coercive nature that explains why systems of law have enabled us to live peacefully together in every society that has one.

APPENDIX

Can There Be a System of Municipal Law with Only Compensatory Damages?* The Coercion Thesis asserts that it is a conceptually necessary condition for the existence of a legal system that it backs some mandatory norms prohibiting acts breaching the peace with the threat of a coercive sanction. The Coercion Thesis hence describes two conceptually necessary conditions for something to count as a legal system: the first is that something counts as a system of law only insofar as it includes mandatory norms governing non-​official behavior that prohibit a sufficient number of assaults on persons and property to enable rationally competent subjects to live together in a state of comparative peace; the second is that something counts as a system of law only insofar as it includes norms governing official behavior that authorize courts to impose coercive sanctions for non-​compliance with court orders and norms governing non-​official behavior that prohibit assaults on persons and property. Systems of municipal law in our world authorize different types of coercive sanctions than those authorized by the U.N. Charter system of international regulation. While the U.N. Charter system relies primarily on the threat of economic and diplomatic sanctions and secondarily on the threat of aggressive military operations, municipal systems rely ultimately on the threat of incarceration (i.e. forcible detention); a municipal legal system can neither execute, imprison, nor torture someone without forcibly detaining her.1 The Coercion Thesis, thus, entails that it is a conceptually necessary condition for the existence of a system of municipal law that it backs some mandatory norms governing non-​official behavior that prohibit acts breaching the peace with the threat of incarceration. This seems to entail that systems of municipal law are foundationally concerned, by nature, with prohibiting acts properly characterized as crimes in virtue of being properly characterized as punishable with something involving incarceration. While it might be conceptually possible to have a system of municipal law without civil norms like those typically found in existing legal systems that govern torts and contracts, it is not conceptually possible * I am indebted to my partner, Maria Elias Sotirhos, for an objection to something in Chapter 3 that made me realize I had to write an appendix in response. 1 There is some pathology to worry about here: one might think a court-​ordered assassination of a criminal defendant convicted in absentia counts as a coercive sanction insofar as an aggressive military operation ordered by the appropriate U.N. tribunal counts as a coercive sanction. (Military operations authorized for purely defensive or pacific purposes do not count as coercive sanctions as that notion has been defined in this volume; see Chapter 9). While I think it implausible to characterize most of the relevant norms of the Charter system as criminal norms, this nonetheless presents a puzzling conceptual issue about which I have conflicting intuitions: on the one hand, I suppose we can call it punishment without too much violence to the concept; on the other, this seems more properly characterized, strictly speaking, as a police operation concerned with eliminating a threat. It does not differ much from someone’s attempting to gain a reward for capturing a subject “dead or alive” by shooting the subject, which also seems properly characterized, strictly speaking, as an assassination or something else that does not count as punishment. For the purposes of this Appendix, I will assume court-​ordered assassination of a defendant convicted in absentia is not properly characterized as punishment. Either way, nothing of real substance turns on this; the Coercion Thesis and some of the supporting arguments are easily tweaked to accommodate this possibility.

256 Appendix to have a system of municipal law without some mandatory norms prohibiting violent acts that count as crimes in virtue of being backed up by the threat of something that counts as punishment. Insofar as the most basic task of a legal system is to keep the peace enough for rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours to live together in something that counts as a community, it is foundationally concerned with deterring violent crime by punishing it.2 The Coercion Thesis requires both that some of these acts are prohibited by systems of municipal law as crimes and that some of these legally prohibited acts are backed ultimately by the authority of the court to incarcerate.3 What ultimately drives the Coercion Thesis is the general claim that it is a conceptually necessary condition for anything to be properly characterized as an artifact “A” that it is reasonably contrived, when not in a state of disrepair, to do what As are characteristically used and supposed, as a functionally normative matter, to do. Applied to a legal system, which is indisputably an artifact insofar as there cannot be a legal system in a world without beings like us to manufacture it, this claim entails that it is a conceptually necessary condition for something to count as a legal system that it is reasonably contrived to do what legal systems are characteristically used and supposed to do, as a functionally normative matter. Even if one takes the absurdly minimalist view that the most that can be said about what law is characteristically used and supposed to do, as a functionally normative matter, is to efficaciously regulate non-​official behavior because one thinks that talk of conceptual functions is problematic or that there are just too many things that law can do to pick out any as especially important, it should be clear that law can do none of these things among beings like us in worlds of material scarcity like ours unless it keeps the peace enough for us to live together in a community. If the peace is not kept because people are generally running amok with violence, then their behavior is not being efficaciously regulated by a legal system. It might be true that coercive enforcement mechanisms are not needed in a society of real angels to keep the peace, but this claim implies only the claim that something that counts as a system of law is not needed to do what systems of law are foundationally needed, characteristically used, and supposed to do, as a functionally normative matter, in worlds of acute material scarcity like ours. It does not imply, as argued in Chapter 10, the claim that there can be a legal system without sanctions. Admittedly, it is not absurd to think that something like a legal system is needed, even among such beings, to authoritatively resolve or mediate disputes about what to do; however, the only disputes that legal systems are reasonably contrived to authoritatively resolve or mediate are disputes that arise under law about what the law requires. But there cannot,

2 Systems of international law are likewise, as a conceptual matter, foundationally concerned with keeping the peace among nations enough for them to exist in a community, but they characteristically, if not necessarily, deploy different coercive mechanisms for enforcing these norms. Because international law is concerned, as a conceptual matter, with regulating relations among nations, which are complex abstract objects constructed by the activities of people who function as officials but are not simply sets of those people, one would expect that the regulatory mechanisms, including those contrived to enforce the relevant norms, would have to be very different from those of municipal law. This is yet another reason that systems of international law in our world, such as they are and have been, do not constitute paradigms of law that ground conceptual theorizing. 3 By allowing for the conceptual possibility of mandatory norms prohibiting certain assaults on persons or property that are not backed by the threat of something necessarily involving incarceration, I mean to leave open the possibility, though this strikes me as counterintuitive, of something counting as a mandatory legal norm that is not backed by anything counting as a coercive sanction. While I am inclined to think, there could not be a legal norm properly characterized as mandatory that is not backed by the threat of a coercive sanction, I am agnostic on the issue for purposes of this volume.

Appendix  257 as a conceptual matter, be a dispute arising under law about what it requires unless there is a legal system; there are no laws without a legal system. Some theorists take the view that the basic function of a legal system is to authoritatively resolve disputes about what one should do, all things considered—​or, as the matter is sometimes put, disputes about what one should do according to the requirements of right reason. But this view simply does not get out of the starting blocks: there is no reason to think that the practices constituting something as a system of law are reasonably contrived to do that because there is no reason to think that judges are inherently qualified, as a conceptual matter, in virtue of being judges to decide what anyone has most reason to do, all things considered. But, more tellingly, there is no reason whatsoever to think that judges are inherently qualified, as a conceptual matter, in virtue of being judges to decide what even the law requires people to do. The fact that someone is appointed or elected to the judiciary does not entail that they are even competent at figuring out what law requires—​much less what right reason requires. Just as the voting public can elect a narcissist, ignoramus, or crook to the office of the presidency, someone without a law degree can, as far as federal law in the U.S. is concerned, be appointed to the Supreme Court by the President and confirmed by the Senate.4 There is nothing in our conceptual practices that precludes the possibility of utterly unqualified legislators, judges, and executives—​a fact that is becoming increasingly evident in these frighteningly troubled times. The idea that the law should even authorize a judicial tribunal to resolve disputes about what right reason requires—​which would require it to resolve disputes about what morality requires because what right reason requires is, as a conceptual matter, largely determined by what morality requires—​is problematic, at least in worlds like ours, both from the standpoint of our collective prudential interests and from the standpoint of political morality. It is problematic from the standpoint of our collective prudential interests because there is always a risk that we appoint or elect someone utterly unqualified to make responsible decisions; and it does not conduce, absent exceptional circumstances, to our collective prudential interests to have utterly unqualified people deciding what the rest of us should do according to right reason and ordering the enforcement of those decisions. It is problematic from the standpoint of political morality because even if we could appoint someone who is infallible with respect to discerning the requirements of right reason, people are entitled, as a matter of political morality, to fair notice of what the law requires of them; and the requirements of right reason are especially contentious in the kind of hard cases giving rise to angry disputes that can get out of hand—​precisely because the requirements of right reason in those cases are passionately contested. It is indisputably true that to ensure that the peace is kept we need some authoritative mechanism for resolving disputes among persons, but there is nothing in the practices constituting something as a legal system that suggests that the persons we select as judges are necessarily better qualified to resolve disputes about what law, right reason, morality, or prudence requires. But even if there were a legal system in which we had an epistemically justified true belief that judges were omniscient with respect to the requirements of right

4 The U.S. Constitution says nothing that entails that one must have even a law degree to serve as a federal judge; and, as far as I can tell, there are no federal statutes creating such a requirement. Kashmir and David Lat, “You Don’t Need No Stinkin’ Law Degree to Be on the Supreme Court” Above the Law (May 14, 2010); available at https://​abovethelaw.com/​2010/​05/​you-​dont-​need-​no-​stinkin-​law-​degree-​ to-​be-​on-​the-​supreme-​court/​. Either way, it should be clear that it is conceptually possible for there to be a legal system in which persons can serve as a judge without having a legal education of any kind.

258 Appendix reason,5 it might be morally legitimate to authorize them to state an opinion as to what right reason requires; however, it would not be morally legitimate to enforce their judgments with the coercive machinery of the law. Even God gives people fair notice of what they must do, depending on one’s view, in the form of the Ten Commandments, Sharia Law, or the Two Great Commands so that they cannot complain when God punishes them with profound detriment—​whether on earth or in hell—​for not doing what they know they should do.6 Either way, one cannot ground plausible counterexamples to a theory that attempts to explain our conceptual practices in a society of angels who are not enough like us to require the kind of behavioral regulation rationally competent self-​interested subjects like us need in worlds of acute material scarcity like ours. It is true that claims describing the metaphysical nature of law, no matter what methodology one chooses, are true in all conceptually possible worlds; however, what is conceptually possible given our practices is constrained to what is possible for rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours. There is nothing that can be learned about our social practices, conceptual or otherwise, as they pertain to regulating the behavior of beings like us in worlds like ours by speculating about a society of beings who are either nothing like us or inhabit worlds that are nothing like ours. Law might not be distinctively normative in any rigorous sense, but the institution of law, as we conceive and practice it, is distinctively human. This distinctively human institution is needed, as a conceptual matter, to keep the peace among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours because the peace is always fragile in worlds where there is not enough for all of us to satisfy our needs and desires. It should be clear that the practices constituting something as a system of law are not reasonably contrived to do that unless they back, whether directly or indirectly, mandatory norms prohibiting assaults on persons and property with the threat of non-​trivial detriment characteristically regarded, as a descriptive matter of contingent fact, by rationally competent subjects as having deterrent force because they should, as an objective matter of normative practical rationality, regard it as such. Even so, there remain difficult issues with respect to the nooks and crannies. It is not obvious, for instance, whether there could be something that counts as a system of municipal law, on our conceptual practices, that prohibits the right kind of behaviors but backs those prohibitions only with compensatory damages of the sort backing tort law in our world. Tort law in our world characteristically restricts many of the same behaviors restricted by criminal law, affording civil liability for behaviors also subject to criminal liability. If you commit an assault against my person or property prohibited by criminal law, I can sue you in tort for compensatory damages as a means of restoring me to the position I would have been in

5 This would not logically imply that we are omniscient with respect to what right reason requires; if something we knew to be an omniscient, morally perfect God thundered from the sky that some set of judges are omniscient with respect to what right reason requires and perfectly honest about telling us what right reason requires, we would have an epistemically justified true belief that they are omniscient without being able to infallibly discern for ourselves the requirements of right reason. While we would have an infallible decision procedure for deciding such questions (ask the perfectly honest omniscient judges), that does not make us omniscient. The notion of infallibility requires only that someone always get the answers right while the notion of omniscience requires that they get the answers right only in certain ways. Infallibility is thus conceptually necessary but not conceptually sufficient for omniscience. 6 The Ten Commandments are revealed in the Torah and Old Testament in Exodus 34:27–​28. Romans 1:18–​20 of the New Testament states: “The wrath of God is being revealed from heaven against all the godlessness and wickedness of people, who suppress the truth by their wickedness, because God has made it plain to them. For since the creation of the world God’s invisible qualities—​his eternal power and divine nature—​have been clearly seen, being understood from what has been made, so that people are without excuse.”

Appendix  259 but for your having committed the assault; your being punished is reasonably contrived to do nothing for me beyond satisfying my overwhelming lust to see you suffer. Legal systems like ours characteristically provide civil and criminal law protections against acts breaching the peace. Things get a little tricky here because the critical issue is sociological and not metaphysical: whether such a system counts as a legal system depends on the sociological issue of whether it is reasonably contrived to keep the peace among beings with the psychological characteristics of beings like us in worlds like ours where we must compete for the things we want or need with the kind of passionate urgency that endows them with a nearly hypnotic seductive appeal. There are no free rides in this world unless one happens to be born into considerable wealth—​and, even then, one’s wealth is just an additional bargaining chip that makes it easier to win what is nonetheless properly characterized as life-​long competition to get what one wants. The rich are not insulated in virtue of their wealth from having to compete; no one is exempt from competing in this world. Their wealth simply makes it much easier to win the competition for anything they might want. There is only so much I can say, as a philosophically trained lawyer, to resolve issues best left to people trained in sociology and psychology; however, I think the following hypothesis is plausible—​and this is properly characterized as a hypothesis precisely because it is empirically predictive: an institutional normative system like the one described above can succeed, as a practical matter, in keeping the peace only if it backs court orders requiring payment of compensation for injuries to persons and for damages to property caused by prohibited assaults with the threat of incarceration.7 We are simply not the kind of beings whose anti-​social impulses—​and every last one of us harbor dangerously anti-​social impulses we can learn to control only with the emphatic help of others—​can be adequately constrained by a legal system that, in essence, does no more than encourage us to pay damages to remedy our wrongs. But the claim that something can do A only if it does B does not logically imply the claim that something can do A if it does B—​and it is the latter claim that defines the relevant issue for my purposes. It might be true that an institutional normative system like the one described above could not, as a practical matter, keep the peace among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours without backing court orders for payment of compensatory damages with the threat of incarceration. But it does not follow that such an institutional normative system could, as a practical matter, keep the peace among rationally competent subjects like us in worlds like ours. The best I  can do here is to hazard an informed guess:  an institutional normative system that backs only court orders to compensate victims of violations with the threat of incarceration—​and does not punish violations with incarceration—​is not reasonably contrived to keep the peace among rationally competent subjects like us. Such a system simply puts a price on crime that only wealthy people can afford. I would hypothesize that class resentment would eventually lead many of us to act out in numbers that would threaten the very stability of society. I think it reasonable to surmise that, although there are surely worse things that can happen to people, it would be utterly awful to live in perpetual fear of either (1) being injured by people who can afford to pay for the privilege and walk away as though

7 A legal system backing court orders to pay compensation with the threat of incarceration is different from the way in which legal systems like ours attempt to induce compliance with the threat of coercive sanctions. See Chapter 3. Such a system of municipal law is clearly conceptually possible insofar as systems of municipal law in our world once did something very like that in virtue of allowing indigents to be locked up in debtors’ prisons.

260 Appendix nothing ever happened or (2) accidentally hurting people whose injuries we cannot afford to buy. Legal systems like that of the U.S. do not back court orders to pay compensatory damages with the threat of incarceration—​although they do back court orders to employers to garnish employee wages with the threat of a contempt sanction; however, that is because the civil law mechanisms for protecting persons against assaults against persons and property supplement the criminal law mechanisms. The civil law mechanisms protecting persons against assaults and property are not adopted to keep the peace; they are adopted to ensure that people are compensated for injuries caused by assaults prohibited by criminal law. The idea that an institutional normative system that is as much like the legal system of the U.S. compatible with its not prohibiting assaults on persons or property with criminal punishment is reasonably contrived to keep the peace among rationally competent self-​ interested subjects like us in worlds of acute material scarcity like ours is facially implausible. Given our collective infatuation with mass incarceration, it is not absurd to hypothesize that my intuitions on the issue mirror that of most people, theorists and laypersons alike. I would guess that few people of modest means—​the unconscionably rich, of course, can buy armed protection to defend their privilege against the rest of us—​would be willing to bet their well-​being and that of their loved ones on experimenting with such practices. The vast majority of us must rely on the ability of the law to make good on threat of incarceration to sleep well at nights. It is important not to overreach here: you need not believe that things would turn out so badly to refrain from making the bet; it is enough—​or should be enough—​that you fear that things would turn out that badly. In hypothetical cases like this, where the best we can do is make informed guesses about catastrophically bad worst-​case outcomes that satisfy some threshold test of intuitive plausibility, we are likely, as a descriptive matter of contingent fact, to try to protect ourselves against those outcomes. I have no idea whether the belief that the worst-​case outcome is sufficiently likely to warrant protecting against such outcomes by adopting coercive regulatory practices is justified according to shared standards of epistemic justification; however, it strikes me as prudentially preposterous, as an objective matter of normative practical rationality, to bet your safety on that belief if you cannot buy your own army. What explains why we adopt the practices we do to keep the peace in any event is as much grounded in prudentially rational fear as it is in epistemically justified belief. It might be true, though I doubt it, that such fear is not justified as an objective matter of normative practical rationality; however, it is rational. My intuitions here might be off the mark, but I cannot fathom that anyone would sincerely respond to my reactions to this thought experiment by saying that I am just being silly. We insist on backing mandatory norms prohibiting assaults on persons and property because we are sensibly afraid of what would happen if we do not. The practices constitutive of law are contrived to meet force with force—​the assumption being that the costs to our collective well-​being of punishing violence with violence are well exceeded by the benefits of punishing violence with violence. While that, by itself, does not imply that any legal system is morally legitimate, it does imply that law has something fairly characterized, from the standpoint of political morality, as a legitimizing purpose in the following sense: whatever else a legal system must do to be morally legitimate, it must optimally realize its legitimizing purpose of keeping the peace. Optimally keeping the peace, other things being equal, is a conceptually necessary condition for moral legitimacy but not a conceptually sufficient condition; the former Soviet Union might have had much lower

Appendix  261 levels of violent crime than we do, but those lower crime rates came at a profound cost to fundamental moral rights. The basic function of law thus has an inextricably moral dimension insofar as it is concerned with bringing about a state of affairs that is morally desirable: if it makes sense to assign moral value to states of affairs rather than actions, then a state of affairs in which the peace is kept through our efforts is not only prudentially preferable to one in which the peace is not kept through our efforts; it is also morally preferable. Although we adopt systems of law to keep the peace for prudential reasons and not for moral reasons, law’s basic task, on such an assumption, is to bring about a state of affairs that has positive moral value. But, either way, it should be clear that (1) the practices constituting something as a system of law are not reasonably contrived to do anything we adopt them to do unless they are reasonably contrived to keep the peace among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours and (2) those constitutive practices are not reasonably contrived to keep the peace among rationally competent self-​interested subjects like us in worlds of acute material scarcity like ours unless they back some mandatory legal norms prohibiting assaults against persons and property with the threat of a coercive sanction—​regardless of whether the imaginary institutional normative system that backs these norms with only the threat of having compensatory damages coercively enforced against non-​complying subjects in some way is properly characterized as a legal system. Claims (1) and (2) are the claims that ground the Coercion Thesis as I describe it—​and they are true regardless of how one comes out on that issue. If, on the one hand, it is true that the system above is not properly characterized as a system of law because it is not reasonably contrived to keep the peace among beings like us in worlds like ours, then the Coercion Thesis stands as it is. If, on the other hand, it is true that the system is properly characterized as a system of law because it is reasonably contrived to keep the peace among beings like us in worlds like ours, then the Coercion Thesis as I described it would have to be modified; however, the required modifications would concern only minor matters. The unhappy idea remains the same: as we construct the notion, law is, by nature, coercive because, in essence, we cannot—​and should not—​simply trust other people to control their base instincts: everyone is capable of violence, if pushed, in this world. Conceptual theorizing among post-​ Hartian positivists has all-​ too-​ frequently been grounded in preposterously optimistic assumptions about what rationally competent self-​ interested subjects like us can be in worlds of acute material scarcity like ours. This does not wholly vitiate the substance of conceptual theories of law grounded in these assumptions; the problematic consequences arise only at the margins—​such as in claims asserting that law is conceptually possible in a society of angels. But insofar as that claim implies the negation of the Coercion Thesis, the damage is theoretically significant—​even for theories purporting to explicate the wide usage of law that applies to things like systems of religious law and chess rules: every such system in our world backs various mandatory norms with the threat of detriment reasonably contrived to deter and punish non-​compliance. Perhaps there could be chess rules and systems of something like religious law lacking sanctions in a society of angels, but that has nothing to do with us: we are not and could never be like them as long as we have to compete for food, water, shelter, sex, love, status, etc., because all those desperately desired goods are acutely scarce in our world. Eliminate the conditions of scarcity for all of these goods—​including the immaterial goods of love and respect we value most once we are adequately fed, watered, and sheltered—​and perhaps there is no need for law because there is nothing that we cannot have without fighting for it. But even that might not be enough to eliminate violence among beings like us if we cannot

262 Appendix radically change what seems to be one of the most basic features of our psychologies: the need to be acknowledged as better than other people. But that is the stuff of utopian fantasies and has little to do with the harsh realities we face in this world because our needs and wants for material and immaterial goods grow ever more quickly than the means of satisfying them. Without the coercive mechanisms enabling law, as a practical matter, to induce us to abstain from committing assaults against persons and property enough for us to live together, things would eventually get badly out of control: life would be nasty, brutish, and short for all of us.

Index Note on index: For the benefit of digital users, indexed terms that span two pages (e.g., 52–53) may, on occasion, appear on only one of those pages.   artifacts artifactual norms  186–87 characteristic uses, basic purposes, and conceptual functions of  75–78 basic characteristic use defining conceptual function 77 capable, meaning of term  77–78 capable of doing what an artifact is supposed to do  77 functional normativity  76–77 idiosyncratic use  75 non-idiosyncratic use  75, 76 practically capable, relevance of  78 concept of  74–75 artificial objects as by-products of activities 74–75 object contrived/used by people to do certain things  75 object counts as artificial  74 contrived to perform the conceptual function of an artifact  80–81 dictionary definition  74n1 legal norms as artifacts  81 legal systems as artifacts  81 artifacts manufactured by social practices, as 81 problem artifact is needed to solve defining its conceptual function  78–80 identifying foundational problem artifact is needed to solve  79 identifying relevant problem  79 specification of artifact’s ultimate end  79–80 civil law and practices civil contempt sanctions  vn8, 9–10, 56, 66 coercive sanctions, as  57 importance of  68 coercion thesis, and  66–67 coercive sanctions, central role of  68–69 compensatory damages  58–61 breaches of duty under the civil law, for  9–10, 58

coercive aspect to civil law practice  60–61 damages not a coercive sanction for contract law breaches  59 damages as a coercive sanction for tort law breaches 59–60 intentional and non-intentional torts 59–60 purpose of  58 see also punitive damages court’s substantive holding, significance of  68 instruments, improperly formed see nonenforcement of improperly formed instruments coercion dictionary definition  4n3, 55 nature of coercion  4–7 descriptive usage  4–5 deterring non-compliance by threatening a detriment 5 detriment deterring non-compliance to enable norm’s success  7 forward-looking 7–8 increasing probability that the subject will comply 5 normative force of threat determined by perceived probability of imposition  6–7 normative force of threat determined by undesirability of detriment  6 rational competence, nature of  4–5n4 thick usage  4–5 threat grounded in reasonable apprehension of detriment  5n5 unsuccessful attempts to coerce  5n5 coercion thesis  1, 4–16, 21 ambiguities of concept-term law see under law applying only to narrow usage of law as denoting legal system  2 coercion, nature of see coercion coercive sanctions see under sanctions legal normativity, and see legal normativity legal systems in purely descriptive sense, and 3

264 Index coercion (cont.) morality, and  16–18 core prohibitions of law incorporating mandatory legal norms  16–17 law prohibiting acts which escalate into cycle of violence  17 law prohibiting aggressions which are likely to escalate  16–17 mandatory legal norms prohibiting acts violating moral norms  16 mental illness and moral competency 17–18 moral norms protecting bodily integrity and property 16–17 requirements of morality as an important concern for people  17–18 motivations of subjects, and  18–20 deterrent value of coercive sanctions  19 motivations for obeying the law  19–20 prospect of coercive sanctions as normatively relevant  19, 129, 133–34 subjects obeying the law to avoid detriment 18–19 nature of  1, 2 nature of a legal system that it is coercive  4 system of law backing certain mandatory legal norms  2, 4 need to keep the peace, and see coercion thesis and need to keep the peace prima facie case for  70–72 sanction, nature of see sanctions coercion thesis and content problem of legal normativity 179–202 capacity to give rise to objective intrinsic motivating reasons  199–201 law uniquely well-suited to regulate behaviour 200 law’s conceptual capacity to rationally induce compliance  201 law’s conceptual normativity, distinctive nature of  199 practices applying and enforcing mandatory legal norms  200–1 practices constituting something as a system of law  199–201 coercion thesis and problems of legal normativity 201–2 intrinsic and instrumental motivating reasons 180–82 instrumental motivating reasons  181 intrinsic and instrumental value  181 intrinsic motivating reasons  181–82 practical reasons  181–82 reasons for action expressing values  180–81

law as a source of objective intrinsic motivating reasons to obey  190–98 legal norms only reproducing mandatory moral norms, effect of  198 no plausible source of objective intrinsic motivating reason to obey  190–92 objective norms of practical rationality, effects of  192–97 law as source of subjective intrinsic motivating reasons to obey  188–90 objective motivating reason to obey law 188–90 morality, conventions, and intrinsic motivating reasons to comply  184–88 artifactual norms  186–87 doing the morally right thing  184–85 mandatory moral norms  185–86 morality protecting something intrinsically valuable 185–86 norms of morality and norms of law  187–88 order and value: relationships among the various kinds of reason  182–84 intrinsic and instrumental motivating reasons, distinguishing 182–83 intrinsic motivating reasons  183–84 coercion thesis and how problem of legal normativity 127–53 authorized coercive sanctions solve the how problem of legal normativity  128–39 coercion thesis explains law’s presumed conceptual normativity 138–39 coercive sanction threat gives rise to motivating prudential reasons 128–29 motivating prudential reasons are defeasible 129–33 motivating prudential reasons are objective 133–34 objective motivating prudential reasons as novel, content-independent  134–37 relevant practices reasonably contrived to provide objective motivating prudential reasons to comply  137–38 coercion thesis and the problems of legal normativity 201–2 coercive sanctions and how problem of legal normativity 153 no other plausible explanation for law’s conceptual normativity 139–46 only coercion thesis explaining law’s conceptual normativity 145–46 system capable of instantiating morally legitimate authority 142–44

Index  265 system metaphysically capable of motivationally/epistemically guiding behaviour 144–45 system satisfies Hart’s minimum conditions for a legal system  139–40 system’s norms reproducing minimum content of natural law  140–42 objective motivating prudential reasons, and 146–49 objective motivating reason to comply not being governing norms  149–53 coercion thesis and need to keep the peace 73–97 characteristic uses, basic purposes, conceptual functions of artifacts  75–78 basic characteristic use defining conceptual function 77 capable, meaning of term  77–78 capable of doing what an artifact is supposed to do  77 functional normativity  76–77 idiosyncratic use  75 non-idiosyncratic use  75, 76 practically capable, relevance of  78 concept of an artifact  74–75 artificial objects as by-products of activities 74–75 object contrived/used by people to do certain things  75 object counts as artificial  74 conceptual function of a legal system  81–91 conceptual function of law  82–83 end for which legal systems regulate behaviour 83 foundational problem legal system is contrived to solve  83–91 legal norms as artifacts  81 legal systems as artifacts  81 legal systems as artifacts manufactured by social practices  81 legal systems regulating behavior  82–83 state of nature  84–91 keeping the peace and authorized coercion  91–97, 195 coercive sanctions reducing frequency of assaults  92, 93 essential that law prohibits assaults on persons and property  94–97 incidence of violent assaults  93 institutional normative system lacking coercive sanctions not a system of law 92 no viable alternatives to coercive sanctions 94

non-trivial prudential disincentives, effects of 94 peace without coercive mechanisms of law 91–92 socially destructive behavior, people’s capacity for  96 problem artifact is needed to solve defining its conceptual function 78–80 identifying foundational problem artifact is needed to solve  79 identifying relevant problem  79 specification of artifact’s ultimate end 79–80 reasonably contrived to perform the conceptual function of the artifact 80–81 coercive sanctions  9–10 central role of authorized coercive sanctions  67–68, 69, 152 civil law see civil law and practices compensatory damages see compensatory damages contempt sanctions see contempt sanctions criminal law see criminal law deployed against or imposed on a party  64 deterring and punishing non-compliance  9–10, 19 disapproval counting as threat of coercive sanction 14 enforcement, importance of  69 formal and informal coercive sanctions 14n11 imposition for non-compliance authorized by norms 10–12 international law, and see international law and coercive sanctions nature, meaning of term  54–55 no moral judgment applying in coercive sanction 10 non-compliance, authorized imposition of coercive sanctions for  10–12 factors in legal systems deterring non-compliance 11 legal norms authorizing judges to impose detriment 10–11 totalitarian states deterring all non-compliance 11–12 worst effects of material scarcity, need to address 11–12 prospect of coercive sanctions as normatively relevant  19, 129, 133–34 social consequences of guilty verdict  67–68 ubiquitous nature of authorized coercive sanctions  67, 68, 69

266 Index compensatory damages  58–61 breaches of duty under the civil law, for  9–10, 58 coercive aspect to civil law practice  60–61 compensatory damages for contract law breaches not coercive sanctions  59 compensatory damages for tort law breaches as coercive sanctions  59–60 court’s ability to enforce mandatory legal norms dependent on officials  60–61, 67 intentional and non-intentional torts  59–60 purpose of compensatory damages  58 system of municipal law with only compensatory damages  255–62 conceptual analysis see under methodology and nature of law conceptual inquiry, nature of  33 contempt sanctions authority to incarcerate, fundamental nature of 56–57 civil and criminal contempt sanctions  9–10, 9–10n8, 56, 66 coercive sanctions, as  57 criminal penalties, and  49 criminal punishment and court’s authority to hold party in contempt  56–58 detriment, regarded as  58 incarceration, notion of  56–57n13 orders enforced by threat of incarceration  56–57, 67 criminal law coercion thesis, and  66–67 criminal contempt sanctions  9–10, 9–10n8 coercive sanctions, as  57 criminal penalties, and  49 criminal punishment and court’s authority to hold party in contempt  56–58 incarceration see incarceration nature of a norm of criminal law  66 sanctions see sanctions deterrence backing mandatory norms prohibiting assaults 135 basic point of any institutional normative system, as  135 coercive sanctions deterrent value of  19 deterring and punishing noncompliance  9–10, 19 detriment deterring non-compliance by threatening a detriment 5 deterring non-compliance to enable a norm’s success  7

factors in legal systems deterring non-compliance 11 forward-looking dimension of punishment, as 9 incarceration, deterrent effect of  94 totalitarian states deterring all non-compliance 11–12 empirical inquiry, nature of  33 enforcement of law central role of enforcement  69 coercive sanctions see coercive sanctions court’s ability to enforce dependent on officials  60–61, 67 enforcement mechanism distinguished from coercive sanction  209 incarceration see incarceration international law, enforcing see international law and coercive sanctions legislative acts, challenging  117 loss of reward as means of enforcing the law 224 mechanisms reasonably contrived to provide reasons to comply  138 rule of recognition defining standards for imposing coercive sanctions  117 sanctions see sanctions free will  40–42 metaphysical nature of  41 incarceration deterrent effect of  94 fundamental nature of authority to incarcerate 56–57 mandatory legal norms, backing  50, 94 notion of  56–57n13 orders enforced by threat of  56–57, 67 instruments, improperly formed see nonenforcement of improperly formed instruments international law and coercive sanctions  203–30 acts withholding a benefit as sanctions  224–27 effects of sanctions  226–27 loss of reward as means of enforcing the law 224 normative force of threats to withhold a financial benefit  225–26 threats to withhold benefits as sanctions 226 diplomatic enforcement mechanisms  222–23 enforcement measures, as  222 severance of diplomatic relations as coercive sanction 223

Index  267 economic enforcement mechanisms  214–22 conditions for coercive sanctions  215–16 dictionary definition of sanction  214–15 economic measures authorized as response to non-compliance 217 economic measures to deter and punish misbehavior 218–20 economic measures to succeed in helping keep the peace  220–22 UN Charter authorizing economic ‘enforcement measures’  215 enforcement mechanisms authorized by U.N. Charter 208–23 diplomatic enforcement mechanisms 222–23 economic enforcement mechanisms 214–22 enforcement mechanism distinguished from coercive sanction  209 military enforcement mechanisms  209–14 Security Council  208–9 military enforcement mechanisms  209–14 coercive element to pacific measures 211–12 creation and deployment of forces to end armed hostilities  210 military interventions to ensure peace 209–10 peacekeeping missions, ability to deter  212 peacekeeping missions not sanctions  211, 212–14 normative system defined by U.N. Charter, coercion thesis and  230 objections and replies  224–30 whether acts withholding a benefit are properly sanctions  224–27 whether veto power implies enforcement mechanisms, not sanctions  227–30 ordinary usages of term international law 204–7 dictionary definition of international law 204 international law, concept of  205–7 international law as unified system of legal norms 207 private international law  205 public international law  205–7 public international law governing behaviour of national officials  206–7 supranational law  207 whether international law is really law 204–5 system of international regulation defined by U.N. Charter 207–8

UN Charter defining body of institutional norms 207–8 judges see under officials law ambiguities of concept-term law  1–3, 51–52 descriptive and evaluative content of think usage 3 law as denoting legal systems, wide and narrow usage of  2 legal norms and legal systems  1, 2 positive law  3 purely descriptive usage  3 conceptual capacity to rationally induce compliance 201 conceptual function of law  82, 83 society of angels, and  239–40 conceptual normativity  139–46 distinctive nature of  199 dictionary definitions of  51–53 international law see international law and coercive sanctions metaphysical nature of see metaphysical nature of law/legal systems municipal law see municipal law nature of law, methodology and see methodology and nature of law natural law see natural law presumed normativity  99–100 coercion thesis explaining presumed conceptual normativity 138–39 public institution, as  113, 120 religious law  2, 3, 52 sanctions see sanctions uniquely well-suited to regulate behaviour  200 whether there can be law in a society of angels see law in a society of angels law in a society of angels  231–53 evaluating normative commitments of the angels 242–50 improbably oddness of angels’ normative commitment 242–48 practical rationality of angels’ deference to authority 248–50 law and Razian angels  235–38 law and real angels  232–35 revising existing legal systems  250–52 society of angels and canons of ordinary usage 238–39 society of angels and conceptual function of law 239–40 society of angels and content problem of legal normativity 241–42

268 Index law in a society of angels (cont.) society of angels and how problem of legal normativity 240–41 legal normativity content problem see coercion thesis and content problem of legal normativity how problem of see coercion thesis and how problem of legal normativity order problem of see order problem of legal normativity: law and exclusionary reasons for action problems of see legal normativity, three problems of legal normativity legal normativity, three problems of  99–125 character of the reasons to which the practices constitutive of law are reasonably contrived to give rise  112–24 law must be reasonably contrived to provide the relevant reasons  122–24 relevant motivating reasons must be novel and content-independent  122 relevant practical reasons are defeasible not conclusive 115–16 relevant reasons are motivating and not justifying reasons  116–19 relevant reasons are objective and not subjective reasons  119–22 relevant reasons are practical not epistemic reasons  112–15 concept of normativity  100 law’s presumed normativity  99–100 logical space of reasons  101–12 defeasible and conclusive reasons  102–6 epistemic and practical reasons  101–2 motivating and justifying reasons for action 108–12 subjective and objective reasons  106–8 three problems of legal normativity  124–25 legal norms artifacts, as  81 authorizing removal of officials  15 criminal law, norms of  66 law referring to legal norms and legal systems  1, 2 legal systems, and  1, 2 mandatory legal norms  1 backing with coercive sanction threat  12–13 core prohibitions of law incorporating mandatory legal norms  16–17 court’s ability to enforce dependent on officials  60–61, 67 free will, and  40–41 giving rise to objective motivating exclusionary reasons 165–67

non-official behaviour, governing  2, 3, 4, 12, 13, 18–19, 23, 162 official behaviour, governing  13 practices applying and enforcing  200–1 prohibiting acts that breach the peace, sanctions for  15–16, 18–19, 20–21 prohibiting acts violating moral norms  16 system of law backing certain mandatory legal norms  2, 4 threat of incarceration, backing  50 unjust mandatory legal norms  29 metaphysically capable of motivationally/ epistemically guiding behaviour 144–45 reproducing minimum content of natural law 140–42 legal systems authorizing judges to impose a detriment  10–11, 12 backing mandatory legal norms governing non-official behavior  2, 3, 4, 12, 13, 15, 18–19 backing some mandatory legal norms with coercive sanction threat  12–13, 137–38 capable of instantiating morally legitimate authority 142–44 conceptual function of  81–91 artifacts, as  81–83 artifacts manufactured by social practices, as 81 conceptual function of law  82, 83 end for which legal systems regulate behaviour 83 foundational problem legal system is contrived to solve  83–91 regulating behavior  82–83 state of nature, and  84–91 conceptual normativity  139–46 only coercion thesis explains  145–46 system capable of instantiating morally legitimate authority 142–44 system metaphysically capable of motivationally/epistemically guiding behaviour 144–45 system satisfies Hart’s minimum conditions for a legal system  139–40 system’s norms reproducing minimum content of natural law  140–42 deterring non-compliance, factors in  11 existence conditions for  193 Hart’s minimum conditions for  139–40 legal norms, and  1, 2 metaphysical nature of see metaphysical nature of law/legal systems

Index  269 regulating behavior  1–2, 64, 82–83 keeping the peace  86–91 limiting permitted or required violence  136 sanctions, and see sanctions mandatory legal norms see under legal norms metaphysical nature of law/legal systems  42–43, 49, 51, 72 authorizing judicial imposition of coercive sanctions 53 backing mandatory legal norms governing non-official behaviour  23 norms backed by coercive sanctions  73, 231 capable of instantiating morally legitimate authority 142–44 coercive in some sense  4, 23 determined by our practices, as  42 mandatory norms, metaphysical nature of  147 metaphysically capable of motivationally/ epistemically guiding behaviour 144–45 philosophical assumptions about  72, 73–74, 238, 239 theories of, requirements  139, 149, 166–67, 178, 198 ultimate end to which systems of law are put 82–83 methodology and nature of law  23–47 conceptual analysis as a descriptive enterprise 28–30 describing content of a concept  28–29 not concerned with what content of concept should be  29–30 distinguishing task of conceptual analysis 30–33 identifying the properties that constitute a concept 30–32 metaphysical nature of a thing  32 properties that are conceptually necessary for membership of class  32 logical structure of a modest conceptual analysis 46–47 conceptual analysis, function of  46 steps in modest analysis of a concept  46–47 modest and immodest conceptual analysis 33–38 differences between  35, 36 immodest conceptual analysis, nature of  34–35, 36 methodologies of both partly empirical 35–36 modest conceptual analysis as only epistemically viable approach  37–38

modest conceptual analysis counting as conceptual analysis  38 modest conceptual analysis, nature of  33–34, 36 putatively a priori character of modest and immodest approaches  36–37 modest conceptual analysis and objective truth 43–46 concepts, role of  45 conceptual claims, nature of  44–45 objective claims, nature of  44 nature, dictionary definition of  23–24n2 philosophical and sociological methodology distinguished 23n1 relevant philosophical assumptions are ours 38–43 free will  40–42 linguistic conventions are ours  38–40 philosophical views informing linguistic practices 40 three types of modal claims  23–28 claims are possible only relative to set of favoured claims  26, 27 empirical claims  25 existence conditions  24–25, 27 inherent properties  23–24, 28 logical modal claims  26 metaphysical claims  25 metaphysical modal claims  26 nomological modal claims  27 morality coercion thesis, and  16–18 core prohibitions of law incorporating mandatory legal norms  16–17 law prohibiting acts which escalate into cycle of violence  17 law prohibiting aggressions which are likely to escalate  16–17 mandatory legal norms prohibiting acts violating moral norms  16 mental illness and moral competency 17–18 moral norms protecting bodily integrity and property 16–17 requirements of morality as an important concern for people  17–18 legal norms only reproducing mandatory moral norms, effect of  198 legal systems capable of instantiating morally legitimate authority 142–44 moral properties of  165 morality, conventions, and intrinsic motivating reasons to comply  184–88

270 Index morality (cont.) artifactual norms  186–87 doing the morally right thing  184–85 mandatory moral norms  185–86 morality protecting something intrinsically valuable 185–86 norms of morality and norms of law  187–88 no moral judgment applying in coercive sanction 10 motivating moral reasons  130–31, 160 exclusionary motivating moral reasons  v prohibiting acts violating moral norms  16 trumping considerations of self-interest  160 motivation legal system metaphysically capable of motivationally guiding behaviour 144–45 motivating reasons see under reasons motivations of subjects, coercion thesis and 18–20 deterrent value of coercive sanctions  19 motivations for obeying the law  19–20 prospect of coercive sanctions as normatively relevant  19, 129, 133–34 subjects obeying the law to avoid detriment 18–19 reasons see reasons retributive motives  135 revenge as motivation for imposition of detriment 8 municipal law and legal systems non-official behaviour in  13n10 sanctions see sanctions as paradigmatic feature of municipal law system of municipal law with only compensatory damages  255–62 natural law minimum content of natural law  135–36, 153 nature of  135n5 reproduced by norms governing nonofficial behaviour  140–42 non-enforcement of improperly formed instruments 61–65 non-enforcement not deployed against or imposed on a party  64 norms governing official behaviour of courts 62 norms not governing parties’ behaviour  61–62 sanction of nullity coercive sanction, as  61, 64–65 incentive to form instruments properly, as  61, 63, 64 not a coercive sanction  61–63

non-official behaviour governing norms reproducing minimum content of natural law  140–42 international law, in  13n10 mandatory legal norms governing non-official behaviour  2, 3, 4, 12, 13, 18–19, 23 legal systems backing  2, 3, 4, 12, 13, 15, 18–19 reproducing minimum content of natural law 140–42 meaning of  2n2 municipal law, in  13n10 sanctions need be authorized only with respect to 13–15 disapproval of behavior  14 rule of recognition  14–15 normative inquiry, nature of  33 normativity concept of  100 dictionary definition of normative  100n2 epistemically normative  100 functional normativity of artifacts  76–73 law’s presumed normativity  99–111 legal normativity see legal normativity practically normative  100 types of normativity  100 official behavior mandatory legal norms governing  3, 13 meaning of  2n2 removal of officials for non-compliance with norms of  14–15, 118 officials coercive removal of  14–15, 118 co-operation in sustaining legal system 193–94 court’s ability to enforce mandatory legal norms dependent on  60–61, 67 judges coerced removal sanctions  176–77 decision-making 161–62 legal systems regulating official behaviour 177–78 mandatory legal norms governing behaviour 174–75 need to enact measures to protect against corruption 176 vulnerable to corrupting influence  175–76 refusal to comply with court’s order, punishment for  117 order problem of legal normativity: law and exclusionary reasons for action 155–78

Index  271 basic sources of objective motivating reasons for action  162–65, 190–91 basic considerations motivating rationally competent subjects  162–63 law’s capacity to give rise to objective motivating reasons  164–65 relationships among three classes of objective motivating reasons  163–64 can coercion thesis explain how law creates objective exclusionary motivating reasons? 167–73 exclusionary motivating prudential reasons to comply with mandatory legal norms 173 motivating reasons trumping another  169–70 objective norms of practical rationality 167–68 peremptory motivating prudential reasons 170–71 prudential considerations  168–69 prudential reasons barring moral motivating reasons  171–73 claim that law provides objective exclusionary motivating reasons to comply coerced removal sanctions  176–77 explains nothing that needs to be explained 174–78 judges vulnerable to corrupting influence 175–76 legal systems regulating official behaviour 177–78 mandatory legal norms governing judges’ behavior 174–75 need to enact measures to protect against corruption 176 first-order, second-order, and exclusionary reasons 156–60 exclusionary motivating reasons  158–60 first-order and second-order motivating reasons 157–58 inclusionary motivating reasons  158–59 nature of first-order and second-order considerations 156–57 peremptory motivating reasons  159 subjective and objective second-order motivating reasons  158 only coercion thesis explains how mandatory legal norms give rise to objective aesthetic reasons to comply with mandatory legal norms  165–66, 191 moral properties of a system of law  165, 191 motivating exclusionary reasons  165–67 prudential reasons to comply with mandatory legal norms  166–67, 191

sanctions, exclusionary reasons, and problems of legal normativity 160–62 judge’s decision-making  161–62 legal norms prohibiting acts as mandatory 160–61 mandatory legal norms governing nonofficial behaviour  162 peace, keeping the see coercion thesis and need to keep the peace positive law  3 prudential rationality  29, 147–48, 170, 234, 245, 249 consenting to coercive regulation  90–91 risk-taking  131–32, 225 punishment see sanctions punitive damages  58, 60 coercive sanctions, as  58–59 court’s ability to enforce mandatory legal norms dependent on officials  60–61, 67 detriment, regarded as  58–59 purpose of  58–59 reasons adjudication, requirement for reasons in  117 character of the reasons to which the practices constitutive of law are reasonably contrived to give rise  112–24 law must be reasonably contrived to provide the relevant reasons  122–24 relevant motivating reasons must be novel and content-independent  122 relevant practical reasons are defeasible not conclusive 115–16 relevant reasons are motivating and not justifying 116–19 relevant reasons are objective and not subjective 119–22 relevant reasons are practical not epistemic 112–15 conclusive reasons contingently conclusive and necessarily conclusive 104–5 nature of  104–5 defeasible reasons defeasible motivating prudential reasons  129–33, 138 defeasible practical reasons  103–4, 106, 115–16 defeasible relevant practical reasons 115–16 defeated by countervailing reason  103–4 types of  102–3 epistemic reasons conclusive epistemic reasons  105

272 Index reasons (cont.) defeasible epistemic reasons  102–3, 104 nature of  101 practical reasons, and  102 practices constituting system of law as giving rise to  113 subjective epistemic reasons  106–7 exclusionary reasons  156–62 exclusionary motivating reasons  158–60 justifying reasons conclusive justifying reasons  109 defeasible justifying reasons  109 motivating reasons, and  109–12 nature of  109 logical space of reasons  101–12 motivating reasons conclusive motivating reasons  108–9 defeasible motivating prudential reasons  129–33, 138 exclusionary motivating reasons  158–60 first-order and second-order motivating reasons 157–58 inclusionary motivating reasons  158–59 instrumental motivating reasons  180–83 intrinsic motivating reasons  180–88 justifying reasons, and  109–12 motivating moral reasons  130–31 motivating reason as a practical reason  108 motivating prudential reasons  128–29 motivating prudential reasons, coercive sanction threat giving rise to  128–29 nature of  108 objective exclusionary motivating reasons  165–73, 174–78 objective intrinsic motivating reasons  190–98, 199–201 objective motivating prudential reasons  133–34, 137–38, 146–49 objective motivating prudential reasons, novel and content-independent 134–37 objective motivating reasons, basic sources of 162–65 objective second-order motivating reasons  158 peremptory motivating reasons  159, 170–71 practices constituting system of law as giving rise to  116–19 relevant motivating reasons as novel and content-independent 122 relevant reasons as motivating reasons 116–19 subjective intrinsic motivating reasons 188–90 subjective second-order motivating reasons 158 nature of reasons  101

objective reasons nature of  107 objective exclusionary motivating reasons  165–73, 174–78 objective intrinsic motivating reasons  190–98, 199–201 objective motivating prudential reasons  133–34, 137–38, 146–49 objective motivating prudential reasons, novel and content-independent 134–37 objective motivating reasons, basic sources of 162–65 objective second-order motivating reasons  158 relevant reasons, as  119–22 subjective reasons, and  107–8 practical reasons  181–82 conclusive practical reasons  105–6 defeasible practical reasons  103–4, 106, 115–16 defeasible relevant practical reasons 115–16 epistemic reasons, and  102 importance of qualitatively  104 justifying reasons, as  109 motivating reasons, as  108 nature of  101 practices constituting system of law giving rise to  112–15 relevant practical reasons  112–15 subjective practical reasons  106–7 qualitative importance of reasons  104 relevant reasons to which law gives rise motivating and not justifying  116–19 objective and not subjective  119–22 practical not epistemic  112–15 relevant motivating reasons as novel and content-independent 122 relevant practical reasons as defeasible not conclusive 115–16 subjective reasons  108 good and bad subjective reasons  107 nature of  106–7 objective reasons, and  107–8 subjective intrinsic motivating reasons 188–90 subjective second-order motivating reasons 158 recognition, rule of see rule of recognition religious law  2, 3, 52 retaliation 8–9 dictionary definition  8–9n6 lacking necessary forward-looking dimension 8–9 revenge as motivation for imposition of detriment 8

Index  273 retribution see under sanction rule of recognition  14–15 adjudication, requirements of  117 endowing sovereign with authority to enact legal norms  117–18 enforcement of law, defining coercive sanctions standards for  117 international law, and  13n10 legislative acts, challenging  117 norms authorizing coercive removal of officials  14–15, 118 norms authorizing impeachment  118 social rule, as  14 informal coercive sanctions backing  14n11 sanctions 7–9 centrality of sanctions to law  72 coercive sanctions see coercive sanctions dictionary definition of  55–56, 214–15 international law, and see international law and coercive sanctions municipal law, and see sanctions as paradigmatic feature of municipal law nature of  54–55 punishment deterrence as forward-looking dimension of punishment 9 distinguished from retaliation  8 forward and backward looking, as  7–8 free will, and  40–42 norm-governed 8–9 notion of sanction conceptually linked to 7–8 retribution expressing backward-looking element of  9, 40–41 types of punishment  56–58 retaliation nature of  8–9 punishment, distinguished from  8 retribution definition of  9n7 expressing backward-looking element of punishment 9 retributive motives  135 sanctions as coercive and punitive  7–8 sanctions authorized as judicial response to non-compliance with some mandatory norms 12–13 sanctions authorized for some mandatory norms prohibiting breach of peace 15–16 sanctions need be authorized only with respect to non-official behavior  13–15 disapproval of behavior  14

rule of recognition  14–15 withholding a benefit as sanctions  224–27 effects of sanctions  226–27 loss of reward as means of enforcing the law 224 normative force of threats to withhold a financial benefit  225–26 threats to withhold benefits as sanctions 226 sanctions as paradigmatic feature of municipal law 49–72 meaning of term law as defined by canons of ordinary usage  50–54 dictionary definitions reporting not establishing correct meanings  53–54 lexical meaning of law as reported by dictionaries 51–53 prima facie case for the coercion thesis  70–72 types of coercive sanction in modern municipal law  54–67 coercive, nature of  54–55 compensatory damages as coercive sanctions see compensatory damages contempt sanction  56–58 criminal law, civil law, and coercion thesis 66–67 criminal punishment and court’s authority to hold party in contempt  56–58 improperly formed instruments see nonenforcement of improperly formed instruments sanction, nature of  54–55 ubiquity and centrality of coercive sanctions in existing municipal legal systems 67–69 actionable grievance, concept of  69 central role of authorized coercive sanctions  67–68, 69 enforcement, central role of  69 social consequences of guilty verdict  67–68 ubiquitous nature of authorized coercive sanctions  67, 68, 69 state of nature  84–91 comparison between life with and without law 84 features of  84 legal systems needed to keep the peace  86–91 scarcity of material resources leading to violence 84–85 submission to an absolute sovereign to end violence 85–86 UN Charter see under international law and coercive sanctions