Justice in Robes 9780674269118

Citation preview

JUSTICE IN ROBES

RO NA L D DWO R K I N

JUSTICE I N RO B E S

T H E BELK NA P P R ES S O F H A RVA R D U NI V ER S I TY P R ES S

Cambridge, Massachusetts London, England

Copyright © 2006 by Ronald Dworkin All rights reserved Printed in the United States of America First Harvard University Press paperback edition, 2008. Library of Congress Cataloging-in-Publication Data Dworkin, R. M. Justice in robes / Ronald Dworkin. p. cm. Includes bibliographical references and index. ISBN 978-0-674-02167-9 (cloth : alk. paper) ISBN 978-0-674-02727-5 (pbk.) 1. Judicial process—Philosophy. 2. Law—Interpretation and construction. 3. Law and ethics. 4. Justice. 5. Jurisprudence. I. Title. K290.D89 2006 340′.112—dc22 2005056114 Designed by Gwen Nefsky Frankfeldt

For Irene Brendel

Contents

Introduction: Law and Morals

1

1

Pragmatism and Law

36

2

In Praise of Theory

49

3

Darwin’s New Bulldog

75

4

Moral Pluralism

105

5

Originalism and Fidelity

117

6

Hart’s Postscript and the Point of Political Philosophy

140

7

Thirty Years On

187

8

The Concepts of Law

223

9

Rawls and the Law

241

Notes 263 Sources 297 Index 299

JUSTICE IN ROBES

introduction

Law and Morals

When Oliver Wendell Holmes was an Associate Justice of the Supreme Court he gave the young Learned Hand a lift in his carriage as Holmes made his way to the Court. Hand got out at his destination, waved after the departing carriage, and called out merrily, “Do justice, Justice!” Holmes stopped the cab, made the driver turn around, and rode back to the astonished Hand. “That’s not my job!” he said, leaning out of the window. Then the carriage turned and departed, taking Holmes back to his job of allegedly not doing justice. How should a judge’s moral convictions bear on his judgments about what the law is? Lawyers, sociologists, philosophers of law, politicians, and judges all have answers to that question: these range from “nothing” to “everything.” I have answers of my own, which I have defended in books and articles over the past thirty years. In the essays now brought together in this volume I discuss the theories of several scholars who disagree with me and each other in different ways and at different levels. In this Introduction I offer a capsule summary of my own views and a chart of the different levels and ways in which I might be wrong and the critics right. Unfortunately the English word “law” and parallel words in other languages are used in so many different ways, we have so many distinct concepts that we use those words to deploy, and the interrelationships among these concepts are so problematic and controversial, that different theories about the connections between law and justice are often answers to very different kinds of questions. This semantic misfortune has caused a good

2

Introduction

deal of confusion in legal theory. The essays in this collection are mainly about law in what I shall call the doctrinal sense. They explore the concept of “the law” of some place or entity being to a particular effect: we use that doctrinal concept when we say, for example, that under Rhode Island law a contract signed by someone under the age of twelve is invalid or, more controversially, that American constitutional law permits the President to order foreigners suspected of terrorism to be tortured. We all make claims of this kind, about what the law requires or prohibits or permits or creates, and we share a great many assumptions about the kinds of argument that are relevant in defending such claims and also about the consequences that follow when such claims are true. We understand as part of that shared understanding that what the Rhode Island legislature has enacted and what Rhode Island judges have written in the past are relevant in deciding whether it is true that contracts signed by young children are invalid in Rhode Island law, and we also understand that whether a Rhode Island judge ought to order a twelve-yearold defendant to pay damages for breach of contract very much depends on whether that proposition is true. Propositions of law play an important role in a complex network of such assumptions and beliefs and inherit their sense from that role. It is obviously an issue of capital practical importance whether moral tests—asking whether it would be wise and just policy to disable young children from making contracts, or whether torture is always morally wrong, for example—are among the tests that judges and others should use in deciding when such propositions are true. We may frame that issue somewhat more formally. It is important to decide whether moral criteria are ever, and if so when, among the truth conditions of propositions of law: the conditions that must hold to make such a proposition true. This is particularly important in political communities like our own in which important political decisions are made by judges who are thought to have a responsibility to decide only as required or licensed by true propositions of law. It matters particularly in such communities whether and when judges must look to morality to decide which propositions are true. We must take great care, however, not to confuse this doctrinal concept of law with other closely allied but nevertheless different concepts.1 We

Introduction

3

also have a sociological concept of law: we use “law” to name a particular type of institutional social structure. We might ask, for instance, using that sociological concept, when law first appeared in primitive tribal societies, or whether commerce is possible without law. Different social theorists use somewhat different tests for identifying law in this sociological sense. Max Weber said that law does not exist where there are no specialized institutions of coercive enforcement,2 for example; and Lon Fuller said that there is no law unless certain minimal requirements of procedural justice are met.3 It might be helpful or even essential for various purposes to stipulate a precise definition of what kind of social structure counts as a legal system: to facilitate predictive social science, or to organize a research project, or to illuminate history in some way, perhaps by showing correlations among various social patterns, or to emphasize the moral importance of certain practices or constraints. However, we must not make the mistake of thinking that there is some natural distinction of social kinds that marks off legal structures as having, just in themselves, some essential nature that these distinctions try to capture. As I explain in Chapter 6, our concepts of different forms of social institution, like bureaucracy, meritocracy, marriage, and law, are not concepts of natural kinds whose essential nature is given by physical or biological structure or something comparable. We— experts and non-experts alike—do share a rough sociological concept of law: we would almost all make certain assumptions if astrozoologists reported that a group of intelligent non-human animals they had discovered on a distant planet had a kind of legal system. But we would think it silly to argue about whether they really had a legal system when we discovered that they had no distinct enforcement institutions, or that ex post facto legislation was the norm there rather than a rare exception, or that their officials never claimed morally legitimate authority. We normally think it unnecessary to ask for a more precise definition of “legal system” than our rough working ideas provide. As I said, anthropologists and sociologists or moralists might find a more refined definition useful to achieve research or classificatory efficiency, but there is enough leeway in our rough understanding to allow them to stipulate more refined definitions without outraging ordinary usage. We can say, for example, without conceptual or se-

4

Introduction

mantic error, either that the Nazis did or did not have law so long as we make plain what sociological or moral point we are making in saying what we do.4 I do not mean to deny that the sociological concept has boundaries. Someone who said that a lollipop or the game of Go Fish is an example of a legal structure would not be making an ordinary kind of mistake. He would either be using some different concept or wholly misunderstanding ours. The doctrinal concept of law figures among the boundaries of the sociological concept in this way: nothing is a legal system in the sociological sense unless it makes sense to ask what rights and duties the system recognizes. That is an important interconnection between the two concepts. But that relation is not reciprocal: not every set of norms that deploys rights and duties is a case of law. Go Fish is structured in that way: it certainly makes sense to ask when that game requires a player to draw a card from the stack. We might deny, for various reasons, that the Nazis had a legal system, but we can nevertheless answer the question what rights and duties the putative Nazi law recognized. So the availability of the doctrinal concept underdetermines the sociological concept. The sociological concept of law has a very different philosophical status, then, from the doctrinal concept. Little normally turns on whether and how the indistinct boundaries of the former are resolved; but of course, as I said, a great deal does turn on how precisely we understand the latter. We must also distinguish the doctrinal concept from a different concept used primarily by a few legal philosophers. This is a taxonomic concept of law: it supposes that any political community that has law in the sociological sense also has a collection of discrete rules and other kinds of standards that are legal standards as opposed to moral or customary or some other kinds of standards.5 Legal philosophers use this taxonomic concept when they ask whether certain moral principles are also principles of law. The idea of law as a set of discrete standards, which we might in principle individuate and count, seems to me a scholastic fiction.6 In any case, we do not need that idea in order to address our main question whether and when morality figures in the truth conditions of propositions of law. The principles of arithmetic plainly figure among the truth conditions of some propositions of law—the proposition that Cohen has a legal obligation to pay

Introduction

5

Cosgrove exactly $11,422, including interest, for example—but it would be at least odd to say that mathematical rules are also legal principles. The taxonomic question is usually a red herring: the important question is whether and how morality is relevant to deciding which propositions of law are true, not how we label whatever moral principles we do take to be relevant. Finally, we share what we might call an aspirational concept of law, which we often refer to as the ideal of legality or the rule of law. For us this aspirational concept is a contested concept: we agree that the rule of law is desirable, but we disagree about what, at least precisely, is the best statement of that ideal. Some philosophers hold that the rule of law is a purely formal ideal: that legality is fully secured when officials are required to and do act only as established standards permit. Other philosophers argue for a more substantive conception of the ideal: they think that legality holds only when the standards that officials accept respect certain basic rights of individual citizens. The debate between these two views is the theoretical substrate of the long argument among American constitutional lawyers about whether the “due process” clauses of the Fifth and Fourteenth Amendments to our Constitution impose substantive as well as procedural constraints. Like the doctrinal concept, but unlike the sociological and taxonomic concepts, a great deal turns on what we take to be the correct conception of the aspirational concept. We need not ask, however, whether political morality is relevant to deciding what the best conception is. That just is a question of political morality.

A Brief Catalogue of Possible Intersections Our main question is about the nature of the doctrinal concept of law. We ask whether moral considerations figure among the truth conditions of propositions of law and, if so, how. We must first notice in how many different ways they might figure. Of course we must reject the simplistic idea that no law that is unjust can be valid. The rates of taxation in the United States are now manifestly unjust, but the propositions that describe these rates are nevertheless true. However, the content of law might well be thought to depend on justice in less dramatic ways. For one thing, some

6

Introduction

nations—America is one—have constitutions that might properly be interpreted as setting moral limits to what law can validly be created in that nation. A constitution may provide, for example, that any putative law enacted by the legislature that denies “equal protection of the laws” to any group is unconstitutional and void. In that case, whether it is the law that men but not women are subject to compulsory military service might be thought to depend on whether that distinction is unfair. What the law is might be thought to depend on what it should be in other ways as well. Much of the law in mature nations is set out in statutes, regulations, and other forms of written enactment, and the text of these enactments may be abstract, vague, or ambiguous. The text may provide, for example, that abortions are legally permissible only when “necessary to protect the health of the mother.” In that case, it might be that whether the law permits an abortion to protect the emotional stability of a woman rather than her physical health depends on whether the law ought to distinguish between mental and physical health in such matters. The text of a regulation might seem clear enough but capable of a surprising result if interpreted literally. There is a wonderful old example: a statute in Bologna made it a crime to “let blood in the streets.” Did this statute have the unintended result of outlawing the then common practice of outdoor dentistry? That may be thought to depend on how unjust that would have been. In Anglo-American legal systems (and, in effect, much of the rest of the world as well) the truth or falsity of propositions of law depends on past judicial decisions as well as statutes. If the correct interpretation of the past decisions of suitably senior courts is that people who are injured by the negligent behavior of others can sue them for their damages, then that principle is part of the law. But it is sometimes unclear what the correct interpretation of a series of judicial decisions is: a string of decisions may be consistent with the principle that someone who is negligent is legally responsible for all the damage caused by his act, but also consistent with the more limited principle that he is responsible only for those damages he could reasonably foresee. It might be that the correct interpretation of those decisions depends on whether the law ought to limit damages in that way.

Introduction

7

We must attend to these different ways in which morality might figure in deciding what the law is. I took care just now to say only that the truth of propositions of law might be thought to depend on the truth of moral claims in these ways because, as we shall see in the course of this book, each of these putative interconnections is controversial. An influential group of American constitutional judges and scholars denies, for example, that the abstract clauses of the American constitution are best understood as making the validity of laws depend on moral issues, or that morality has anything to do with how imprecise statutes should be interpreted. I offer this catalogue now only to remind us that we must be sensitive to the manifold different ways in which law is thought by some lawyers to depend on morality.

Sorenson’s Case We must also be sensitive to the different points in a general theory of law at which the question of morality’s role might arise. I shall illustrate these different points by describing an imaginary case and showing the different stages of a general theory that could offer a comprehensive account of what propositions of law are true in that case. I shall compare my own views about how that case should be decided, according to the general theory I have myself tried to construct, with the views of other theorists I discuss in this book. I describe that imaginary case at some length in Chapter 6 of this book: it arises because Mrs. Sorenson has for many years taken medication whose generic name is inventum, but which was manufactured under different proprietary names by a variety of manufacturers. Inventum had serious side effects, which its manufacturers were negligent in not discovering, and Mrs. Sorenson has suffered grave heart damage. But she cannot prove which drug company manufactured the pills she took: she undoubtedly took pills made by one or more of the many companies that manufactured inventum, but also undoubtedly did not take pills manufactured by some of them. She simply cannot remember and cannot now discover which was which. Mrs. Sorenson’s lawyers have sued together all the drug companies that made inventum during the period in which she took that drug; they argue

8

Introduction

that since it cannot be determined how many, if any, of the pills that she took were made by each manufacturer, the law should be understood to make each of them liable to her for a share of her damages in proportion to their market share of inventum sales during the pertinent years. The drug companies’ lawyers reply that, on the contrary, the law holds none of the companies liable for any damages at all unless she can prove that that company is responsible for her injuries. They argue that she is therefore entitled to no recovery at all. Both sides make claims about what the law is, not what it ought to be. Neither proposes that the judges who hear the case ignore the law and rule in their favor on the ground that that is what justice requires. If we ourselves are to have a view about which side’s claims (if either side’s) are right, we must ourselves decide what the law in Mrs. Sorenson’s jurisdiction actually is. What kind of a question is that? We might say it is a legal question, to be answered by looking in law books. But how do we know what conclusions to draw from what we find in the books? Suppose we find many past judicial decisions in which judges stated that no one is liable for injuries he did not cause and no past decisions in which the judge awarded anyone damages based on market share rather than direct causation. The drug companies’ lawyers would argue that it follows from these historical facts that their view of the law is right: the law makes no one liable who has not been shown to have caused the injury. But Mrs. Sorenson’s lawyers would disagree: they would say that since her situation is different from that of any plaintiff in any of these past law suits, we cannot simply assume that the general principles that judges cited in the past, about liability and causation, make up a fully correct statement of the law. They would argue that the law is best stated in a more general principle that requires a showing of causation in most cases in which one party sues another to recover damages for an injury, but not in all such cases, and not, in particular, in Mrs. Sorenson’s case. True, they say, judges have not yet announced or even identified that deeper principle, but that hardly shows that it does not exist. How shall we decide which of these two very different methods of deciding what the law says is the correct method? Lawyers over the course of their careers fall into ways of arguing about law that they take for granted.

Introduction

9

But suppose we wish to be more reflective and to offer a deeper and more general answer than most lawyers have time or inclination to construct. Then we must develop what I call a general theory of law: a general account of the doctrinal concept of law that would be helpful in answering our question.

The Semantic Stage It is not immediately clear, however, what this general theory would be like and how we should go about building it. That depends on what we suppose the function of the doctrinal concept to be for those to whom we address the theory: concepts can be put to very different kinds of uses, and our theory of any of the concepts of law must be sensitive to the role we are supposing it to play. The key question is this: What assumptions and practices must people share to make it sensible to say that they share the doctrinal concept so that they can intelligibly agree and disagree about its application? The following distinctions will be useful in answering that question. Criterial concepts. People share some concepts only when they agree on a definition—rough or precise—that sets out the criteria for the correct application of the associated term or phrase. People share the concept of bachelorhood only when they know that a bachelor is an unmarried male, for example, and they share the concept of an equilateral triangle only when they know that such a triangle has sides of equal length. Equilaterality is a precise concept. Bachelorhood is reasonably precise, though it has some imprecision: is an eighteen-year-old never-married male a bachelor? Other criterial concepts are much less precise. The concept of marriage is a moderately imprecise criterial concept: we call many different forms of legal and social arrangements found in different societies marriages. The sociological and taxonomic concepts of law are also moderately imprecise concepts. Developing a theory of this kind of concept means proposing a more precise definition for some particular purpose. But it would be a mistake to claim that any one more precise definition better captures the essence of the concept than others. It is a mistake to say, for example, as many now do, that the essence of marriage is a

10

Introduction

union between a man and a woman so that “gay marriage” is an oxymoron. Natural kind concepts. People share some concepts whose instances have a natural physical or biological structure—metals and animals, for example—even though they do not agree about the essential nature of the examples or in the criteria they use to identify those examples. Some experts know a great deal about a tiger’s DNA; many other people know that tigers have DNA and that the structure of that DNA makes them what they are; many more people have never heard of DNA but nevertheless suppose that all tigers have the same biological structure that scientists can in principle identify, whatever it might be; and a vastly much greater number have no understanding of biological structure at all but know that a tiger is a special kind of animal: a large, dangerous, striped beast found either in zoos or in jungles. They all nevertheless have the same concept of tigerhood: a member of a primitive society who believes that tigers are manifestations of evil spirits and an evolutionary zoologist who probes their genetic history can agree about how many tigers there are in a room, and their disagreement about how tigers came to exist is genuine, not spurious. Concepts of natural kinds permit a type of analysis that concepts that are only criterial do not: science can claim to have discovered a true essence of the former, in the form of its molecular or biological organization, that would make no sense for the latter. But of course it does not follow that a philosophical theory of some natural kind must take that scientific route. Anyone trying to explain Blake’s poem would ignore the Tyger’s DNA to focus entirely on the animal’s more superficial properties. Interpretive concepts. Some of our concepts function differently still: they function as interpretive concepts that encourage us to reflect on and contest what some practice we have constructed requires. People in the boxing world share the concept of winning a round even though they often disagree about who has won a particular round and about what concrete criteria should be used in deciding that question. Each of them understands that the answers to these questions turn on the best interpretation of the rules, conventions, expectations, and other phenomena of boxing and of

Introduction

11

how all these are best brought to bear in making that decision on a particular occasion. As I explain in Chapter 6 of this book, the central concepts of political and personal morality—the concepts of justice, liberty, equality, democracy, right, wrong, cruelty, and insensitivity—function for us as interpretive concepts as well.7 People share the concept of justice in spite of sharp disagreements both about the criteria for identifying injustice and about which institutions are unjust. Convergent linguistic practice determines the correct application of both criterial and natural kind concepts, though in different ways for these two types of concepts. People who share a criterial concept can of course disagree and be mistaken about whether the criteria for its application hold in a particular case: whether some man was ever married, for instance, and hence whether he is a bachelor or a widower. People who share a natural kind concept can be mistaken more fundamentally: some or all can be mistaken about the essential nature of the concept’s properties, as people were mistaken for centuries who did not know that sounds are waves.8 They can also be mistaken about instances: whether shiny ore in a stream is gold or only fool’s gold or whether a whale is a fish. But identifying such mistakes presupposes an underlying convergent practice that ties the concept to a particular natural kind. If ordinary people generally took both chemist’s gold and iron pyrite to count as gold even though they knew that a jeweler would distinguish them as different materials and discard the latter, we would presumably say not that ordinary people were mistaken but that chemists had developed a technical term of art for the valuable kind of gold. If there were two potable fluids on our planet that behaved in all ways exactly as water does, and people used the word “water” to refer indiscriminately to both though scientists knew that the two substances had different molecular compositions, we would have to study linguistic practice more carefully to see whether in popular use “water” referred to water and was mistakenly applied to the other substance as well or whether it referred to water and the other substance collectively so that no mistake was made. Interpretive concepts also require that people share a practice: they must converge in actually treating the concept as interpretive. But that does not mean converging in the application of the concept. People can

12

Introduction

share such a concept even when they disagree dramatically about its instances. So a useful theory of an interpretive concept—a theory of justice or of winning a round—cannot simply report the criteria people use to identify instances or simply excavate the deep structure of what people mainly agree are instances. A useful theory of an interpretive concept must itself be an interpretation, which is very likely to be controversial, of the practice in which the concept figures. In my view the doctrinal concept of law functions as an interpretive concept, at least in complex political communities. We share that concept as actors in complex political practices that require us to interpret these practices in order to decide how best to continue them, and we use the doctrinal concept of law to state our conclusions. We elaborate the concept by assigning value and purpose to the practice, and we form views about the truth conditions of the particular claims that people make within the practice in the light of the purposes and values that we assign.9 That is the view I defended in my book Law’s Empire10 and in this book, particularly in Chapters 6 and 8. It is hardly a decisive objection that very few people would identify their own practice in that way: we are engaged in philosophical explanation, not vicarious semantic introspection. My hypothesis explains the kinds of agreement and disagreement about law that we actually find and the alternate hypotheses—that the doctrinal concept is criterial or that it is a natural kind concept—do not. Other legal philosophers do treat the concept in one of these two other ways, however. So we must count, as part of any such theory, an initial semantic stage at which that choice is made or (much more commonly) just assumed.

The Jurisprudential Stage At the next stage of legal theory, which we might call the jurisprudential stage, a theorist must construct the kind of theory of law that is appropriate given his answer at the semantic stage to the question what kind of a concept the doctrinal concept is. Since I believe that the doctrinal concept is an interpretive concept, I try at the jurisprudential stage to interpret the practices in which that concept figures in a general way: I offer a general

Introduction

13

account of the mix of values that best justifies the practice and that therefore should guide us in continuing the practice when at the next stage we frame truth conditions for discrete propositions of law. I argue, in Chapter 6, that we must find these values by studying the aspirational concept of law to determine which values supply the best conception of that concept—which other values, that is, best explain the rule of law as a political ideal. At this stage, reflections on the doctrinal and the aspirational concepts come together. So at this stage the project is inevitably one in which morality figures, because any theory about how best to understand an explicitly political value like the aspirational value of law must be an exercise in political morality. I believe that any adequate account of the aspirational concept—of the values of legality and the rule of law—must give a prominent place to the ideal of political integrity, that is, to the principle that a state should try so far as possible to govern through a coherent set of political principles whose benefit it extends to all citizens. Recognizing and striving for that dimension of equality is, I think, essential to the legitimization of state coercive power. But other theorists who at the semantic stage agree with me that the doctrinal concept of law is an interpretive concept and also agree that we must find the general value of legal practice in the aspirational concept of legality might nevertheless defend very different accounts from mine of the values captured in that aspirational concept. They might well think, for example, that the political and social value of legal order lies in the ability of that order to facilitate citizens’ planning and coordinate their activities in the interests of individual and collective efficiency.

The Doctrinal Stage Once we deploy a theory of law’s value at the jurisprudential stage, we move to a third, or doctrinal stage, at which we construct an account of the truth conditions of propositions of law in the light of the values identified at the jurisprudential stage. If the drug company lawyers in Mrs. Sorenson’s case take the view I just described—that the general value of legal practices lies in facilitating personal and collective efficiency—they might then argue that this value is best served through a doctrinal theory

14

Introduction

that makes the truth of particular propositions of law depend exclusively on what designated legal officials have declared in the past, because that practice would help to make the content of legal rules uncontroversial and so promote efficient coordination. In that way they could support their doctrinal claim that morality is not relevant to judging the truth of Mrs. Sorenson’s claim even though they had acknowledged the relevance of morality at the jurisprudential stage. Since my own opinions at the jurisprudential stage emphasize integrity, not efficiency, I argue for a very different theory at the doctrinal stage. In my view, the best way to enforce the integrity-based interpretation of legal practice is by adopting at the doctrinal stage truth conditions that make the question of what the law is on any issue itself an interpretive question. A proposition of law is true, I suggest, if it flows from principles of personal and political morality that provide the best interpretation of the other propositions of law generally treated as true in contemporary legal practice. The question whether the law entitles Mrs. Sorenson to market-share damages from all the drug companies is to be settled, on this view, by asking whether the best justification of negligence law as a whole contains a moral principle that would require that result in her circumstances. That formula does not automatically decide the issue in her case either way. It might be that the best justification of negligence law includes a moral principle to the effect that it is unfair to hold anyone liable in damages for an injury he did not cause. If so, then the law probably favors the drug companies. But it might be that the best justification would reject that general principle in favor of a different set of principles that includes the idea that those who profit from a risky enterprise should share in the risk. In that case, the law would probably favor Mrs. Sorenson. I am cagey about each of those conclusions because, as we shall see, any general interpretation of a substantial body of law must be more complex than the argument so far indicates. I mean only to suggest the kind of reasoning that my view would require at the doctrinal stage and also to emphasize that the questions that view sets are explicitly moral questions. If I am right, morality is implicated in the identification of law not only at the jurisprudential stage of legal theory but at the doctrinal stage as well. One interpretation of a body of legal doctrine like negligence law is

Introduction

15

better than another—it shows legal practice as better fulfilling the ideals of law proposed or assumed at the jurisprudential stage of analysis—if it provides a better moral justification for that body of doctrine. We can therefore distinguish two dimensions on which we might measure the success of a proposed justification. First, a justification must at least roughly fit what it purports to justify: it would not be a competent justification of contemporary legal practice to say that it serves the value of enforcing a god’s will as this is revealed in some specified biblical document. Even if that were a legitimate and important goal for legal practice to adopt, we cannot claim that it is the goal of our legal practice because that claim would not even begin to fit what lawyers and judges actually do. Second, a justification of a practice must do more than roughly fit that practice; it must also describe some sufficiently important value that the practice serves. It would not justify the institutions and practices of law to say that these practices provide a very good living for many lawyers. That consequence, though certainly real, is not important or valuable enough to justify a political practice of such enormous consequence. The distinction between these two dimensions of interpretation is not meant to capture how interpreters actually think. Any lawyer has built up, through education, training, and experience, his own sense of when an interpretation fits well enough to count as an interpretation rather than an invention, and he will deploy that sense unselfconsciously. I intend the distinction not as a phenomenological report but an analytic device to help understand the logic of interpretation and the different ways in which an interpretation may be challenged. In any case it is important not to misunderstand the distinction by supposing, as some commentators have, that the test of fit is only a mechanical test of consistency. On the contrary, the two dimensions of fit and value represent different aspects of a single overall judgment of political morality, and how we apply the two tests and bring them together in a final assessment of interpretive success at the doctrinal stage will reflect our judgment at the prior jurisprudential stage.11 If we take the overall political justification of law to include some requirement of integrity, and if that is our reason for choosing an interpretive test for propositions of law, then we must elaborate the dimension of fit to reflect our more refined sense of what integrity is and of where its

16

Introduction

value lies. Shall we insist, for instance, that the principles we propose to justify the law of negligence in Mrs. Sorenson’s state fit not just the actual decisions that the courts of her state reached in the past but the opinions that the judges who decided those cases wrote to support their decisions? We must answer that and parallel questions by trying to become clearer about why it is important that a political community extend to everyone the same regime of principle. In my own view, the proper explanation of this requirement fixes on how the community actually uses its powers to intervene in citizens’ lives, not on the reasons that different officials have given for such intervention in the past. An overall legal interpretation is complex in yet another important way. It seeks principles that would justify the substantive claims about legal rights, duties, and the rest that a particular legal practice recognizes and enforces, but it must also justify the great army of constitutional and procedural practices in which these substantive claims are embedded. So any overall justification of legal practice must give a prominent place to principles of political morality that assign law-making powers to particular institutions and to other principles that limit those powers in various formal and informal ways. Suppose the legislature of Mrs. Sorenson’s state had adopted a statute providing explicitly that no one may recover damages for an injury caused by a dangerous drug except against people or institutions that had been shown to have caused that injury. Any competent interpretation of the law of that state would then end in the doctrinal judgment that Mrs. Sorenson has no right to damages. It would be wrong to think that morality would then play no role in the judgment. It plays an important role in the interpretation of the constitutional role of the legislature. The reasons we have for supposing that a body constituted as that body is constituted has the power to make law are reasons of political morality, and if lawyers disagree about the precise character of those moral reasons they will inevitably disagree at least on some occasions about what law the legislature has in fact made. That role of morality is particularly evident in nations like the United States (and, increasingly, the other mature democracies) where legislative power is created in constitutions that also limit that power. Suppose the constitution in force in Mrs. Sorenson’s jurisdiction contained the clause I

Introduction

17

mentioned earlier, about “equal protection” of the laws. Then it might be an important constitutional question whether a statute directed only against users of dangerous drugs is a case of unfair discrimination. That is of course a moral question. Even when it seems obvious that a legislative enactment does not offend any constitutional moral standard—that traffic laws do not deny anyone due process, for example—morality plays a negative role in that judgment, like Sherlock Holmes’s dog who did not bark. It may be obvious that traffic laws are unobjectionable in morality; but that too is a moral judgment. Even when the question is one of legislative interpretation, not legislative power, the political principles that are taken to justify legislation remain powerful because they justify interpretive strategies. Suppose a legislature has adopted the law I described earlier, which permits abortion when the health of the mother is threatened. How should we decide whether the law permits abortion to protect mental health? We might, as I said, ask whether it would be morally arbitrary to draw a distinction between mental and physical health. Or we might ask whether those who drafted the law intended that distinction: we might find, for example, that the legislators plainly thought they were protecting physical health only. It will be decisive which of these two ways of interpreting the statute—or which of many other available ways—we choose. But we must defend our choice as the best justification of the complex practice of legislation, and that will require us to defend it in a particular conception of democratic or other political morality—a particular account of the point and value of representative majoritarian government. So an interpreter deciding Mrs. Sorenson’s case must find principles not only to justify the uncontroversial substantive law of her state—that people whose injuries can be shown to have been caused by a particular negligent drug manufacturer have a right to damages from that manufacturer—but also to justify the settled arrangements of power and authority in that state and in the nation, and the latter principles may limit the force of the former in deciding what concrete legal rights Mrs. Sorenson has. I said earlier that the drug company lawyers might reject integrity as an important legal value at the jurisprudential stage in favor of a more efficiency-based conception of legality and therefore adopt, at the doctrinal

18

Introduction

stage, tests for propositions of law that exclude political morality at that stage. But they might also serve their client by accepting integrity at the jurisprudential stage and also accepting something like my view of interpretive analysis at the doctrinal level: they could argue that the best interpretation of the full range of pertinent legal practice, including its constitutional and procedural dimension, assigns considerable weight to the principle that the law should respect settled expectations that past legal practice has encouraged, like the settled expectations of the drug companies and their insurers reflected in their various budgetary provisions that manufacturers are liable only for injuries they can be shown to have caused. Mrs. Sorenson’s lawyers would resist giving that principle so much weight; they would insist that it provides a better justification of contemporary practice to suppose that commercial expectations should anticipate, not thwart, improvements in substantive justice. If the argument between the two groups of lawyers unfolds at the doctrinal stage in this way we cannot say that one side appeals more than the other to political morality as among the sources of law. We must rather say that since their moral claims differ, their legal judgments differ as well.

The Adjudicative Stage Most people believe that certain kinds of political decisions—in particular those made by judges deploying the state’s monopoly of coercive power— should be taken only as required or permitted by true propositions of law. For most people this is a nearly absolute constraint. But it is not absolutely absolute: they accept that in very rare cases judges may have a moral obligation to ignore the law when it is very unjust or perhaps when it is very unwise, and to use their political power to prevent injustice or great inefficiency. We must therefore recognize a fourth stage of legal analysis: the adjudicative level, at which the question arises what political officials who are generally expected to enforce the law should actually do in particular cases. Of course that is a political and therefore a moral question. It is not a question about how morality figures in identifying law but a question about when, if ever, morality requires judges to act independently of or even contrary to law.

Introduction

19

Most people think that the adjudicative question is almost always settled by the doctrinal question: that only in the rarest cases will judges in a reasonably decent democracy be required to do any more than identify and apply true propositions of law. But this apparently very reasonable view in fact depends on assumptions—conscious or unconscious—that belong to what I have described as earlier stages of legal theory. I suggested, in discussing the semantic stage, that the doctrinal concept of law is an interpretive concept. It functions for us as embedded in a practice that supposes that propositions of law have implications for the exercise of power. It is an interpretive rather than some other kind of concept because our sense of its correct application is sensitive to our judgment about how such power ought to be exercised. If we accept that account of the concept at the semantic stage, then we cannot sensibly contradict it at the adjudicative level. We may grant that on some exceptional occasions judges should disown or ignore the law in what they do, but we must suppose a standing expectation that they will not do this, that they will decide in accordance with what they take true propositions of law to require or permit. For us, a theory of doctrine is an indispensable part of, and nearly exhausts, a theory of adjudication not just because we believe in a moral obligation to obey the law but because, given our sense of how the doctrinal concept of law functions, we must suppose this in order even to construct a theory of legal doctrine. But if we had taken a different decision at the semantic stage, we might well have a different story to tell when we reach the adjudicative stage. Suppose that we had decided, back at that early stage, that the doctrinal concept of law is more like the concept of bachelorhood or more like the concept of water than it is like the interpretive concepts I described. We might then have decided that we explicate that concept simply by identifying the tests lawyers all use in deciding what the law is on some matter, if we think that the doctrinal concept is like bachelorhood, or by laying bare the true essence or nature of what the lawyers mainly agree is law, if we think that that concept is like the concept of water. We would then have analyzed “It is the law that . . .” without assuming in our analysis, even as a general proposition, anything about how the state of the law ought to figure in a judge’s determination of the case before him. We would have left

20

Introduction

that adjudicative question entirely untouched. True, we might then say that as a matter of political morality judges should normally enforce the law. But we might also say, with equal consistency, that they should never enforce the law unless they independently find the law just or wise or efficient. In that case, the adjudicative stage would be an entirely independent, self-contained stage of analysis. We might even say that a theory of adjudication is not part of a theory of law at all. Indeed, legal philosophers who take that view of the philosophical character of the doctrinal concept of law can say, as some of my critics do say, that my integrity-based doctrinal theory is only a theory of adjudication, not a theory of law. That claim is made possible by the decision they have taken at the first, semantic, stage of legal theory. What a legal theory provides at the adjudicative stage depends not only on decisions taken at the basic semantic stage, but on decisions taken at later stages as well. Suppose, for example, that Mrs. Sorenson’s lawyers accept at the semantic stage that the doctrinal concept of law is interpretive, but insist at the later jurisprudential stage that law should be understood to serve values of efficiency and coordination, and therefore urge a doctrinal theory that makes only past explicit declarations of legal officials, and nothing else, relevant to determining the truth of any proposition of law. They might then argue that although it is true that no past official declaration grants their client a legal right to market-share damages, no past declaration denies such a right either. They might conclude that in this case (as in untold other cases) there is what some legal theorists call a “gap” in the law. No proposition of law dictates a result either way, so that even if judges accept that they must follow the law when there is law to follow, they must develop an independent theory of adjudication that tells them how to decide cases when there is no law to follow and, according to Mrs. Sorenson’s lawyers, they should then do justice by forcing the drug companies to pay her damages according to market shares. (Perhaps no lawyer would ever make that argument quite that way in court—he might be disciplined if he did—but judges might understand that when the lawyers say that the law favors their client that is what they really mean.) Morality plays a role in a legal theory within this new structure at two points: at the jurisprudential stage, when value is attributed to legal practice; and at the adjudicative stage, when judges are urged to do justice and

Introduction

21

told that justice requires market-share damages. But the two infusions of morality are distinct. In my own view, to the contrary, the value of integrity that we should attribute to legal practice flows through the doctrinal stage into the adjudicative stage because, I argue, integrity requires judges to look to morality in some cases, including this one, to decide both what the law is and how to honor their responsibilities as judges. Once again, the difference is not between theories that include and those that exclude morality, but between theories that introduce morality at different stages of analysis with different consequences for the final political judgment in which a complete legal theory terminates.

Legal Pragmatism This anatomy of a legal theory, which divides any full theory into semantic, jurisprudential, doctrinal, and adjudicative stages, is of course artificial: legal philosophers do not articulate their theories in this stylized way. But the artificial anatomy provides a useful schema for identifying and distinguishing a variety of types of legal theories. In this book I begin with a theory that is both radical in the history of legal thought and of very great importance in contemporary legal practice. This theory has taken different forms and attracted different names. I shall call it “legal pragmatism.” Pragmatism is most easily and generally described as a theory of adjudication: it holds that judges should always decide the cases before them in a forward-looking, consequentialist style. They should make whatever decision is best for the community’s future with no regard for past practice as such. Any more precise version of pragmatism must specify some particular conception of consequentialism: it must specify how to decide which consequences of a decision would be best. This might be an act-utilitarian conception, which holds that individual political decisions should each aim to maximize the average expected welfare of a specified population according to some specified conception of welfare: happiness, for example, or desire-satisfaction. Or it might be a non-welfare conception that defines the best consequences in terms of economic efficiency or wealth maximization, for example. In any case, a pragmatist judge must nevertheless accept instrumental

22

Introduction

constraints that require him to have an eye to what legislatures have enacted or what judges have decided in the past. These constraints are not exogenous to his chosen conception of best consequences but rather emerge from it. According to pragmatism, judges must on the whole obey the legislature and keep faith with past judicial decisions because the power of legislative and judicial institutions to coordinate future behavior is of great benefit in securing efficiency or any other goal, and that power would be undermined if judges characteristically ignored past declarations in new decisions. But there can be no other, less instrumental, constraints on what judges can do, so that when efficiency or some other community goal is actually better served by ignoring or rewriting past declarations, that is what a pragmatist judge should do. Our anatomy suggests different ways in which a legal theorist might arrive at that position. He might, for example, work his way through the four stages of legal theory I described and might indeed accept roughly the answers I give to the questions posed at the semantic and jurisprudential stages. He might hold at the semantic stage that the doctrinal concept of law is an interpretive concept and at the jurisprudential level that the best account of law’s value includes an ideal of integrity. He might then hope to provide, at the doctrinal stage, the best interpretation of contemporary practice but disagree with me by supposing that the best interpretation would justify setting only forward-looking, consequential truth values for law, so that the proposition that Mrs. Sorenson has a legal right to marketshare damages would be true if, but only if, a judge’s awarding her such damages would do more good than harm on the whole for the community at large. The important intellectual movement called “law and economics,” which took root initially in American law schools but has spread to other nations, has spawned a variety of sects. One of these is plainly jurisprudential and interpretive: scholars in this wing of the movement claim that common-law courts have characteristically made decisions that maximized economic efficiency somehow defined. (I have discussed and criticized different aspects of the movement in several essays and books.)12 If so, then integrity is best protected by developing the law in hard cases like Mrs. Sorenson’s in whatever way would maximize that forward-looking goal.

Introduction

23

But other pragmatists argue in a very different way, starting in a much more radical position at the initial semantic stage of analysis. I have assumed, in summarizing my own views in this Introduction, that propositions of law have truth conditions—that it makes sense to ask whether the proposition that Mrs. Sorenson has a legal right to market-share damages is true and, if so, why—and that a philosophical study of the doctrinal concept of law should be directed at identifying those truth conditions. These assumptions have for many decades been challenged by some academic lawyers who call themselves “realists” about law and who insist that propositions of law are neither true nor false but only expressions of the subjective preferences of judges or other officials, so that the project of seeking truth conditions for propositions of law is a pointless waste of time. They hold that the doctrinal concept of law is not criterial or a natural kind concept or interpretive, but that it is, rather, bogus. A more sophisticated version of these skeptical claims has been defended, in a more organized and philosophical fashion, by other writers who called themselves pragmatists. In Chapter 1, I study the claims and arguments of two such writers: philosopher Richard Rorty and literary scholar Stanley Fish. Though neither is a lawyer, both have written critically about my own legal views. I argue, in that chapter, that the version of pragmatism they defend is philosophically confused and cannot be stated except in metaphors that cannot be unpacked without making their arguments self-refuting. The most influential version of pragmatism in academic law is different still. It is on the surface less philosophical and more practical than the views discussed in Chapter 1, but its central thesis is nevertheless an important and controversial position in political philosophy. It holds that anyone with political power should use that power to try to make things better in whatever way is possible given his institutional position and the degree of his power. On this view, any concern that judges might show for the truth of propositions of law is a wasteful distraction from the goal that they should pursue single-mindedly, which is the improvement of their political community. This form of pragmatism unfolds entirely at the adjudicative stage of legal theory: it has no need of any of the prior stages. So it takes no view about the nature of the doctrinal concept of law, or about how the contemporary practices of law are best justified, or about

24

Introduction

the truth conditions of legal propositions. All the paraphernalia of traditional jurisprudence are swept away in the fresh wind of forward-looking, instrumental, means-end calculation. The most influential legal pragmatist who takes this view has been Richard Posner, who is both a prominent federal judge and a hugely prolific legal scholar and writer. Chapters 2 and 3 discuss his views. (In an appendix to the latter chapter, I discuss Posner’s attempt to deploy his pragmatist theories to justify the Supreme Court’s egregious decision in Bush v. Gore that made George W. Bush president and fatefully changed the world’s history.) I argue throughout my discussion of Posner’s theories that his form of pragmatism comes to nothing, that it is empty, because though he insists that judges should decide cases so as to produce the best consequences he does not specify how judges should decide what the best consequences are. That is an important criticism, because specifying a concrete conception of consequentialism would make any version of pragmatism immediately less attractive. Many American legal scholars who now proudly announce themselves pragmatists would be embarrassed if they had to embrace an actual theory instead of only rhetoric. Utilitarianism, for all its well-known defects, would seem the most likely moral theory for a legal pragmatist to adopt, but Posner has rejected it without indicating what moral theory he would adopt in its place. His only response to the argument that his pragmatism remains empty is that American judges sufficiently agree about the proper goals of their society so that no academic definition or discussion of these goals is necessary. It is enough simply to let the judges get on with doing what they all agree is best. No one who has followed recent political and congressional battles about President Bush’s nominations of judges to the federal courts can credit that claim. In fact actual and potential judges disagree profoundly about the whole range of political issues that matter in law: from the relative importance of economic efficiency, safety, and environmental protection to racial justice and gender equality. In Chapter 2, I bracket Cass Sunstein with Posner as both members of an “anti-theory” school of thought. I now regard that claim as an exaggeration because though both scholars are critical of the role that political and moral philosophy play in my account of law I now believe that the dif-

Introduction

25

ferences between Sunstein’s views and my own are, as he says,13 much less profound than my differences with Posner. Nevertheless the differences between Sunstein’s views and my own are significant. He believes that the requirements I make of judges by supposing that propositions of law are true when they are supported by the best interpretation of legal practice overall are too demanding and for that reason are likely to produce bad decisions. I believe, as I explain in Chapter 2, that the interpretive question of what the law holds on some particular subject is in principle an openended question. Judges can normally reach competent answers by consulting only the legal materials of their own jurisdiction that fall into the doctrinal neighborhood of their immediate problem, guided by what I call the principle of “local priority.”14 But their arguments, limited in that way, are finally sound in my view only if they would be sustained by a much more general interpretation that embraces legal materials as a whole and is grounded in a more basic jurisprudential conception of law. How far a lawyer or judge must himself venture into this more general territory before announcing a conclusion about the state of the law is essentially a practical question: it depends, among other things, on the challenges to his view that have actually been offered by other lawyers or legal officials. It may be that these challenges cannot be answered from the materials in the immediate neighborhood and that some “theoretical ascent,” as I call it, is necessary. Sunstein thinks that judges should resist this theoretical assent even in principle: they should decide cases in a more traditional way, seeking only partially articulate justifications of their decisions in search of “incompletely theorized agreement” with their colleagues. I criticize these claims in Chapter 2. Sunstein has recently published with coauthors an empirical study of adjudication whose results, he generously says, tend to support my views rather than his.15

Moral Pluralism My thesis can be criticized at the jurisprudential stage and therefore at later stages as well in a very different way. My interpretive account of legal practice supposes that the values that justify legal practice, even if varied and complex, form an integrated whole and in that way can both require

26

Introduction

and guide integrity at the doctrinal and adjudicative stages. But there is a powerful philosophical tradition, now resonant in the opinions of many lawyers, that politically important values like liberty and equality are in deep conflict with one another so that compromise among them is necessary. That compromise cannot itself be guided by some more fundamental value because the conflict is among the most fundamental values. In the end some unguided and subjective choice among values is necessary, and that fact challenges my assumption that one interpretation of overall legal practice or even of some local area of law can sensibly be defended as overall best. The most powerful champion of fundamental conflict in political values has been Isaiah Berlin. Berlin’s theory of deep conflict in moral values challenges the ideal of integrity at a very basic level. I explicate and criticize his claims and arguments in Chapter 4.

Political Doctrinal Positivism I provoked unnecessary controversy by my early attempts to characterize the jurisprudential doctrine of legal positivism: nothing of any importance turns on how that name is used or, indeed, on which writers are called positivists.16 It will be helpful, however, as a matter of exposition, to distinguish different jurisprudential positions, each of which might be called a positivistic position. These are different from one another because they are theories of different concepts of law. Doctrinal positivism holds that moral facts cannot figure, at least fundamentally, in the truth conditions of propositions of law. In his book The Concept of Law, H. L. A. Hart defended doctrinal positivism. “According to my theory,” he said, “the existence and content of the law can be identified by reference to the social sources of the law (e.g., legislation, judicial decisions, social customs) without reference to morality except where the law thus identified has itself incorporated moral criteria for the identification of the law.”17 Sociological positivism holds that moral tests do not figure among the proper tests for distinguishing law from other forms of social or political organization. Taxonomic positivism holds that moral principles and legal principles are distinct and therefore that the law does not include any moral principles. I have already suggested reasons for doubting whether

Introduction

27

either sociological or taxonomic positivism is an important philosophical position. If, as I believe, the sociological and taxonomic concepts of law are imprecise criterial concepts, no theory that insists on imposing sharp boundaries on these concepts can have philosophical, as distinct from methodological, importance. My writings about positivism have taken doctrinal positivism as their target, and unqualified references to positivism in this book should be understood as referring to that position. Doctrinal positivism plainly has philosophical importance. Does it have practical importance as well? That depends on the connection we suppose between the doctrinal and adjudicative levels of legal analysis. Both legal pragmatism and the integrity-based view of legal doctrine that I defend have practical implications: they unfold against the background assumption that it matters how judges decide the cases that come before them. They are therefore among the important theories about law that are actually at work, often on the surface, in legal analysis—in judicial opinions and in lecture halls and seminar rooms where public and private law incubates. Some versions of legal positivism, which I shall refer to collectively as political positivism, are important in the same way: they too speak to the question of how judges should decide cases. They assume that what judges should do is decided largely by what propositions of law are true, and they appeal, as I do, to political morality to justify their views about the truth conditions of propositions of law. As I say in Chapter 7, prominent American lawyers, including both Oliver Wendell Holmes and Learned Hand, defended legal positivism because they believed that judges should obey the decisions of popular legislatures and not themselves attempt to criticize or supplement those decisions out of their own perhaps different (and, those lawyers thought, typically more conservative) moral convictions. These political positivists assumed that the best justification of legal practice includes what they took to be a democratic principle: that the people as a whole and not judges, whom the people cannot discharge, should decide what laws should govern them. Some contemporary academic lawyers defend legal positivism in a similar, explicitly practical and political, way: Liam Murphy, for example, believes that if judges appeal to morality in reaching their decisions about what the law is, the general population will be led to think that whatever is

28

Introduction

law is just.18 That seems counterintuitive and implausible, but Murphy’s arguments are nevertheless a good illustration of the character of political positivism. H. L. A. Hart’s version of doctrinal positivism, which has been influential in academic legal philosophy though not in legal practice, is now generally understood to be an example not of political but of analytic positivism, a type of theory I describe below. But there are at least hints of political positivism in his early work. He once also argued, for example, that it would facilitate moral criticism of law firmly to separate legal from moral argument.19 The most formidable contemporary version of political positivism, however, which has very great practical importance indeed, is a theory of how the United States Constitution should be read. As I said earlier, the legal constitutions of many countries limit the powers of legislatures and executive officials in abstract language that is either explicitly moral—no punishment may be “cruel and unusual”—or open to moral interpretation— government may not deny “due” process of law. I have argued that these legal constraints on official power must be read as moral constraints, and that lawyers and judges called upon to test legislative and executive acts against these standards must therefore themselves explore moral issues.20 In order to decide whether a men-only draft is prohibited by the “equal protection” clause of the Constitution’s Fourteenth Amendment, judges must decide whether the distinction is justified in political morality, even though that is not the only matter they must decide. This view is firmly rejected by those lawyers and judges who call themselves “originalists”: they believe that even the abstract and apparently moral clauses of the American Constitution should be interpreted to prohibit only legislation or other acts that those who long ago wrote or enacted the clauses or their long-dead constituents would have expected it to prohibit. They insist that only historical facts about the beliefs or wishes or expectations of historical people provide the truth conditions of propositions of constitutional law: the equal protection clause forbids a men-only draft only if the politicians who enacted the Fourteenth Amendment after the Civil War, in which a men-only draft was used, would have expected their clause to prohibit such a draft. Morality has nothing to do with it. Proponents of this view are mainly not legal philosophers, and it is of-

Introduction

29

ten difficult to discern the full structure of their arguments from what they say. They often present their opinion as flowing from the very definition or nature of law, which might suggest that they reject my view at the semantic stage that law is an interpretive concept. But they often buttress their claim with appeals to the political principle that it is more consistent with democratic theory that the nation be governed by principles that were once enacted by the people and their representatives, even if that happened long ago, rather than by principles that are favored by a few lawyers who enjoy judicial office now. They believe, that is, that contemporary judges should not use moral reasoning to interpret the Constitution’s commands not because by definition law has nothing to do with morality but on the contrary because the best arguments of political morality require that contemporary judges be constrained to follow the moral opinions of some earlier generation until these are replaced not by them but by the people as a whole through a constitutional amendment. So they agree with me that law is an interpretive concept—why else would an appeal to political morality be necessary?—and take the political principles they cite to be indispensable to a justification of American legal practice. (There is sometimes room for skepticism about how sincere the appeal to those principles is. As I argue in Chapter 5, in discussing Justice Antonin Scalia’s jurisprudence, some of the Supreme Court justices who have embraced originalism do not follow its principles consistently in their own judicial behavior. Their votes in controversial cases might be better explained by a very conservative political agenda that does not depend on the success of any overall interpretation of our legal practice.) I have argued against this “originalist” theory of constitutional interpretation for many years; in Chapter 5 I discuss a philosophical confusion that I believe infects the arguments that conservative lawyers and judges have offered on its behalf. I distinguish between two very different ideas: semantic originalism, which insists that the words in a constitutional text must be given the meaning that those who enacted the text intended them to have; and expectation originalism, which argues that these words must be given the force in law that they expected it to have. Semantic originalism seems to me unassailable: it simply applies general philosophical understandings of the idea of meaning to legal texts. If “cruel” meant

30

Introduction

in the eighteenth century what “expensive” means now, we would misunderstand the Eighth Amendment to read it as condemning cruel punishments. (Professor Laurence Tribe, who is a superb constitutional lawyer and advocate, seemed to me to miss the importance of semantic originalism in his own comments on my writings about constitutional interpretation, and I take the occasion of Chapter 5 to explain why I think his criticism mistaken.) Semantic originalism is irresistible, but expectation originalism must be resisted because if the framers of the Constitution meant to lay down abstract moral standards, as I believe that they did, then we are faithless to what they meant to say if we take them to have declared that legislation should be tested against their own opinions of fairness rather than against fairness itself. I believe that Justice Scalia’s arguments about statutory and constitutional interpretation, which I discuss in Chapter 5, have particularly suffered from this confusion.

Analytic Doctrinal Positivism Analytic doctrinal positivism claims that the independence of law from morals does not depend on any political or moral interpretation or justification of legal practice or any political doctrine at the adjudicative stage of legal theory but follows directly from the correct analysis of the very concept or idea or nature of law. Once we properly understand the doctrinal concept of law, analytic positivists claim, we see that it is a necessary truth of some kind that moral facts cannot figure in the truth conditions of such propositions. This is the position that Hart’s best-known book, which he called The Concept of Law, is generally taken to defend, although there are traces even in that book of political rather than analytic positivism.21 I had assumed it to be the position of Hart’s many followers in academic legal philosophy, including the legal philosophers whose work I discuss in Chapter 7; though in Chapter 8, in which I discuss various articles written after Chapter 7 was published, I offer reasons for wondering whether some of Hart’s followers should now be understood as defending only taxonomic positivism. Analytic doctrinal positivists disagree with my views at the most fundamental semantic stage of legal theory because they insist that, contrary to

Introduction

31

my opinion, the doctrinal concept of law is not an interpretive concept whose elucidation requires taking a stand on issues of political morality but a concept of some other kind whose elucidation is entirely a descriptive or conceptual project in which substantive morality has no place. My difficulty in understanding analytic positivism lies mainly in understanding that suggestion. In Law’s Empire I suggested a way of understanding it: I said that Hart assumed, in effect, that the doctrinal concept of law is a criterial concept and that analyzing that concept means bringing to the surface the criteria that lawyers actually use, even if unselfconsciously, in applying it. That methodology of conceptual analysis was familiar in the Oxford philosophical world in which Hart prepared his book. I thought Hart meant to argue, in the spirit of a maxim of his colleague J. L. Austin that Hart himself quoted in his book, that we can learn about the nature of law by reflecting on distinctions inherent in the language with which we describe and in which we conduct our legal practices. I did not mean that Hart took himself to be offering definitions in the ordinary sense. “The philosophers I have in mind,” I said, describe “the sense of propositions of law—what these mean to those who use them—and this explanation takes the form either of definitions of ‘law’ in the older style or accounts of the ‘truth conditions’ of propositions of law—the circumstances in which lawyers accept or reject them—in the more modern style.”22 I criticized the assumption that the doctrinal concept of law is a concept of this sort, that it can be elucidated by calling attention to the different circumstances in which “we” deem the use of the doctrinal concept and allied concepts appropriate. I said that the failure to understand that the doctrinal concept is an interpretive rather than a criterial concept constitutes a fallacy that I called the “semantic sting,” which I discuss in more detail in Chapter 8. In new material published only after his death Hart rejected my description of his earlier methodology: he said I had misunderstood him. I continue to think, as do other legal philosophers,23 that my original diagnosis was correct, and it has received fresh support, I believe, from Nicola Lacey’s excellent recent biography of Hart.24 However, we must take Hart’s later disclaimer as at least a statement of how he had by then come to think of his work, and I therefore consider in Chapter 6 alternate explana-

32

Introduction

tions of his methodology. But without success: I can find no other explanation of how Hart could have thought that a theory of the truth conditions of propositions of law, which he wished to provide, could be politically and morally neutral. The fate of analytic doctrinal positivism hinges entirely on its ability to propose an account of the doctrinal concept of law that shows that concept to be amenable to substantial and morally neutral philosophical analysis. If we cannot make sense of the role of the doctrinal concept in legal practice without taking it to be an interpretive concept, then there can be no helpful analysis of that concept that is wholly innocent of political morality. Chapter 7 examines analytic positivism from a more contemporary perspective. In the 1970s a debate began, prompted by an article in which I suggested that Hart’s version of analytic positivism, which claimed to be an account of the concept of law itself and therefore to apply to all fullfledged legal systems now and in the past, falsified a substantial part of law’s phenomenology and official written record.25 Hart had said that all such legal systems contain a fundamental, if sometimes complex, social “rule of recognition,” which is accepted by substantially all legal actors and which operates for them as a decisive pedigree test of true propositions of law. I argued that this claim neglected the important role of moral principles within legal reasoning. Such principles figure in judges’ explanations of why the law is as they claim it to be, but they are not themselves identified by any broadly accepted master test of pedigree. On the contrary, the identity, character, and weight of the principles that figure in legal argument, I said, is controversial, and any lawyer’s or judge’s opinion depends on his own convictions of personal and political morality. Hart did not himself publish any substantial reply to my arguments in his lifetime. Other analytic positivists did reply, however. One of them, Jules Coleman, has said that my article acted as a catalyst, spurring them to refine and develop analytic positivism to meet my objections. In Chapter 7, I describe their attempts and assess their success by considering recent theories of Coleman and of Hart’s most prominent disciple, Joseph Raz. Coleman replied to my arguments by suggesting, in a doctrine now called “inclusive” positivism, that morality is relevant in legal reasoning only when and because rules of law that do meet the pedigree tests of a

Introduction

33

rule of recognition incorporate morality by reference—only when, for example, a constitutional provision explicitly provides that “unfair” laws are invalid. I argue that his arguments are mistaken in several ways and that his claim ends not in victory for his version of analytic positivism but in surrender of positivism altogether. Raz replied to my arguments in a very different way: by insisting that it is essential to the concept of law that law be capable of acting as an authority and then developing a special account of the idea of authority according to which no standard can act as an authority if moral reasoning is required to identify the content of that standard. I argue that Raz’s special theory of authority is arbitrary and designed only to produce that result and that, on any normal conception of authority, standards can be authoritative even if interpretation that draws on moral conviction is necessary to determine what they require. In Chapter 8 I review the argument between analytical positivists and myself in a different way. I discuss recent essays by Coleman, Raz, Michael Stephen Green, and Scott Shapiro that I believe demonstrate how profoundly the arguments of contemporary analytic positivists depend on ignoring the crucial distinctions among the doctrinal, sociological, and taxonomic concepts of law that I discussed earlier in this Introduction.

Legal Philosophy Courses called “Jurisprudence” or “Legal Philosophy” have for centuries been staples in law schools and legal academies across the world. But the content of these courses and their importance in legal education have changed frequently in response to shifts in the character and apparent practical importance of the debates among scholars who claimed legal philosophy as their subject. When I was a law student courses in jurisprudence were occupied with a traditional canon of conceptual issues about the nature of law that were deemed very different, in subject matter and in the skills needed for their study, from the issues of substantive and procedural law studied in other law school classes. We debated questions about the role of morality in legal reasoning by asking whether legal positivism or natural law offered a better understanding of law’s essential nature.

34

Introduction

That scene has changed in two important ways. First, classes called “Jurisprudence” are no longer devoted only or even mainly to those conceptual issues: they take up a great variety of more political issues about, for examples, the role of economics in law, legal sociology, feminism, and what is revealingly called “critical race theory.” Second, legal philosophy has migrated to the heart of many other academic courses and subjects, thoroughly blurring the distinction between jurisprudence and substantive law. Many more academic lawyers have had philosophical training as undergraduates or even graduate students, and some of the most philosophically sensitive and valuable work in legal theory has been done by academic lawyers who classify themselves not as legal philosophers but as constitutional lawyers or experts in contract or tort or environmental or some other branch of private or public law. True, as I suggest in Chapter 7, some legal philosophers—chiefly analytic positivists—continue to treat their conceptual investigations of law as independent of both legal substance and political philosophy. But they talk mainly to one another and have become marginalized within the academy and the profession. Indeed much of the most important work in legal theory is now done not by lawyers but by political philosophers and economists working both within law schools and within their own academic departments. No theorist has made a greater contribution to legal philosophy in modern times than the political philosopher John Rawls. I discuss his contributions in the last chapter of this book.

A Final Suggestion My discussion has so far not challenged the traditional understanding that “morality” and “law” name departments of thought that are in principle distinct, though perhaps interdependent in various ways. I want now to suggest that this traditional understanding, which encourages us to chart relations between two different intellectual domains, is unsatisfactory. We might do better with a different intellectual topography: we might treat law not as separate from but as a department of morality. We understand political theory that way: as part of morality more generally understood but distinguished, with its own distinct substance, because applicable to

Introduction

35

distinct institutional structures. We might treat legal theory as a special part of political morality distinguished by a further refinement of institutional structures. My suggestion has no independent substantive force: I can say everything I wish about the interconnection between law and morality in the classic vocabulary that assumes that these are sensibly regarded as in the main distinct intellectual domains. But the shift I recommend would organize our subject matter in a more pellucid way. It would encourage us to see jurisprudential questions as moral questions about when, how far, and for what reason authoritative, collective decisions and specialized conventions should have the last word in our lives. We would no longer doubt that justice plays a role in fixing what the law is. We could then concentrate on the more complex and important issue of precisely what that role is.

chapter one

Pragmatism and Law

For more than a decade American legal theory has been too occupied in metatheoretical debates about its own character or possibility. Creditable political aims inspired part (though only part) of this preoccupation. But, in the event, nothing came even of those political aims; those who discussed nihilism and deconstruction with social justice in mind could have done more for that cause by dealing with its problems more directly. We should now set aside, as a waste of important energy and resource, grand debates about whether law is all power or illusion or constraint, or whether texts interpret only other texts, or whether there are right or best or true or soundest answers or only useful or powerful or popular ones. We could then take up instead how the decisions that in any case will be made should be made, and which of the answers that will in any case be thought right or best or true or soundest really are.

The New Pragmatism Some lawyers who call themselves pragmatists mean only that they are practical people, more interested in the actual consequences of particular political and legal decisions than in abstract theory. But “pragmatism” is also the name of one kind of abstract philosophical theory. Professor Rorty, who says he is a philosophical pragmatist, includes within that tradition not just William James, Charles Sanders Peirce, and John Dewey, but also Ludwig Wittgenstein, W. V. O. Quine, and Donald Davidson,

Pragmatism and Law

37

though the latter three philosophers have not so much supported as refuted Rorty’s version of that tradition. Rorty says that we must give up the idea that legal or moral or even scientific inquiry is an attempt to discover what is really so, what the law really is, what texts really mean, which institutions are really just, or what the universe is really like. We should give up the idea that one vocabulary of concepts, one collection of propositions, can be more faithful than another to some independently existing “reality.” Instead, we should accept that the vocabulary we have is just the one we have, the one that seems to suit us, or to be useful to us. We should also accept that when that vocabulary of ideas and propositions no longer seems to be useful—no longer seems to suit us—we can and should change it, to see “how we get on” with a different one. Inquiry, so understood, is experimental. We try out new ideas to see how they work out, to see which ideas or vocabularies prove to be useful or interesting. This sounds exciting, but it is philosophically a dog’s dinner, as many philosophers have by now pointed out. I quote a succinct restatement of the point by Bernard Williams summarizing Hilary Putnam’s devastating critique. “[Rorty’s views] simply tear themselves apart. If, as Rorty is fond of putting it, the correct description of the world (for us) is a matter of what we find it convenient to say, and if, as Rorty admits, we find it convenient to say that science discovers a world that is already there, there is simply no perspective from which Rorty can say, as he also does, that science does not really discover a world that is already there, but (more or less) invents it.”1 The point applies equally to law and morals. Ordinary lawyers practicing their profession think that some judicial opinions really get the law right or straight and that others do not. Ordinary citizens think that the war in the Persian Gulf really was just or unjust. They don’t mean that it is amusing or interesting or helpful or useful to say it was, but that it really was, because expelling an invading army really is a just thing to do or because killing innocent civilians really is always unjust. It would be an understatement to say that this distinction—between what the law really is or what justice really requires and what it would be useful in some way to say or think—is important to us. It is crucial: We could not “get on” at all, let alone well, without it. If we thought the pragmatist was asking us to give

38

Pragmatism and Law

up that distinction, we would reject his advice as pragmatically self-defeating: Taking that advice would make our “vocabulary” not more but much less useful to us. So pragmatism self-destructs wherever it appears: It offers advice it tells us not to take. It must have surprised some readers, therefore, that Rorty says that in law, at least, we have already made the changes his brand of pragmatism demands, that pragmatism and its allies have all but swept the field, that the long battle they fought is now largely won, and that in legal theory, at least, we are all pragmatists now.2 How can that be, because we still talk as if lawyers’ statements of law are statements about what the law is, not what it would be useful to say it is, and because we still suppose that lawyers’ statements can get the law right or wrong? The explanation lies in a diagnosis I have offered before, at some length, but will summarize now.3 Rorty and his followers apparently all distinguish, though without making this clear, between two levels at which people supposedly think and speak. The first is the internal level at which some practical enterprise like law or science or literary activities or moral engagement is carried on. That is the level at which people use the vocabulary that is useful to them: the level at which people rightly say, because that is useful, that science describes how the world really is and that the law is not just what it would be useful to think it is. The second is the external level at which philosophers and other theorists talk about these enterprises rather than participate in them. That is the level at which, according to Rorty and the others, some bad philosophers of science claim that science discovers how the world really is, and bad legal philosophers say that lawyers and judges try to discover, even in hard cases, what the law really is. This is the level Rorty means to occupy: He wants to say, himself now occupying that external level, that these external claims are metaphysical, foundational, and other bad things. Refuting these mistaken external descriptions, he thinks, will not change thought or speech at the internal level—the level of actual science and actual legal practices—except to free it from whatever confusion and obscurity has leaked into the practice from the bad external theories. So Rorty says that the triumph of pragmatism has only cleared the conceptual ground so that actual practice can continue liberated from that kind of confusion. The difficulty with this defense, however, is that the external level that

Pragmatism and Law

39

Rorty hopes to occupy does not exist. There is no external philosophical level at which the statement “science tries to describe the world as it is” can mean something different from what that statement means in the internal world of science, and no external jurisprudential level at which “the law, properly understood, allows affirmative action” can mean something different from what it means in court. Language can only take its sense from the social events, expectations, and forms in which it figures, a fact summarized in the rough but familiar slogan that the key to meaning is use. That is true not only of the ordinary, working part of our language, but of all of it, the philosophical as well as the mundane. Of course, we can use part of our language to discuss the rest. We can say, for example, what I just said: that meaning is connected to use. And, certainly, ordinary words can acquire technical meaning in the special practices of a particular profession: Lawyers use “consideration” in a very special way, for example. But we cannot escape from the whole enterprise of speech to a different and transcendent plane where words can have meanings wholly independent from the meaning any practice, ordinary or technical, has given them.4 So it is not enough for Rorty simply to appeal to a mysterious philosophical or external level. He needs to locate the bad philosophical statements in some context of use; he must show them to have some special technical or other sense, so that when a legal philosopher says that legal propositions are true or false in virtue of what the law really is, he is not merely saying, in a more general way, what an ordinary lawyer says when he says that a particular judicial opinion got the law wrong. Neither Rorty nor other pragmatists have actually tried to do that, however. It is difficult to see how they could succeed if they did try. They would have to paraphrase the philosophical statements in some way to bring out their supposedly special meaning, and in doing that they would have to fall back on other words and ideas that also have a perfectly ordinary and clear use, and they would then have to tell us how those words mean something different from what they do in that ordinary use. Suppose the pragmatists tell us, for instance, that the bad philosophers’ theories have a special meaning because these theories claim that the content of the real, external world is independent of human purposes, or independent of culture and history, or something of the sort. The dif-

40

Pragmatism and Law

ficulty is that these new phrases—about the independence of reality from purpose—also have ordinary meanings, and if we give the philosophers’ claims that ordinary meaning, then what they turn out to be saying is ordinary, too. It is perfectly true, for example, that, using all these words in their ordinary way, the height of Mt. Everest is not relative to human purposes or history or culture, though the metric measures we use to describe its height, and the fact that we take any interest in its height at all, certainly does depend on purposes and cultures. So a pragmatist would then have to supply special meanings for such phrases as “independent of purpose,” special meanings that once again try to explain why when the philosopher says that reality is independent of purpose he says something different from what ordinary people mean when they say it. And anything the pragmatist then said—any new paraphrase or translation he offered—would encounter the same difficulty, and so on and on. Would it help if the pragmatist said that though, for example, it is true that the height of a mountain is independent of our purposes, that is true only given how we go on, and that the bad philosopher denies or doesn’t understand that? No, because once again given how we go on—that is, as a statement drawing its sense and force from the practices we have in fact developed—this claim is false. Given how we go on, the height of the mountain is not determined by how we go on but by masses of earth and stone. I hope no one will think, incidentally, that I am now claiming that pragmatism is not skeptical enough, or that it is, in some paradoxical way, swallowed up in its own skeptical success. Let me repeat: Philosophical claims, including skeptical claims of different sorts, are like any other kind of proposition. They need to be understood before they can be embraced, and they can only be understood against the background of how the concepts they employ are used. So understood, the pragmatist claims we have been discussing are not triumphantly true but only, in a straightforward and pedestrian way, false. Given how we go on, it is not true but false that there is no reality for scientists to discover, for example, or that law is only a matter of power, or that there is no difference between interpretation and invention. These announcements sound fascinating, radical, and liberating. But only until we ask whether they actually mean, in the only language we have, what they seem to say. I said, a moment ago, that Rorty’s new pragmatists, their predecessors

Pragmatism and Law

41

and allies, have made no genuine effort to answer the question I posed: What is the difference in sense between the philosophical or theoretical claims they reject and the parallel ordinary ones they accept? How can that be? How can they believe themselves to have refuted positions they have not described? Never underestimate the power of metaphor and other devices of self-deception. The pragmatists use scare-quotes and italics like confetti: They say that the bad philosophers think not just that things really exist but that they “really” or really exist, as if the quotes or italics change the sense of what is said. Metaphor is their heavy artillery, however. They say that the bad philosophers think that reality or meaning or law is “out there”; or that the world, or texts, or facts “reach out” and “dictate” their own interpretation; or that law is “a brooding omnipresence in the sky.” These metaphors are meant to suggest, as it were, that the bad philosophers are claiming a new, different, metaphysically special kind of reality, reality beyond the ordinary, a new, supernatural, philosophical level of discourse. But it is only the pragmatists who, in fact, ever talk that way. They have invented their enemy or, rather, tried to invent him. For if the pragmatist explained his heated metaphors, he would have to fall back on the mundane language of ordinary life, and then he would not, after all, have distinguished the bad philosophers from the ordinary lawyer or scientist or person of conviction. If saying that law is “out there” means that there is a difference between what the law is and what we would like it to be, for example, then most lawyers think that the law is out there, and the pragmatist has no perspective from which he can sensibly say that it is not.

The Right Answer Farrago My thesis about right answers in hard cases is, as I have said, a very weak and commonsensical legal claim. It is a claim made within legal practice rather than at some supposedly removed, external, philosophical level. I ask whether, in the ordinary sense in which lawyers might say this, it is ever sound or correct or accurate to say, about some hard case, that the law, properly interpreted, is for the plaintiff (or for the defendant). I answer that, yes, some statements of that kind are sound or correct or accurate about some hard cases.5 (In fact, I say that some such statement is

42

Pragmatism and Law

characteristically or generally sound in hard cases. But we can ignore that more ambitious statement in this discussion about the kind of claim I am making.) The most natural way to support that legal claim is therefore to try to show what the right answer is in some particular hard case. I can only do that, of course, by making an ordinary legal argument. I have in fact made many such arguments about very hard cases: I argued, for example, that a correct understanding of the United States Constitution required the Supreme Court to reverse the Missouri Supreme Court in the Cruzan case.6 Four members of the Court agreed with that conclusion. Five disagreed: They thought the best available arguments required the opposite answer— that they were required to affirm the Missouri court. I have now mentioned ten very different lawyers all of whom thought (or at least said) that there was a right answer in the Cruzan case, as a matter of ordinary legal judgment. And, of course, many thousands of other lawyers thought the same thing. Now it’s your turn. Have you yourself found any ordinary legal argument on balance the soundest, in any kind of hard case? Then you, too, have rejected the no-right-answer thesis I take to be the target of my own claim. Legal theorists have an apparently irresistible impulse, however, to insist that the one-right-answer thesis must mean something more than is captured in the ordinary opinion that one side had the better argument in Cruzan. They think I must be saying not just that there are right answers in some ordinary way, as an unselfconscious lawyer might say that, but that there are really right answers, or really real right answers, or right answers out there, or something else up the ladder of verbal inflation. Their mistake is just Rorty’s mistake: Thinking that they can add to or change the sense of the position they want to attack by inserting these redundancies or metaphors in it. There is no perspective from which these inflated and decorated claims can have a sense different from their sense uninflated and undecorated, and that is the sense they have in ordinary legal life. So there is nothing in what I have said for them to deny except what most of them would think it perverse to deny. If the skeptical no-right-answer thesis has any practical importance at all, therefore, it must be treated as itself, not a metaphysical but a legal claim. It claims that, contrary to ordinary lawyers’ opinion, it is a legal

Pragmatism and Law

43

mistake to think there are right answers in hard cases. So understood it stands or falls by legal argument. Philosophy and morality are certainly, and in many ways, pertinent to that legal argument. Legal positivists, for example, have argued that the one-right-answer thesis must be wrong, in law, as a matter of logic or semantics. (I tried to answer their arguments in an early article.)7 Members of the Critical Legal Studies movement point to what they take to be pervasive internal contradictions in legal doctrine that, if they exist, would rule out right answers. (I have tried to show that this suggestion confuses contradiction with competition, however.)8 Moral skeptics, including John Mackie, defend a kind of internal moral skepticism that, if sound, would also defeat the possibility of right answers.9 No doubt other arguments with legal bite can and will be deployed in favor of the internally skeptical view. But these are legal arguments; if successful they call for reform, and if successful they can be made without the crutch of inexplicable metaphor. They are not like the pragmatist’s objection, which cannot be made except by redescribing what I say in metaphorical terms, trying to hijack me to some mythical philosophical level where external skeptics hang out, vultures desperate for prey.

Fish and the Subtlety of Practice Professor Fish has been (as he might put it) on my case for a very long time. He has written no less than three highly critical articles about my work,10 which accuse me, among other vices, of “slipperiness” and “spectacular confusion”; he refused to allow a commissioned reply to one of these to be published; and he ends his enthusiastic review of Judge Posner’s book by gratuitously reporting his “somewhat churlish” criticism of me in casual conversation.11 I have no wish further to provoke so energetic an opponent. But his many articles about interpretation, including those critical of me, illustrate so starkly the features of pragmatism I have been discussing that it would be cowardly of me not to call your attention to them. I said that pragmatists invent their opponents through bizarre metaphorical transformations of ordinary statements, and then defend that move by insisting that these supposed opponents are not talking in the ordinary way but are trying to occupy some special, external level of dis-

44

Pragmatism and Law

course that the pragmatist cannot actually describe but insists is there anyway. Fish’s oeuvre confirms that diagnosis, but he adds a new and important twist: There must be a second, external level to interpretation, he says, because nothing of any interest can possibly be said from within an intellectual practice about it. A priori claims are always unbecoming to a self-announced antitheorist; but this one is a particularly serious mistake, because anyone who is blind to the critically argumentative and reflexive character of intellectual practices will understand almost nothing else about them. That fear is realized in Fish’s account of what he thinks the cardinal enemy of pragmatism—foundationalism—really is. “By foundationalism I mean any attempt to ground inquiry and communication in something more firm and stable than mere belief or unexamined practice.”12 Notice the contrast: Mere unexamined practice—doing what comes naturally— on the one hand and “something more firm and stable” on the other. The contrast self-destructs in the way Rorty’s parallel pronouncements do, because it is part—it is an indispensable part—of mere unexamined practice to think that some inquiry and some communication is indeed grounded in something more substantial than mere belief: facts, for example. Fish obscures the point by immediately producing the familiar list of bad ideas that someone who believes in “something more substantial” supposedly must take on. The usual suspects are all there: “ground . . . invariant across contexts and even cultures”; a “‘brute fact’ world”; a “set of eternal values”; “the free and independent self ”; a method of inquiry that “will produce, all by itself, the correct result” (emphasis in original). But the fact that none of that nonsense is part of our ordinary practice doesn’t mean that the distinction between mere belief and something more substantial isn’t; it rather means that Fish doesn’t grasp, or rather is trying to forget, what the distinction really does come to as a matter of “how we go on.” His first article on my work exploited the now-familiar metaphor strategy. He told his readers that in my view meanings are “just there” or “selfexecuting” or “already in place” or “just given” in the text, that literary works “announce their own affiliation” to form and genre, and that novels have an “uninterpreted core” that guides their own interpretation. He ended, however, by scrupulously reporting the curious fact that I myself had taken care to deny everything these metaphors might be thought to

Pragmatism and Law

45

suggest; indeed that I might be thought to have anticipated everything he himself had said. But he said that my disclaimers, far from showing that his frightful metaphors were out of place, only revealed confusion. Someone who says that there is a difference between interpreting a text and inventing a new one, he said, must be assuming a “just there” or “uninterpreted core” picture of meaning, whatever he later says he is doing or assuming or thinks. In his second article the two-level device became explicit. My slipperiness and spectacular confusion, he said, consisted in switching between two levels of discourse without warning my reader that I was doing so. The first is the internal level of a practice like interpreting or judging, the level at which ordinary scholars and judges just have beliefs and make decisions. The second is the external, more “general and abstract,” level at which we might try to “characterize judicial activity in a decisive and illuminating way,” or make “prescriptive or normative” claims about it. He applied this distinction to my claim that there is a difference between judges’ following precedent and ignoring it, a distinction he had earlier flatly denied. Thus while there is, at the level of practice, a distinction between continuing the legal history and striking out in a new direction, it is a distinction between methods of justifying arguments and not between actions whose difference is perspicuous apart from any argument whatsoever. The difference, in short, is interpretive, and because it is interpretive, it can’t be used to settle anything, for it is itself what is continually being settled. Dworkin is thus in a perfect bind: he can stick with the original . . . form of his distinction [which Fish, remember, took to mean texts fighting back, in spite of my protestations], in which case he fails to distinguish meaningfully (in a way that can be consulted or used) between judicial activity or anything else; or he can invoke it as a distinction within . . . practice, in which case it has no prescriptive or normative force because it is a distinction between contestable modes of self-description or accusation.13

We should look at this remarkable passage in some detail. The opening denial that the distinction between interpreting and inventing is not “perspicuous apart from any argument whatsoever” is the usual red herring, more “just there” stuff. No one ever thought the distinction was perspicu-

46

Pragmatism and Law

ous apart from any argument, whatever that might mean. The related assertions—that the distinction between following and ignoring precedent itself involves an interpretive claim, that accusing some judge of having ignored precedent is a “contestable” accusation, that the distinction doesn’t settle anything, and is always itself being settled—only mean, I suppose, that lawyers often disagree about whether a particular form of argument counts as interpreting or inventing, and that both lawyers’ and legal philosophers’ opinions about these matters are constantly shifting. No one has ever denied that either.14 But so far nothing in the argument is pertinent to the question Fish is meant to be addressing: whether the distinction between interpreting and inventing can be used in an illuminating and critical way within interpretive practice, that is, giving the distinction only the sense Fish now agrees it has within that practice. Can it make sense to say, using the ordinary distinction, that some judge is not interpreting precedent but is striking out on his own? Can that count as a criticism of that judge? Of course it can. If the ordinary distinction can’t be used in that descriptive and critical way, then how can it be used? Of course we characterize judicial practice in an “illuminating” way when we say (if it is true) that judges accept a responsibility to interpret precedent rather than to ignore it. And, of course, it is an important normative claim to say that, whether or not they do accept that responsibility, they should. How can it lessen the force or cogency of these claims that they are themselves, as of course they are, interpretive claims? Or that they are inherently controversial and often unlikely to be “settled” in the sense of commanding a consensus? Why can’t interpreting the practice be part of an interpretive practice?15 Fish’s two-level claim seems a casebook example of Wittgenstein’s diagnosis of philosophical bewitchment: Theorists puzzling themselves out of common sense by some hidden a priori commitment. Fish’s crucial assumption that an interpretive practice cannot be self-conscious and reflexive, an assumption presupposed in each of his countless charges about my moving in a confused way from one level to another, is undefended, counterintuitive, pervasive, and crippling. The force of the assumption is consistently down-market: It makes interpretive practice seem unreflective and automatic.16 It generates serious misunderstanding of both the activities it wrongly separates. It abandons

Pragmatism and Law

47

interpretive theory to the external metalevel of invented enemies, and it leaves actual interpretive practice flat and passive, robbed of the reflective, introspective, argumentative tone that is, in fact, essential to its character. Both consequences are conspicuous in Fish’s third article about my work. He repeated, first, that the battle between us must be understood, whatever I say, as taking place on an external logical plane wholly independent of interpretive practice. I try, he said, to occupy an Archimedean point outside all practice; my “law as integrity” is just a “stand-in for the general claim of philosophy to be a model of reflection that exists on a level superior to, and revelatory of, mere practice.” His announced new argument for this description was that Law’s Empire tries to give lawyers advice that they do not need, because they could not possibly act contrary to that advice anyway. The argument fails even in its own terms.17 But it wouldn’t work, for Fish’s purposes, even if it did work in its own terms. Even if my claims about adjudication were all otiose and unnecessary, it wouldn’t follow that they were in any way Archimedean or external; banality is all too internal and worldly. Fish has to show (as I said earlier any pragmatist must show) that he can assign the statements he finds offensive a sense sufficiently different from the sense they have within ordinary interpretive practice to justify his claim that they are deployed at a different, strange, and disengaged level of discourse. I am not aware that he has even tried. Fish’s second assumption, about the passive, unreflective character of interpretive practices, dominates his complaint that I am not content to report that judges think “within” a practice but insist that they should think “with” one: To think within a practice is to have one’s very perception and sense of possible and appropriate action issue “naturally”—without further reflection— from within one’s position as a deeply situated agent . . . To think with a practice—by self-consciously wielding some extrapolated model of its working—is to be ever calculating just what one’s obligations are, what procedures are “really” legitimate, what evidence is in fact evidence, and so on. It is to be a theoretician.18

But as any lawyer knows, there is no difference, in the case of law, between thinking in and with the practice: These are the same thing. A good

48

Pragmatism and Law

judge will “naturally” and “without further reflection” see that it is part of his job to be self-aware and self-critical, to ask what his “obligations” really are, what “evidence is really evidence,” and so forth. He will naturally see that he must be, in Fish’s terms, a theoretician as well as, and in virtue of, occupying his role as a participant. That doesn’t mean (I had better say) that lawyers or judges construct theories of their enterprise from scratch every time they speak. It rather means what I said in discussing Grey’s views about “overarching” theory: that they recognize the argumentative character of even the views they hold unreflectively and that they understand that even these are, in principle, vulnerable to a theoretical challenge they have a responsibility to meet, if and when it arises, the best they reasonably can. Here as elsewhere, Fish dramatically underestimates the complexity of the internal structure of practices that people can quite naturally fall into; he doesn’t see that, in some jobs, theory itself is second-nature. Some things we do are more argumentative than throwing a forkball: Denny Martinez never filed an opinion. Even in baseball, moreover, theory has more to do with practice than Fish acknowledges. The last player who hit .400, fifty years ago, was the greatest hitter of modern times, and he built a theory before every pitch.19

chapter two

In Praise of Theory

Introduction I am going to address the role of theory in legal reasoning and legal practice. Examples are better than anything, so I will start with some. Suppose that a woman has taken generic pills that turned out to have very damaging side effects. Many different manufacturers manufactured the pills, and she does not have any idea who made the actual pills she bought and took from one year to the next, and therefore no idea whose pills caused her injuries. Can she sue any or all of the drug manufacturers? Or do we insist that no one is liable in tort for damage that he or she or it did not cause? Lawyers have argued both sides. Some, including the California Supreme Court, have said that the drug manufacturers are jointly and severally liable.1 Others insist that none of them is liable, and the woman’s loss, sadly, is an uncompensable loss in law. Suppose (to provide a different example) that people burn American flags by way of political protest, and the question arises whether the government can make that a crime consistently with the First Amendment. Again, as you know, lawyers and others have taken different views. The Supreme Court replied “no,” but many lawyers continue to think it made a mistake of constitutional law. There are thousands of other examples of deep controversies about what the law is. The Supreme Court is about to hear a case on appeal from the Ninth Circuit in which an even more daunting question is raised: whether the Constitution grants some right, at least in principle, to assisted suicide.2 Judges and lawyers and ordinary people answer that question in sharply different ways. Now I can state the main issue. What kind of a statement is a statement

50

In Praise of Theory

that, for example, the drug manufacturers are in law jointly and severally liable? Or that the First Amendment protects flag-burning? Or that the Fourteenth Amendment grants a right to assisted suicide? These are not straightforward historical statements, not just descriptive reports of events that took place in the past. Nor are they just prediction: Someone who says that the Constitution protects assisted suicide may predict (as I do) that the Supreme Court will decide the other way. So what in the world makes a claim, about what the law is on some matter, true or false? Here is what I believe to be another way to put much the same question. What is an appropriate way to reason or argue about the truth of claims of law? Let us distinguish two very general answers to that question. I shall call the first the “theory-embedded” (or just the “embedded”) approach. Legal reasoning means bringing to bear on particular discrete legal problems, such as those I described, a vast network of principles of legal derivation or of political morality. In practice, you cannot think about the correct answer to questions of law unless you have thought through or are ready to think through a vast over-arching theoretical system of complex principles about the nature of tort law, for example, or about the character of free speech in a democracy, or about the best understanding of the right to freedom of conscience and of personal ethical decisions. The second answer—I will call it the practical as opposed to theoretical approach—might be put this way. All that I just said about large, general, over-arching theories is misplaced. A judicial decision is a political occasion, and judges and lawyers and everyone else who thinks about the law should be directing their attention to the immediate practical problem posed by any political occasion. The only question should be: How can we make things better? You do need to know a lot about the consequences of different decisions—and perhaps also some economics in order to gauge these consequences—in order to answer that practical question helpfully. But you do not need volumes of political philosophy. I dare say that as I described these two approaches, you immediately knew which one was yours. The practical approach seems so down-toearth, so sensible, so American. The theory-embedded approach, on the contrary, seems abstract, metaphysical, and wholly out of place when there is real work to be done. As you have by now gathered, I am going to at-

In Praise of Theory

51

tempt to argue exactly the opposite. I am going to argue that the theoryembedded approach (which I described as its enemies do, but will redescribe in a more qualified way in a moment) is not only attractive but inevitable. The practical alternative, I will argue, suffers from one commanding defect: it is wholly impractical. I will begin by trying to describe in somewhat more detail how I understand the theory-embedded view of legal reasoning. I will say something, in the course of that exposition, about Hercules and other titans. Then I will consider two recent attacks on the theory-embedded view so understood. The first is offered by Judge Richard Posner3—you know, the lazy judge who writes a book before breakfast, decides several cases before noon, teaches all afternoon at the Chicago Law School, and performs brain surgery after dinner. The second is by an almost equally prolific colleague of his, Cass Sunstein, who also teaches at the Chicago Law School.4 Together these scholars form a Chicago School of anti-theoretical, nononsense jurisprudence. Both criticize the embedded conception of legal reasoning and endorse the practical one, and both describe my own account of the former as a paradigm of the errors they hope to correct. I will therefore use their work to test my argument that we really do not have a choice between the so-called theoretically over-arching, despicably abstract view they denounce and the practical view they tout.

The Embedded View I asked a question a while ago. What kind of a claim is the claim that drug manufacturers are (or are not) liable jointly and severally for harm some of them did not cause? We do best, I suggest, to regard that as an interpretive claim: it asserts that principles are embedded in our legal practice such that, when you apply those principles to the case at hand, they entitle (or do not entitle) the plaintiff to a decision against the drug manufacturers as a group. The phrase “principles embedded in practice” is, of course, a metaphor, and though metaphors have their appeal, in jurisprudence they have too often been substitutes rather than spurs to thought, and it is best to get rid of them as soon after they appear as possible. My metaphor is meant to suggest that we justify legal claims by showing that principles

52

In Praise of Theory

that support those claims also offer the best justification of more general legal practice in the doctrinal area in which the case arises. Of course, lawyers will disagree about which set of principles provides the best justification for the general shape of any considerable part of the law. Someone might offer, for example, as providing the best justification for the law of unintentional harm, the principle that people are responsible for the harm they cause negligently even though unintentionally, but not for any harm they do not cause. If we take that principle as providing the best justification, then the drug manufacturers win and the plaintiff loses because she cannot prove that any of them caused her any harm. But other lawyers would argue that tort law in this area is better justified by a very different principle—that when misfortunes happen as an almost inevitable consequence of some valuable commercial enterprise, like pharmaceutical research, development, and marketing, then the loss should not fall only on particular unfortunate victims, but should be distributed among the class of those who profit from the enterprise. That principle would presumably argue for the opposite result. Of course other pertinent principles might be formulated, some of them more persuasive and much more complex, but two are enough for our example. We can also construct, for the sake of another example, two rival principles pertinent to the flag-burning case. The first holds that the special protection our practice gives freedom of speech is justified by the instrumental importance of that freedom in the functioning of our democracy. The second holds that free speech practice is better justified by the rather different principle that it is part of equal citizenship—and therefore a principle constitutive of, rather than instrumental toward, democracy—that no one may be denied the expression of a conviction or opinion or preference simply because it is offensive. The former of these principles, I believe, would better support a decision against a right to burn a flag, and the latter would better support the opposite decision. A claim of law—either that the drug victim wins or loses or that flag burning can or cannot constitutionally be prohibited—is tantamount to the claim, then, that one principle or another provides a better justification of some part of legal practice. Better in what way? Better interpretively—better, that is, because it fits the legal practice better, and puts it in a better light.5 In that case, any legal argument is vulnerable to what we

In Praise of Theory

53

might call justificatory ascent. When we raise our eyes a bit from the particular cases that seem most on point immediately, and look at neighboring areas of the law, or maybe even raise our eyes quite a bit and look in general, say, to accident law more generally, or to constitutional law more generally, or to our assumptions about judicial competence or responsibility more generally, we may find a serious threat to our claim that the principle we were about to endorse allows us to see our legal practices in their best light. For we may discover that that principle is inconsistent with, or in some other way sorts badly with, some other principle that we must rely on to justify some other and larger part of the law. For example, we might be prepared to accept that people or institutions can be held responsible for compensation in tort without showing that their acts caused any of the injury that they are asked to help to compensate. But someone might then press on us the possibility that that principle has been rejected elsewhere—that it has been implicitly rejected, for example, in those cases that deny liability on the ground that the defendant’s action was too remote in the causal chain that produced the plaintiff ’s injury. Or we might come to press that possibility on ourselves. Of course we might well be able to turn away the threat by showing how those latter decisions can, after all, be reconciled with the principle that we hold applies to enterprise liability. But we cannot simply ignore the threat, because the character of the interpretive argument we are making—which we must make to sustain a legal claim—makes any such threat relevant. We cannot simply ignore the claim that our purported justification would in fact show our legal practice to be unprincipled, because it appeals to a particular principle in justifying coercion against some citizens and rejects the same principle in denying compensation to other people. If that claim is justified, our proposed decision would be objectionable, not just as a matter of theoretical elegance, but also as a matter of how a community committed to equal citizenship should govern itself.

Hercules and Minerva Of course I do not mean, by calling attention to the constant threat of justificatory ascent, that the threat will always or even often materialize.

54

In Praise of Theory

Most of the time it will not, at least in a serious and time-consuming way, and we can cheerfully proceed on the footing of what we might call very local priority—in effect, looking no further in our interpretive arguments than the statutes or cases directly dealing with the matter at hand.6 But justificatory ascent is always, as it were, on the cards: we cannot rule it out a priori because we never know when a legal claim that seemed pedestrian and even indisputable may suddenly be challenged by a new and potentially revolutionary attack from a higher level. I tried to capture that vulnerability in principle in my picture of the heroic Judge Hercules, who, given his talents, might well proceed in the opposite direction from the one I just described. He might think, not inside-out, from more specific problems to broader and more abstract ones, as other lawyers do, but outside-in, the other way around. Before he sits on his first case, he could build a gigantic, “over-arching” theory good for all seasons. He could decide all outstanding issues of metaphysics, epistemology and ethics, and also of morality, including political morality. He could decide what there is in the universe, and why he is justified in thinking that is what there is; what justice and fairness require; what freedom of speech, best understood, means, and whether and why it is a freedom particularly worth protecting; and when and why it is right to require people whose activity is connected to other people’s loss to compensate them for that loss. He could weave all that and everything else into a marvelously architectonic system. When a new case arises, he would be very well prepared. From outside—beginning, perhaps, in the intergalactic stretches of his wonderful intellectual creation—he could work steadily in toward the problem at hand: finding the best available justifications for law in general, for American legal and constitutional practice as a species of law, for constitutional interpretation, for tort, and then, finally, for the poor woman who took too many pills and the angry man who burned his flag. Ordinary people, lawyers, and judges cannot do much of that. We reason from the inside-out: we begin with discrete problems forced upon us by occupation or responsibility or chance, and the scope of our inquiry is severely limited, not only by the time we have available, but by the arguments we happen actually to encounter or imagine. A judge reasoning from the inside-out will rarely find either the time or the need to under-

In Praise of Theory

55

take long, laborious research or argument. Sometimes, however, he will. Benjamin Cardozo felt that necessary in MacPherson v. Buick Motor Co.,7 and he changed the character of our law. We can all think of other decisions in which judges found themselves drawn upward in a justificatory ascent they may not have anticipated when they began to think about the case at hand. That ascent may be rare. But the absolutely crucial point is that there is no a priori or wholesale test for deciding when it will be required. A lawyer or judge must be well along in thinking about an issue before he knows whether he will be tempted or drawn into a more theoretical argument than he first thought or hoped. There is no inconsistency in these two pictures—of Hercules thinking from outside-in and of the mortal lawyer reasoning from inside-out. I stress the compatibility of the two descriptions because many of the critics of an embedded approach to law make a point of saying that real judges are not Hercules. They do not mean only that judges are not superhuman creatures: they mean that my biographies of Hercules are beside the point. Analogies are always dangerous—almost as dangerous as metaphors—and I hope to keep the one I am about to make on a very short leash. But an analogy to science may help to show how an outside-in view of an intellectual domain can be helpful even to those who think within it from the inside-out. We think—or at least hope—that the body of knowledge we call, compendiously, science is very much of a seamless web. There are still seams, and scientists and philosophers worry about those seams. But we have no trouble with the ambition that our physics must be at least consistent with our chemistry, our cosmology, our microbiology, our metallurgy, and our engineering. We hope, indeed, for something more, which we believe we have partly realized—not only that each of these conventionally distinct bodies of knowledge is consistent with the others, but that they can be hierarchically arranged so that physics, perhaps, is taken to be the most abstract, and the others can be seen as drawn from it as progressively more concrete departments of thought. We might illustrate these theoretical and structural ambitions by imagining, in the style of Hercules, a goddess Minerva who spent the centuries necessary to master the biography of space and time and the fundamental forces of particle theory before she undertook to build a single bridge. Then, when someone asked

56

In Praise of Theory

her whether a particular metal would bear a certain weight, she could deduce the answer from her wonderful and complete theory. We understand that picture because it captures how we think about the body of our science. But of course no scientist could even begin to follow Minerva’s example. An engineer who builds a new kind of bridge works from the insideout. She does not know what problems she will discover until she discovers them, and she cannot tell, at least until then, whether the problems she will inevitably discover will require her to rethink some principle of metallurgy, or whether, if they do, her excursion into metallurgy will require her—or someone else—to rethink particle physics. Minerva’s story (grasping the possibility of that goddess’s life) is one way of appreciating the basic assumptions that in turn explain the very different engineer’s story— that explain why the ladder of theoretical ascent is always there, on the cards, even when no one is tempted to take even the first step up it. That is what I hoped to capture, for law, in the story of Hercules. My claim, to repeat, is that legal reasoning presupposes a vast domain of justification, including very abstract principles of political morality, that we tend to take that structure as much for granted as the engineer takes most of what she knows for granted, but that we might be forced to reexamine some part of the structure from time to time, though we can never be sure, in advance, when and how. The theory-embedded view I have been trying to explain is an account of legal reasoning—of how we properly argue toward claims about what the law is. It is also an account of what truth in such claims consists in. It is not automatically an argument about the responsibilities of judges in ordinary cases or even in constitutional cases. Though that may be obvious, I say it here because so many people have resisted the embedded view on the ground that it licenses judges to engage in, as they often put it, vast “excursions” of theory. But it does not automatically follow from the fact I have been stressing—that the correct identification of any kind of law involves an interpretative exercise and is therefore vulnerable to justificatory ascent—that any particular kind of official should be given responsibility to conduct that exercise on any particular kind of occasion. If the commu-

In Praise of Theory

57

nity says to a judge, “The Constitution is the highest law, and your job is to say what the Constitution means,” then, as I have often tried to argue, that instruction will turn out to require a very considerable “excursion” into political morality. But we do not have to instruct our judges that way. It is perfectly intelligible to insist that our judges should not be charged with final and authoritative interpretation of the Constitution. If you fear too great judicial power, that is what you should say. It is a serious confusion to disguise your dislike of judges having great power, which can be remedied, in theory, by changing their jurisdictional power, as a false theory of legal reasoning. I should make one other remark on the side of caution. I do not for a moment mean to suggest that lawyers or judges or anyone else will agree about whatever large theoretical issues justificatory ascent throws their way. Of course they will not agree. That is why we have dissenting opinions and good classroom arguments. I only mean that law is theory drenched, and that reflective lawyers understand that even though they do not agree on what theory it is drenched in.

The Chicago School I turn, finally, to the critics I have been promising you. But first just a word about the spirit of our age that prompts so many people to complain about theory. Our century’s adolescence was bathed in ideology, and ideology did not serve the century well. At century’s end our intellectuals distrust theory perhaps more than any earlier age has. We hear, wherever we turn, the injunctions and disclaimers of the post-modernists, the prestructuralists, the deconstructionists, the critical legal students, the critical race scholars, and a thousand other battalions of the anti-theory army. Some say that theory is phony, and others that it is oppression, and many that it is both. I shall concentrate, however, not on the, let’s say, more erudite and fantastical members of the anti-theory horde, even within law schools, but on relatively mainstream critics. That is why I have taken the Chicago School, and particularly Judge Posner and Professor Sunstein, as examples. The arguments that they and others of similar opinion make against the use of

58

In Praise of Theory

moral or abstract theory in legal argument can usefully, I think, be shepherded under the following three headings: metaphysical, pragmatic, and professional. metaphysics

First, the metaphysical. The theory-embedded approach, I said, sometimes requires lawyers and judges to ask themselves complex issues of political morality—to try to puzzle out, for example, whether it is ever fair to hold someone liable in damages who did not cause any harm, or to try to identify the different reasons of policy and principle why freedom of speech deserves special protection in a democracy of equal citizenship. But there is just now, among us, a lively and remarkably influential opinion—it is the core of what I called the intellectual spirit of our age—that there are no objectively correct answers to such questions, that there is no objective truth about political morality “out there” in the universe for lawyers or judges or anyone else to discover. On this view, all our convictions on these matters—and more fundamental issues, including, for example, whether genocide is wicked, or whether racial discrimination is unjust, or whether freedom of speech is a fundamental right at all—are simply creatures of (now to use a phrase that Wittgenstein, please forgive him, made popular) “language games.” We, in our society, have adopted, for our own purposes and out of our own needs, a particular way of speaking according to which it is true that genocide is an abomination, racial discrimination is appalling, and freedom of speech is special. Through that game, we have constructed the “moral reality” to which we appeal. Freedom of speech is a basic right in our local language game. It is not objectively or transcendentally a basic right: there is no such right “out there” in the fabric of the universe. To paraphrase Judge Posner’s beau ideal, Oliver Wendell Holmes, if different societies differ sufficiently on matters of great enough importance, one may have to destroy the other, but neither should think that its own opinions are any more valid, from the perspective of the universe, than the opinions it hates. Judge Posner has flirted with this amazing thesis. In his book Overcoming Law, he talks of language games and seems warmly disposed to the view that language creates rather than aims to report our moral universe.8

In Praise of Theory

59

In any case, that is now an extremely popular view across much contemporary academic discourse, with the exception of philosophy. If that popular view is also persuasive, then the theory-embedded approach to legal reasoning is profoundly misguided, and should be abandoned, for two reasons. First, legal reasoning, according to the embedded approach, presupposes that one interpretive claim will, at least ordinarily, be superior to its rivals, not just superior in the opinion of its proponent, but actually superior, and if there is no objective moral truth, no such claim can be actually superior in any genuinely difficult case. Second, the case I made for that approach is itself a moral case—lawyers, I said, must be ready to offer a theoretical justification for their judgments because it is unfair to subject some citizens to a regime of principle the community disavows in other circumstances—and that moral case itself claims objective status. It would not be enough, to sustain the theoretical approach, to say that it aims at, and is justified by, not objective truth but truth according to our community’s language games. Contrary to the apparent assumption of those who believe in language games, if these exist at all in contemporary democracies, they do not unite but divide us. We disagree, if not at the most general level of moral conviction, at almost that level, and it would be absurd to suppose that a single answer to complex questions about compensatory justice or free speech or racial justice can form the way we all talk or think. So if the argument that there is no objective truth about moral matters is sound, its consequence is not that there is nevertheless a truth for our community, but rather that there is a distinct truth for each of us, and we cannot sustain a theoretical approach to adjudication on that basis. In spite of its current popularity, however, this skeptical metaphysical thesis is not coherent. Suppose I say to the metaphysical critic, “Genocide is wicked,” or “Racial discrimination is unjust.” He replies, “Yes, that’s true, I agree with you. But please do not make the mistake of thinking that these propositions are objectively true or that their truth is grounded in reality. You have only expressed your own opinion, with which I and others of our speech or interpretive community happen to agree.” A prominent proponent of the metaphysical thesis, Richard Rorty, drew that distinction in the following way. He said that, of course, as we all know, mountains exist. They existed before human beings did, and they will

60

In Praise of Theory

probably continue to exist long after human beings perish. But he then added that if you ask him a different question—whether mountains exist as part of Reality As It Really Is, with very big capital letters on these phrases—he would reply no, that is ridiculous. The existence of mountains is not part of Reality As It Really Is; their existence just flows from a language game that we play. But this distinction requires that we be able to distinguish the meaning of the following two propositions. The first is that mountains would have existed even if human beings never had. That is the statement that Rorty says is true. The second is that mountains are part of Reality As It Really Is. That is the statement that he says is false. But I cannot for the life of me see what sense you can make of the second proposition, no matter how many capital letters you pack into it, that makes it mean something significantly different from the first proposition. If that is right, Rorty’s thesis collapses. But some of you will think that we can construct a more successful thesis if we confine our skepticism about objectivity to justice and leave mountains out. But we cannot, and for the same reason. Suppose we say that genocide is wicked, or racial discrimination is unjust, or clitoridectomy is appalling, or freedom of speech is essential. Then we add that each of these judgments is just our opinion; that none of them is objectively true. We must be supposing that there is a difference in meaning between the following two propositions. Racial discrimination is unjust. Racial discrimination is objectively unjust. But we cannot find one. I am not going to make my argument for that claim here, because I have devoted an article (“Objectivity and Truth: You’d Better Believe It”) to the general issue of external skepticism, in which I report that argument at considerable length.9 pragmatics

But though, as I said, there are echoes of the metaphysical argument in Posner’s writings, he has said that he would not want to rest his own recommendations on any philosophical thesis: he regards his views of adjudication as free-standing. His views, he says, are best expressed not in any general theory, but rather in attitudes, and he offers his most formal account of those attitudes in a passage that strikes close to home. “The adjectives that I have used to characterize the pragmatic outlook—practi-

In Praise of Theory

61

cal, instrumental, forward-looking, activist, empirical, skeptical, anti-dogmatic, experimental—are not the ones that leap to mind when one considers the work of, say, Ronald Dworkin.”10 So Posner’s attitudes are presumably meant to contrast with those represented by the embedded approach to legal reasoning, though it is hard to grasp how from his Polonian list of virtues. He urges us not to lock out strange ideas, to attend to the consequences of decisions, and otherwise to conduct our intellectual and legal activities in a sage way. That is valuable advice: Dogmatism is a grievous fault, and, if we succumb, grievously must we answer it. But this is not the stuff on which a jurisprudence is built, and though Posner makes plain that he does not approve of my account of adjudication, he says very little that is precise about why or how his is different. Still, two items in his catalogue of the virtues I lack seem particularly substantial. He says, first, that the pragmatic approach is forward-looking. It is important to distinguish, however, between two very different contrasts he might have in mind. He might mean that legal reasoning should be consequential rather than deontological, or that it should be welfarist rather than consequential in some other way. I shall explain and consider each of these possibilities in turn. It is a central question in moral theory whether it is ever mandatory to do what will produce a worse state of affairs—whether we must always tell the truth, for example, even when by lying we could prevent a state of affairs worse in every way, including worse because more lies are being told. A consequentialist argues that we are never morally required to act in a way that produces worse consequences, and a deontologist that we sometimes are so required. (The argument is more complex than this description captures, but that is enough to make my point.) If Posner has this contrast in mind, he has misunderstood the embedded approach I defend, which is plainly consequential rather than deontological. It is consequential in its overall aim: it aims at a structure of law and community that is egalitarian in the sense I tried to describe in Law’s Empire.11 And it is consequential in detail: each interpretive legal argument is aimed to secure a state of affairs that is superior, according to principles embedded in our practice, to alternatives. So it cannot be an objection to the embedded approach that it is not sufficiently forward-looking if forward-looking means consequential.

62

In Praise of Theory

It is a further, almost as central, issue in moral theory whether, when we compare the goodness of states of affairs, we should look only to people’s welfare in those states of affairs—that is, only to whether and by how much they are better-off in one than in others. A welfarist must choose some function of welfare—some way of measuring whether and by how much a group is better-off—and the most popular such function is utilitarianism. A utilitarian welfarist argues that a law or judicial decision improves the state of affairs only if people are, in the aggregate or on average, better off as a result. Someone who rejects utilitarianism supposes that at least sometimes one state of affairs is better than another even though people are not better off on average or in the aggregate—because rights are better respected, perhaps, or because the situation is fairer or more just in some other way. We might, without too much eccentricity, use “forward-looking” to describe the utilitarian side of that argument, so we could say that Posner is recommending that legal reasoning and argument should be devoted to finding decisions that are better from the utilitarian point of view. The embedded account of adjudication is not necessarily anti-utilitarian in detail. Someone who accepts it might argue (as Posner, in effect, often has argued) that the best interpretation of legal practice shows the principle of utility at its core. But neither is the embedded theory committed to utilitarianism as a guide to adjudication—and, at least in my view, much of our law, including our constitutional law, cannot be justified on utilitarian grounds, but, on the contrary, must presuppose principles of equality and fairness that are not utilitarian in spirit or effect. And the overall aim of the embedded theory, which is egalitarian, is plainly not utilitarian. So if Posner uses “forward-looking” to mean utilitarian, then he is justified in accusing the embedded account of not being forwardlooking enough. But in that case he owes us an argument for utilitarianism, or at least an answer to the many serious objections that have been raised to it.12 It is hardly self-evident that progress consists in making people on average happier, or even, as Posner has sometimes suggested in the past, richer. So we do not make much of a case for the pragmatic alternative to the embedded approach if we take the former only to endorse utilitarian cal-

In Praise of Theory

63

culation. We should therefore turn to the other muscular adjective in Posner’s catalog: he says that the pragmatic approach is experimental. There is a sense in which the embedded approach is plainly experimental—indeed, more not less so than its principal rivals. It recommends that adjudication be imaginative with principle, so that a judge might propose, as providing the best interpretation of an area of the law, a principle that had not been recognized in the past, as Cardozo did in MacPherson, for example.13 So if Posner means to condemn the embedded approach as insufficiently experimental, he must have a different sense of experimental in mind: He must mean, not experimental in theory, but experimental in place of theory. If so, then we might paraphrase his advice in this way: Lawyers and judges should try different solutions to the problems they face to see which work, without regard to which are recommended or endorsed by some grand theory. They should concentrate on the practical problems before them and ask which of the available solutions would actually make things better. Let’s consider when that advice would be useful. Suppose your car has broken down on a lonely, winter night, far from help. The engine has gone dead and it will not start. It might be very good advice to say, “Don’t reflect on the physics of the internal combustion engine, just try a lot of things and see if one of them works.” For example, it may be that if you put your hat on backward, shut your eyes, and turn the key, the car will start, and if it does, do not argue with it, just drive away. In these circumstances, Posner’s advice might seem helpful. But now suppose you are a cosmologist asking about the age of the universe. You would then be very puzzled if Posner told you not to worry about what is really true, but only about what works. It is disconcerting to be told not to worry about the truth when that is exactly what you are worrying about. The advice would not be dangerous, however, because you know what “works” means in this context: a cosmological thesis works if it fits well with the rest of what we believe, and generates predictions about evidence and discoveries that turn out to be warranted. It is, in my view, a philosophical confusion to say that a scientific proposition’s being true just consists in its furnishing reliable predictions, but this does no harm in practice because the two ideas, as we might put it, march hand-in-hand: a scientist looking for reli-

64

In Praise of Theory

able predictions is, at least normally, a scientist well-placed to discover what is true. But now suppose a very different situation: You are a judge trying to decide whether the drug companies are really jointly and severally liable for the damage the patient suffered in spite of the fact that most of them did not cause that damage, and you are told not to worry about what is really true but just to see what works. The advice is now entirely useless, because you have to decide what is true about a variety of issues—about, for example, what is fair—before you can decide what “works,” because now—unlike the case of the stalled car or even the elusive big bang—you have no independent standard at all as to what “working” means. Suppose it appears, for example, that a decision for the drug companies would both promote more research and keep the price of drugs lower than would a decision the other way. Even that would not prove that this former decision “works” better than the latter, because it remains to be decided whether a decision that achieves these desirable results, but at the cost of depriving someone injured through defective drugs of compensation, is desirable. The emptiness of advising lawyers and judges to seek the decision that “works” is even more manifest when we consider a socially more divisive issue: abortion. Many judges and lawyers, as well as philosophers, have thought it important, in puzzling over this agonizing issue, to try to answer deeply theoretical questions about, for example, whether a fetus has interests of its own during the first two trimesters of pregnancy. How would it help to suppose that we should stop worrying ourselves about such difficult issues and only ask what solution would work? Suppose we say to some devoutly pro-life group that we should be experimental and try an extremely permissive policy for a while to see whether the social tension the issue has produced would then dissolve. If the tension does dissolve, we might say, and people no longer seem to care about the issue, that would prove that the permissive solution had worked for us. The prolife group will reply, in horror, that that prospect would not show that the permissive policy had worked; it would show, to the contrary, that it was an even more hideous disaster because it had made the community terminally insensitive. In law and morals, particularly, the admonition to avoid

In Praise of Theory

65

thorny questions by seeing “what works” is not just unhelpful. It is unintelligible.

professionalism

I will call the third vehicle of attack on the embedded approach to legal reasoning the professional objection. “We’re just lawyers here. We’re not philosophers. Law has its own discipline, its own special craft. When you go to law school, you are taught what it is to think like a lawyer, not a philosopher. Lawyers do not try to decide vast theoretical issues of moral or political theory. They decide particular issues at retail, one by one, in a more limited and circumscribed way. Their vehicles of argument are not the grand ones of the philosophical treatise, but the more homespun and reliable methods of close textual analysis and analogy.” The most distinguished and influential versions of this perspective are those of the great legal philosophers of the positivist tradition: Bentham, Austin, and, above all, H. L. A. Hart, who brought that tradition to a new level of sophistication and elegance.14 As I interpret him, Hart said that legal reasoning consists, in its core, in applying special legal rules developed in a political community for that purpose, so that general theoretical considerations, including moral or philosophical theory, are relevant to identifying law only to the degree to which those special legal rules make them relevant by explicitly incorporating theoretical standards. If so, then legal reasoning is properly understood as embedded in more general theoretical assumptions only to the contingent degree that conventional legal practice has so decreed, and, at least in Hart’s view, convention has not much so decreed in most contemporary legal systems. In 1994, two years after his death, a new edition of his famous book, The Concept of Law, was published with a new Postscript on which he had been working, off and on, for some years, but which he had never completed. That Postscript in some respects clarifies—though in other ways raises fresh questions about—the nature of Hart’s opposition to the embedded account of law. I hope to publish a substantial response to the Postscript in the near future, but cannot begin that daunting task here. Instead I will concentrate on less philosophical, and ostensibly more

66

In Praise of Theory

practical, versions of the professional challenge to integrity. Several decades ago, Edward Levi, once Dean of the Chicago Law School and Attorney General of the United States, published a thin but influential book, called An Introduction to Legal Reasoning, in which he set out a highly professionalized account of legal argument.15 He said that thinking like a lawyer consists not in applying large structures of theory to discrete legal issues, but rather in reasoning by analogy from one group of concrete legal decisions to another. Another Chicago Law School professor, Cass Sunstein, has now taken up and elaborated this view, which he describes as the “incompletely theorized” approach to law, and he has particularly stressed the contrast between it and the embedded view of legal reasoning. Sunstein is a recent convert to the anti-theory camp, but he is as zealous in its pursuit as converts characteristically are.16 He makes a variety of claims for “incomplete” theory. We must take care to distinguish these, and therefore to distinguish different responsibilities of citizens and officials.17 We have, first, responsibilities of judgment: we must each decide for ourselves what political positions and decisions to support and make. Some of us also have, second, responsibilities of coordination: we must decide whether and how to cooperate with others in furthering the policies or making the decisions we support. The form of that responsibility of coordination depends, of course, on role: for legislators it is a matter of forming legislative alliances, for ordinary citizens a matter of joining parties and ad hoc interest groups, and for judges in a multi-judge court, a matter of seeking a majority for a favored decision. Some of us—officials—have a third responsibility: of exposition. Officials must often provide a formal accounting of the decision they have made. The form of that accounting, again, is sensitive to role, and, when it takes the form of a joint document, like a legislative report or a judicial opinion signed by more than one judge, it speaks for people whose grounds of judgment may have been different from one another. Sunstein proposes an “incompleteness theorem” with respect to each of these responsibilities. Two of these theorems—about responsibilities of coordination and exposition—are unsurprising and, except in extreme circumstances, unexceptional. He uses the Rawlsian apparatus of overlapping consensus to argue that we should be willing to work with those who favor the policies or decisions we do even when our grounds are different

In Praise of Theory

67

from theirs. There are circumstances, he would agree, in which we should resist that advice: I have declined invitations to campaign with neo-Nazis against laws that make it a crime to declare that the Holocaust never occurred. But in the normal circumstances of politics, including judicial decision, the advice is wise. Sunstein’s second theorem, about exposition, suggests that compromise might also be wise in preparing a joint public accounting for an official decision, and that, too, seems good advice in many circumstances. It is of course possible for each of the judges of a multi-judge court who make up an “overlapping consensus” in favor of a particular decision in, say, the drug manufacturers’ case to write a separate opinion describing his own theoretical ground. But it might well sometimes be better, for a variety of reasons, for the majority to settle on a single, more superficial, opinion that each can join, and Sunstein is certainly right that that option should not necessarily be dismissed. Neither of these two theorems, about incompletely theorized political coordination and incompletely theorized joint exposition, contradicts the theory-embedded view of legal reasoning. But Sunstein offers a third theorem, about our initial responsibilities of individual judgment. The Rawlsian model of overlapping consensus supposes that each party to the consensus has made an individual judgment on theoretical grounds drawn from what Rawls calls “comprehensive” ethical schemes that differ from party to party. But Sunstein suggests that lawyers and judges should abstain, even in the exercise of the responsibility of individual judgment, from venturing into the more abstract reaches of political moral theory. He means to claim, that is, not just that political and judicial alliances may be forged out of concrete agreement even in the face of theoretical divergence, but that the individual judgments that produce the concrete agreement should themselves be superficial. He presents this view of “ordinary” legal reasoning as in contrast to my own. But Hercules, in Dworkin’s view, “shows us the hidden structure of ” ordinary “judgments and so lays these open to study and criticism.” Of course Hercules aims at a “comprehensive theory” of each area of law, whereas ordinary judges, unable to consider all lines of inquiry, must aim at a theory that is “partial.” But Hercules’s “judgments of fit and political morality are made on the same material and have the same character as theirs.” It is these points that I am denying here.18

68

In Praise of Theory

In fact, as we shall see, it turns out that he does not deny those points after all.19 But it is important to see what his initial statement might mean in order to see why he must finally abandon it. In Law’s Empire, and earlier in this article, I described the differences between Hercules and ordinary judges as differences in the direction and ambition of their reflections but not in the material on which they reflect or the character of the reflection. Though “ordinary” lawyers and judges reason about concrete legal issues from the inside-out, like an engineer reasoning about the feasibility of a new structure, they can set no a priori limit to the justificatory ascent into which a problem will draw them. It is the character of the inquiry itself— the problems it generates as it unfolds—that dictates the level of theory that must be explored, and this cannot be known or stipulated in advance. A lawyer or judge need not pursue a legal inquiry beyond the point at which he may responsibly assume that integrity has been satisfied as well as can be, and he must take into account, in deciding when it is responsible to assume this, his practical circumstances, including the need for a decision and the press of other responsibilities.20 If Sunstein really means to “deny” this account of legal reasoning, he must suppose that a lawyer or judge should refuse to confront problems about integrity when these are manifest, or should close his mind to such problems so that he is unaware of them. Imagine a judge forced to decide one of our sample problems. He cannot simply appeal to precedent, either because there is no precedent on point, or because the precedents most directly on point seem to him inconsistent with principles recognized elsewhere. Integrity asks him to continue his inquiry, expanding its theoretical scope if and as necessary. What contrary advice would Sunstein give? Should the judge try to decide whether the drug manufacturers are jointly liable without asking whether it is fair, according to standards embedded in our tradition, to impose liability in the absence of any causal connection? Should he try to decide whether women have a constitutional right to an abortion without asking whether a fetus is a person within the meaning of our constitutional structure, or whether the Due Process Clause is properly used to protect fundamental liberty, or whether freedom to control reproduction through abortion is a fundamental liberty? If so, why should he refuse to consider these obviously pertinent issues? If

In Praise of Theory

69

not, which “theoretical” issues should he decline to consider, and why are these more or differently theoretical? On what different, and less “completely” theoretical, basis should he decide? Sunstein’s answer to the last of these questions is unhelpful. He offers Levi’s solution: He says judges should decide hard cases not by turning to more abstract levels of theory but in a more lawyer-like way—by analogy. But that is a false contrast because (to paraphrase Kant) analogy without theory is blind. An analogy is a way of stating a conclusion, not a way of reaching one, and theory must do the real work. Would holding the drug manufacturers all liable be more like holding people liable who actually do cause damage or more like seeking out people who had nothing to do with an accident at all and making them pay its costs? Is burning your own flag more like making a speech on Hyde Park Corner or assaulting people with offensive insults? Is abortion more like infanticide or appendectomy? We cannot even begin to answer those questions without a deep expedition into theory: without asking basic questions about the connection between causation and responsibility, or why the liberty of speech is of special importance, or how the intrinsic value of human life is best understood and expressed. Sunstein understands this. He concedes that the method of analogy requires recourse to general principles, but he insists that this concession does not destroy the distinction between his views and the embedded account because, he says, analogy requires appeal only to “mid-level” principles, not to the high-level principles that integrity might require lawyers sometimes to invoke. But this is a peculiarly troublesome distinction, not only because “mid-level” is so uninformative a classification (is a political theory explaining why speech is particularly important a “midlevel” theory or something higher or lower?) but also because the very idea of an a priori constraint on legal reflection, defined as a boundary of abstraction such reflection must not cross, is so bizarre phenomenologically as well as logically. Lawyers (like other people) discover the scope of reflection they need to pursue in the course of inquiry, by finding where inquiry leads before a responsible resting place is reached. They do not— cannot—accept a methodology that stipulates in advance where they must stop no matter how inconclusive or unsatisfying their reflection to that point.

70

In Praise of Theory

So Sunstein’s appeal to analogy does not, after all, discriminate his method from the theory-embedded account he means to oppose. That makes the question why he thinks that judges should avoid theory all the more important. But his answers to that question are also puzzling, because they actually state, not reasons why integrity should be abandoned or even relaxed as a goal, but the demands of integrity itself. He calls attention, for example, to the importance of precedent in legal practice, and fears that judges anxious to impose new theoretical structures over the law would be too quick to overrule these. But that is a concern that Hercules noticed as well, and that led him, precisely out of respect for integrity with that feature of our practice, to adopt what I called the principle of “local priority,”21 a principle that Sunstein does not mention, but which seems equivalent to his own suggestion that judges should adopt a “presumption” that revisions in the law should be “local.”22 He also suggests that judicial commitment to large scale theories—such as, for example, a “personal autonomy” account of free speech—would rigidify law and make change more difficult. But, as he also points out, theoretical explicitness may make it easier to identify mistakes, and may also facilitate large-scale changes when the declared theories of the past are themselves identified as mistakes, as in the case, for example, of the precedents of the Lochner era. Integrity opens the way to useful change in a variety of ways: by separating the fact of precedent from its previously announced theoretical basis, for example, and through the device of gravitational force. Sunstein’s most intriguing defense of theoretical “incompleteness” is, however, more explicitly political. “For reasons of both policy and principle,” he says, “the development of large-scale theories of the right and the good is a democratic task, not a judicial one. These remarks should suggest the ingredients of an account of legitimacy of which incompletely theorized agreements would be a part.”23 But it is mysterious how “democracy” could generate “large-scale theories of the right and the good” unless judges accepted that it was part of their responsibility to identify which such theories were latent in legislation and other political events. There is little prospect of legislation explicitly enacting a general declaration of abstract principle: explicitly enacting, for example, that natural wonders have intrinsic value or that the risks imposed by beneficial enterprises

In Praise of Theory

71

should be born by the class of people who benefit from them. It is only through interpretation of more concrete enactments that we can identify the principles which we have together embraced. Sunstein may have in mind only constitutional adjudication, in which judges guided by what I have called the moral reading of the Constitution might try to impose their own “large-scale” theories on a public that would reject them. Even so, his argument wrongly conflates adjudication and jurisdiction.24 Perhaps judges should not have the responsibility to interpret constitutional restraints—perhaps that power should have been left, in some way, more with the people. But it hardly follows that a judge who is charged with deciding what follows from the First Amendment should refrain from asking, as one among other questions, why a democracy would have reason to protect speech in that special way. As I said, Sunstein finally concedes that he is not after all offering an alternative to the embedded account. He says, immediately after declaring that he denies my position, that he will qualify that denial later, and the qualifications turn out to leave little if any disagreement standing. “In brief,” he says, introducing the qualifications, “some cases cannot be decided at all without introducing a fair amount in the way of theory. Moreover, some cases cannot be decided well without introducing theory. If a good theory is available and if judges can be persuaded that the theory is good, there should be no taboo on its judicial acceptance. The claims on behalf of incompletely theorized agreements [he must mean incompletely theorized individual judgments] are presumptive rather than conclusive.”25 In the pages following, he describes the advantages of judges seeking integrity in ways that parallel much of my own account:26 He explains, for example, the importance of not insulating even the most uncontroversial rules and practices from theoretical examination.27 He then insists that he is not arguing that “general theory is always illegitimate in law. What makes sense is a more modest point . . . Judges should adopt a presumption rather than a taboo against high-level theorization.”28 But the embedded account gives the same advice: it recommends that judges ascend to more abstract theory only when they have special reason to do so. At one point Sunstein announces advice that the embedded account does not give. He says that, “Judges should adopt a more complete theory

72

In Praise of Theory

for an area of law only if they are very sure that it is correct.”29 (I would have thought, on the contrary, that we want most to be protected from judges who are “very sure” that their moral theories are the correct ones.) But he cannot really mean this, because, as he had pointed out, some cases cannot be decided at all, and others cannot be decided well, without introducing theory, which means that judges will often have to make theoretical judgments that bring conviction, or at least greater conviction than their rivals, even when this falls short of certainty. When, after all, is a judge right to think that he confronts a case that cannot be decided “at all” or “well” without some theoretical reflection? Isn’t it enough, to satisfy that standard, that without theoretical reflection the judge lacks conviction as to which answer is the one that, all things considered, best comports with his responsibilities? And isn’t it then sensible for him to carry his theoretical reflection to the point at which conviction is reached? If so, then there is no difference—none at all—between the embedded view, with its demand for adjudicative integrity, and Sunstein’s “modestly incomplete theorization.”

Summary: In Defense of Theory I shall close by returning to the question I postponed at the beginning. We face a revolt from theory, in law and across the rest of the intellectual landscape. Posner and Sunstein are only two examples: America’s foremost academic litigator, for example, Professor Laurence Tribe, announced that he had no general theory of constitutional adjudication, and did not intend to try to develop one.30 What explains this trend? Earlier I described a form of philosophical relativism that is now popular outside academic philosophy: it holds that truth in general, and truth about political morality in particular, is created by our practices, and that there is no truth about these matters that is independent of a particular culture or language. The popularity of that deeply confused philosophical position is not the explanation of the symptoms I described, however, but rather another symptom to be explained. Perhaps part of the answer lies in the great appeal, at the close of our century of ideological and technocratic disasters, of a becoming modesty.

In Praise of Theory

73

Intellectual modesty seems the opposite of a variety of vices: of racism and sexism, which presuppose superiority, of the ambitions of the metaphysicians and system-builders, which seem hubristic, and above all of the elitism of mandarin intellectuals, which seems undemocratic. We have now seen the trap in mistaking an anti-theoretical posture for modesty, however. Posner’s apparently innocent experimentalism ends in one of the most ambitious and technocratic absolutisms philosophers have ever devised, which is utilitarian consequentialism, and Sunstein’s counsel of judicial abstinence, if it were feasible at all, would produce not more democracy but the paralysis of a process essential to democracy. Modesty is an attitude, not a calling. We are modest, not when we turn our back on difficult theoretical issues about our roles and responsibilities as people, citizens, and officials, but when we confront those issues with an energy and courage forged in a vivid sense of our own fallibility. Our reflective judgment may charge us with self-restraint in a hundred dimensions, but accepting these is an act of modesty only if that judgment was itself truly and thoroughly reflective. I agree with the critics that not all judges are trained in philosophy. But if my arguments are sound we have no choice but to ask them to confront issues that, from time to time, are philosophical. The alternative is not avoiding moral theory but keeping its use dark, cloaked under all the familiar legal phlogistons like the mysterious craft of lawyer-like analogical reasoning. The other day, for the first time in my life I ate ostrich. Those beasts belong in the desert and perhaps on the table, though I am not yet convinced. But they do not belong on the bench. My praise of theory has so far been, you might think, negative. I have replied to critics but not said much positive for integrity in law. So my last words hope to remind you of why integrity is so important. Every contemporary democracy is a divided nation, and our own democracy is particularly divided. We are divided culturally, ethnically, politically, and morally. We nevertheless aspire to live together as equals, and it seems absolutely crucial to that ambition that we also aspire that the principles under which we are governed treat us as equals. We must strive, so far as we can, not to apply one theory of liability to pharmaceutical companies and a different one to motorists, not to embrace one theory of free speech when we

74

In Praise of Theory

are worried about pornography and another when we are worried about flag burning. We cannot pursue that indispensable ambition unless we undertake, when necessary, to ascend high enough in our collective deliberations, including our adjudicative deliberations, to test our progress in that direction. We must undertake that sovereign duty if we claim a rule of law that is not just an instrument for economic achievement and social peace, but an emblem and mirror of the equal public regard that entitles us to claim community.

chapter three

Darwin’s New Bulldog

Urgent Questions Richard Posner’s jeremiad is his third attack in recent months1 on what he calls “moral theory.”2 One of his earlier essays was a response to a lecture of mine,3 and I in turn responded to that essay,4 so this comment continues an already long discussion. Posner’s Lectures are characteristically entertaining, hasty, picaresque, and punchy. They are packed with a great variety of relevant and irrelevant excursions, references, and insults. The arguments he offers for his main claims are so spectacularly unsuccessful, however, as to make urgent a question he himself raises. What actually explains his fierce hostility—he calls it a “visceral dislike”5—toward the academic work he has set himself against? I will suggest an answer to this question: in spite of his assurances to the contrary, Posner may himself be in the grip of a substantive, noninstrumentalist moral theory that he does not fully acknowledge or perhaps even recognize. This theory is very different from the “moral relativism” that he formally embraces, but much better explains his various positions and passions. I must first, however, defend my harsh judgment of the quality of his arguments, and I am aware of a danger in doing so. Posner’s bad arguments may well be traps, for one of his central claims is that judges are not good at philosophical reasoning, and he may be tempting critics to help prove his claim by showing it true of at least one—particularly intelligent and eminent—judge. If that is his strategy, then he has caught one more woodcock in his springes. This Response considers Posner’s various claims in considerable detail. I

76

Darwin’s New Bulldog

have two reasons for troubling readers with a painstaking response to what I have described as evidently bad arguments. First, Posner’s arguments are in service of a populist anti-theoretical movement that is now powerful in American intellectual life—the disastrous view of science that he contemplates in the course of his argument is only another example of the trend.6 That movement wastes our time, because much of it, as in the case of Posner’s argument, consists of empty calls for action in pursuit of goals it cannot even describe, let alone justify. Worse, the anti-theorists mock an idea that they do not pause to understand, but that is a crucial predicate of any responsible search for social justice. Any moral principle, no matter how thoroughly embedded in our culture, language, and practice, may yet be false—or, no matter how thoroughly rejected, may yet be true. I do not know when the anti-theory episode in our intellectual history will have run its course; it may already have begun its overdue decline. But we should leave no prominent statement of its creed unchallenged.

Morality’s Independence My second reason for examining Posner’s arguments in depth is a tactical one. His Lectures illustrate an important philosophical point, because though he aims to show that both ordinary people and judges can dispense with moral theory, his own arguments again and again fall back on just such theory. He fails to see the contradiction because he fails to recognize the crucial difference between moral philosophy, on the one hand, and moral sociology, anthropology, and psychology, on the other. We can raise a great variety of questions about moral judgments in general or about any moral proposition—for example, that clitoridectomy is everywhere wrong—in particular. These questions fall into different intellectual domains. One is the domain of moral sociology. Do most people around the world agree about important moral convictions? If not, how great is the diversity of opinion? How many people, for example, think that clitoridectomy is wrong, and how many that it is morally permissible or even obligatory? A second domain belongs to moral anthropology. What best explains how human beings developed the disposition to make judgments of moral right and wrong? Do human beings have the capacity

Darwin’s New Bulldog

77

to perceive special moral fields or particles, or otherwise to establish sensory contact with moral elements “out there” in the universe? If not, what best explains why people have the opinions they do—why most people in some cultures think clitoridectomy wrong, and most in other cultures do not? A third domain belongs to moral psychology. What leads people to change moral opinions once these are formed, or to develop new ones? How far, for example, can arguments or other inducements shift people’s opinions about the rightness or wrongness of clitoridectomy? A fourth domain is that of morality itself. Is clitoridectomy morally wrong? Is it wrong everywhere, or nowhere? Or only in cultures without certain traditions or special needs or circumstances? There are important connections among these various domains and issues, but it is of capital importance to recognize that the fourth domain is conceptually distinct from any of the others. It is certainly possible, for example, for someone to think consistently that clitoridectomy has been widely accepted in many cultures, that its acceptance in some cultures and rejection in others, including his own, reflects only the different economic and other needs of the two societies, that no argument will ever change anyone’s views about the practice, and that the practice is everywhere morally odious. At one point Posner seems to acknowledge the difference between the first three of these domains and the last. He distinguishes questions “about” morality, which include the first three domains, from questions “of ” morality, which constitute the fourth, and he declares that his Lectures are concerned only with the former.7 If that were true, his Lectures would be much less objectionable. But it is not true, because Posner’s main claims, as we shall see, are “of ” rather than “about” morality; indeed, if they were only “about” morality, they would in no way contradict the opinions of his academic targets, whose work, so far as he objects to it, is entirely of the “of ” variety. Posner’s failure adequately to distinguish domains is evident throughout his Lectures. He provides page after page of unimpeachable reports of moral diversity and citations to now-familiar evolutionary explanations of altruism and other moral attitudes. (Do the “academic moralists” on his hit list really need to be told, or told so often, that societies, subcultures, and individuals have different moral opinions? Or that a warthog, if capa-

78

Darwin’s New Bulldog

ble of such thoughts, might think another warthog beautiful?)8 He seems unaware, however, of the need to show how any of this sociology or anthropology or science fiction bears on his “strong thesis,” which is a substantive claim of morality, or on the various substantive moral positions, including moral relativism, that he endorses. He may assume (as I fear many lawyers and legal scholars have) that the sociological facts and anthropological speculations he offers in themselves entail some relativist or other anti-objectivist moral position. But he must know that the writers he criticizes do not accept this entailment and that he should therefore have tried to explain and defend it. It is particularly odd that he feels no responsibility to defend his assumption, because he says that he means to pay “particular attention” to my own views,9 and I have recently published a long article explaining why any such assumption is a mistake and why the only kind of argument that could support a substantive moral position, including those to which Posner says he is attracted, is a moral argument.10 I wrote that article to combat confusions about moral theory that I believe are particularly popular now in law schools and certain other university departments, and that fuel the populist anti-theory movement. Posner has read the article, and cites it (often incorrectly)11 on several occasions. But, though he repeats the arguments I called fallacious, he makes no effort to respond to, or even to acknowledge, what I said. I will not attempt to summarize here the arguments I made in that article, though I am assuming those arguments in what follows. No doubt Posner believes he has grounds for rejecting my arguments, and I hope and assume that he will use part of his Reply to this Response to explain in some detail what these grounds are.12

What Is “Moral Theory”? Posner’s announced target is not morality, with which he says he has no quarrel, but something he calls “moral theory.”13 He may think that even if morality cannot be subverted by nonmoral arguments alone, “moral theory” can be. But if so, his strategy fails, because his distinction is itself confused. He thinks that the difference between moral judgment or reasoning, which occupies ordinary people, and moral theory, which tempts only a

Darwin’s New Bulldog

79

caste of deracinated academics, is a difference in kind. But the difference can be defended, if at all, only as an elusive matter of degree. People typically take up moral reasoning as a response to felt uncertainty or vulnerability in their moral convictions. Many people have initial opinions about moral and political issues: whether an early fetus has moral rights of its own; whether there is a moral difference between a doctor’s honoring a patient’s request to remove life support and honoring a patient’s request for lethal pills; whether a political community should try to reach collective decisions about such matters as abortion and euthanasia and enforce them on every citizen through law or, rather, should permit individuals to reach and act on their own convictions. Almost no one, except a few philosophers of “metaethics,” worries about the philosophical status of these opinions. Few people puzzle about whether their convictions are reports of mind-independent facts or only projections of emotions onto a morally neutral world. But many people do worry about whether their convictions are sound: they think it very important to arrive at the truth of these matters and to act out of a sense of that truth. They are, in short, morally responsible people, and their interest in moral reasoning is a natural consequence of that sense of responsibility. They want to reflect on their convictions and to satisfy themselves that these convictions are not inconsistent with the more general principles or ideals that they endorse on other occasions. They might well ask themselves, for example, whether their views about abortion presuppose some more general position about the connection between sentience and interests or rights, and whether, when this more general position is exposed, they can honestly endorse it or subscribe to its other implications. Or they might ask themselves if their views about whether it is proper for a state to enforce one view of euthanasia on everyone are consistent with their views about whether it is right for the state to enforce one view of abortion on everyone. I do not mean, of course, that people care only about the consistency of their convictions, as if truth didn’t matter. They care about the integrity of their convictions because they are anxious to do what is right. True, many other people are impatient with reflection of that character: they know what they think and do not wish to be troubled with doubts or suggestions that they are inconsistent or lack principle. They do not want

80

Darwin’s New Bulldog

any “theory” before they march or vote for war, and they ridicule those who do. Posner sides with them, but he shouldn’t misunderstand the motives and assumptions of more reflective people. He assumes that their sole motive is to convince everyone else in the world that they are right— time and again he says that disagreement proves that moral theory has failed.14 But that is much too crude. Of course, in ordinary life as well as in academic philosophy or journalism, people often hope to persuade others as well as themselves. But that goal does not exhaust the point of moral reflection on any occasion, and often is not even central to it. Reflective people want to satisfy themselves. They also want to satisfy other people, whose interests are affected by what they do, that they are acting out of tested conviction and with integrity. So they try to explain their convictions in a way that displays reflection, sincerity, and coherence, even when they have no hope of converting others to those convictions. Nor do reflective people insist on constructing an entire moral or political philosophy, like utilitarianism or some variant of Kantian metaphysics, before they believe or decide anything. Rather, they reason, as I have put it on other occasions, from the inside out.15 They begin with a particular concrete problem, and with reasons to worry whether they can defend their position against objections that it is arbitrary or inconsistent with their other views or convictions. Their own sense of intellectual, moral, and professional responsibility, therefore, dictates how general a “theory” they must construct or entertain to put these doubts to rest. When their responsibility is particularly great—as it is for political officials—they might well think it appropriate to test their reflections against the more comprehensive and developed accounts of other people, including moral and legal philosophers, who have devoted a great deal of time to worrying about the issues in play. People turn to these sources not with the expectation of finding definitive answers—they know that the sources will disagree among themselves—but rather for rigorous tests of their convictions, for fresh ideas if they find that their convictions need repair, and, often, for theoretical guidance they can follow in reworking their opinions into more accurate and better-supported convictions. I have described this process of reflection as one of “justificatory ascent.”16 It is impossible for anyone to stipulate in advance, through some a priori distinction between “reasoning” and “theory,” how far the process

Darwin’s New Bulldog

81

of ascent should continue: it should continue until the puzzles or conflicts that prompted it have been resolved, and that point is not knowable in advance.17 So it is in principle impossible to say where moral judgment ends and moral theory begins. Supporting a judgment that seems shaky or arbitrary after ordinary moral reflection by tracing its links to larger principles, visions, or ideals is part of moral reasoning, not something different added on, just as playing extra innings when a game is tied after nine is part of baseball. Moreover, principles and ideals have been produced at almost every level of generality. They include the overarching utilitarian thesis that whatever increases pleasure is good; John Rawls’s sharply limited political conception of justice;18 the theory of political democracy that I have tried to defend;19 Thomas Scanlon’s explanation of the point and value of freedom of speech;20 Herbert Hart’s observations about punishment in criminal law and the moral foundations of the law of negligence;21 the observations about personal autonomy at the heart of the three-justice concurring opinion in the Casey abortion case;22 the remarks about the fairness of market-share liability in several recent judicial opinions and law review articles;23 editorial-page observations about the separation of church and state; dinner-table remarks about a nation’s responsibility for protecting the human rights of people in other countries; a schoolteacher’s lessons about the environmental obligations one generation owes to another; a parent’s efforts to change a child’s opinions by asking, “How would you feel if he did that to you?” These specimens of moral “theory” differ only in their level of generality or abstraction, and any flat categorization of a moral argument as concrete or theoretical would be hopelessly arbitrary. Some of the “moralists” on Posner’s hit list write at relatively higher levels of abstraction and some at relatively lower levels; his indiscriminate bundling of them all together only compounds the arbitrariness of his position. He has missed both the complexity of the motives that animate moral reasoning and the complexity of the interaction between reflection and conviction as moral phenomena.

The “Strong” Thesis Posner’s “strong” thesis holds that no moral theory can provide a “solid basis” for a moral judgment.24 This thesis is itself, of course, a moral judg-

82

Darwin’s New Bulldog

ment of a theoretical and global kind, for whether any kind of moral claim provides a “solid basis” for another one is itself a moral question. It is a moral question, for example, whether a principle that condemns racial discrimination provides a sound basis for condemning affirmative action: that depends on whether the principle is sound, how it is best interpreted, and whether, so interpreted, it has that consequence. These moral questions must be carefully distinguished from the empirical issues that occupy the great bulk of Posner’s Lectures, because the latter bear mainly on the psychological question whether a theoretical defense of any moral judgment, particularly one constructed by someone who teaches in a university, can persuade someone else to alter his own contrary conviction. That is plainly a different matter: it is perfectly consistent for someone to insist that particular moral principles—for example, those at the heart of an unpopular religious tradition—are absolutely true and do indeed provide a “solid base” for a large variety of more concrete claims about moral rights and duties, and yet still concede that it is extremely unlikely that anyone else can be brought to see the truth of those principles or to accept them as a ground for those more concrete judgments. So Posner could defend his strong thesis only with a substantive moral theory of his own. One such theory is moral nihilism, which claims that nothing is morally right or wrong; nihilism would obviously justify the claim that no theoretical argument can supply a good reason for thinking an act either right or wrong. But Posner denies that he is a nihilist.25 Instead he describes himself as a moral “relativist” who believes that there are valid moral claims, namely those that meet “the criteria for pronouncing a moral claim valid.”26 Those criteria are “local, that is, are relative to the moral code of the particular culture in which the claim is advanced.”27 In fact, as we shall see,28 Posner is not successful in stating a coherent version of this relativism. But that does not matter now, because no even minimally plausible form of relativism, understood as a substantive moral theory and not just a piece of (here irrelevant) moral sociology, could justify his strong thesis. If relativism were true, then, together with general information about “the moral code of a particular culture,” it would provide a “solid basis” for moral claims within that culture. Indeed, several of

Darwin’s New Bulldog

83

the academic philosophers and lawyers who are Posner’s targets make arguments that many relativists would accept as providing a basis for at least part of what the philosophers claim. For example, John Rawls often characterizes his arguments as showing the implications of principles and ideals latent in the public culture of modem democracies.29 My own arguments about constitutional law are also interpretive of a particular political culture.30 Obviously, neither Rawls nor I, nor anyone else on Posner’s list, is a moral relativist on any plausible conception of that school, but a relativist would not deny that at least some of our arguments provide a “solid basis” for discrete moral judgments.31 Posner, then, offers no defense at all of his strong thesis. Most of his Lectures are taken up with a different—but also remarkably implausible— claim: that no general moral theory or argument can persuade someone to accept a moral judgment that he initially rejects. I agree that no moral argument can persuade anyone unless it finds a grip in his or her imagination. But imagination can take many forms, and many people’s imaginations include a yearning for ethical and moral integrity. They want their lives to display their convictions, and they want their convictions to be true. That alone is enough to explain how different levels of reflection, and even different kinds of academic moral theory, can have an impact on them. Some people want more than integrity: they want a vision of how to live—and of how to live together—that can inspire as well as justify, and that desire explains why the best moral philosophy has survived for centuries and even millennia. It is part of our folk-wisdom, as well as the opinion of formidable historians, that ideas sometimes do, in the end, move mountains and armies. Posner offers no real empirical evidence to the contrary. He cites crude versions of some of the “just so” stories of fashionable evolutionary biology. But that a moral sense has survival value in our species hardly shows that that moral sense does not include an ambition for integrity and coherence; it might as readily suggest the contrary. Posner offers his own reflections about how and why the training of all academic lawyers and philosophers in Western universities (whose personalities, backgrounds, histories, temperament, literary style, and argumentative strategies differ from one another, even within a single university, in countless ways)

84

Darwin’s New Bulldog

somehow makes them, as a monolithic group, unfit for moral persuasion. And he points out that some educated people have been guilty of moral crimes. This won’t do. If a decent general account of the impact of moral argument or theory on conduct and belief were possible, it would be inordinately complex and discriminating. It would distinguish between immediate and delayed impact; it would chart the variety of ways in which the latter might be mediated through the various institutions of popular culture; it would take account of the quality of argument as well as the skill and reputation of the arguer; and it would notice the thousands of other cultural and psychological variables on which any such impact would depend. “Does moral argument ever change people’s minds?” is, in short, a very poorly framed question, and much work would be needed to make it into a respectable one. Even so, we can confidently reject either of the following two answers to the question in its current crude form: “Always” and “Never.” Only a cosmically stupid optimist could think that good moral arguments always defeat initial self-interest or contrary inclination. Only a dogmatic cynic could insist that moral arguments never make any difference at all, no matter how good the argument or how mediate the impact. Posner seems tempted by the latter opinion, but he offers nothing but a bit of a priori psychological rationalism and a few anecdotes to show us why.

The “Weak” Thesis Posner’s weak thesis holds that, whatever force moral theory might have in ordinary life or politics, judges should ignore it, because they have better devices available for their special purposes.32 Once again, he ignores the independence of morality. If judges are not required to make moral decisions in deciding hard cases, then of course they are not required to consult moral theory. But if they do face moral issues, it would be a category mistake, like advising someone who has trouble with algebra to try a can opener, to tell judges to resolve those issues through history or economics or any other nonmoral technique. Posner hopes to persuade us that judges do not confront moral issues.

Darwin’s New Bulldog

85

In a recent article, he accused me of ignorance because I supposed that judges are interested in questions of fairness.33 I replied by citing several discussions of fairness in judicial opinions in the area of law I had been discussing and in the law review literature discussing those opinions.34 (I might also have cited a 1991 study of products liability decisions that concluded that “fairness was developed 18% more frequently than efficiency, and fairness controlled in the decision 24% more frequently.”)35 Early in his Lectures, Posner repeats his claim: moral theory, he says, doesn’t “mesh with the actual issues in cases.”36 Much later, however, he shifts his attack. He admits that “it may seem that the judge is plunged into the domain of moral theory,”37 but insists that a careful study of the record will show that I (and presumably all the scholars I cited) have been tricked into thinking that judges use moral terms in a moral way. But in what other way could they use them? He doesn’t say, and he undermines his suggestion that judges use moral terms in a nonmoral way immediately after he has made it, when he explains why judges do appeal to morality so often. They do so, he says, “to be impressive,” in order to “speak a language that the laity . . . is more likely to understand,” and, finally, because there is a “considerable overlap between law and morality.”38 All of these explanations assume, of course, that judges use moral terms in the same sense that you and I do, that is, to refer to moral concepts. Directly after these apparently decisive concessions, Posner switches to the offensive. He accuses “Dworkin and his allies” of trying to bring the law fully into accord, in every respect, with what we take the moral law to require, and compares us, charmingly, to the Afghan fundamentalists who decapitate people who affront Islamic religious law.39 But, of course, no one has to subscribe to the absurd claim that the law should enforce all and only moral obligations in order to think that moral convictions are sometimes—even often—relevant in deciding what the law is. Posner next tries to show that certain Supreme Court cases that I have discussed raise no moral issues. But his own examples show the contrary. He says, for example, that in the assisted suicide cases the Supreme Court justices “ducked the philosophical issue” that the “philosophers’ brief ” had discussed.40 The main moral claims of that brief were, first, that competent dying individuals have, in principle, a right to decide for themselves how

86

Darwin’s New Bulldog

to die, and, second, that even if recognizing that right would to some degree increase the risk that other patients would be pressured into choosing death against their will, that increased risk does not justify refusing to recognize the right at all. None of the Justices “ducked” both of these claims—three of them decided against our position on the first and five on the second.41 He says that the Supreme Court also “ducked the moral question” in Roe v. Wade, and adds, in this context, that his “entire point is that the courts are not capable of balancing ‘moral costs.’”42 (That comment follows a misguided though revealing complaint about an argument I had made involving the “moral cost” of delaying recognition of a right to abortion.)43 But the Court did not—it could not—“duck” the moral issue whether states should respect individual autonomy in matters of personal morality. (That issue was even more prominent in various Justices’ opinions in a later abortion case, Casey.)44 The Court could not have reached its decision in favor of autonomy, moreover, without also having decided a further moral issue—that an early fetus does not have interests of its own that entitle it to constitutional protection—because an issue cannot be framed in terms of individual autonomy if fundamental rights of others are at stake. (The Court would not hold, after all, that a mother has the right to decide for herself whether to kill her infant child.) There is no doubt that in the abortion cases the Court decided the moral issue whether a fetus is a person with interests and constitutional rights of its own. In fact, at several points Posner actually concedes that the judges who decide great constitutional issues often make controversial decisions of moral principle; indeed, he declares that they should, and even tells them which moral theories they should rely on. For example, he offers a moral argument to explain how he would have decided the assisted-suicide cases: the Supreme Court should have upheld the anti-assisted-suicide statutes because, given the balance of power on the issue in the nation, the values of democracy are best served by allowing the euthanasia battle to be fought out in ordinary politics. Of course, Posner must hide his appeal to moral theory, and he does so in a breathtaking way. He declares that convictions about political morality, including his own convictions about the proper workings of a democracy, are not moral judgments at all: they are

Darwin’s New Bulldog

87

only, he says, claims “about the political or judicial process.”45 (Similar statements about the difference between “moral” and “political” claims are sprinkled throughout his essay.)46 But these convictions are not “political” in any descriptive sense. They are normative claims about how political and judicial institutions should work. Nor are they normative in the strategic sense, as Posner sometimes suggests: they are not judgments about how best to achieve a stipulated goal, but rather highly controversial claims about what goals should be pursued. They are moral judgments about how the powers of government should be distributed and exercised, and when, if at all, these powers should be limited out of respect for individual moral rights.47 Posner’s discussion of racial segregation is even more seriously marred by his need to disguise his own appeals to moral theory. He first declares that the arguments the courts used (and others he says they should have used) were actually “nonmoral” arguments.48 For example, he says that the Brown Court rested its judgment on the “nonmoral” finding by psychologists “that segregation impaired the self-esteem” of blacks.49 But that “nonmoral” argument presupposes that blacks are the political equals of whites and deserve equal respect, and in the same discussion he declares that this presupposition was, in 1954, a highly controversial moral claim.50 He next tries a different tactic: he says that some of the moral principles the Court cited were not really controversial after all, but were merely part of the agreed “backdrop” of Brown, like the facts of the case.51 But he gives as his example the principle that “government should have a good reason . . . for allocating benefits or burdens on the basis of race,”52 and, as he has just conceded, it was certainly controversial whether blacks were politically inferior to whites, and therefore whether their inferiority was “a good reason” for segregating them on the basis of their race. Finally, he seems altogether to abandon his claim that judges should decline to make controversial moral judgments in constitutional cases; he rather advises them not to advertise that they are doing so. He says that the Supreme Court was “disingenuous” in its Brown decision, but that its course was wiser than embarking on any discussion of political morality.53 He adds that the Court’s opinion would have been even more effective—and more “honest”—if it had simply declared that “everyone knows” that “the Equal Protection Clause was in some sense intended, or should be used, to prevent” racial

88

Darwin’s New Bulldog

segregation in schools and other public places.54 That would of course have been a lie—the Justices were aware that many very distinguished jurists did not “know” that the Equal Protection Clause “was in some sense intended, or should be used, to prevent” segregation—and it is unclear how that lie would have contributed to “honesty.” It is even less clear why Posner thinks it would have been better for the Court, having announced that official segregation is inconsistent with equal citizenship, to refuse to defend that proposition. The Court would not have found it necessary, after all, to publish a dissertation on whether a Kantian understanding of basic human equality is superior to a Benthamite one, or on any other comparable philosophical topic. It would have been enough to satisfy the requirements of intellectual and political responsibility for the Court to have set out some principle identifying the kind of equality that the Equal Protection Clause should be understood to embody, a principle the Justices would have been willing to accept in its other implications. We can easily construct one such principle, though it might not have commended itself to the Brown Court. The Equal Protection Clause forbids legal constraints or institutions that can be justified only on the ground that some citizens are inferior to others, or that their fates are of less than equal concern, and official racial segregation cannot be adequately justified on any other supposition. (I believe that the Court has in fact been moving toward that principle, in fits and starts, between Brown and Romer.)55 Of course, that is not a self-enforcing principle: an argument must be deployed to justify each of its features, including its application to school segregation. That is just to say, however, that moral judgment, even as it figures in constitutional adjudication, is intrinsically open-ended and controversial. Posner would prefer judicial decisions to be based on algorithms or science, so that they could always be, in his special sense of the term, “resolved” by consensus. But, for better or worse, judges face moral issues, and railing at moral theory can’t change those issues into mathematical or scientific ones.

The New Pragmatism If I am right, Posner has provided no arguments at all for his “strong” claim that moral theory offers no ground for moral judgment, or for his

Darwin’s New Bulldog

89

“weak” claim that judges can and should avoid moral theory. On the contrary, in spite of several heroic but contradictory efforts to disguise the fact, he continually appeals to moral theory himself. We must therefore return to my original question. If his arguments are so bad, what does explain his fierce hostility to “academic moralism”? Set aside his own playful suggestion that age has dulled his skill at theory and therefore his taste for it.56 (I have no liking for the premise of that explanation.) Perhaps the answer lies only in temperament. William James describes a person whom he calls a philosophical “amateur,” who wishes to be “tender-minded” and “tough-minded” at the same time, and so combines contradictory attitudes.57 Perhaps we should be content to notice the same appetite in Posner, though he wants to cram together not tough- and tender-minded sentiments, but tough- and tougher-minded ones. But I hinted at a different explanation, which I shall now explore. Posner’s claims, I insisted, could only be sustained by relying on a large, substantive moral theory of his own. Perhaps Posner is in fact moved by such a theory. If so, it is not the theory he explicitly embraces in these Lectures. He says that he is a “moral relativist,” who believes that “the criteria for pronouncing a moral claim valid are local, that is, are relative to the moral code of the particular culture in which the claim is advanced.”58 But it is hard to take that self-description seriously as anything other than camouflage, for several reasons. For one thing, his account of moral relativism quickly produces a contradiction. If the moral code of China demands an abortion whenever a mother of two children conceives, and the code of Ireland forbids abortion in any circumstances, then, according to Posner’s definition, the Chinese produce a “valid” moral claim when they say that anyone in China or Ireland has a duty to abort in certain circumstances, and the Irish also produce a “valid” moral claim when they declare, for people in both countries, exactly the opposite. Perhaps he meant to say “in which the conduct takes place” rather than “in which the claim is advanced.” On that reading, Chinese and Irish speakers need not contradict each other: each can say that all women have a duty to obey their own community’s code. Then an Irish speaker who accepts relativism, so defined, should agree that Chinese women do nothing wrong when they abort. But Posner says that he rejects the “vulgar” version of relativism that has that consequence.59 The confu-

90

Darwin’s New Bulldog

sion thickens a page later, moreover, when he declares his sympathy for what he calls moral “subjectivism”: someone who rejects the moral code of his own community, he says, is not, after all, “morally wrong,” because there are no “transcultural moral truths.”60 But a relativist, as Posner has just defined this view, denies that we need “transcultural” moral truths to make “valid” moral judgments. Posner may sense that he has contradicted himself, because he adds that he really means that valid moral norms are relative not to a community of people, but to a single individual.61 But only a page earlier he denies that he is “a moral relativist in the ‘anything goes’ sense more accurately described as moral subjectivism.”62 That seems to be exactly the sense in which he later expressly says, however, that he is a relativist: he says that he would not, after all, call someone immoral who “sincerely” believed that it is right to kill infants.63 (That opinion, in turn, seems flatly to contradict another position he has recently announced elsewhere: it “is not my view,” he says, “that if someone said to me that it was okay to torture children all that I could say, in reply, was that I disagreed but that every person is entitled to his own opinion.”)64 The darkness becomes total later in the Lectures, when the “relativist” declares: A moral principle may be unshakable at present without being “right.” The fact that no one in a society has questioned a taboo against, say, racial intermarriage would not make that taboo morally right. To think it would, would be to embrace vulgar relativism, the idea that a society’s acceptance of a moral principle makes that principle morally right.65

Posner has announced, remember, only pages earlier, that “the criteria for pronouncing a moral claim valid are . . . relative to the moral code of the particular culture in which the claim is advanced.”66 Isn’t that the very view he now declares “vulgar”? All this makes it excusable to think that Posner is actually moved by a different “moral stance” from any of those that he formally endorses. He offers a glimpse of this different stance when he declares that “[r]elativism suggests an adaptationist conception of morality, in which morality is judged—nonmorally, in the way that a hammer might be judged well or poorly adapted to its function of hammering nails—by its contribution to the survival, or other goals, of a society.”67 We might be tempted to dismiss this statement at once: how can an “adaptationist” conception of morality

Darwin’s New Bulldog

91

be nonmoral? We cannot evaluate a morality by asking whether it helps a society to “survive,” because the morality a society adopts will almost always determine not whether it survives, but the form in which it does so. Nor does the reference to “other goals” help. If it were uncontroversial within some particular society what its “goals” were—if it were settled, for example, that it should aim to maximize its aggregate wealth according to some specified measure of wealth—then the hammer analogy might be apt. But of course that is not uncontroversial for our own community; on the contrary, our most divisive arguments are over which goals we should pursue—should we aim to be rich at the cost of serious inequality, for example? So “nonmorally” looks like whistling in the dark: if Posner completes his “adaptationist conception” by stating what goals he thinks appropriate, he has only offered, in a baroque form, a standard moral theory, and if he does not, he has said nothing at all. In his (greater) youth, Posner tried to defend a particular suggestion about the goals we should collectively pursue, and he did not shrink from labeling his suggestion a moral theory. Our goal, he said, should be to make our community, on average, richer.68 He now agrees, apparently, that he made a mistake,69 but he offers no substitute account of proper social goals. So far, this looks like the standard pragmatist dilemma. Pragmatists argue that any moral principle must be assessed only against a practical standard: does adopting that principle help to make things better? But if they stipulate any particular social goal—any conception of when things are better—they undermine their claim, because that social goal could not itself be justified instrumentally without arguing in a circle. So, typically, they decline to say what making things better means: Richard Rorty and the other leaders of Posner’s anti-theory army seem to assume, contrary to all political experience, that it is obvious to all when a situation is improving or, in a word they believe useful, whether a particular strategy “works.”70 But moral disagreements necessarily include disagreement about what counts as “working.” “Prolife” and “prochoice” activists would give a very different account of what form of abortion regulation “works,” for example. So moral pragmatism has seemed to many critics an empty theory: it encourages forward-looking efforts in search of a future it declines to describe. The vogue for Darwinian moral biology might seem to offer new hope

92

Darwin’s New Bulldog

for pragmatism, provided that pragmatism is sufficiently laissez faire. Suppose we were persuaded that through evolution human beings came to develop attitudes and dispositions that helped them not only to survive, but to flourish. We might then put our faith not in our own ability to identify appropriate norms and attitudes, but in nature’s ability to do this through natural selection or some analogue to it. We need not say, that is, that we know what is best for ourselves and our communities; we need only have faith in the process that has made certain inclinations, attitudes, sympathies, and dispositions natural in different communities. We need not be prepared to state, certainly not in the detail expected in a theory of the matter, what goals we should collectively pursue, or what counts as an improvement. We can say, instead, that natural inclinations must be assumed to be wise, and that the goals to which they direct us must be assumed to be appropriate, unless we are persuaded, of course, that our unreflective assumptions are based on false factual information. We might call this roughly sketched attitude “Darwinian pragmatism.” It is important to understand that it is, at bottom, a substantive and noninstrumentalist moral attitude, because it presupposes that certain kinds of human lives and certain states of human societies are intrinsically superior to others. It is instrumental only in the sense that it proposes a particular means—a combination of factual investigation and otherwise noninterventionist quietism—for identifying and reaching those states. That instrumentalism would make no sense—it would be barren tautology—but for the further assumption that there are intrinsically superior states of human life and society for nature to discover, that is, that these states are not “superior” only because they are what nature, corrected only for factual error, produces. The assumption that Darwinian pragmatism is Posner’s intuitive but hidden conviction dissolves all the mysteries I have been describing. It explains his reluctance to condemn anything as immoral that is the natural and unexamined expression of genuine moral convictions. It explains his fascination with the biologists’ stories of how altruism and morality arose. It explains why he is such a fickle moral relativist. He is strongly drawn to the spirit of relativism, which instructs us to respect moral codes in force, but not to its condemnation of our natural impulse to call moral codes that differ from our own wrong. He is also unwilling to denounce moral

Darwin’s New Bulldog

93

rebels as immoral: their rebellion is also natural and, as he insists, may have Darwinian value. (“In fact,” he says, “we need some immoralists, or at least amoralists.”71 Of course, the immoralists’ or amoralists’ victims do not need them. The “we” in question is the human race, which is still evolving and needs mutations for the purpose.) The hypothesis explains why he wants “elite” judges to use their constitutional power to strike down what offends them viscerally—what they cannot “stomach”—but not to attempt to pale this over with any sickly cast of theory.72 It explains why he suddenly abandons his relativism when he declares that “unshakable” moral codes may not settle what is morally right: the opposite view would block the kind of “adaptationist” process he thinks is the only true progress. It explains why he resists any skepticism or cynicism about morality: he understands that morality has its natural, evolutionary force only so long as it feels like morality. So he declares that there is a local “fact of the matter” about moral claims, as there is about the temperature in a particular city, and that he is a “kind” of moral realist.73 But he is at the same time wary of any but an instrumental justification for the morality people create, and so, pages later, he insists that he is not “sliding into” the very realism that he here embraces.74 Darwinian pragmatism helps to explain, moreover, his deep admiration for Oliver Wendell Holmes, whose own essay “The Path of the Law” has puzzled legal philosophers anxious to categorize that Justice, but which is also best understood, as its title suggests, as a celebration of the inexorable lava flow of nature cutting its path through history.75 The Darwinian hypothesis explains, above all, the distinction we found so troubling earlier between “ordinary” untutored moral reasoning, on the one hand, and “academic” moralism, on the other. Posner is anxious to protect what strikes him as natural, and “unreflective” means “natural” to him. He is equally anxious to ward off anything that smells not of nature but of the lamp: he thinks that academic theory is unnatural, interventionist, written by people who haven’t really lived, and (however he might protest its innocuousness) in the end dangerous.76 He calls for the death of moral theory, but, like all of philosophy’s would-be undertakers, he only means the triumph of his own theory. For his arguments show the opposite of what he intended: they show that moral theory cannot be eliminated, and that the moral perspective is indispensable, even to moral

94

Darwin’s New Bulldog

skepticism or relativism. Posner is himself ruled by an inarticulate, subterranean, unattractive but relentless moral faith.

Appendix: Pragmatism and Bush v. Gore By far the most popular defense of the Supreme Court’s decision in Bush v. Gore is that the Court saved the nation from a further, and perhaps protracted, period of legal and political battles and of continuing uncertainty about who the next president would be. On this view, the five conservative justices knew they could not justify their decision on legal grounds, but they decided, heroically, to pay a price in damage to their reputation as jurists in order to save the nation from those difficulties: They “took a bullet,” as it has sometimes been put, for the rest of us. In a book of comments about the election and the Court’s decision that I edited, Richard Posner, with his characteristic incisiveness and vigor, sets out the argument for this view more clearly than anyone else has.77 Posner says that at least sometimes judges should take a “pragmatic” approach to their work and make decisions they believe will have the best results overall, even if these are not decisions that past legal doctrine would authorize. That pragmatic approach, he believes, would have recommended deciding Bush v. Gore in the way the conservatives did; he compares their decision to Abraham Lincoln’s defiance of the Constitution in suspending habeas corpus during the Civil War and, more ominously, to the Supreme Court’s decision permitting the internment of Japanese-Americans during World War II.78 He does not, however, think that judicial pragmatism should be reserved for exceptional emergencies like full-scale war: on the contrary, he advocates pragmatism as a general style of adjudication, right for judges in quotidian cases as well as in constitutional emergencies. We should therefore look at his formal statements of what pragmatism is and means. I have taken one of these from his earlier book: “Pragmatic” as an adjective for anything to do with the judicial process still causes shudders. It seems to open up vistas of judicial willfulness and subjectivity and to mock the rule of law; it seems to equate law to prudence, and thus to be Machiavellian. All that pragmatic adjudication need mean, how-

Darwin’s New Bulldog

95

ever—all that I mean by it—is adjudication guided by a comparison of the consequences of alternative resolutions of the case rather than by an algorithm intended to lead the judges by a logical or otherwise formal process to the One Correct Decision, utilizing only the canonical materials of judicial decision making, such as statutory or constitutional text and previous judicial opinions. The pragmatist does not believe that there is or should be any such algorithm. He regards adjudication, especially constitutional adjudication, as a practical tool of social ordering and believes therefore that the decision that has the better consequences for society is the one to be preferred.79

In his discussion of Bush v. Gore, Posner puts this account in a more complex philosophical context. He distinguishes “everyday” pragmatism, which is the consequentialist, “hard-nosed,” cost-benefit approach to legal reasoning described in the paragraph just quoted, from two more philosophical forms of pragmatism: “orthodox” and “recusant.”80 The judge who is a pragmatist in this everyday, consequentialist sense does not despise precedent and technical legal argument: on the contrary, he is aware of, and takes into account, both the good consequences that flow from systematic judicial respect for traditional legal argument and doctrine, which include encouraging people to plan their affairs with confidence, and the bad consequences that might flow from a judge’s ignoring traditional doctrine on particular occasions, which include defeating such expectations and weakening the general benefit of systematic respect for them. But the pragmatic judge is also aware of the dangers of a slavish deference to orthodox legal reasoning; he knows that in some circumstances he can achieve better consequences, even in the long run, by reaching the decision that will produce some particularly important benefit or avoid some particularly grave danger, even though this decision flies in the face of established doctrine. So pragmatic judges must balance the long-term benefits of respecting doctrine against the long-term benefits, from time to time, of ignoring it. As Posner says, “There is no algorithm for striking this balance . . . He or she should try to make the decision that is reasonable, all things considered, where ‘all things’ include the standard legal materials . . . but also the consequences so far as they can be discerned of the decision in the case at hand.”81 In Bush v. Gore, Posner says, the Supreme Court had to strike that bal-

96

Darwin’s New Bulldog

ance. Would it produce “the better consequences for society” in the long run to follow precedent and doctrine, which recommended dismissing Bush’s appeal and therefore allowing the Florida recount to continue, or to endorse an unpersuasive legal argument in order to halt the recount so that Bush would become president-elect at once? It was predictable, he says, that if the five conservative justices voted for the second choice, they would be thought to have made a rank partisan decision, and that the Court’s reputation for honesty and impartiality, which is important, would suffer. That counted for the first choice. But the possibility of what he calls a “worst-case scenario” following that decision argued more powerfully for the second choice. Here is the worst-case scenario Posner says the Supreme Court justices might have contemplated in December of 2000, when they had to make their decision. The recount might have shown Gore the winner in Florida, and the Florida court might then have ordered the state’s electoral votes certified for Gore. Since the Supreme Court’s decision was to be handed down on December 12, that recount would not have been completed by the “safe harbor” deadline of that very day, which immunizes a state’s certification of electors from congressional challenge; indeed a responsible recount could not have been completed even by December 18, the date on which electors are required to cast their votes. In the meantime, the Florida legislature, dominated by Republicans, might have chosen its own slate of electors pledged to Bush. Congress would then have had to choose between the two slates, but Congress might be divided: the Republican House siding with the Bush slate, but the equally divided Senate, still presided over by Vice President Gore, who would have cast the deciding vote, endorsing the Gore slate. If Congress was unable to agree, the slate certified by the Florida governor, who is Bush’s brother, would be seated. But what if the Florida supreme court had ordered the governor to certify the Gore slate, the governor had refused, and the Florida court had declared him in contempt? Who would decide what the governor’s official verdict was? Suppose in the end no Florida votes were counted at all. Gore would then have had a majority of the electoral votes cast, but not a majority of the overall number of votes, and then the presidency would depend on the unsettled issue of whether he would need only the former to win. The Supreme Court might refuse

Darwin’s New Bulldog

97

to decide that question on the ground that it is a political question, in which case the impasse would drag on indefinitely. An acting president would be needed and, under certain assumptions, that would be Lawrence Summers, then Secretary of the Treasury (and now the president of Harvard University). Would Summers have been an effective acting president? Posner sometimes makes modest, negative claims for the plausibility of this worst-case scenario. He says that it “is by no means fantastic, or even highly improbable,” and later that it is “not inevitable,” but “it could not be thought phantasmal.”82 For the most part, however, he argues as if his scenario was, if not inevitable, at least so probable that a pragmatist judge should assume it to be the consequence of his decision to allow the recount to continue. (He called his book on the 2000 election Breaking the Deadlock, not Breaking a Not Phantasmal Deadlock.) In fact, assignments of probability are indispensable to any genuine consequentialist analysis. It would be irrational for a pragmatist to compare two alternatives by comparing only the worst possible consequences of each, or only the best, or even only the most likely. He must compare the various possible consequences of each decision, taking into account their gravity, but discounting each by its probability. Posner’s pragmatic argument becomes strikingly less impressive, even in its own terms, when we reformulate it in that spirit. Posner begins by assuming that the manual recounts ordered by the Florida supreme court could not possibly have been completed by the “safe harbor” deadline of December 12 and were extremely unlikely to have been completed even by December 18. But this assumption supposes something that a pragmatic consequentialist cannot assume: that something other than a nice balance of long-term consequences required the Supreme Court to stay the Florida recounts on December 9 and also required it to declare, in its December 12 ruling, that the manual recounts under way were defective because they did not stipulate uniform recount standards. A consistent pragmatist would have had to ask himself, in considering whether to issue a stay on December 9, whether the long-term consequences of a decision not to intervene at all, for which the Court would have had ample doctrinal cover, would be better than those of staying the recount on that date and deciding a few days later that recounts

98

Darwin’s New Bulldog

were unconstitutional unless subject to uniform standards. He could not have assumed, in asking himself how much disorder the first choice would produce and the second avoid, that the recounts would for some unrelated reason stop until December 13, begin again only after whatever time it took to debate and choose uniform standards, and allow time for adjudication of arguments about the application of those new standards.83 So the pertinent consequentialist question was whether, if the Florida supreme court’s recount had been allowed to proceed in the form that court originally ordered, free from Supreme Court interference of any kind, it would have finished in time for Florida to certify a winner on December 12, or in any case by December 18. If so—and there is no reason to doubt it—then the rest of the “worst-case scenario” is irrelevant. This is the fallacy not just in Posner’s argument, but in the more informal and widely embraced argument that the Court did us a favor by avoiding a crisis. All these arguments ignore the fact that the Court immeasurably worsened the risks before it avoided them. If the Court took a bullet for the rest of us, it also fired the gun. Even if we set this crucial objection aside, moreover, and assume that the recount process had to begin again, under new standards yet to be chosen, on December 13, the argument that the Court saved the nation from crisis is still much weaker than Posner or the popular view assumes. A pragmatist would have to assume, even under that assumption, that there was a fifty-fifty chance that a recount, whenever finished, would show Bush the winner, in which case the controversy would end. Even if it showed Gore the winner, none of the steps that would provoke the deadlock Posner describes was inevitable. The Florida legislature might not have elected an alternate slate—there were genuine political risks even for some of the Florida Republicans in doing so. Even if it did, Congress might not deadlock in choosing between the rival slates. Some Republican congressmen might have thought that the case for preferring electors chosen by the people of Florida to those chosen by political officials was too strong to overcome. Some Democratic senators from states carried by Bush might have yielded to political pressure to vote for the Florida electors pledged to him. One or the other candidates might have stepped aside. The moral case for Bush withdrawing would have been very strong:

Darwin’s New Bulldog

99

He had lost the national popular vote and knew that many more Florida voters intended to vote for Gore than for him; if, on top of that, a recount had shown that more Florida voters actually had voted for Gore, public opinion might have swung so decisively against his becoming president in a brokered deal, or a congressional power play, that he would have thought it best to yield. (Many Europeans were surprised that he did not withdraw after the butterfly ballot episode made it apparent that he was not morally entitled to win on any reckoning.) Or, as the mini-crisis continued, public opinion might have begun to grow, in a self-fueling way, against Gore, and he may have decided that his political future would be enhanced by yielding at once. Or some deal might have emerged more quickly in Congress than Posner imagines. Or, if it became clear that Florida’s votes would not be counted, the Supreme Court could have agreed to decide whether a majority of the electoral votes cast or all the electoral votes was necessary to win the presidency: it could plausibly have argued that this was a straightforward question of interpretation that, in the absence of a political resolution in Congress, it had the responsibility to decide. Or, if all the other steps in the worst-case scenario had indeed materialized, Summers might have governed well while acting president. Some of these possibilities had distinctly lower probabilities than others, of course, and some had very low probabilities indeed. But if we begin by supposing that the probability of a recount showing Gore the victor was only about fifty percent to begin with, and then take into account the compound probabilities of Posner’s other speculations, his worst-case is exceedingly improbable. Which is only a fancier way of saying that in politics you never know. Even Posner’s worst-case scenario would not have been a national tragedy: it would not have been as bad (to recall Posner’s analogies) as a Southern victory in the Civil War or a Japanese victory in World War II. So far, then, Posner’s pragmatic defense seems a failure. But we have not yet reached the most serious problem of all. I said that, as of December 9, the balance of pragmatic considerations plainly tilted against Supreme Court intervention. But that judgment, I must now concede, ignores the single most important pragmatic consideration of all. If a pragmatist Supreme Court justice set himself to evaluate, on December 9, the overall consequences of halting the Florida recount compared to the conse-

100

Darwin’s New Bulldog

quences of letting it continue, it would be irrational of him to ignore the fact that halting the recount would insure a Bush presidency for at least four years, while letting the recount continue would mean a substantial— let us say a fifty-percent—chance that Gore would be president instead. The question of which of these two politicians finally became president would swamp all other factors in a genuinely pragmatic comparison of the two decisions. The justices no doubt each had opinions about whether the country would be better off if Bush or Gore became president in early 2001. Presidents have enormous power for good or bad, as the Bush presidency has dramatically demonstrated, and the justices knew that the difference in consequences for the country of one rather than the other being president for four years would be very great. They knew that none of the costs and benefits of a decision either way in Bush v. Gore that Posner describes would be on the same scale of importance. Someone who welcomed a Bush presidency would have thought the Supreme Court’s decision a consequentialist triumph: it achieved the desired result without the trouble and risks of continuing the post-election battle. But someone who thought Bush dangerous for the nation would have thought the decision a pragmatic disaster: he would have thought—and this is the crucial point—that a continued post-election battle was a small price to pay for a substantial chance of finally avoiding a Bush presidency. A pragmatist justice would have had to decide for himself which of these two consequentialist judgments was right, and that means deciding whether Bush or Gore was better for the country. Of course, that is just what the severest critics think the five conservative justices actually did: they decided that Bush would be a better president and they acted accordingly. Indeed, almost everyone thinks that if the identical case had come to the Court with only the difference that the candidates’ position was reversed—Gore had been certified the winner, Bush had persuaded the Florida courts to order recounts, and Gore was asking the Supreme Court to halt those recounts—the five justices would have voted not to intervene at all. (Even Posner suspects that, though he suggests that the influence of politics and self-interest on the five justices was only to make them “more sensitive to” arguments that they might otherwise have overlooked.)84 But

Darwin’s New Bulldog

101

almost everyone—including, I assume, Posner—believes this fact to be regrettable. It is thought to be a devastating criticism of the five justices, if true, that it made a crucial difference to them which candidate would win if they stopped the recounts. But if Posner is right that the justices had a responsibility to reach a “pragmatic” result in this case, then it would have been irresponsible of them not to have allowed that to make a crucial difference. Posner acknowledges this difficulty: In his discussion of Bush v. Gore he makes the surprising admission that it poses “perhaps the ultimate challenge to pragmatic adjudication,” meaning, apparently, that pragmatism would be an unacceptable theory of adjudication if it recommended that judges should sometimes decide election cases so as to elect the best candidate for the nation.85 He declares, however, that though a good pragmatist would take everything else into account in his assessment of consequences—would take into account, for example, the risk that some “rogue” nation would be tempted to take advantage of a protracted presidential contest to injure us—he would not take account of which candidate’s overall policies would be better for us over four years. He describes that result as “fortunate,” but his defense of it is unconvincing.86 He falls back on a strategy known to philosophers as “rule-consequentialism.” This strategy supposes that people often produce the best consequences in the long run by following a rule carefully constructed so that following that rule in every case, whether or not it produces the best consequences in that case, considered on its own, produces the best consequences over time. He insists that judges should follow a strict rule, for that reason, not to make partisan political judgments. It is unclear why Posner thinks that following this rule, instead of taking into account who would be a better president in those rare cases in which a judicial decision will decide a presidential election, would indeed produce the best consequences, even in the long run. Of course it would damage the Court’s reputation and therefore its effectiveness if people generally thought that the justices had made a partisan decision. But people do think that anyway about the decision in Bush v. Gore; that was, as Posner himself emphasizes, an important and inevitable cost of the Court’s intervening in the election, at least once it was clear that only the conservative

102

Darwin’s New Bulldog

justices would decide for Bush. He does not think, in any case, that a pragmatic judge should openly admit to pragmatism. He suggests that the Bush v. Gore majority should have decided on pragmatic grounds, but constructed the best cover story of doctrine they could to hide that fact. But then why shouldn’t the actual pragmatic grounds for decision, which are to be hidden from the public anyway, include the decision’s single most important consequence? Posner says that that would derange the balance of powers in the American government. If he means that it would have bad long-term consequences for that reason, he must explain why. Why is it not better in the long run to allow a very occasional minor derangement, hidden as well as possible, when that will save the nation from a calamitous presidency? Posner’s argument seems driven more by the need at all costs to deny that a pragmatist judge would ever base his decisions on partisan political grounds than by any actual pragmatic case for that denial. He is right, of course, that judges should follow a rule never to rely on such partisan judgments. But he is wrong to imagine that the conservative justices’ decision in Bush v. Gore could be defended as a decision promoting, in their view, the best overall consequences if they had followed that rule. Rule-consequentialism provides an argument for not judging consequences case by case, but deciding in accordance with fixed rules instead. But Posner is now proposing something very different: a hybrid process in which judges decide by assessing consequences, case by case, but adopt a rule that requires them to leave the most important consequences out. That is perverse. It would make good sense to say that, because judges should not be partisan, they should decide cases involving presidential elections strictly on principle and doctrine, not on any calculation about which result will be overall better for the nation.87 It also makes sense— though unattractive sense—to say that because judges should aim to produce the best consequences, they should make partisan political judgments in cases, including cases involving presidential elections, in which an overall assessment of consequences is impossible without them. It makes no sense at all to say that judges should decide such cases pragmatically by assessing the costs and benefits of a decision either way but taking no account of the likely winner of that struggle, which is a fact of absolutely decisive importance in assessing those costs and benefits. How could

Darwin’s New Bulldog

103

anyone sensibly decide the true cost of risking even Posner’s worst case without considering how good or bad a Bush presidency—the certain outcome of the most certain way of avoiding that worst case—would be? Posner’s settled recommendation is a formula not for pragmatic decision but for a parody of pragmatism, like asking a doctor to choose between alternate medicines for a patient by comparing their price, availability, and ease of administration without also asking which will cure and which will kill him. So Posner’s good-consequences defense of the Supreme Court’s action collapses on careful examination. But since Posner offers a more general defense of judicial pragmatism in his article on the Court’s decision, we should also consider his more general arguments. “Pragmatism” is a buzzword among lawyers now: it appears everywhere and in the oddest contexts.88 But since judges, like everyone else, disagree about the relative value of different possible consequences of their decisions, telling them to decide by weighing consequences is only—as Posner conceded many people think it is—an invitation to lawlessness. The difference between judicial pragmatism and more orthodox theories of adjudication is easy to grasp in dramatic cases when it is clear that a conventional legal analysis would recommend a decision that almost everyone in the community would accept is bad. Posner’s example of Lincoln’s decision to defy the Constitution by suspending habeas corpus during the Civil War is that sort of case. There is room for doubt, of course, whether Lincoln’s assumption that this was necessary to protect the nation’s security was sound as a matter of consequential analysis. But the goal he aimed to serve—that the nation’s security be protected—was not controversial within the community for whom he acted. In many—probably most—hard cases, however, it will not help simply to say that judges must think about consequences, because the nerve of the controversy is how those consequences should be assessed. The abortion cases provide a dramatic example that I have used before to make this point. Would it produce the best consequences for society to forbid early abortions, or to permit them? Citizens, lawyers, and judges who disagree about the morality of abortion also, and just for that reason, disagree about which consequences would be best. One side thinks abortion murder and that any society that permits murder is degraded. They therefore believe that the

104

Darwin’s New Bulldog

consequences of permitting abortion are disastrous. The other side thinks that forbidding abortion dooms thousands of women to a miserable life for no adequate reason, and therefore that the consequences of that decision would be dreadful. If a judge undertook to decide the constitutionality of prohibiting abortion by asking whether the consequences of permitting or banning abortion were better on the whole, he would have to choose between these dramatically opposed convictions and he would have no choice but to rank the consequences himself, according to his own lights, relying on his own convictions, so that if he himself believed that abortion was murder or otherwise deeply immoral he would uphold the constitutionality of laws prohibiting it. He would tell himself that, on his own best judgment, and needing no other authority, he deemed the consequences of prohibiting abortion better for society than the consequences of permitting it. That would be a genuinely pragmatic judgment, though almost all lawyers and citizens (and perhaps even Posner) would think it wrong and even irresponsible. It is therefore no surprise that Posner’s kind of pragmatism would issue a similar, and similarly irresponsible, command to the justices who decided Bush v. Gore: to decide whether all things considered a Bush presidency would be sufficiently better than a Gore one to outweigh the damage the Court incurred by accepting the case and deciding on flimsy doctrinal grounds for Bush. I should add—though I hope it goes without saying—that judges need not choose between weighing consequences in that personal way and ignoring consequences completely. No one supposes that judges could or should decide cases “by an algorithm intended to lead [them] by a logical or otherwise formal process to the One Correct Decision, utilizing only the canonical materials of judicial decision making, such as statutory or constitutional text and previous judicial opinions.” That account of adjudication is a straw man and always was. Of course judges must take the consequences of their decisions into account, but they may only do so as directed by principles embedded in the law as a whole, principles that adjudicate which consequences are relevant and how these should be weighed, rather than by their own political or personal preferences.

chapter four

Moral Pluralism

I believe that Isaiah Berlin’s ideas are growing in influence and will continue to do so. It’s mainly in political philosophy, and in his idea of value pluralism, that I detect this growing and continuing influence. I’m going to quote a few sentences from his work, not connected sentences, but nevertheless adequate to suggest the considerable originality and interest of his thesis. He begins: What is clear is that values can clash. Values may easily clash within the breast of a single individual. And it does not follow that some must be true and others false. Both liberty and equality are among the primary goals pursued by human beings through many centuries. But total liberty for the wolves is death to the lambs. These collisions of value are the essence of what they are and what we are. If we are told that these contradictions will be solved in some perfect world in which all good things are harmonized in principle, then we must answer to those who say this that the meanings they attach to the names which for us denote the conflicting values are not ours. If they are transformed, it is into conceptions not known to us on earth. The notion of the perfect whole, the ultimate solution in which all good things coexist seems to me not merely unobtainable—that is a truism—but conceptually incoherent. Some among the great goods cannot live together. That is a conceptual truth. We are doomed to choose, and every choice may entail an irreparable loss.

Toward the end of his most famous essay, Berlin took this theme up again, but in a more minatory way. He recognized the appeal of the view that I

106

Moral Pluralism

just quoted him as declaring false, the appeal of the ideal of the perfect whole. He recognized its appeal as enduring and important. But he said that we must not yield to this impulse because “to allow it to determine one’s practice is a symptom of an equally deep and more dangerous moral and political immaturity.” Those are strong words, and they accuse me for what I am about to say. I shall try, nevertheless, to defend the holistic ideal that Berlin condemned in that grave way. But before I do, I should like to comment on his suggestion that this ideal is not only false but dangerous. There are indeed dangers in the hedgehog, but we must not forget that there are dangers in the fox as well. Just as tyrants have tried to justify great crimes by appealing to the idea that all moral and political values come together in some harmonious vision so transcendently important that murder is justified in its service, so other moral crimes have been justified by appeal to the opposite idea, that important political values necessarily conflict, that no choice among these can be defended as the only right choice, and that sacrifices in some of the things we care about are therefore inevitable. Millions of people in this extraordinarily prosperous country are without decent lives or prospects. They have no health insurance, no adequate shelter, no jobs. How often have you heard it said, in answer to the charge that we must do something about this, that we cannot do too much because equality conflicts with liberty? That if we were to raise taxes to the level necessary to address poverty in any serious way, then we would be invading liberty? Or when we raise our eyes and look abroad, and see that in many parts of the world democracy is a joke, and we say there may not be much that we can do about this, but perhaps there is something, or when we look at the policies of the Taliban that deny medical care to women, and we recoil in horror, and we ask whether economic sanctions might do something about that, how often are we told that different cultures have different values, that it is a form of imperialism for us to insist that only our values are the right ones and that different values are mistaken? That we have our way of organizing a society, and the Taliban and other fundamentalist societies have theirs, and that at the end of the day all we can say is that a single society cannot incorporate all values, that they have made their choices among these, and that we have made ours?

Moral Pluralism

107

The hedgehog need not be a tyrant—it’s a great mistake, as Thomas Nagel pointed out, to think that because value monism may serve as the banner of tyranny it must always do so. Nor, of course, does value pluralism inevitably lead either to selfishness or indifference. But there are dangers on both sides, and whether the danger of the hedgehog is greater than the danger of the fox, as Berlin thought, seems very much to depend on time and place. In the mid-Fifties, when he wrote his famous lecture, Stalinism was rampant and the corpse of fascism still stank. It may very well have seemed, then, that civilization had more to fear from the hedgehog. But in contemporary America, and in other prosperous Western democracies, that doesn’t seem so plain: the fox may be the more threatening beast. Perhaps there’s a pendulum that swings between these dangers. Danger is not our main story, however. We want to consider how far Berlin was right, not about the good effects of his value pluralism, but about that doctrine’s truth. I said that his is an original, powerful view, and I shall now try to explain why, not only because we must try to identify Berlin’s claims accurately, but because the difficulties in his view only emerge once it is separated from more familiar claims. Berlin did not just insist, as so many writers now do, on the anthropological platitude that different societies are organized around very different values, and have difficulty in understanding one another. Nor does he simply combine that platitude, as so many others have, with the further skeptical claim that it makes no sense to speak of “objective” values at all. It is all too common, in the so-called postmodernist age, for scholars to claim that all values— liberal or fundamentalist or “Asian” values—are just subjective reactions or social creations, so that to think of these values as true or false is a deep philosophical mistake. Berlin’s view is more complex and interesting. He believed that values are indeed objective, but also that there are irresolvable conflicts among the true values. He argued, that is, not just that people are in conflict about what the truth is, but that there is conflict in the truth about these matters. That is why he spoke, as I quoted him, of conflict within a single breast, and we can capture his view most accurately by putting it in the first person. Suppose we ourselves set out to imagine a life that had everything in it that an ideal life should have. Or to construct a political consti-

108

Moral Pluralism

tution that would respect and enforce every important political value. We would be doomed, Berlin says, to fail in either of those projects. The doom, he adds, is conceptual not contingent, and I must try to explain what he means by that further distinction, though I am not sure that I can show it to be as crisp as he thought. Obviously there are circumstances in which, for various accidental reasons, or because of injustice or wickedness, we cannot meet all our obligations to everyone. We might well not be able to rescue all the victims of a natural catastrophe before some die, for example. Churchill, in the exigencies of war, thought he had to sacrifice the citizens of Coventry, by not warning them of an impending air raid, in order to keep secret the fact that the Allies had broken a German secret code. If a nation has suffered from an unjust system of economic class, it might be necessary to limit liberty by abolishing private schools, at least for a generation, in order to help restore equality. These are cases in which, for different kinds of contingent or historical reasons, we cannot do all that we should. Our values conflict, Berlin insists, in a deeper way than that, which is why he says that the ideal of harmony is not just unobtainable but “incoherent” because securing or protecting one value necessarily involves abandoning or compromising another. Our values conflict, that is, even if we get all the breaks. His examples help to clarify this distinction. You may feel that a life given over to spontaneity—following the urges and impulses of the moment—would be a glorious life to lead. But you may also feel the demands of the very different value of prudence: you may feel that a life committed to forethought, particularly for the needs and interests of others, would be a splendid life. But if you felt these twin appeals, you would have to cheat on one of them. You could not organize a life that made spontaneity dominant and yet left adequate room for prudence, or vice versa. If you tried to bring these two values together in a single life the result would be a terrible mess: imagine the man who sets his wristwatch alarm to remind him when it is time for an hour of spontaneity. It won’t work, and this is not just a matter of historical accident. The two values cannot be combined because they are, in the nature of the case, at war with each other. We can easily find other examples that might have a greater purchase in

Moral Pluralism

109

your own life. Many people here, I expect, feel both the need for total absorption in some work or project and also a commitment to family responsibilities and pleasures that almost always pull in the opposite direction, and they know from experience how wrenching that conflict can sometimes be. Any choice someone in that position makes deprives his life, he thinks, of something essential to a good life, at least for him. That last idea—that a conflict in important values involves some genuine and important damage—is central to Berlin’s idea. He doesn’t just mean that we cannot have everything that we want—that we cannot cram all the adventures and treats that we might want into a single life. That, as he says, is a truism. He means that we cannot bring into a single life everything that we think it makes a life defective not to have. What is the political analogue of that kind of failure? Of course a political community cannot achieve all the economic or cultural success that its citizens dream of, and of course its policies must sometimes disappoint some citizens through policies that benefit others. But political values name distinct responsibilities that a community has to its citizens, responsibilities that it is not simply disappointing but wrong to ignore or violate. If we accept equality as a value, and we think that equality means that every citizen must have access to decent medical care, then we think that a prosperous community that allows some citizens to die for want of such care does them a grave wrong. If we accept liberty as a value, and we think that liberty is violated when rich people are taxed to provide more money for the poor, then we think that such taxes not only inconvenience the rich but wrong them. If we accept both equality and liberty, and think they have those implications, then we must think that a political community violates its responsibilities no matter what it does. It must choose, that is, not whether to wrong some group, but which group to wrong. That is the kind of conflict in political values that Berlin had in mind: the inevitability not of disappointment but of irredeemable moral stain. His claim is a positive one—that values of different kinds do conflict in that tragic way—and we must take care to distinguish that positive view from the different and much less troubling observation that we are sometimes uncertain what our values demand of us. Thoughtful people are often uncertain about important political issues, and sometimes swing be-

110

Moral Pluralism

tween opposite positions. We might begin thinking about the troubling issue of hate speech, for example, persuaded that any government that deprives anyone of free speech just because it disapproves of what he says, or because what he says is offensive to other people in the community, is acting illegitimately. And then we might listen a bit to what some people actually say—we might hear someone calling a black woman a nigger or telling some Jewish boy that Hitler was right and Jews should be gassed. And then we might have second thoughts: we might wonder whether free speech is really as important as we thought it was. Does a nation really compromise its legitimacy when it protects its most vulnerable citizens from that kind of attack? We might come first to one decision and then to the other: we might twist and turn and swing back and forth. Or we may collapse into indecision and find that, once we see the appeal of each side of the question, we simply cannot say, certainly with any confidence, what we think. Berlin’s claim has nothing to do with uncertainty, however, even that terminal kind of uncertainty. He claims, not that we often do not know what is the right decision, but that we often do know that no decision is right, which is a very different matter. So we must concentrate on the following question. When are we entitled, not simply to the negative idea that we do not know what it is right for us to do, but to the positive claim that we know that nothing that we do is right because, whatever we do, we do something wrong? The latter is an extremely ambitious claim: it purports to see to the bottom of a dilemma and see that there is no escape. Are we ever entitled to so ambitious a claim? That depends on how we conceive the source of our responsibilities. Imagine yourself in the position of Abraham holding a knife over the breast of his son, Isaac. Suppose you believe that you have an absolute religious duty to obey your God, no matter what, and also an absolute moral duty not to injure your own child, no matter what, and you conceive these as duties independent in their source. Your theology insists both that God’s authority in no way stems from the morality of his command and that morality’s authority in no way stems from God’s command. So long as you hold these convictions, you will be certain that you cannot avoid doing wrong. You are, as it were, subject to two sovereigns—God and mo-

Moral Pluralism

111

rality—and in the tragic difficulty that, at least as you understand the situation, the command of each counts for nothing in the eyes of the other. You must choose, and each choice is a final and terrible disloyalty. But is that really our situation in politics? I just said that we might be uncertain whether a government does wrong when it prohibits racist speech, or, on the contrary, it does wrong when it permits such speech. What further argument or reflection could replace this indecision with the positive conviction that government does wrong in either case? Our situation is very different from Abraham’s: we are not beholden to two independent sovereign powers one of whom commands freedom of speech and the other of whom commands prosecution for racial insult. On the contrary, we are drawn to each of the rival positions through arguments that, if we were finally to accept them as authoritative, would release us from the appeal of the other one. If we really believe that citizens have a right to speak out even in ways that offend certain other citizens, then it would be odd also to believe that certain citizens have a right not to be offended by what other citizens say. And vice versa. We reach a political conviction in cases like this one, that is, not by discovering what some deity or authoritative body has commanded, which may easily produce a deep conflict, but by reflecting on and refining our own sense of the needs and values in play, and it is mysterious how that process could produce the kind of conflict Berlin claims. It seems puzzling how we could be persuaded, at one and the same time, that citizens have a right that racial insults not be uttered and that citizens have a right to utter racial insults. But unless we can finally accept both of these claims, and at the same time, we cannot claim the positive view that we violate citizens’ rights whatever we do about racist speech. Berlin would reply, I believe, that my account of how we might become uncertain about racist speech misses an important and pertinent fact— that we come to particular political controversies, like that one, encumbered by prior commitments to two abstract political values—liberty and equality—and that these values can and do act in the fashion of independent and competing sovereigns whose dictates may conflict. Indeed, as you know, he thought that the conflict between liberty and equality was a paradigm of value conflict, and that is also, as I suggested earlier, the alleged

112

Moral Pluralism

conflict that seems the most troublesome and dangerous in contemporary politics. Do liberty and equality, now considered as abstract values, conflict in some way that explains why a political community might find itself not merely uncertain about what to do but certain that it must do wrong whatever it does? That depends on what we mean by liberty and by equality: it depends on how we conceive these abstract values. Berlin makes plain, in his famous essay about liberty, and in several other places as well, how he understands liberty. Liberty, he says, is freedom from the interference of others in doing whatever it is that you might wish to do. Now if that is how we understand liberty, then it’s immediately apparent, isn’t it, that liberty for the wolf is, as he says, death to the lamb. If that is what we mean by liberty, and we are committed to liberty so understood, then it is very plausible that this commitment will often conflict with other commitments, including even minimal egalitarian ones. But are we committed to liberty understood in that way? Here is a rival conception of liberty, which I introduce now just to show that our commitment to liberty is not automatically a commitment to liberty as Berlin understood it. We might say: liberty isn’t the freedom to do whatever you might want to do; it’s freedom to do whatever you like so long as you respect the moral rights, properly understood, of others. It’s freedom to spend your own rightful resources or deal with your own rightful property in whatever way seems best to you. But so understood your liberty doesn’t include freedom to take over the resources of someone else, or injure him in ways you have no right to do. The lamb would be happier with that account of liberty, even though the wolf might not be. In any case, it is far from obvious that liberty understood in this different way would produce an inevitable conflict with equality. On the contrary, it seems unlikely that it would: if higher taxes are necessary to provide poorer citizens with what equality entitles them to have, then taxation for that purpose cannot count as an invasion of the liberty of the rich, because the property taken from them in taxes is not rightfully theirs. You might object—I expect that many of you will object—that I have begged the question against Berlin by defining liberty so that conflict is excluded from the start. But are you assuming that the only

Moral Pluralism

113

successful account of liberty is one that makes liberty independent of other values? One that allows us to decide what liberty requires, and when it has been sacrificed, without considering what rights—to equality or anything else—other people have? That begs the question in the opposite way—it assumes the picture of values as rival and independent sovereigns that makes conflict inevitable. In fact, we might say, the large question of value pluralism and conflict in politics that Berlin introduced just is the question of whether our political values are independent of one another in the way his definition of liberty insists, or whether they are interdependent in the way the rival conception of liberty I sketched suggests, and that is a question, as I shall now argue, not of dictionary definition or empirical discovery but of substantive moral and political philosophy. Berlin warned us, in the passage I quoted at the beginning, against such rival conceptions of liberty. If people come forward with ideas of liberty that don’t produce conflicts, he said, we must tell them that the values they offer are not our values. But what does that mean? How do we decide that Berlin’s definition of liberty, which produces conflict, is our notion of liberty and that rival accounts of liberty are foreign to us? He’s not, of course, making a semantic point: he doesn’t mean that the dictionary is decisive for his account. Indeed he recognizes that people use the words that name political concepts in many different ways. Nor can we conduct any analog to a laboratory experiment or investigation to see what liberty really is, the way we can conduct tests to decide what a lion is really made of. We can’t conduct a DNA analysis of liberty. So how do we test Berlin’s claim about what conception of liberty is ours, and which other conceptions are foreign? Can we look to history? I suspect that some readers are already beginning to think that I am not giving history its due. I agree that the history of ideas is often crucial, and of course I agree that it was of the first importance for Berlin. But we must go beyond simply stating that history is crucial and try to see why and how it is crucial. I don’t quite see how history can be decisive at this point in our argument. History may of course teach us that many societies whose reigning ideology denied any conflict among important values ended in some form of disaster, and that should no doubt put us on our guard. But history can’t, it seems to me, help us further. We’re trying to de-

114

Moral Pluralism

cide how better to understand the value of liberty—a value to which we take ourselves to be committed—in order to see whether we do wrong when, for example, we tax the well-off to redistribute to the poor. I see no substitute for treating that, at least in the main, as a moral rather than an historical issue. How should we proceed? I said, you recall, that for Berlin a conflict in fundamental values, either personal or political, is not merely an inconvenience or disappointment but a kind of tragedy. When we face such a conflict, in his view, we inevitably suffer or commit some injury: we deprive our life of something we think it wounds a life not to have, or we wrong someone by denying him what it wrongs him not to have. We should begin there. We need an account of our political values that shows us why, in the case of each of our important values, that is so. Our leading political values—liberty, equality, democracy, justice, and the rest—are general ideals we agree in endorsing in the abstract. That abstract agreement is important: we agree that it is essential that citizens not be coerced by government in offensive ways, that an economic structure treat people with equal concern, that the people govern themselves, and so forth. When we try to make these very abstract values more concrete, by deciding what forms of coercion are offensive, what distribution of resources does treat people with equal concern, what form of self-government is possible, and so forth, we must respect and preserve that first understanding. We must formulate more precise conceptions of our values so as to show, more precisely, what the value we identified in the abstract really is. We need an account that shows us what is good about liberty or equality or democracy, so that we can see why any compromise of these values is not merely inconvenient but bad. Of course we will disagree at this point: we will each defend somewhat different, and perhaps very different, conceptions of liberty, equality, and the rest. But it is crucial that we each defend conceptions that, for us, carry the abstract value forward into the controversial conception, so as to make plain why what we regard as a compromise of fundamental values is, just in itself, something grave or at least bad. We can therefore test a proposed conception of liberty—or of any other value—in the following way. We should ask whether the various actions that the proposed conception defines as violations of liberty are really bad

Moral Pluralism

115

or wrong—are really breaches of some special responsibility for which a state should feel remorse even when those breaches are necessary in order not to violate some supposed competing value. If not—if a state does not wrong any citizen when, according to the proposed definition, it invades his liberty—then the proposed conception of liberty is inadequate. It declares a violation when a violation is no wrong, and it therefore does not show us what the special importance of liberty is. Does Berlin’s account of liberty pass that test? Suppose I want to murder my critics. The law will stop me from doing that, and the law will therefore, on Berlin’s account, compromise my liberty. Of course, everyone agrees that I must be stopped: those who defend Berlin’s definition say that although my liberty has been invaded, the invasion is justified in this case, because the wrong done to me is necessary to prevent a greater wrong to others. In this case, they say, liberty is in conflict with other values and those other values must prevail. But I am asking whether the account of liberty that produces that alleged conflict is a successful account: if nothing wrong has taken place when I am prevented from killing my critics, then we have no reason for adopting a conception of liberty that describes the event as one in which liberty has been sacrificed. We are not, to repeat, required to describe it that way in virtue of what the word “liberty” means, or in virtue of any scientific discovery about the composition of liberty. A conception of liberty is an interpretive theory that aims to show us why it is bad when liberty is denied, and a conception of liberty is therefore unsuccessful when it forces us to describe some event as an invasion of liberty when nothing bad has happened. So I ask again: is there anything even pro tanto wrong in forbidding me to kill my critics? Of course, it would be better if no one wanted to injure anyone else or anyone else’s property: it would be better if the criminal law was unnecessary. That is not the question, however. Given that some people do want to kill on some occasions, is any wrong done to them by preventing them from doing so? Do we have any reason to apologize to the wolf who is denied his leg of lamb? Certain philosophers would answer that question: yes. Something important is lost, they say, whenever people of extraordinary spirit and ambition are thwarted by the laws of moral pygmies. I’m not asking whether anyone could think that. I’m asking what

116

Moral Pluralism

you think. And if you, like me, think that nothing wrong is done through such laws, then you will have that reason for rejecting Berlin’s account of liberty. If his view that a conflict between liberty and equality is inevitable depends on that account of liberty, then you will have that reason for rejecting that view as well. Of course I have not shown that conflicts between liberty and equality are not inevitable. Perhaps there is a more subtle account of liberty than Berlin’s, which is not open to the objection I made, but which would still guarantee conflicts between liberty and equality. My point has been a limited one. I have tried, first, to clarify Berlin’s important thesis about value pluralism, to show its originality, interest, and ambition, and, second, to show how difficult it is to sustain that ambitious thesis. Berlin said that the value conflicts he described were all about us, and evident to all except the immature. I do not think that he sustained that very broad claim; indeed, as I have just argued, I do not think he sustained it even in the case he took as paradigm: the supposed conflict between liberty and equality. That does not mean the defeat of value pluralism. But it does suggest, I believe, that the argument necessary to defend pluralism must be a very long and complex one. That argument must show, in the case of each of the values it takes to be in some kind of conceptual conflict with one another, why the understanding of that value that produces the conflict is the most appropriate one. Nothing is easier than composing definitions of liberty, equality, democracy, community, and justice that conflict with one another. But not much, in philosophy, is harder than showing why these are the definitions that we should accept. There is no short cut to that demonstration. Perhaps, after all, the most attractive conceptions of the leading liberal values do hang together in the right way. We haven’t yet been given reason to abandon that hope.

chapter five

Originalism and Fidelity

In the first state of the union address of his second term, President George W. Bush promised to appoint judges to the federal courts and the Supreme Court who adhered to the intent of the Constitution’s framers, not their own personal convictions, and he cited the 1857 Dred Scott decision, in which the Supreme Court upheld slavery, as an example of the kind of decision the judges he would appoint would avoid. Bush is not a constitutional scholar and his mistake was evident: Dred Scott was an example of justices not ignoring but enforcing the framers’ intention because the original constitution contemplated slavery. But his meaning was clear enough: he meant he would not appoint the kind of judges who voted in the majority in Roe v. Wade, the 1973 decision in which the Supreme Court held that states could not make early abortion a crime. Indeed, many commentators thought he was signaling his intention to appoint judges who would overrule that decision as the Supreme Court rejected Dred Scott after the Civil War. He invited his audience to assume that that is what keeping faith with the Constitution would mean. That appeal to the idea of fidelity was, however, an even more profound mistake. Those the president had in mind as good judges are actually judges for whom fidelity to the Constitution actually counts for little. And those that he would count as bad judges are, in my view, the true heroes of fidelity. In any case, that is the argument of this chapter. I must start with a distinction, however: between fidelity to the Constitution’s text and fidelity to past constitutional practice, including past ju-

118

Originalism and Fidelity

dicial decisions interpreting and applying the Constitution. Proper constitutional interpretation takes both text and past practice as its object: lawyers and judges faced with a contemporary constitutional issue must try to construct a coherent, principled, and persuasive interpretation of the text of particular clauses, the structure of the Constitution as a whole, and our history under the Constitution—an interpretation that both unifies these distinct sources, so far as this is possible, and directs future adjudication. They must seek, that is, constitutional integrity. Fidelity to the Constitution’s text does not exhaust constitutional interpretation, and on some occasions overall constitutional integrity might require a result that could not be justified by, and might even contradict, the best interpretation of the constitutional text considered apart from the history of its enforcement. But textual interpretation is nevertheless an essential part of any broader program of constitutional interpretation because what those who made the Constitution actually said is always at least an important ingredient in any genuinely interpretive constitutional argument. So I shall concentrate on textual interpretation here. That seems appropriate, because constitutional lawyers often think that fidelity to the Constitution means fidelity to its text. That is the kind of fidelity President Bush apparently had in mind. It is the kind of fidelity demanded by selfstyled constitutional “originalists,” like Supreme Court Justice Antonin Scalia, and rejected by critics of originalism like Professor Laurence Tribe. I shall argue that even if we concentrate exclusively on textual fidelity, we reach radically different conclusions from those that Bush, Scalia, and other “originalists” expect.1 Indeed, textual fidelity argues so strongly in favor of a broad judicial responsibility to hold legislation to direct moral standards that a great many constitutional scholars, including those who call most loudly for “originalism,” actually argue against textual fidelity as a constitutional standard. They rely on other standards and values as substitutes for fidelity. They rarely put it that way. But if you listen carefully, you will find that substitution for fidelity is a hidden subtext. Some scholars will argue that we should try to discover not what those who wrote or ratified the Constitution and its various amendments meant to say, but what they expected or hoped would be the consequence of their saying what they did, which is

Originalism and Fidelity

119

a very different matter. Others will argue that we should ignore the text itself in favor of how most people understood its import over most of our history: they argue, for example, that the fact that many states have made homosexual sodomy a crime shows that the Constitution does not prohibit that piece of injustice. These are both ways of ignoring the text of the Constitution. Why do distinguished scholars work so strenuously to avoid the Constitution? I attempt an answer to that question later in this chapter, by identifying various grounds people might think they have for ignoring constitutional fidelity. But I should now say that I do not mean to assume that there are no such grounds. It is true that most citizens expect the Supreme Court to cite the Constitution to justify its constitutional decisions. But different departments of our government make very important and consequential decisions for which no argument of fidelity to any text or tradition is required. We send men and women to war, adopt foreign policies or monetary strategies, and shoot missiles at Mars, and we justify these decisions on the ground that good will come of them in the future—that we will be more secure or more prosperous or more at home in our universe. We shouldn’t rule out, at the start, that such forward-looking justifications would be more appropriate in constitutional adjudication than the backward-looking argument of textual fidelity, particularly since no less distinguished a judge than Richard Posner has argued with great passion that they are more appropriate. Nor should we assume that the different kind of backward-looking justification some of the scholars I mentioned endorse, which appeal to history apart from constitutional text, would not be more appropriate. Perhaps it is justifiable, at least in some circumstances, to disregard fidelity. Before considering that question, however, I must first justify my initial claims about what fidelity to our Constitution’s text means. I must be careful to distinguish the question my claims answer from a different question with which it is often confused: the institutional question of what bodies—courts, legislatures, or the people acting through referenda—should be assigned the final responsibility to decide what fidelity requires in particular cases. It is perfectly possible for a nation whose writ-

120

Originalism and Fidelity

ten constitution limits the power of legislatures to assign the final responsibility of interpreting that constitution to some institution other than a court; this could include the legislature itself. My question concerns an issue prior to that of institutional design. No matter what or who is given final interpretive responsibility, what does our Constitution really mean? We have a constitutional text. We do not disagree about which inscriptions comprise that text; nobody argues about which series of letters and spaces make it up. Of course, identifying a canonical series of letters and spaces is only the beginning of interpretation. For there remains the problem of what any particular portion of that series means. Hamlet said to his sometime friends, “I know the difference between a hawk and a handsaw.” The question arises—it arises for somebody playing the role, for example—whether Hamlet was using the word “hawk” that designates a kind of a bird, or the different word that designates a Renaissance tool. Milton spoke, in Paradise Lost, of Satan’s “gay hordes.” Was Milton reporting that Satan’s disciples were brightly dressed or that they were homosexual? The Constitution says that a president must be at least “thirty-five years of age.” Does that mean chronological age, or does it mean (what would be alarming to several contemporary politicians) emotional age instead? The Eighth Amendment of the Constitution forbids “cruel” and unusual punishment. Does that mean punishments that the authors thought were cruel or (what probably comes to the same thing) punishments that were judged cruel by the popular opinion of their day? Or does it mean punishments that are in fact—according to the correct standards for deciding such matters—cruel? The Fourteenth Amendment says that no state shall deny any person “equal protection of the laws.” Does that mean that no state may deny anyone the equality of treatment that most states have accorded over our history? Or does it mean that no state may perpetuate any distinctions that contradict genuine equal citizenship, whether Americans have understood that contradiction before or not? We must begin, in my view, by asking what—on the best evidence available—the authors of the text in question intended to say. That is an exercise in what I have called constructive interpretation.2 It does not mean peeking inside the skulls of people dead for centuries. It means trying to make the best sense we can of a historical event—someone speaking or writing in a particular way on a particular occasion. If we apply that stan-

Originalism and Fidelity

121

dard to Hamlet, it’s plain that we must read his claim as referring not to a bird, which would make the claim an extremely silly one, but to a Renaissance tool. Hamlet assured his treacherous companions that he knew the difference between kinds of tools and knew which kind he was dealing with in them. In the case of Satan’s gay hordes, there’s a decisive reason for thinking that Milton meant to describe them as showy, not homosexual, which is that the use of “gay” to mean homosexual postdated Milton by centuries. In my view, we have as easy a job in answering the question of what the people who wrote that a president must be thirty-five years of age meant to say. It would have been silly of them to have conditioned eligibility for the presidency on a property so inherently vague and controversial as that of emotional age, and there is no evidence of any such intention. We can make sense of their saying what they said only by supposing them to have meant chronological age. When we come to the word “cruel” in the Eighth Amendment, the equal protection language of the Fourteenth, the freedom of speech language of the First, and the due process language of the Fifth and the Fourteenth, however, we have more difficult problems of translation. We have to choose between an abstract, principled, moral reading on the one hand— that the authors meant to prohibit punishments that are in fact cruel as well as unusual, or meant to prohibit whatever discriminations are in fact inconsistent with equal citizenship; and a concrete, dated reading on the other—that they meant to say that punishments widely thought cruel as well as unusual at the time they spoke, or discriminations then generally understood to reflect unfair distinctions, are prohibited.3 If the correct interpretation is the abstract one, then judges attempting to keep faith with the text today must sometimes ask themselves whether punishments the framers would not themselves have considered cruel—capital punishment, for example—nevertheless are cruel, and whether discriminations the framers themselves thought consistent with equal citizenship—school segregation, for example—are nevertheless a denial of equal protection of the laws. If the correct interpretation is the dated one, these questions would be out of place, at least as part of an exercise in textual fidelity, because the only questions a dated understanding would pose is the question of what the framers or their audience thought. If we are trying to make best sense of the framers speaking as they did

122

Originalism and Fidelity

in the context in which they spoke, we should conclude that they intended to lay down abstract, not dated, commands and prohibitions. The framers were careful statesmen who knew how to use the language they spoke. They presumably meant to say what people who use the words they used would normally mean to say—they used abstract language because they intended to state abstract principles. They made a constitution out of abstract moral principles, not coded references to their own opinions (or those of their contemporaries) about the best way to apply those principles. But that answer to the question of how the apparently abstract rightsbearing constitutional provisions should be understood makes the task of adjudicating contemporary constitutional disputes much more difficult than it would be if the concrete, dated understanding were the correct one. If the standards that we—citizens, legislators, judges—must try to apply out of fidelity to the text of the abstract provisions are abstract moral standards, the questions we must ask and the judgments we must make must be moral ones. We must ask: What is really cruel? What does equal citizenship really require? What legislation is consistent with due process of law, given that legal integrity is of the essence of law’s process and that integrity requires that the liberties our culture recognizes in broad principle— freedom of conscience, for example—must be respected in individual legislative decisions about, for example, freedom of choice in dying?4 These are difficult questions. Citizens, lawyers, and judges should not try to answer them on a clean slate, ignoring the answers that others, particularly judges, have given to them in the past. As I said, any strategy of constitutional argument that aims at overall constitutional integrity must search for answers that mesh well enough with our practices and traditions—that find enough foothold in our continuing history as well as in the Constitution’s text—so that those answers can plausibly be taken to describe our commitments as a nation. If I were trying to answer the question of what equal citizenship means as a philosophical exercise, for example, I would insist that citizens are not treated as equals by their political community unless that community guarantees them at least a decent minimum standard of housing, nutrition, and medical care. But if the Supreme Court were suddenly to adopt that view and to announce that

Originalism and Fidelity

123

states have a constitutional duty to provide universal health care, it would have made a legal mistake, because it would be attempting to graft into our constitutional system something that (in my view) doesn’t fit at all. Very often, however, controversial decisions that seem novel do satisfy that test of fit. When the Supreme Court decided in 1954 that official segregation by race was illegal despite generations of practice to the contrary, it did not just announce an academic political truth. It called attention to general standards of equality that were firmly fixed in our history though selectively ignored in our practice, standards that condemned arbitrary discriminations serving no legitimate governmental purpose. The Court was able convincingly to argue that the practice of racial segregation was inconsistent with a broader reading of principle. I think the same could be said for the Court’s abortion decision, Roe v. Wade. In that case, the Court had to ask itself whether the idea that certain basic liberties are in principle immune from government regulation, an idea that is embedded by precedent in the Fourteenth Amendment, guarantees a more concrete right to an early abortion. I offer these examples to make plain that though I think that the moral judgment required to apply the abstract moral principles of the Constitution is constricted by history and precedent, in virtue of the commands of legal integrity, it is plainly not preempted by that history. Fresh questions of moral principle—for example, is a right to abortion sufficiently fundamental to fall within the basic liberties that define due process of law?— inevitably remain. Why isn’t it more faithful to the Constitution to assign such questions not to contemporary judges but to those who made and ratified that document, at least so far as we can discover or surmise the answers they would give? I have already insisted that we must look to the authors’ semantic intentions to discover what the clauses of the Constitution mean. Why doesn’t it follow that we should also defer to their political intentions—their assumptions and expectations as to how the clauses they wrote would be applied? If the authors of the Equal Protection Clause didn’t think that racially segregated schools denied equality of citizenship, why doesn’t that end the question of what fidelity to that clause requires? In fact, however, taking that further step is a serious intellectual confusion and constitutional mistake. It is a fallacy to infer from the fact that

124

Originalism and Fidelity

the semantic intentions of historical statesmen inevitably fix what the document they made says that keeping faith with what they said means enforcing the document as they hoped or expected or assumed it would be enforced. Imagine that you are the owner of a large corporation that has a vacancy in one of its departments. You call in your manager and say to her, “Please fill this vacancy with the best candidate available. By the way,” you add, without winking or nudging, “you should know that my son is a candidate for this position.” Assume you are honestly convinced that your son is the best qualified candidate. Assume, moreover, that you wouldn’t have given the manager those instructions unless you were sure that it was obvious to everyone, including her, that your son was the best candidate. Assume, finally, that your manager knows all of this: she knows that if the choice were yours you would conscientiously appoint your own son as the most qualified candidate. Nevertheless, you didn’t tell her to hire your son. You told her to hire the best candidate. And if, in her judgment, your son is not the best candidate but someone else is, then she would be obeying your instructions by hiring that other candidate and disobeying your instructions by hiring the candidate you intended and expected would get the job. You might—I hope you wouldn’t—fire her if she obeyed your instructions in that way. But you couldn’t deny that she had been faithful to your instructions and that she would not have been faithful had she deferred to your view about the best candidate instead of her own. An agent is unfaithful to an instruction unless she aims to do what the instruction, properly interpreted, directs. If the instruction sets out an abstract standard, she must decide what meets that standard, which is of course a different question from the question of what some person—any person—thinks meets the standard. Contemporary legislators and judges are subject to the same rigorous demand. The question of fidelity dominated a conference some years ago at Princeton, at which Justice Scalia delivered two Tanner Foundation Lectures, four commentators replied to those lectures, and Scalia responded to those replies. The proceedings were published (after some participants edited and expanded their original remarks) in a volume called A Matter of Interpretation.5 The comments of two of the participants, Scalia and

Originalism and Fidelity

125

Tribe, illustrate both the difficulty and the importance of the distinction that I have just been emphasizing—the distinction between semantic intention (what the framers meant to say) and political or expectation intention (what they expected would be the consequence of their saying it). In my own remarks at the conference, I used that distinction to contrast two forms of what Scalia called “originalism”: semantic originalism, which takes what the legislators meant collectively to say as decisive of constitutional meaning; and expectation originalism, which makes decisive what they expected to accomplish in saying what they did. I said that in his first Tanner lecture Scalia had subscribed to the former version of originalism, but that in his second lecture, in his remarks about constitutional interpretation, he had relied on expectation intention instead—as indeed, I said, he has done in his career on the Supreme Court. In his published reply to my comments, Scalia accepted the distinction and declared himself a semantic rather than an expectation originalist. He denied the inconsistency I pressed, but in a way that confirms my suspicion that his constitutional practice has abandoned the fidelity he preaches. I had used one of the examples I have used here—the interpretation of “cruel” in the Eighth Amendment—to illustrate the difference between semantic and expectation originalism. Scalia argued in his lecture that the fact that the framers of that amendment contemplated the possibility of capital punishment elsewhere in the Bill of Rights—by declaring, for example, in the Fifth Amendment that “life” could not be taken without due process of law—was clear proof that they did not intend to prohibit it in the Eighth Amendment. I said that if Scalia were a true semantic originalist, he would be assuming, in that argument, something that seems very odd: that the framers intended to say, by using the words “cruel and unusual,” that punishments generally thought cruel at the time they spoke were to be prohibited—that is, that they would have expressed themselves more clearly if they had used the phrase “punishments widely regarded as cruel and unusual at the date of this enactment” in place of the misleading language they actually used. Scalia replied that his argument about capital punishment presupposes no such thing, and he called my suggestion that it does a “caricature” of his view. Only a very few lines later, however, he stated his own view: that

126

Originalism and Fidelity

what the Eighth Amendment enacts is “the existing society’s assessment of what is cruel. It means . . . what we [that is to say, the framers and their contemporaries] consider cruel today.” He then draws the appropriate conclusion: “On this analysis, it is entirely clear that capital punishment, which was widely in use in 1791, does not violate the abstract moral principle of the Eighth Amendment.”6 Unless these latter passages have a deep meaning I cannot fathom, they endorse exactly the view that Scalia, only a few lines earlier, had rejected as caricature. So Scalia, confronted with an account of what his constitutional position would presuppose if he were a semantic originalist, rejects that account as preposterous. But when, immediately after, he tries to state a view that is both faithful to the Constitution’s text and consistent with his own constitutional attitudes, he is forced to state the very view he had just rejected. His theoretical stance is therefore contradictory. It cannot inhibit his constitutional adjudication, which in fact has little to do with keeping faith with the Constitution. Could there be a more dramatic illustration of the difficulty under which constitutional lawyers in Scalia’s mode labor? They must profess fidelity but feel bound in practice to disown it. Though Tribe is the most prominent, unabashed, and uncompromising liberal among well-known constitutional practitioners, his own remarks show that he faces a dilemma surprisingly similar to Scalia’s, though it will take me longer to identify it. Tribe begins his much expanded published comments by declaring that he disagrees with both Scalia and myself: “I do not agree with either Professor Dworkin or Justice Scalia,” he says, “that one can ‘discover’ . . . empirical facts about what a finite set of actors at particular moments in our past meant to be saying.” He criticizes me in particular for thinking that constitutional interpretation depends on retrieving such empirical facts, and also for thinking that novel constitutional interpretations “actually represent nothing new at all but . . . are all merely inferences that emerge by a straightforward . . . process of reasoning our way to the right answers to questions of principle that we can be sure the Constitution’s authors or ratifiers actually put to us ages ago.” (I have never held or defended anything like that view.) I take these remarks to mean that Tribe rejects any requirement of fidelity to the Constitution’s text when different interpretations of that text

Originalism and Fidelity

127

are possible. It is true that his remarks can be read in a much more innocuous way, but if they are, then the disagreement he announces between himself and both Scalia and me would disappear. We can read them, for example, as objecting only to the confidence we seem to show in the particular judgments about semantic intention that we make. He sprinkles epistemic intensifiers over the views he attributes to Scalia and me. He says, for example, that though he himself holds many views about constitutional interpretation, “I do not claim these to be rigorously demonstrable conclusions, or confuse them with universally held views.” He elsewhere describes, as the “Dworkinian reading” of the First Amendment, that it was “self-evidently intended to enact a broad moral principle,” and he denies that that claim of self-evidence “is demonstrable either.” He spurns what he identifies as Scalia’s and my “no doubt sincere (but nevertheless misguided) certitudes.”7 Of course, neither Scalia nor I would accept these reports of our opinions. It may be useful polemically to decorate opponents’ views with such phrases, but that degrades the argument because no one would defend the views so decorated. Constitutional interpretation is not mathematics, and no one but a fool would think his own constitutional judgments beyond any conceivable challenge. We argue for our constitutional interpretations by offering the best and most honest case we can for their superiority to rival interpretations, knowing that others will inevitably reject our arguments and that we cannot appeal to shared principles of either political morality or constitutional method to demonstrate that we are right.8 We can also read Tribe’s comments as innocent by supposing that all he means to reject, in Scalia’s views and mine, is bad metaphysics of interpretation. When he denies that it is possible to discover “empirical” facts about what people meant in the past, he may mean just that it is not possible to establish direct mental links to some dead person’s or persons’ phenomenology. But of course that is not the enterprise that either Scalia or I have in mind. As I said earlier, people translate what other people have said—across the dining table as well as across the centuries—by a process of constructive interpretation that aims not at intracranial peeks but at making the best sense possible of their speech and other behavior.9 That is a normative, not an “empirical,” process. It is a particularly complex pro-

128

Originalism and Fidelity

cess when the object of interpretation is a political act and we cannot settle on a translation except by attributing political principles or purposes to a group whose members may have been politically divided. But the essence of constructive interpretation remains the same in that complex case as in individual ones: we must find a translation that better explains—and in the political case better justifies—the speech event than any rival translation can. That was my ambition in arguing that we make much more sense of the great abstract constitutional clauses by supposing them to state great abstract principles. It is also Scalia’s ambition in arguing, so far as he does, for his contrary view, as everything he says in its favor shows. The pertinent question is therefore not whether we can “demonstrate,” or establish “certitudes,” or “self-evident” propositions about what the framers intended to say, or whether we can fish mental states from history and subject them to a merely “empirical” examination. It is whether, in spite of the fact that we cannot do any of that, we must nevertheless decide which view of what the framers said is, on balance, the better view, even though it is controversial. If Tribe really does disagree with Scalia and me, he must think that since it is at best uncertain what the framers intended to say, we should and need not reach any conclusion about that matter at all, at least in the case of the great abstract clauses, but conduct constitutional interpretation in some way that does not include such conclusions—through a direct appeal to our own political morality, perhaps, or to what we take to be the dominant political morality of our time. But though that is a coherent position, and appeals, as I said, to many commentators, it seems flatly to contradict other claims that Tribe makes with equal fervor. Elsewhere in his comments on Scalia’s Tanner Lectures, for example, Tribe endorses a very strong form of textual fidelity. “I nonetheless share with Justice Scalia the belief that the Constitution’s written text has primacy and must be deemed the ultimate point of departure, that nothing irreconcilable with the text can properly be considered part of the Constitution; and that some parts of the Constitution cannot plausibly be open to significantly different interpretations.”10 That is a very strong statement of textual fidelity—stronger that I would myself endorse, because, as I said, precedent and practice over time can in principle supersede even so basic a piece of interpretive data as the Constitution’s text

Originalism and Fidelity

129

when no way of reconciling both text and practice in an overall constructive interpretation can be found. I agree with Tribe, however, that the text must have a very important role: we must aim at a set of constitutional principles that we can defend as consistent with the most plausible interpretation we have of what the text itself says, and be very reluctant to settle for anything else. But a text is not just a series of letters and spaces: it consists of propositions, and we cannot give a text “primacy”—or, indeed, any place at all— without a semantics, that is, an interpretation that specifies what (if anything) the letters and spaces mean. Until we have interpreted the letters and spaces in that way, we can have no idea what is or is not “irreconcilable” with the text, or whether the text is vague or ambiguous, or whether it can “plausibly be open to significantly different interpretations.” So Tribe’s remarks about the text’s primacy presuppose a semantic strategy. What is it? The natural strategy is the one I just described: we decide what propositions a text contains by assigning semantic intentions to those who made the text, and we do this by attempting to make the best sense we can of what they did when they did it. But if Tribe really means to disagree with Scalia and me, he rejects that strategy. What are the alternatives? Someone might say that we should interpret an old text like the Constitution not by attempting to understand what those who made it said but by assuming that it means what it has been taken to mean in the past. There are regression problems in that strategy, but we need not explore them because the strategy is not in any case a way of giving the text the independent primacy Tribe intends. On the contrary, it denies that the text has any independent role in contemporary interpretation. Consider, therefore, this different possibility (perhaps suggested by Tribe’s remark about “plausible” alternatives). Someone might say that textual primacy requires only that no constitutional interpretation be deemed acceptable unless the strings of characters and spaces that make up the text could be used, in some circumstance or other, to express the proposition it deploys. That extremely weak constraint would permit judges to declare a broad “moral” constitutional right of free speech whether or not the framers intended to lay down an abstract moral principle in the First Amend-

130

Originalism and Fidelity

ment, because the string of letters and spaces they used could have been used to declare a broad principle. (On a parallel view of the primacy of text in poetry, we could interpret Paradise Lost as homophobic, because the letters that make up “gay” are now used to make a word that refers to homosexuals.) But that odd interpretive strategy is arbitrary and unmotivated in legal or political principle. Why should it matter whether the inscriptions found in the Constitution could or could not be used to state a particular proposition, unless we think that that is the proposition that they were used to state? Would it not be equally sensible to say instead that the text must be primary in the anagram sense—that it can be understood as forbidding anything that the letters in it can be rearranged to forbid? It is thus extremely difficult to reconcile Tribe’s vows of textual fidelity with any interpretive strategy except the natural one. But that is the one he must reject if he really does disagree with Scalia and me. There is more trouble still. In many other passages Tribe plainly assumes that that natural strategy is the correct one and that it does make sense to suppose that the framers intended to say one thing rather than another when the words they used might be used to say either. He has added to his own expanded text, for example, the argument I made at the conference and repeat here: that Scalia’s reliance on other parts of the Bill of Rights as evidence that the Eighth Amendment does not forbid capital punishment is inconsistent with his professed semantic originalism.11 But my charge of inconsistency absolutely depends on a claim about what the framers meant to say in the Eighth Amendment. If they meant to set out the proposition I said Scalia assumed they did—that punishments unusual and cruel according to the general opinion of the day are forbidden—then Scalia’s textual argument would not be inconsistent; on the contrary, it would be persuasive. We can declare it inconsistent only if we can be confident, as Tribe in this context certainly appears to be, that the framers did not mean to set out that dated provision, but an abstract one instead. In adopting the argument, he relies on exactly the kind of judgment that he criticizes Scalia and me for supposing we can make. Even these passages do not exhaust the mystery. Tribe is a skilled litigator, and he does not lightly set aside arguments in favor of his positions.

Originalism and Fidelity

131

He says he agrees that the First Amendment states a principle of political morality. Why then does he reject the argument that, on the best interpretation, it was intended to express such a principle? Why does he claim a disagreement with Scalia and me that he must struggle to make genuine, that makes much of the rest of what he says incoherent if he succeeds, and that forfeits apparently strong arguments for many of his constitutional convictions? The answer is strongly suggested in the beginning of another passage: he says he wants to prevent constitutional interpretation “from degenerating into the imposition of one’s personal preferences or values under the guise of constitutional exegesis. . . .”12 Though, as I said, judges who interpret an abstract moral constitution do not simply put moral issues to their own conscience, that is certainly part of what they must do. Tribe wishes he could disavow any reliance on what the framers meant to enact because he believes, as I do, that on the most persuasive view they meant to enact a Constitution that left faithful judges and lawyers no choice but to do what he worries is illegitimate. But he has no way out of that conundrum. Notice how he finishes the sentence whose beginning I just quoted: “. . . one must concede how difficult the task is; avoid all pretense that it can be reduced to a passive process of discovering rather than constructing an interpretation, and replace such pretense with a forthright account, incomplete and inconclusive though it might be, of why one deems his or her proposed construction of the text to be worthy of acceptance, in light of the Constitution as a whole and the history of its interpretation.”13 The end of the sentence, in other words, in no way mitigates the worry of the beginning, but only confirms it. Even judges who look to “the Constitution as a whole and the history of its interpretation,” as integrity requires them to do, must nevertheless deploy their own convictions of political morality in arguing why their interpretations are more “worthy of acceptance” than other constructions that would also pass that test, and though “forthright” candor is of course a virtue, it does not alter what one is candid about. Tribe’s difficulties, though parallel to Scalia’s, run in the opposite direction. Scalia wants to be seen to embrace fidelity, but he ends by rejecting it. Tribe wants to reject fidelity, but ends by embracing it. Whose ending is

132

Originalism and Fidelity

the right one? That is the question I postponed and must now take up. Is constitutional fidelity, properly understood, a political virtue or a political vice? Even a fidelity to the abstract Constitution that is disciplined by integrity requires judges, lawyers, legislators, and others who interpret the Constitution to make fresh moral judgments about issues that deeply divide citizens, such as abortion, assisted suicide, and racial justice. Any official’s opinion on these issues is certain to be not only controversial but hated by many. Perhaps our judges would do better to set fidelity aside. Perhaps they would serve us better if they aimed not to enforce the Constitution we have but to make up a better, or in any case, a different one. Perhaps if Bush really meant the opposite of what he said, he was right. But what grounds could we offer for setting fidelity aside? What could trump fidelity? I can think of three virtues that might be thought more important in the constitutional context. First, in some circumstances fidelity might be trumped by justice. A political society might find itself saddled with a constitutional settlement that many of its members now deem very unjust, and in some circumstances judges of that community might properly decide simply to ignore that constitution. They might lie by claiming publicly that their decisions were dictated by fidelity, though they knew, and perhaps privately conceded, that the opposite was true. Or they might (if their situation allowed) simply declare that they regarded the old constitution, or at least part of it, as no longer binding on them. Many people now think, for example, that the pre–Civil War judges who were asked to declare the Fugitive Slave Act unconstitutional should have done so in spite of the fact that (in these critics’ opinion) fidelity demanded the contrary result. We can easily imagine such a judge telling himself that though the Constitution did not condemn laws requiring citizens to return escaped slaves, such laws were nevertheless too monstrous to be enforced. I mention justice as a possible trump over the demands of fidelity, however, only to set it aside. For the supposed problem we have identified is not that fidelity requires judges to uphold laws they think immoral. It is close to the opposite: that since the Constitution contains abstract moral principles, fidelity gives judges too much leeway to condemn laws that

Originalism and Fidelity

133

seem unjust to them though they have been endorsed by a properly elected legislature. A second possible trump over fidelity is much more in point now. It is democracy itself. Fidelity to a moral constitution does not entail that judges should be the final arbiters of what that constitution requires in concrete controversies. Still, in our political arrangement, federal judges are the final authorities of constitutional law, and many scholars, judges, and citizens believe that is undemocratic for judges to make the kinds of moral judgment that true fidelity would require, because in a genuine democracy the people ought to decide fundamental issues of political morality for themselves. If we care about our democracy, according to this objection, we will turn a blind eye to fidelity, at least in the case of the great abstract clauses of the Constitution, and insist on a more modest role for our courts. Is that argument sound? It all depends on what you mean by democracy. For we can distinguish two concepts of democracy, one of which certainly would justify this complaint, and the other of which would not. The first is majoritarian. The essence of democracy, according to that view, is that all issues of principle must be decided by majority vote: democracy, in other words, is majority rule all the way down. If that is what democracy means, then of course a scheme of judicial review that gives judges power to set aside judgments of political morality that a majority approves is antidemocratic. But suppose we define democracy differently: that democracy means self-government by all of the people acting together as members of a cooperative joint venture with equal standing. In my view, that is a much more attractive understanding of democracy’s nerve than the majoritarian claim. Majority rule is democratic only when certain prior conditions—the democratic conditions of equal membership—are met and sustained. What conditions are these? I tried to define them at some length in my book Freedom’s Law: The Moral Reading of the Constitution.14 First, there can be no democracy, conceived as a joint venture in self-government, unless all citizens are given an opportunity to play an equal part in political life, and that means not only an equal franchise but an equal voice both in formal public deliberations and in informal moral exchanges. That is the

134

Originalism and Fidelity

right guaranteed in principle by the First Amendment. Second, there can be no democracy so conceived unless people have, as individuals, an equal stake in the government. It must be understood that everyone’s interests are to be taken into account, in the same way, in determining where the collective interest lies. I believe that requirement underlies the Equal Protection Clause, properly understood. Third, there can be no democratic joint venture unless individuals are granted a private sphere within which they are free to make the most religious and ethical decisions for themselves, answerable to their own conscience and judgment, not that of a majority. No one can regard himself as a full and equal member of an organized venture that claims authority to decide for him what he thinks self-respect requires him to decide for himself. That is the basis of the First Amendment’s guarantee of religious freedom, and also the due process clause guarantee—so far only imperfectly realized—of independence in the fundamental ethical choices that define each individual’s sense of why his life is valuable and what success in living it would mean. According to the alternative view of democracy—you might call it partnership democracy—majority rule isn’t even legitimate, let alone democratic, unless these conditions are at least substantially met. So if you adopt that partnership view of democracy the argument that judicial review is in its nature inconsistent with democracy fails. I do not mean, as I have already said, that constitutional democracy positively demands a structure like ours, a structure that records the democratic conditions in a written, foundational document and assigns final interpretive authority to courts over whether those conditions have been met. You might well think that it would have been better to have given that responsibility to some special elected body, that the decision we either made or ratified in the nineteenth century to give unelected judges that interpretive responsibility was unwise. But that’s a different matter. You can’t say that the majority has an automatic, default, title to make those interpretive decisions without begging the question, because of course the majority has no title to govern unless the conditions are satisfied. It begs the question to think that the concept of democracy can dictate which institutions should or should not be given final interpretive authority. That decision, as I argue in Freedom’s Law, must be made on other grounds.

Originalism and Fidelity

135

We have now considered two putative trumps over fidelity. We set aside the first—that a constitution might be so unjust that it forfeits fidelity—as not germane to the present argument. We studied the second—that under our structure of government, which assigns final interpretive authority to judges, fidelity to the abstract moral clauses of the Constitution gives undemocratic power to those officials. Now I want to discuss, though briefly, a third ground for trumping fidelity, which we might call legal pragmatism.15 Pragmatism has exerted some influence on American legal theory for many decades, and is now enjoying something of a renaissance, particularly in constitutional theory. Legal pragmatism—or at least one prominent strain of it—argues that judicial decisions should be small, careful, experimental ones. It makes that claim not out of a concern for democracy, but rather out of a conviction that lawyers and judges will do better for society by trying to discover what really works in practice rather than by attempting to deduce concrete decisions from large, broad, abstract statements of principle of the kind that fidelity to the Constitution’s text would require. Judges, pragmatists claim, should concentrate on the actual and limited circumstances of particular cases, trying only to find accommodations of issues and interests that are successful in that limited frame. Above all, judges should be wary of deciding more than they have to decide at any one time. It is better that the law, including constitutional law, grow slowly, incrementally, by analogy rather than grand principle, testing its steps one by one, attempting bit by bit to make the law work better. Fidelity to our abstract Constitution commands the opposite. It commands judges to construct large-scale interpretations of grand moral principles. It points us, on this view, in exactly the wrong direction. Legal pragmatism sounds wonderfully wise, doesn’t it? It sounds very American, empiricist, hard-headed even—particularly when you take account of the main protagonists of the view, very Chicago. Broad-shouldered hog butcher to the world, and all of that. But its voice is also, with appropriate change of diction and nuance, the voice of Oliver Wendell Holmes, the patron saint of the approach, who said that the life of the law is not logic but experience. It is the voice of the American legal realists, as they came to be called, who transformed American legal education begin-

136

Originalism and Fidelity

ning in the 1930s by insisting that academic law should become more practical, more fact-drenched, less theoretical and abstract. Up to a point, that pragmatist voice offers sound (if unsurprising) advice. It reminds us that it is well to be as informed as possible and to have an eye to consequences when doing or deciding anything. But does it really provide an argument for trumping constitutional fidelity and ignoring the great questions of abstract political principle that constitutional adjudication in the spirit of fidelity requires? We must distinguish two situations in which a political community might find itself. In the first, it has a pretty good idea of what goals it wants to pursue through its constitutional and other law—a good sense, we might say, of where it wants to end up. It wants to keep inflation down and nevertheless have sustained growth, for example. It wants a lively political discourse, a lower crime rate, and less racial tension. It would know when it has achieved these goals, but it is now uncertain how to pursue them. It might indeed be helpful, in some such cases, to tell such a community not to try to solve its problems by first constructing grand economic or moral principles and then proceeding in their light, but instead to be experimental—to try one thing after another, just to see what works. In the second situation, however, the community’s problem is not that it doesn’t know which means are best calculated to reach identified ends. Rather, it doesn’t know what goals it should pursue, what principles it should respect. It wants to be a fair society, a just society, but it doesn’t know whether that means increased liberty for people to make intimate sexual decisions for themselves, for example, or whether it means giving preferences in hiring and education to minorities. Its difficulty is not that it doesn’t have the factual basis to be able to predict the consequences of granting increased sexual liberty or of adopting affirmative action programs—or at least that is not its only problem. Its deeper problem lies in not knowing whether these consequences would be overall improvements or further defects in the justice and fairness of its structures. In these circumstances, isn’t it evident that the pragmatist’s earnest advice—his practical, empirical counsel of caution and theoretical abstinence—is worthless? When we are in the second situation, we can’t avoid general principles

Originalism and Fidelity

137

by asking ourselves whether any particular step “works.” We can’t do that because no one can have any opinion about what working is until he has endorsed, however tentatively, a general principle that identifies a step as a step forward rather than a step backward. Whether increased sexual liberty or affirmative action makes a society more just cannot be decided without deciding what denials of liberty or distinctions of treatment are unjust and why. We can, it is true, offer ourselves a substitute test. We can say that a social or legal program is working if it reduces social tension, if it seems to help us live together with less apparent strain. But that strategy assumes that a reduction in tension is a sign of an improvement in social justice, and the opposite may be true. Perhaps, if we keep a tighter lid on the social and professional aspirations of minorities, as we did for many decades, there will be no fire this time or next time. It might be said that when Jim Crow no longer worked, after the Second World War, the nation abandoned it in the spirit of pragmatism. But those long decades of acquiescence were decades of injustice, and we can’t explain why without constructing an account of equal citizenship that must, in the nature of the case, be a general principle of constitutional dimension. Nor would it help to suggest that constitutional cases should be decided by analogy rather than principle because (paraphrasing Kant) analogies without principles are blind. Which analogy should the Supreme Court have adopted when asked whether women have a right to an early abortion? An abortion is in some ways like infanticide, in others like an appendectomy, and in others like the destruction of a work of art. Which, if any, of these comparisons is in fact appropriate depends on a vast network of argument: there is nothing there just to “see.” We can use analogies in stating conclusions reached through principles, but not as different routes to those conclusions. What about that old staple of conventional legal wisdom that judges should never attempt to develop a constitutional principle further than what is required by the case at hand—that they should not “reach out” to decide issues of principle not immediately before them? If we read the Constitution as I recommend, as a charter of principle, and if we insist that the responsibility of judges is to name these principles and de-

138

Originalism and Fidelity

fine their measure and scope, must we advise them to dishonor that old counsel? Not necessarily. For judges acquit their duty to keep faith with the Constitution by reaching the result in the case before them that, in their considered opinion, follows from the best interpretation of an abstract principle of constitutional morality. Fidelity does not in itself require them to state the principle more broadly than is necessary for that limited purpose. But we should notice that stating a principle broadly enough to show why it does apply to the question at hand will often mean stating it broadly enough to justify conclusions about other cases, and at least sometimes there is no evident gain and some evident cost in judges refusing to call attention to that fact. Consider, for example, Romer v. Evans, the 1996 case in which the Supreme Court declared unconstitutional Colorado’s state constitutional provision disabling its cities and subdivisions from granting homosexuals civil rights against discrimination. The decision for the Court didn’t even mention Bowers v. Hardwick, its own shameful decision of a decade earlier, in which it declared constitutional Georgia’s law making homosexual sodomy between consenting adults a crime. Many people applauded the Court’s reticence in refusing expressly to overrule the earlier decision, in spite of the fact that the principle it relied on in Evans was patently inconsistent with the earlier decision.16 They said that the issue in the Georgia case was technically independent, even though the same principle would cover both. The Court did overrule Bowers later, in 2003, in Lawrence v. Texas. Perhaps there was something to be said for the Court’s waiting seven years. But when constitutional rights are in play, there is a standing and great risk to be set against any reasons there may be for courts postponing recognizing the full implications of their decisions of principle. That is the risk of injustice to a great many people until the day of ripeness is reached. Constitutional timing affects not only doctrine but how people live and die. Waiting those seven years meant that homosexual citizens were forced to live for another, irretrievable segment of their lives as second-class citizens. Lives don’t pause while the passive, pragmatic virtues drape themselves in epigrams and preen in law journal articles. Our great experiment as a nation, our country’s most fundamental con-

Originalism and Fidelity

139

tribution to political morality, is a great idea triangulated by the following propositions. First, democracy is not simple majority rule, but a partnership in self-government. Second, that partnership is structured and made possible by a moral constitution guaranteeing to individuals one by one the prerequisites of full membership. Third, we are committed by our history to an institutional strategy of asking judges—men and women trained in the law—to enforce those guarantees of equal citizenship. Of course there are many risks in that great political adventure, as there are in any great political ambition. We have been envied for our adventure, however, in spite of its risks, and we are now increasingly copied all over the world, from Strasbourg to Capetown, from Budapest to Delhi. Let’s not lose our nerve, when all around the world other people, following our example, are gaining theirs.

chapter six

Hart’s Postscript and the Point of Political Philosophy

Archimedeans hart’s project

When Professor H. L. A. Hart died, his papers contained a draft of a long comment about my own work in legal theory, which he apparently intended to publish, when finished, as an epilogue to a new edition of his best-known book, The Concept of Law. I have no idea how satisfied he was with this draft; it contains much that he might well not have found fully satisfactory. But the draft was indeed published as a Postscript to a new edition of the book. In this chapter I discuss the Postscript’s central and most important charge. In The Concept of Law, Hart set out to say what law is and how valid law is to be identified, and he claimed, for that project, two important features. First, he said, it is a descriptive rather than a morally or ethically evaluative project: it aims to understand but not to evaluate the pervasive and elaborate social practices of law. Second, it is a philosophical rather than a legal project. It is the business of lawyers to try to discover what the law is on particular subjects—whether it is against the law of England to parade a lion in Piccadilly, for example. But identifying what law is in general is not just a particularly ambitious legal exercise, but a philosophical one, calling for entirely different methods from those lawyers use day by day. I challenged both these claims. I argued that a general theory about how valid law is to be identified, like Hart’s own theory, is not a neutral description of legal practice, but an interpretation of it that aims not just to

Hart’s Postscript and the Point of Political Philosophy

141

describe but to justify it—to show why the practice is valuable and how it should be conducted so as to protect and enhance that value.1 If so, then a legal theory itself rests on moral and ethical judgments and convictions. I also argued that ordinary legal argument has the same character: a judge or citizen who has to decide what the law is on some complicated issue must interpret past law to see what principles best justify it, and then decide what those principles require in the fresh case. So a legal philosopher’s theory of law is not different in character from, though it is of course much more abstract than, the ordinary legal claims that lawyers make from case to case. Hart insists, in the Postscript, that I was wrong on both counts: I had no right, he declared, to deny his project the special philosophical and descriptive character he claimed for it. My own ruminations about how judges should decide hard cases at law are moral and engaged, he said, because I am criticizing and evaluating their activities. But he, on the contrary, simply describes these activities in a general and philosophical way, and describes them from outside, not as an active participant in the legal wars but as a disengaged scholar of those wars. There is room in jurisprudence for both of these projects, he said, but they are different projects. Hart’s view of his own methodology is typical of much contemporary philosophy. Specialist areas of philosophy like meta-ethics and the philosophy of law flourish, each supposedly about but not participating in some particular type or department of social practice. Philosophers look down, from outside and above, on morality, politics, law, science, and art. They distinguish the first-order discourse of the practice they study—the discourse of non-philosophers reflecting and arguing about what is right or wrong, legal or illegal, true or false, beautiful or mundane—from their own second-order platform of “meta” discourse, in which first-order concepts are defined and explored, and first-order claims are classified and assigned to philosophical categories. I have called this view of philosophy “Archimedean,” and this is Archimedeanism’s golden age. The most familiar of these specialist philosophies is so-called “metaethics.” It discusses the logical status of the “value judgments” that ordinary people make when they say, for example, that abortion is morally wrong, or that racial discrimination is wicked or that it is better to betray

142

Hart’s Postscript and the Point of Political Philosophy

one’s country than one’s friends. Some meta-ethical philosophers say that these value judgments are either true or false, and that if they are true then they correctly report some mind-independent moral fact. Others deny this: they say that value judgments are not reports about an independent reality, but are rather expressions of emotion or personal taste, or recommendations for behavior, or something subjective of that character. But the philosophers in both groups insist that their own theories—the theory that value judgments are objectively true as well as the rival theory that they only express emotion—are not themselves value judgments. Secondorder philosophical theories about value judgments, the philosophers insist, are neutral, philosophical and uncommitted. They take no position about the morality of abortion or discrimination or friendship or patriotism. They are conceptual or descriptive, not substantive and engaged. I argued against this view of meta-ethics in previous work: I believe that philosophical theories about the objectivity or subjectivity of moral opinions are intelligible only as very general or abstract value judgments of their own.2 Hart’s claims about his own methods illustrate a somewhat different, though related, form of Archimedeanism, which is more prominent in political philosophy including legal philosophy than in moral philosophy. The key distinction, once again, is between levels of discourse: in this case between the first-order, substantive “value judgments” of ordinary people about liberty, equality, democracy, justice, legality, and other political ideals, and the second-order, neutral, philosophical analyses of these ideals by political philosophers. Ordinary people—politicians and journalists, citizens and presidents—argue about the relative importance of these ideals. They debate whether legality should sometimes be compromised in order to secure justice, or whether liberty should sometimes be limited in order to achieve equality or preserve community. Philosophers, on the contrary, try to provide accounts of what legality or liberty or equality or democracy or justice or community really is, that is, of what ordinary people are arguing and disagreeing about. Once again the philosophers’ work, in their opinion, is neutral among the controversies. It is a descriptive or conceptual question what liberty and equality are, and why conflict between them is inevitable, and any philosophical theory that answers those second-order questions is neutral about which of these values

Hart’s Postscript and the Point of Political Philosophy

143

is more important than the others, and which should be preferred and which sacrificed in which circumstances. This version of Archimedeanism is also mistaken. I shall argue here that definitions or analyses of the concepts of equality, liberty, law, and the rest are as substantive, normative, and engaged as any of the contending opinions in the political battles that rage about those ideals. Hart’s ambition of a purely descriptive solution to the central problems of legal philosophy is misconceived, as are the comparable ambitions of many leading political philosophers. sorenson’s case

I must describe Hart’s version of Archimedeanism in more detail, and it will be helpful for that purpose to have before us an example of a complex legal problem.3 Mrs. Sorenson suffered from rheumatoid arthritis and for many years took a generic drug—inventum—to relieve her suffering. During that period inventum was manufactured and marketed under different trade names by eleven different pharmaceutical companies. In fact the drug had serious and undisclosed side-effects, of which the manufacturers should have known, and Mrs. Sorenson suffered permanent cardiac damage from taking it. She was unable to prove which manufacturer’s pills she had actually taken, or when, and of course unable to prove which manufacturer’s pills had actually injured her. She sued all the drug companies who had manufactured inventum, together, and her lawyers argued that each of them was liable to her in proportion to its share of the market in the drug over the years of her treatment. The drug companies replied that the plaintiff ’s request was entirely novel and contradicted the long-established premise of tort law that no one is liable for injury he has not been shown to have caused. They said that since Mrs. Sorenson could not show that any particular defendant had injured her or even manufactured any of the inventum she took, she could recover against none of them. How should lawyers and judges decide which side—Mrs. Sorenson or the drug companies—is correct in its claims about what the law actually requires? In my own view, as I said earlier, they should try to identify general principles that underlie and justify the settled law of product liability, and then apply those principles to this case. They might find, as the drug

144

Hart’s Postscript and the Point of Political Philosophy

companies insisted, that the principle that no one is liable for harm that neither he nor anyone for whom he is responsible can be shown to have caused is so firmly embedded in precedent that Mrs. Sorenson must therefore be turned away with no remedy. Or they might find, on the contrary, considerable support for a rival principle—that those who have profited from some enterprise must bear the costs of that enterprise as well, for example—that would justify the novel market-share remedy.4 So on the view I favor Mrs. Sorenson might, but does not necessarily, have the best case in law. Everything depends on the best answer to the difficult question of which set of principles provides the best justification for the law in this area as a whole. Hart’s response to cases like Sorenson’s was quite different. He summed up that response in the Postscript I referred to in these words: According to my theory, the existence and content of the law can be identified by reference to the social sources of the law (e.g. legislation, judicial decisions, social customs) without reference to morality except where the law thus identified has itself incorporated moral criteria for the identification of the law.5

(I shall call this view—about how law is to be identified in hard cases like Sorenson’s case—Hart’s “sources thesis.”) Hart and I disagree, therefore, about how far and in what ways lawyers and judges must make their own “value judgments” in order to identify the law in particular cases. In my view, legal argument is a characteristically and pervasively moral argument. Lawyers must decide which of competing sets of principles provide the best—morally most compelling—justification of legal practice as a whole. According to Hart’s sources thesis, on the other hand, substantive legal argument is normative only when social sources make moral standards part of the law. No legislature or past judicial decision has made morality pertinent in Mrs. Sorenson’s case so, on Hart’s view, no moral judgment or deliberation enters into the question whether she is legally entitled to what she asked. So far as the law is concerned, he would have said, she must lose. Since Hart and I hold opposite opinions about the same issue—how to decide whether Mrs. Sorenson had a valid claim in law—it is hard to credit

Hart’s Postscript and the Point of Political Philosophy

145

his claim that we are not really disagreeing or that we are not trying to answer the same questions. But the issue remains as to how the project we share should be characterized. His account, he declared in his Postscript, “is descriptive in that it does not seek to justify or commend on moral or other grounds the forms and structures which appear in my general account of law.”6 He said that I might conceivably be right and he wrong about how law is to be identified. Perhaps I am right that lawyers and judges must make value judgments to discover the law in all hard cases. But if I were right about that, he insisted, it would only be because my account of first-order legal practice is better as a second-order description of that practice than his is. So we disagree not only about how law is to be identified, but also about what kind of theory a general answer to that question is. He believed that such a theory is only and purely a description of legal practice. I believe that such a theory is an interpretation of legal practice that makes and rests on moral and ethical claims. In one respect, however, we are in the same boat. We both believe that we will understand legal practice and phenomena better if we undertake to study, not law in some particular manifestation, like the law of product liability in Scotland, but the very concept of law. Our different claims about the nature and proper methods of that conceptual study, however, might each be thought mysterious, though for different reasons. Conceptual investigations are generally to be contrasted with empirical ones. How can Hart think that his conceptual study is “descriptive”? What sense of “descriptive” can he have in mind? Conceptual investigations are also normally to be contrasted with evaluative ones. How can I think that a study is both conceptual and evaluative? In what way can deciding what law should be like help us to see what, in its very nature, it actually is? These are sufficiently important questions to justify changing the subject for several pages.

Political Concepts Political philosophers construct definitions or analyses of key political concepts: of justice, liberty, equality, democracy, and the rest. John Stuart Mill and Isaiah Berlin, for example, both defined liberty (roughly) as the

146

Hart’s Postscript and the Point of Political Philosophy

ability to do what you might want to do free from the constraint or coercion of others, and that definition has been popular among other philosophers. On that account, laws that prohibit violent crime are invasions of everyone’s liberty. Almost all the philosophers who accept this proposition quickly add that though such laws do invade liberty they are plainly justifiable—liberty, they insist, must sometimes yield to other values. That further judgment is a value judgment: it takes sides about the relative importance of liberty and security, and some extreme libertarians might actually reject it. But, Berlin insisted, the definition itself, according to which laws against violence do compromise liberty, is not a value judgment: it is not an endorsement or criticism or qualification of the importance of liberty, but just a politically neutral statement of what liberty, properly understood, really is. Some very important conclusions do follow from that supposedly neutral statement: in particular, that the two political virtues of liberty and equality must inevitably conflict in practice. The choice between these, when they do conflict, Berlin said, is a question of value about which people will differ. But that they must conflict, so that some such choice is necessary, was for him not itself a matter of moral or political judgment, but a conceptual fact of some kind. Berlin was therefore an Archimedean about political philosophy: the project of analyzing what liberty really means, he thought, must be pursued by some form of conceptual analysis that does not involve normative judgment, assumptions, or reasoning. Other philosophers insist that liberty is, among other things, a function of money, so that taxation of the rich decreases their liberty. That definition, they insist, leaves fully open the question whether taxation is in principle justified in spite of its impact on liberty. It permits the value judgment that taxation is wicked, but also the opposite value judgment that taxation, like making violence a crime, is a justifiable compromise of liberty. Other political philosophers have treated other political values in a parallel way. It is a very popular idea, for example, that democracy means majority rule. That definition is said to leave open, for substantive decision and argument, such questions as whether democracy is good or bad, and whether it should be compromised by constraints on majority rule that might include, for example, a constitutional system of individual rights against the majority enforced by

Hart’s Postscript and the Point of Political Philosophy

147

judicial review. These latter questions, according to the Archimedean view, are substantive and normative, but the threshold question, of what democracy is, is conceptual and descriptive. These various accounts of liberty and democracy are Archimedean because though they are theories about a normative social practice—the ordinary political practice of arguing about liberty and democracy—they claim not themselves to be normative theories. They claim rather to be philosophical or conceptual theories that are only descriptive of social practice and neutral among the controversies that make up that practice. That claim is embarrassed, however, by two connected difficulties. First, ordinary political argument often includes, not merely as a neutral threshold to substantive controversies but as a central element in those controversies, argument about the very conceptual issues that the philosophers study. Second, the term “descriptive” is ambiguous—there are many ways or dimensions in which a social practice might be “described.” So Archimedeans must choose a more precise sense of description in order to make their position defensible. But they cannot do this: each sense of “description,” considered in turn, proves patently inapplicable. We must examine these independently fatal objections in turn. controversy over concepts

Philosophers’ controversies are often political controversies as well. There is just now a lively argument not only in America but across the world, about whether judicial review is inconsistent with democracy. Lawyers and politicians who argue about this do not just assume that democracy means majority rule, so that judicial review is by definition undemocratic and the only question left to be decided is whether it is nevertheless justified. On the contrary, lawyers and politicians argue about what democracy really is: some of them insist that judicial review is not inconsistent with democracy because democracy does not mean just majority rule, but majority rule subject to those conditions that make majority rule fair.7 Most of those who oppose judicial review reject this more complex definition of democracy and insist that democracy just means majority rule, or, perhaps, majority rule limited only by a few narrow procedural rights, including freedom of speech, rather than by the full set of rights that are now

148

Hart’s Postscript and the Point of Political Philosophy

typically protected in national and international constitutions. Politicians who defend taxation do not concede that taxation invades freedom. On the contrary, they deny this and insist that taxation, in itself, has no impact on liberty whatsoever. Some politicians and polemicists do, I agree, declare that taxation cheats on liberty, but, at least in America, these are all politicians who hate taxation and wish to end it. If the definition of democracy or liberty really is a neutral—threshold—issue, with no implications for substantive debate and decision, then why should politicians and citizens waste time arguing about it? Why hasn’t common sense taught ordinary people simply to converge on a standard definition of these concepts— that democracy means majority rule, for example—so that they can save their energies for the genuinely substantive issues, like the issue of whether democracy should sometimes be compromised for other values? It might be said, in answer, that people are drawn to definitions that seem most naturally to support their own substantive positions. But that reply concedes the objection: if definitions really are neutral, why should any particular definition be thought an argumentative advantage? The Archimedean story ignores the way in which political concepts actually function in political argument. They serve as abstract plateaus of agreement. Almost everyone agrees that the values in question are of at least some importance, and perhaps very great importance, but that agreement leaves open crucial substantive issues about what more precisely these values are or mean. We see this most dramatically in the case of the most abstract political concept of all: justice. People do not much dispute the importance of justice: it is normally a decisive objection to a political decision that is unjust. Disputes about justice almost always take the form of argument, not about how important justice is or when it should be sacrificed to other values, but what it is. That is, we might say, where the action is. It would therefore be most implausible to treat a philosophical theory of that concept as Archimedean: it would be implausible, that is, to suppose that an informative theory about the nature of justice could be neutral among issues of substantive political argument. True, skeptical philosophers of justice—who argue that justice is only in the eye of the beholder, or that claims of justice are only projections of emotion—often suppose that their own theories are neutral. But it would be very surpris-

Hart’s Postscript and the Point of Political Philosophy

149

ing to find a philosopher defending a positive conception of justice—that political justice consists in the arrangements that maximize a community’s wealth, for example—who believed that his theory was not itself a normative theory. Philosophers of justice understand that they are taking sides: that their theories are as normative as the claims about justice and injustice that politicians, leaders, writers, and citizens make. The thicker political concepts of liberty, equality, and democracy play the same role in political argument, and theories about the nature of those concepts are also normative. We agree that democracy is of great importance, but disagree about which conception of democracy best expresses and accounts for that importance. None of those who argue about whether judicial review is inconsistent with democracy would accept that the question of what democracy really is, properly understood, is a descriptive matter to be settled by studying, for example, how most people use the word “democracy.” They understand that their dispute is deeply, essentially substantive.8 I should emphasize the difference between the position I am now defending and the more familiar opinion of several philosophers, which is that the leading political concepts are “mixed” descriptive and normative concepts. On this familiar view the concepts of democracy, liberty, and the rest have both emotive and descriptive components, and philosophers can disentangle these from each other. The emotive meaning is a matter of social practice and expectation: in our political culture declaring that some practice is undemocratic is almost inevitably meant and taken as a criticism, and some stranger who did not understand that would have missed something crucial about the concept. But, on this view, democracy nevertheless has a wholly separable descriptive and neutral sense: it means (according to one account) government according to majority will, and there would be no contradiction, in spite of the surprise it would occasion, in someone’s saying that America is a democracy and much the worse for it. The Archimedean political philosophers who insist that their theories of the central political values are politically neutral are therefore, on this view, making no mistake. They are of course aware of the political force or charge that these concepts carry, but they ignore that charge in laying bare the underlying, and in itself neutral, descriptive meaning. The truth, I am arguing, is different. The concepts of liberty, democracy,

150

Hart’s Postscript and the Point of Political Philosophy

and the rest function in ordinary thought and speech as interpretive concepts of value: their descriptive sense is contested, and the contest turns on which assignment of a descriptive sense best captures or realizes that value. Descriptive meaning cannot be peeled off from evaluative force because the former depends on the latter in that way. Of course it is possible for a philosopher or citizen to insist that there is no value, after all, in democracy or liberty or equality or legality. But he cannot defend that stance simply by choosing one among the many contested accounts of liberty, for example, and then insisting that, so understood, liberty has no value. He must claim, not simply that liberty on some conception is worthless, but that it is worthless on the best defensible conception, and that is a much more ambitious undertaking that does not separate descriptive and evaluative meanings but trades on the interconnection between them. descriptive in what way?

The second difficulty I mentioned becomes stark when we ask in what sense of “descriptive” the supposedly second-order philosophical project of identifying a political value is a descriptive rather than a normative project. Is the supposed project a semantic analysis aiming to uncover the criteria that ordinary people actually use, perhaps all unaware, when they describe something as an invasion of liberty or as inegalitarian or undemocratic or illegal? Or is it a structural project that aims to discover the true essence of what people describe in that way, something like the scientific project of identifying the true nature of a tiger in its genetic structure or the true nature of gold in its atomic structure? Or is it a search for an impressive statistical generalization of some kind—perhaps an ambitious one that depends on the discovery of some law about human nature or behavior that leads people to denounce the same act as illiberal, for example, or perhaps a less ambitious kind of generalization that merely claims that, as a matter of fact, most people do regard a particular kind of political decision as illiberal? We should work our way through this brief catalogue of possibilities. The semantic suggestion assumes a certain factual background. It assumes that the use of “liberty,” “democracy,” and the other names of political concepts is governed—in our language—by shared criteria that determine

Hart’s Postscript and the Point of Political Philosophy

151

whether a use is correct or incorrect, or falls in some borderline area between the two. It may not be obvious at the outset what these criteria are—indeed, if the philosophical project is worth doing, that will not be obvious. But careful attention, aided by thought experiments about what it would seem right to say in particular situations, will bring these hidden criteria to the surface. These semantic assumptions are plausible in some cases: when we are studying the concept of an artifact, for example. If I described a single sheet of paper with print on it as a book, I would be making a mistake because there are shared criteria for the application of the concept of a book, and these exclude a single page. Whether I use “book” correctly depends on how the word is usually used, and if I say that a single page of text is an excellent book I have said something false. Some philosophers have made the mistake, I believe, of supposing that all concepts are governed by shared criteria in that way, or at least of uncritically assuming that the concepts they study are so governed.9 But many concepts, including those of most importance to political philosophers, are plainly not. The shared-criteria background does not hold—to return to our easiest case—for the concept of justice. To be sure, we can imagine claims about justice or injustice that would seem ruled out on semantic grounds. I would be making a conceptual mistake if I insisted, and meant it literally, that seven is the most unjust of the prime numbers.10 But we cannot imagine claims about justice of even the slightest importance or controversy being ruled out in that way. That is also true, as we have already seen, of the thicker concepts of equality, liberty, democracy, patriotism, community, and the rest. Once again we can construct silly examples of linguistic mistakes involving these concepts: the claim, for example, that a country automatically becomes less democratic when its annual rainfall increases. But there are no standard criteria of usage from which it would follow, one way or the other, whether judicial review imperils democracy, or whether all criminal laws invade people’s freedom, or whether taxation compromises liberty. Nor does anyone think that standard usage can settle such controversies. Whether judicial review is inconsistent with democracy does not depend on what most people think or on how most people talk, and people have genuine disagreements about democracy, liberty, and equality, in spite of

152

Hart’s Postscript and the Point of Political Philosophy

the fact that each is using a somewhat different conception of these political values. Indeed, people’s political disagreements are particularly profound when they disagree about what democracy or liberty or equality really is. We should therefore turn to the second possibility in our catalogue. Some of our concepts are governed not by the background assumptions about shared criteria I just described but by an entirely different set of background assumptions: that the correct attribution of the concept is fixed by a certain kind of fact about the objects in question, facts that can be the object of very widespread error. What philosophers call “natural kinds” provide clear examples. People use the word “tiger” to describe a certain kind of animal. But zoologists may discover, through appropriate genetic analysis, that only some of what people call tigers really are tigers; some of them may be a different animal, with a very different genetic composition, that look exactly like tigers. In this way, by identifying the distinctive tiger DNA, scientists can improve our understanding of the nature or essence of tigers. We can tell a parallel story about other natural kinds including, for example, gold. People may be entirely wrong in what they, perhaps uniformly, call gold. A sophisticated chemical analysis may show that some, or indeed all, of what most people now call gold is not really gold at all but only the gold of fools. Are the political concepts of democracy, liberty, equality, and the rest like that? Do these concepts describe, if not natural kinds, at least political kinds that like natural kinds can be thought to have a basic ingrained physical structure or essence? Or at least some structure that is open to discovery by some wholly scientific, descriptive, non-normative process? Can philosophers hope to discover what equality or legality really is by something like a DNA or chemical analysis? No. That is nonsense. We might pretend to such an idea. We might compile a list of all the past and present arrangements of political power that we would agree are democratic arrangements, and then ask which of the features that all such instances share are essential to their counting as democracy and which are only accidental or dispensable. But that pseudo-scientific recasting of our question would not help us, because we would still need an account of what makes one feature of a social or political arrangement essential to its

Hart’s Postscript and the Point of Political Philosophy

153

character as a democracy and another feature only contingent, and once we have rejected the idea that reflection on the meaning of the word “democracy” will supply that distinction, nothing else will. That is true not only of political concepts but of all the concepts of different kinds of social arrangement or institution. Suppose a task force were assembled to compile a long list of the different kinds of legal and social arrangement over the centuries that we would now describe as all instances of marriage, in spite of their great institutional and other differences. Suppose we found that in every case in our enormously long list some dateable ceremony was involved and that in no case was this ceremony performed to unite two people of the same sex. Now the question arises—imagine for the first time—whether a common law marriage is really a marriage or whether homosexuals can, as a conceptual matter, marry. It would be mad, would it not, to suppose that these questions about the very nature of marriage could be settled by staring, however long, at the list we had compiled? So philosophical analysis of political concepts cannot be shown to be descriptive on the model of scientific investigation into natural kinds. Liberty has no DNA. Now turn to the third possibility on our list. We now suppose that Archimedean political philosophy is scientific in a more informal sense. It aims only at historical generalizations, so, just as we might say that as a matter of fact no homosexual marriages have been recognized anywhere in the past, we might also say, if our evidence supported this proposition, that in the past people have always regarded judicial review as inconsistent with democracy. But this seems, not just weaker than the conceptual claims political philosophers make, but too weak to distinguish political philosophy from social history or political anthropology. Isaiah Berlin said not merely that liberty and equality have very often been thought to conflict, but that they do, in their nature, conflict, and he could not have supported that ambitious claim simply by pointing out (even if this were true) that almost no one had ever doubted it. True, we might fortify the interest of such sociological generalizations by attempting to explain them in biological or cultural or economic law or theory. But that would not help much. It provides no effective argument for the proposi-

154

Hart’s Postscript and the Point of Political Philosophy

tion that marriage is by its very nature or essence limited to heterosexual couples to insist that there are good Darwinian or economic explanations why homosexual marriage has been everywhere rejected. conceptual and normative?

Still, just as there is plainly something different between a lawyer’s argument about whether Mrs. Sorenson should win her case and a philosopher’s argument about what law is, so there is something different between the way a politician appeals to liberty or democracy or equality and a philosopher’s studied conception of these ideals. If we can’t distinguish between the two by supposing that the philosopher’s enterprise is descriptive, neutral, and disengaged, then how can we identify the difference? Can we say that the philosopher’s engagement is conceptual in some way that the politician’s is not? How can a normative argument also be conceptual? And if it can, why isn’t the politician’s argument conceptual as well? Return for a moment to the argument I made about natural kinds: in fact there are instructive similarities between natural kinds and political concepts that I ignored in that argument. Natural kinds have the following important properties. They are real: neither their existence nor their features depend on anyone’s invention or belief or decision. They have a deep structure—their genetic profile or molecular character—that explains the rest of their features, including the surface features through which we recognize them whether or not we are aware of that deep structure. We recognize water in part because it is transparent and liquid at room temperature, for example, and the deep structure of water—its molecular composition—explains why it has those characteristics. Political and other values are in almost all those respects like natural kinds. First, political values, too, are real: the existence and character of freedom as a value does not depend on anyone’s invention or belief or decision. That is, I know, a controversial claim: many philosophers dispute it. But I shall assume that it is true.11 Second, political values have a deep structure that explains their concrete manifestations. If progressive taxation is unjust, it is unjust in virtue of some more general, fundamental property of just institutions that progressive taxation lacks. That, too, is a controversial claim: it would

Hart’s Postscript and the Point of Political Philosophy

155

be rejected by “intuitionists” who believe that concrete moral facts are simply true in and of themselves, as they are, in their view, apprehended to be true. But once again I shall assume that it is true. The difference between natural kinds and political values that I emphasized of course remains after we have noticed these similarities. The deep structure of natural kinds is physical. The deep structure of political values is not physical—it is normative. But just as a scientist can aim, as a distinct kind of project, to reveal the very nature of a tiger or of gold by exposing the basic physical structure of these entities, so a political philosopher can aim to reveal the very nature of freedom by exposing its normative core. In each case we can describe the enterprise, if we wish, as conceptual. The physicist helps us to see the essence of water; the philosopher helps us to see the essence of liberty. The difference between these projects, so grandly described, and more mundane projects—between discovering the essence of water and discovering the temperature at which it freezes, or between identifying the nature of freedom and deciding whether taxation compromises freedom—is finally only one of degree. But the comprehensiveness and the fundamental character of the more ambitious study—its self-conscious aim at discovering something that is fundamental by way of explanation—justifies reserving the name of conceptual for it. We cannot sensibly claim that a philosophical analysis of a value is conceptual, neutral and disengaged. But we can sensibly claim it to be normative, engaged, and conceptual. what’s good about it? i

A conceptual claim about a political value aims to show, as I said, the value in it: it aims to provide some account of its value that is comparably fundamental, by way of explanation, to the molecular structure of a metal. So a general theory about justice will try to capture, at a suitably fundamental level, the value of justice: it will try to show justice, as we might put it, in its best light. But how can we do that without begging the question? Wouldn’t that be like trying to explain the color red without referring to its redness? We can say that justice is indispensable because only justice avoids injustice, or that democracy is valuable because it gives people selfgovernment, or that liberty has value because it makes people free, or that

156

Hart’s Postscript and the Point of Political Philosophy

equality is good because it treats people as of the same importance. But these propositions are not helpful, because they use the idea they are meant to explain. How could we hope to do better than that? We might try an instrumental justification—justice is good because injustice makes people miserable, or democracy is good because it generally promotes prosperity, for example. But these instrumental claims don’t answer: we want to know what is distinctively good about justice and democracy, not what other kinds of good they provide. The “mixed” account of political values that I mentioned earlier hopes to evade that difficulty: it allows philosophers to acknowledge the “value” part of democracy’s meaning, as a kind of brute fact, and then concentrate on unpacking the purely “descriptive” part. But, as I said, that won’t do either: if we want to understand what freedom or democracy or law or justice really is, we must confront the difficult question of how to identify a value’s value. We can only hope to do this—I shall argue—by locating that value’s place in a larger web of conviction. I cannot begin that argument, however, without introducing another important distinction. detached and integrated values

We want to understand better what justice, democracy, and freedom are because we think we can all live better, together, if we understand and agree on this. But there are two views we might take about the connection between understanding a value and living better in consequence. We might, first, treat the value as detached from and fixed independently of our concern to live well: we must respect it simply because it is, in itself, something of value that we do wrong or badly not to recognize. Or, second, we might treat the value as integrated with our interest in living well: we might suppose that it is a value, and has the character it does, because accepting it as a value with that character enhances our life in some other way. Orthodox religions take the first view of the central values of their faith: they treat these as detached. They insist that living well requires devotion to one or more gods, but they deny that the nature of these gods, or their standing as gods, in any way derives from the fact that a good life consists in respecting them, or that we can advance our understanding of their na-

Hart’s Postscript and the Point of Political Philosophy

157

ture by asking how, more precisely, they would have to be in order to make respecting them good or better for us. We take the same view of the importance of scientific knowledge. We think that it is better for us to understand the fundamental structure of the universe, but we do not think— unless we are crude pragmatists or mad—that that structure depends on what it would be in any way good for us that it be. We are, we might say, add-ons to a physical world that already and independently had whatever fundamental physical structure it has now when we arrived. So though our practical interests are prods and signals in our science—they help us to decide what to investigate and when to rest content with some claim or justification—they do not contribute to the truth of the claim or the cogency of the justification. Many people take the same view of the value of art. We are add-ons, they say, to the world of that value: we are responsible for discovering what is wonderful in art, and respecting its wonder, but we must take care not to commit the fallacy of supposing that something is beautiful because it makes our life better to appreciate it, or that we can identify and analyze its beauty by considering what it would be otherwise good for us to admire in the way we admire art. G. E. Moore held a very strong form of the view that art’s value is detached: he said that art would retain its full value even if all the creatures that could appreciate it perished never to return. We need not go that far to suppose that the art’s value is detached, however, we can say that a painting would have no value if it could have no meaning for or impact on any sensibility without also supposing that its value depends on the impact that it actually has, or the independent value of that impact for any creature. On the other hand, it would be starkly implausible to treat the personal virtues and achievements that make up a creditable life as having only detached value. Being amusing or interesting are virtues to cultivate and admire, but only because of the contribution they make to the enjoyment of our own and other people’s lives. It is harder to identify the contribution of more complex virtues, like sensibility and imagination, for example, but it is equally implausible that our recognition of these as virtues would survive a general understanding that they make no independent contribution at all. Most people cherish friendship: they think a life with no close con-

158

Hart’s Postscript and the Point of Political Philosophy

nections to others is impoverished. But we do not think that friendship just is what it is, like a planet, and that its only connection with a desirable life is that a desirable life is one that recognizes it, whatever it turns out to be. I do not mean, of course, that relationships like friendship are valuable only for the narrow benefits they bring to friends, like cooperation in achieving goals. But their value is not independent of the way that they enhance life in other ways; we may disagree about exactly which ways these are—friendship is an interpretive concept12—but no one thinks that friendship would remain something of importance if it turned out to do nothing for the lives of friends, except make them friends. But though it would be implausible to suppose that some personal quality or achievement has only detached value, it is often difficult, as some of these examples suggest, to identify the way in which the value of that virtue or achievement is bound up in the more comprehensive idea of a good life. We count integrity, style, independence, responsibility, modesty, humility, and sensibility as virtues, for example, and friendship, theoretical knowledge, and self-respect as important achievements. Some enterprising social Darwinian might one day show that these traits and ambitions had survival value in ancestral savannahs. But that is not how they appear to us: we do not think that sensibility or personal integrity or achieving some understanding of the science of the day is important because a community would be less prosperous or more at risk from enemy invasion if its citizens did not take it to be a virtue or a goal. We rather consider these values as aspects or components of, not instrumental means toward, an attractive, fully successful life. It would make as little sense to treat the political values that we have been discussing, like justice, freedom, legality, and democracy, as detached values. Justice is not a god or an icon: we value it, if we do, because of its consequences for the lives we lead as individuals and together. True, the Archimedean tradition sometimes seems to suppose that liberty, for example, like great art, just is what it is, and that though we must perhaps consult our own needs and interests in deciding how important liberty is, those needs and interests are not relevant in deciding what it is. Or what democracy or equality or legality really means. Nothing but that assumption could explain Berlin’s confident declaration, for example, that liberty

Hart’s Postscript and the Point of Political Philosophy

159

and equality, just in the nature of the case, are conflicting values, or other philosophers’ claim that liberty, properly understood, is compromised by even fair taxation. But it nevertheless seems deeply counterintuitive that important political values, which almost everyone must sometimes make sacrifices to protect, have only detached value, and none of the political Archimedeans, so far as I am aware, has actually made that claim. what’s good about it? ii

That apparently irresistible fact—that the political values are integrated rather than detached—routes us straight back to the difficulty we encountered earlier. How can we explain what’s good about these values without begging the question? That demand is less threatening in the case of detached than integrated values. We might well think it crazy even to imagine that the question of why great art, for example, has value could be answered without begging the question. If the value of art just lies in its own, detached, value, then it really would be just as odd to ask for an account of that value in other terms as to ask for a description of red’s color in other terms. We might of course question whether art actually does have value, after all. But we could not sensibly urge, as evidence that it does not, that it is impossible to specify that value in some non-circular way. We cannot dispose of the difficulty so easily in the case of integrated values, however, for we suppose not only that an integrated value’s existence depends on some contribution it makes to some other, independently specifiable, kind of value, like the goodness of the lives that people can lead, but that the more precise characterization of an integrated value—the more precise account of what liberty, for instance, actually is—depends upon identifying that contribution. Imagine a discussion about some virtue: modesty, for example. We ask whether modesty is, after all, a virtue, and, if so, what the line is between that virtue and the vice of self-abnegation. It would be perfectly appropriate to expect, in the course of that reflection, some account of modesty’s benefits, and if none could be provided, except that modesty is its own reward, to count that fact as fatal to the virtue’s claims. So we cannot avoid, but must now confront, the question how the value of integrated values including political values can be identified. Some integrated values, like charm, might be thought wholly instrumental. But the

160

Hart’s Postscript and the Point of Political Philosophy

more interesting ones, like friendship, modesty, and the political values, are not instrumental in any obvious way. We do not value friendship just for the narrow advantages it might bring, or democracy just because it is good for commerce. If we could arrange these various integrated values in a hierarchical structure, we might be able to explain the contribution of those lower in the hierarchy by showing how they contribute to or enhance those higher. We might be able to show, for instance, that modesty is a virtue because it contributes in some way to a capacity for love or friendship. But this project seems hopeless, for—though it is possible to see some ethical values as supporting others in some way—the support seems more mutual than hierarchical. A modest person might for that reason have a greater capacity for love or friendship, but deep love and friendship might also contribute to making people modest. No one aspect of what we take to be an attractive and successful life seems sufficiently dominant to make it plausible that all the other virtues and goals we recognize are only servants to it. We can, I think, speculate about the general character of a good life. I have elsewhere argued, for example, that we should adopt a challenge model for ethics—living well means performing well in response to a challenge that can be met well without otherwise affecting human history—rather than a model that measures a life’s success by asking how much it has improved human history.13 But no general model for ethics can serve as a final or ultimate test for subordinate virtues or goals. We can accept that living well means responding well to a distinct kind of challenge without thereby deciding whether living with flair is responding well or only preening, or whether humility in certain circumstances is really servility, or whether nobility is soiled by an interest in commerce, or whether democracy is only majority rule. If we are better to understand the non-instrumental integrated values of ethics, we must try to understand them holistically and interpretively, each in the light of the others, organized not in hierarchy but in the fashion of a geodesic dome. We must try to decide what friendship or integrity or style is, and how important these values are, by seeing which conception of each and what assignment of importance to them best fits our sense of the other dimensions of living well, of making a success of the challenge of living a life. Ethics is a complex structure of different goals, achievements,

Hart’s Postscript and the Point of Political Philosophy

161

and virtues, and the part each of these plays in that complex structure can only be understood by elaborating its role in an overall design fixed by the others. Until we can see how our ethical values hang together in that way, so that each can be tested against our provisional account of the others, we do not understand any of them. Two of the most overworked of philosophical images are nevertheless apposite here. In value as in science we rebuild our boat one plank at a time, at sea. Or, if you prefer, light dawns slowly over the whole. Political philosophy that aims better to understand the political values must fold its own work into that large structure. It must aim, first, to construct conceptions or interpretations of each of these values that reinforce the others—a conception of democracy, for example, that serves equality and liberty, and conceptions of each of these other values that serves democracy so understood. It must aim to construct these political conceptions, moreover, as part of an even more inclusive structure of value that connects the political structure not only to morality more generally but to ethics as well. All this sounds, no doubt, impossibly and even perhaps unattractively holistic. But I see no other way in which philosophers can approach the assignment of making as much critical sense as is possible of any, let alone all, parts of this vast humanist structure. If we understand that that is philosophers’ collective responsibility, over time, we will each have a better sense of our own separate marginal and incremental roles. I must concede that this conception of political philosophy stands in opposition to two of the most noted examples of contemporary work in that field: John Rawls’s “political” liberalism and the political pluralism associated with Isaiah Berlin. My recommendation is similar to Rawls’s method of reflective equilibrium, which aims to bring our intuitions and theories about justice into line with one another. The difference with Rawls’s methodology is more striking than the similarities, however, because the equilibrium I believe philosophy must seek is not limited, as his is, to the constitutional essentials of politics, but embraces what he calls a “comprehensive” theory that includes personal morality and ethics as well. If political philosophy is not comprehensive in its ambition it fails to redeem the crucial insight that political values are integrated, not detached. I cannot describe political philosophy so conceived in any greater detail

162

Hart’s Postscript and the Point of Political Philosophy

here. But I offer my book Sovereign Virtue as an example of work at least self-consciously in that spirit.14 I should stress that this comprehensive project is not based on the preposterous premise that truth in political philosophy, or in the theory of value more generally, is a matter of coherence. Elegant and exquisitely coherent theories of political morality may be false, even repulsive. We aim, not at coherence for its own sake, but at both conviction and as much coherence as we can command. Those twin aims may—indeed I think they often must—reinforce one another. It is easier to find a deep sense of rightness in a unified, integrated set of values than in a shopping list. But the two aims may, we must also remember, cause trouble for one another. They may do so, for example, when our initial sense of the character of two values—patriotism and friendship in E. M. Forster’s celebrated example, for instance, or liberty and equality as Berlin explains these—shows these values to be in conflict. We may be able to construct conceptions of patriotism and friendship, or of liberty and equality, that eliminate the conflict. But these conceptions may not grip our soul: they may feel artificial or alien or just not right. We should reflect further, if we have world enough and time, and imagination enough and skill: we should try to find some compelling conception of both friendship and patriotism, for instance, that show them not in conflict. We may not be able to do this, however.15 We must then believe whatever it is that we cannot help believing—that patriotism and friendship are both essential but that we cannot have both in full or even adequate measure, perhaps. But we cannot then think that our reflection has been a success, that we have earned the right to stop. We are only stuck, which is different.

Law hart’s defense

Law is a political concept: people use it to form claims of law, that is, claims that the law of some place or other prohibits or permits or requires certain actions, or provides certain entitlements, or has other consequences. An enormous social practice is built around making, contesting, defending, and ruling on such claims. But their character is elusive. What does the claim that “the law” requires something really mean? What

Hart’s Postscript and the Point of Political Philosophy

163

in the world makes that claim true when it is true, and false when it is false? The law of England requires people to pay taxes periodically, and to pay damages if they break their contracts, except in certain circumstances. These propositions are true, English lawyers will tell you, because of what Parliament has enacted and what English judges have decided in the past. But why do these particular institutions (rather than, for example, an assembly of the presidents of major universities) have the power to make propositions of law true? Lawyers often claim, moreover, that some proposition of law is true—for instance, that Mrs. Sorenson is legally entitled to a share of damages from each of the drug companies—when no legislature or past judges have so declared or ruled. What else, beside these institutional sources, can make a claim of law true? Lawyers often disagree about whether some claim of law, including that one, is true, even when they know all the facts about what institutions have decided in the past. What in the world are they then disagreeing about? We want, moreover, to answer these questions not just for a particular legal system, like English law, but for law in general, whether in Alabama or Afghanistan, or anywhere else. Can we say anything, in general, about what makes a claim of law true wherever it is true? Can there be true claims of law in places with very different kinds of political institutions from those we have? Or no recognizable political institutions at all? Is there a difference, in England or anyplace else, between the claim that the law requires someone to perform contracts he signed and a prediction that officials will punish him if he does not? Or between that claim and the apparently different claim that he is morally obligated to perform his contracts? If a claim of law is different from both a prediction of consequences and a statement of moral obligation, how, exactly, is it different? Hart set out to answer these ancient questions in The Concept of Law. I quoted his own summary of his answer—the sources thesis—earlier. The details of that thesis are well known among legal philosophers. Hart thought that in every community in which claims of law are made the great bulk of the officials of the community all accept, as a kind of convention, some master rule of recognition that identifies which historical or other facts or events make claims of law true. These conventions might be very different from one legal system to another: in one place the master

164

Hart’s Postscript and the Point of Political Philosophy

convention might identify legislatures and past judicial decisions as the source of true legal claims, while in another the convention might identify custom or even moral soundness as the source. What form the convention takes, in any particular community, is a matter of social fact: everything turns on what the bulk of the officials in that community happen to have agreed on as the master test. But it is part of the very concept of law that in every community some master convention exists and picks out what counts as law for that community. Hart’s sources thesis is controversial: my own view of what makes claims of law true when they are true is very different, as I said. What is now important, however, is not the adequacy of Hart’s theory but its character. Ordinary, first-order legal practice may consist in competing value judgments: it will do so, Hart says in his Postscript, if the community’s master rule of recognition uses moral standards as part of the test for valid claims of law. But his own theory, he insists, which describes ordinary legal argument, is not a normative or evaluative theory—it is not a value judgment of any kind. It is rather an empirical or descriptive theory that elucidates the concepts that that ordinary legal argument deploys. Hart’s position is a special case of the standard Archimedean view that there is a logical divide between the ordinary use of political concepts and the philosophical elucidation of them. His position is therefore open to the same objections we reviewed against Archimedeanism in general. First, it is impossible to distinguish the two kinds of claims—to distinguish the first-order claims of lawyers in legal practice from second-order philosophers’ claims about how first-order claims are to be identified and tested—sufficiently to assign them to different logical categories. Hart’s sources thesis is very far from neutral between the parties in Mrs. Sorenson’s case, for example. No “source” of the kind Hart had in mind had provided that people in Mrs. Sorenson’s position are entitled to recover damages on a market-share basis, or stipulated a moral standard that might have that upshot or consequence. So if Hart is right Mrs. Sorenson cannot claim that law is on her side. Indeed, the drug companies’ lawyers made exactly the same argument in court as Hart made in his book. They said that her claim fails because nothing in the explicit law of the state, as identified by settled legal conventions, pro-

Hart’s Postscript and the Point of Political Philosophy

165

vides for such a claim. Mrs. Sorenson’s lawyers argued to the contrary. They denied the sources thesis: they said that general principles inherent in the law entitled their client to win. So Hart’s view is not neutral in the argument: it takes sides. It takes sides, in fact, in every difficult legal dispute, in favor of those who insist that the legal rights of the parties are to be settled entirely by consulting the traditional sources of law. So the first difficulty of political Archimedeanism holds for Hart’s legal version as well. So does the second difficulty. In what way is Hart’s social sources theses supposed to be “descriptive”? Of course, as he and his defenders acknowledge, description is always itself a normative enterprise in some sense: any descriptive theory picks out one explanation of some phenomena as more revealing or salient or useful or something of the sort. Hart agreed that his analysis of law was normative in the sense in which any explanation of anything is normative: he meant that his theory is descriptive as opposed to morally or ethically evaluative. But as we noticed in the case of liberty, equality, and the rest, there are several modes of description, and we must ask in which of these modes he meant his theory to be descriptive. Though he and his followers have energetically protested that my criticism of their work is based on a misunderstanding of their methods and ambitions, it is difficult to find any helpful positive statements of what these methods and ambitions are, let alone any that explains their claim to a descriptive status. In a famously baffling phrase in the original version of The Concept of Law, he said that that book should be understood as “an exercise in descriptive sociology.” But he did not elaborate that bare claim, and it is far from plain, as we shall see, what he could have meant by it. We must, once again, exercise our own imagination. I earlier distinguished three ways in which someone might think that a conceptual analysis of a political concept is a descriptive enterprise, and we must consider each of these again, in this context. Is the sources thesis a semantic claim: does it aim to bring to the surface linguistic criteria that lawyers everywhere, or at least the bulk of them, actually follow when they make and judge claims of law? Hart did not mean, of course, to offer a simple dictionary definition or set of synonyms for any particular word or phrase. But it seems to me plausible that he meant to make a more ambitious philo-

166

Hart’s Postscript and the Point of Political Philosophy

sophical claim, elucidating criteria of application that lawyers and others might recognize, after he had pointed them out, as the rules that they actually do follow in speaking about what the law requires or permits. I proposed that understanding of his enterprise in Law’s Empire; I said that if my understanding was correct his enterprise was doomed because there are no shared criteria, even hidden ones, for endorsing or rejecting propositions of law, even among lawyers in particular jurisdictions let alone everywhere. In his Postscript, Hart vigorously denies that he intended any such thing; he says that I deeply misunderstood his project. I am battered but unbowed: I still think that my understanding of the enterprise in The Concept of Law is the best available.16 Still, since Hart ridiculed this understanding of his project in his Postscript, we must look elsewhere. Could he have thought that propositions of law form a kind of natural kind, like tigers and gold, so that discoveries might be made about them that could contradict what most people think about their truth or falsity? Just as we might discover that many animals labeled “Tiger” in zoos are not actually tigers, so, on this view, we might discover that, whatever people think, nothing is law that does not conform to the sources thesis. Deep discoveries about natural kinds do seem at once conceptual—tiger DNA can plausibly be called the essence of tigerhood—and descriptive. So this hypothesis, if we could accept it, would explain Hart’s apparent belief that a conceptual investigation into law could be descriptive but not semantic. We need not pursue this, however, because Hart could not have thought that true claims of law form a natural kind. If liberty has no DNA, neither does law. We are left with the third possibility I distinguished: that Hart’s sources thesis is meant to be descriptive in the way of an empirical generalization of some sort. Some army of legal anthropologists might conceivably collect all the data that history can provide about the various occasions on which people have made, accepted, or rejected what we regard as claims of law. Some sociologist with a room-sized computer and a huge budget might hope to analyze that Everest of data, not to find the essence or nature of law, but simply to discover patterns and repetitions in the vast story. He might, most ambitiously, aim to identify laws of human nature: if he finds that people accept propositions of law only when the sources

Hart’s Postscript and the Point of Political Philosophy

167

thesis endorses them, for example, he might hope to explain that remarkable fact through Darwinian principles, perhaps, or economic equations, or something of the sort. Or he might be much less ambitious—he might simply point out the regularity, which would certainly be interesting enough in itself, and not try to explain it. Shall we understand Hart’s Archimedeanism to be empirical in either the more or less ambitious of these senses? There is an insuperable threshold objection: neither Hart nor his descendants have even so much as begun on the lifetime-consuming empirical studies that would be needed. They have not produced an anthill, let alone an Everest, of data. There is a further threshold objection, at least in Hart’s own case. It would be exceedingly odd to refer to any such empirical study or generalization as aimed at discovering the concept or nature or very idea of law, and so odd to name a book supposedly reporting those discoveries The Concept of Law. Imagine an economist saying that Ricardo’s laws lay bare the very concept of wage or profit, for example. Behind these threshold difficulties lies a third and even greater embarrassment. If we conceive Hart’s theories—or those of his descendants—as empirical generalizations, we must concede at once that they are also spectacular failures. It would, I said, take a mountain of data to support the sources thesis as an empirical generalization, but it takes only a few counter-examples to refute it, and these are everywhere. There is now a lively debate in the United States about whether capital punishment is constitutional. The argument hinges on whether the Eighth Amendment to the Constitution, which prohibits “cruel and unusual punishments,” incorporates some moral standard for appropriate punishments, which capital punishment might well be thought to fail, or whether, on the contrary, it incorporates no moral standard but instead prohibits only punishments that the statesmen and politicians who made the amendment—or the general public to which it was addressed—thought cruel. If we assume that capital punishment is in fact unacceptably cruel, but that almost no one thought it so in the eighteenth century, then lawyers who accept the first of these interpretations will think that constitutional law prohibits capital punishment and those who accept the second will think that it permits capital punishment. Those who argue for the first, or moral, reading

168

Hart’s Postscript and the Point of Political Philosophy

plainly contradict the social sources thesis, since no social source has directed that the Eighth Amendment be read to incorporate morality. But, since no social source has ruled that morality is not relevant, those who argue against the moral reading also contradict that thesis. Hart said that morality becomes relevant to identifying law when some “source” has decreed that morality should have that role, and he gave the abstract clauses of the American Constitution as examples. But he misunderstood the state of American constitutional law. There is no consensus either in favor of or against the moral reading of the Constitution: it is on the contrary a matter of fierce disagreement. I, among others, endorse the moral reading that Hart apparently has in mind.17 But others, including Justice Antonin Scalia of the United States Supreme Court, and a notorious former judge, Robert Bork, denounce the moral reading as profoundly misjudged.18 There is no convention for or against it, no basic rule of recognition from which either side could hope to support the propositions of constitutional law that it nevertheless claims to be true.

The Value of Legality legality

A fresh start? I said earlier that political concepts are concepts of value, and that political philosophers should aim to show, for each of them, more precisely where its value lies. I said that since political values are integrated rather than detached, this project must find the place of each value in a larger and mutually supporting web of conviction that displays supporting connections among moral and political values generally and then places these in the still larger context of ethics. This picture of political philosophy is not only wildly ambitious—it can only even be imagined in a cooperative way—but it is also, as I conceded, very much against contemporary fashion. It is not in the spirit of modest value pluralism. It aims at a utopian and always unrealized goal—Plato’s unity of value—instead. We should try to approach the ancient puzzles of law in that way. We need to find, however, a political value that is linked to those puzzles in the right way. It must be a real value, like liberty, democracy, and the rest, and it must be widely accepted as a real value, at least if our project is to have

Hart’s Postscript and the Point of Political Philosophy

169

any chance of influence. The value must nevertheless function, within our community, as an interpretive value—those who accept it as a value must nevertheless disagree about precisely what value it is, and must disagree, in consequence, at least to some degree, about which political arrangements satisfy it, or which satisfy it better and which worse. It must be a distinctly legal value so fundamental to legal practice that understanding the value better will help us better to understand what claims of law mean and what makes them true or false. We must be able to see, for example, how a specific conception of the value would generate the sources thesis, and how other conceptions would generate the very different theories of law that are also part of the literature of jurisprudence. We must be able to see how embracing one conception of the value rather than another would mean reaching one decision rather than another in Mrs. Sorenson’s case. It should now be clear what value that is: it is the value of legality—or, as it is sometimes more grandly called, the rule of law. Legality is a real value, and it is a distinctly legal value. Many people think, for instance, that the Nuremberg trials in which Nazi leaders were tried and sentenced after World War II offended legality even though they were justified by other political values—justice or expediency, for example. Legality is, moreover, a very popular value. It has been much more widely embraced, and over many more centuries, than the other values I discussed earlier, and it is very widely regarded as of even more fundamental importance than they are. Classical and medieval philosophers analyzed and celebrated legality long before other philosophers celebrated liberty, let alone equality. From the beginning, moreover, legality was an interpretive ideal, and it remains so for us. There are various ways to state the value abstractly. Legality is engaged, we might say, when political officials deploy the state’s coercive power directly against particular persons or bodies or groups—by arresting or punishing them, for example, or forcing them to pay fines or damages. Legality insists that such power be exercised only in accordance with standards established in the right way before that exercise. But that abstract formulation is, on its own, almost entirely uninformative: it remains to be specified what kinds of standards satisfy legality’s demands, and what counts as a standard’s having been established in the right way in advance. People disagree markedly about those issues. Some say, as I

170

Hart’s Postscript and the Point of Political Philosophy

just noted, that the Nuremberg trials offended legality, whether or not they were finally justified by some other value. But others say that the trials protected or enhanced the true ideals of legality. People disagree now, along similar lines, about trials of deposed dictators for inhumane acts not condemned by local law when they acted, and about the trials of Balkan villains in international criminal courts. These different views represent a common adherence to the value of legality, but different conceptions of what legality is. Nor can there be much doubt about the connection between the value of legality with the problem of identifying true or valid claims of law. Conceptions of legality differ, as I said, about what kinds of standards are sufficient to satisfy legality and in what way these standards must be established in advance; claims of law are claims about which standards of the right sort have in fact been established in the right way. A conception of legality is therefore a general account of how to decide which particular claims of law are true: Hart’s sources thesis is a conception of legality. We could make little sense of either legality or law if we denied this intimate connection between conceptions of legality and the identification of true claims of law. We can sensibly think that though the law rejects Mrs. Sorenson’s claim for damages according to market share, justice supports that claim. Or (less plausibly) the other way around: that though the law grants her that claim, justice condemns it. But it would be nonsense to suppose that though the law, properly understood, grants her a right to recovery, the value of legality argues against it. Or that though the law, properly understood, denies her a right to recovery, legality would nevertheless be served by making the companies pay. We can rescue the important questions of legal philosophy from Archimedean obscurantism by attacking them in this different way. We understand legal practice better, and make more intelligible sense of propositions of law, by pursuing an explicitly normative and political enterprise: refining and defending conceptions of legality and drawing tests for concrete claims of law from favored conceptions. There is no question of taking theories of law constructed in that way to be merely “descriptive.” They are conceptual, but only in the normative, interpretive sense in which theories of justice, as well as theories of democracy, liberty, and equality, are conceptual. They may be, like such theories, more or less am-

Hart’s Postscript and the Point of Political Philosophy

171

bitious. The more ambitious try to find support for their conceptions of legality in other political values—or rather, because the process is not oneway, they try to find support for a conception of legality in a set of other, related, political values, each of these understood in turn in a way that reflects and is supported by that conception of legality. I offer my own book, Law’s Empire, as a more elaborate example of what, at least in effort, I have in mind. I did not emphasize the word “legality” there, but I did appeal to the value: I said that a philosophical theory of law must begin in some understanding of the point of legal practice as a whole. I was not then so concerned with isolating and refining the other values that any persuasive account of law’s point would implicate. But the more ambitious description of jurisprudence I have now described helps me better to understand, and I hope better to pursue, issues underdeveloped or ignored in the book. I said there, for example, that identifying true propositions of law is a matter of interpreting legal data constructively, and that a constructive interpretation aims both to fit and to justify the data. I warned that “fit” and “justification” are only names for two rough dimensions of interpretation, and that further refinement would require a more careful analysis of other, discrete political values through which to understand these dimensions more thoroughly, so that we might see, for example, how to integrate them in an overall judgment of interpretive superiority when they pull in opposite directions. The key political concepts that must be explored in that way, it now seems to me, are those of procedural fairness, which is the nerve of the dimension of fit, and substantive justice, which is the nerve of political justification. Understanding the concept of legality better, that is, means expanding the discussion of adjudication to include a study of these further values, and though it would be surprising if that further study did not alter our understanding of law in some way, it would also be surprising if our understanding of law did not produce at least somewhat different views of fairness and justice as well. A wide-ranging reinterpretation of political values leaves nothing wholly as it was. jurisprudence revisited

Can we interpret the leading traditions or schools of jurisprudence as reflecting (and therefore as different from another in respect of) different

172

Hart’s Postscript and the Point of Political Philosophy

conceptions of legality? That value insists that the coercive power of a political community should be deployed against its citizens only in accordance with standards established in advance of that deployment. What kind of standards? Established in what way? We attack these questions by proposing some reading of the value of legality—some putative point served by constraining the use of political power in that way—and this reading must implicate, as I have several times now said, other values that we recognize. If it is sufficiently ambitious, it will implicate a great many of them in what I called earlier a web of conviction. Nevertheless different conceptions will select different connected values as more important in that mix: conceptions will differ, we might say, in the importance each assigns to different values in creating the local magnetic field in which it places legality. Schools or traditions of jurisprudence are formed by large differences in the character of those choices. Three important traditions have in fact been formed by the rival choices, as locally influential values, of the political values of accuracy, efficiency, and integrity. I shall explore each of these three traditions in that light, but I want particularly to emphasize, in advance, that I am not suggesting that any of the traditions I describe has chosen one of these three values as the exclusive key to legality, and disparaged or neglected all others. I claim that the legal positivist tradition emphasizes the relation between legality and efficiency, for example, but I do not mean that positivists have been insensitive to either good or fair government. Positivists differ among themselves, not only because they hold somewhat different views of what political efficiency means and why it is valuable, but also because they hold different views, reflected in the details of their positions, about the character and force of many other political ideals, and I shall mention some of the other values to which different positivists have appealed to shape and reinforce their dominant reliance on efficiency. My tripartite division distinguishes the centers of gravity of different groups or schools of theory; it is not meant to exhaust the complexity or explain the details of any one theory. Accuracy. By accuracy I mean the power of political officials to exercise the state’s coercive power in a substantively just and wise way. Legality

Hart’s Postscript and the Point of Political Philosophy

173

promotes accuracy if official acts are more likely to be wise or just if they are governed by established standards than if they represent merely the contemporary judgment of some official about what would be just or wise. It is not immediately evident that that will always, or even usually, be the case. Plato said that legality would hinder accuracy if the officials whose power it restrained were people of great knowledge, insight, and character, because they would know more about the immediate case than those who had laid down laws in the past, and they would be sensitive to discrete aspects of the case that might require or justify some different treatment. But there are at least two possible reasons for thinking that nevertheless legality does improve accuracy. The first appeals to institutional, historical, or other contingent reasons for thinking that the judgment of past lawmakers, in spite of their distance from some immediate problem or issue, is nevertheless likely to be better than the contemporary official’s instinct or decision. Plato endorsed legality, in spite of the reservation I just reported, for that sort of reason. Philosopher-kings are rarely in power, he said, and, particularly in a democracy, the people actually in charge are ill-informed, incompetent, corrupt, self-serving, or all of these. In those unfortunate circumstances, he said, it is better that officials be constrained to follow what was laid down in the past, because they cannot be trusted to make a good contemporary decision of their own. Political conservatives, like Burke and Blackstone, often defended legality in much the same way. They thought that established law was a repository of accumulated wisdom and clear thought, and was therefore more to be trusted than the decisions, particularly those made in the heat of some moment, of individuals of limited character, knowledge, and skill. The second reason for supposing that legality improves accuracy is very different: it relies, not on any contingent reason for supposing that established standards are wiser and more just than fresh case-by-case rulings, but on a conception of legality that allows the tests for established standards to promote or even to guarantee that result. The medieval natural lawyers thought that good government meant government in accordance with God’s will, that God’s will was expressed in moral laws of nature, and that divinely inspired priests and rulers were reliable guides to that law. They were naturally attracted, therefore, to a conception of legality that

174

Hart’s Postscript and the Point of Political Philosophy

emphasized these fortunate connections between legality and political virtue, and therefore to tests for law that include a requirement of moral worth or acceptability. There is nothing in the abstract concept of legality that excludes that connection, and if the true value of legality is identified only through a conception that formalizes it, then that conception will seem, for those who accept the sets of understandings into which it fits, irresistible. The natural law tradition, in its various forms and manifestations, is premised on that way of understanding why legality has the value it does. Efficiency. Jeremy Bentham, the founder of at least the British form of legal positivism, was not, however, attracted to either of these two sets of assumptions. He did not suppose that old standards are good ones; on the contrary he was a restless, even radical, innovator. He did not believe that the moral law is evident in God’s nature: he thought, on the contrary, that the very idea of natural rights is nonsense on stilts. His conception of legality’s virtue lay not in accuracy but in efficiency. Political morality, he thought, lies in the greatest good of the greatest number, and that can best be secured, not by different coercive or policy decisions taken by different officials relying on their own immediate and diverse judgments, but by detailed policy schemes whose complex consequences can be carefully considered in advance, and which can be laid down in detail, preferably in elaborate statutory codes, and enforced to the letter. Only in that way can the massive problems of coordination that the government of a complex society confronts be solved. Legal positivism is a natural upshot of that understanding of legality true point and value. Efficiency is compromised or entirely undermined, he thought, when moral tests are included among the tests for law, because moral tests allow citizens and officials who disagree, often strenuously, about what morality requires to substitute their own judgment about what standards have been established: the consequent disorganization will produce not utility but chaos. So Bentham and his followers insisted that law is whatever and only what the sovereign ruler or parliament has decreed: law stops where decree ends. Only that understanding can protect law’s efficiency. Later positivists have been true to that faith: they all stress law’s role in substituting crisp direction for the uncertainties of customary or moral

Hart’s Postscript and the Point of Political Philosophy

175

imprecation. Hart wrote, much in the spirit of Thomas Hobbes, a positivist of an earlier era, that legality cures the inefficiencies of a mythical prelaw state of nature or custom. Joseph Raz argues that the nerve of legality is authority, and that authority is damaged or undermined unless its directives can be identified without recourse to the kinds of reasons for action that citizens have before authority has spoken. Authority cannot serve its purpose, he insists, unless its directives replace rather than only add to the reasons people already have. As I said, efficiency is not the only value that positivists take into account in forming their conceptions of legality, and it is worth noticing some of the others. Bentham, for example, thought it important that the public retain a healthy sense of suspicion and even skepticism about the moral worth of their laws: they should understand the difference between law as it is and as it should be. He worried that if judges could properly appeal to morality in deciding what the law is, then this crucial line would be blurred: people might assume that whatever judges declare to be law cannot be very bad because it has passed that moral test. Liam Murphy, among contemporary legal positivists, has appealed to the importance of public vigilance in defending his own positivistic understanding of legality’s value.19 Hart was concerned, not just about efficiency, but about an independent aspect of political fairness. If a community’s law can be determined simply by discovering what the pertinent social sources—the legislature, for instance—have declared, then citizens are put on fair warning about when the state will intervene in their affairs to help or obstruct or punish them. If, on the other hand, the decisions of those sources are subject to supplement or qualification by moral considerations and principles, citizens cannot as easily or with the same confidence know where they stand. In America, some constitutional lawyers are drawn to a version of positivism for an entirely different reason. If morality is acknowledged to be among the tests for law, then judges whose own moral opinions would then be decisive in constitutional cases have much greater power over ordinary citizens than if morality is understood to be irrelevant to their office. Particularly when judges are appointed rather than elected, and cannot be deposed by popular will, this aggrandizement of their power is undemocratic.20 So legal positivists can defend their conception of legality, which insists

176

Hart’s Postscript and the Point of Political Philosophy

that morality is not pertinent to the identification of law, by showing how well legality so understood serves efficiency, and also these other values. That defense assumes, of course, particular conceptions of these other values, and these conceptions can and have been challenged. It might be argued that political efficiency means coordinating a population’s behavior toward good goals, for example, rather than simply any goals, that fair warning is sufficiently given, at least in some contexts, by the promise or threat that moral standards will be applied in judging particular behavior, that the critical judgment of a populace is sharpened not diminished by a “protestant” understanding of law that allows it to disagree, in part on moral grounds, with official declarations of what the law requires, and that democracy means not just majority rule but majority rule subject to the conditions, which are moral conditions, that make majority rule fair. Positivism rejects these and other alternative understandings—that is, it not only selects which political values to emphasize in constructing an account of legality, but also interprets those other values, controversially in the light of its own conception of legality. There is nothing threateningly circular in this complex conceptual interaction; on the contrary it is exactly what the philosophical project of locating a political value like legality in a larger web of value requires. Integrity. Efficiency of government, on any plausible conception of what that means, is plainly an important product of legality, and any plausible explanation of legality’s value must emphasize that fact. No ruler, even a tyrant, survives for long or achieves his goals, even very bad ones, if he altogether abandons legality for whimsy or terror. But there is another important value that legality might also be seen to serve, not in competition with efficiency, but sufficiently independent of it to provide, for those who take it to be of great importance, a distinctive conception of what legality is for. This is political integrity, which means equality before the law, not merely in the sense that the law is enforced as written, but in the more consequential sense that government must govern under a set of principles in principle applicable to all. Arbitrary coercion or punishment violates that crucial dimension of political equality, even if, from time to time, it does make government more efficient. Integrity has been a popular ideal among political philosophers for cen-

Hart’s Postscript and the Point of Political Philosophy

177

turies, and its connection with legality has often been noted. The connection is sometimes expressed in the rubric that under the rule of law no man is above the law; but the force of that claim, as the various discussions of it make plain, is not exhausted by the idea that each law should be enforced against everyone according to its terms. That stipulation would be satisfied by laws that, by their terms, applied only to the poor, or exempted the privileged, and the philosophers who describe legality in this way have in mind substantial and not merely formal equality before the law. A. V. Dicey, for example, in his classic study of the British Constitution, draws the following distinction: We mean in the second place, when we speak of the rule of law . . . not only that with us no man is above the law, but (what is a different thing) that here every man, whatever be his rank or condition, is subject to the ordinary law of the realm . . .

and he later refers to this as “the idea of legal equality.”21 F. A. Hayek makes much the same claim though, unsurprisingly, he associates it with liberty rather than equality. He wrote in a classic work: The conception of freedom under the law that is the chief concern of this book rests on the contention that when we obey laws, in the sense of general abstract rules laid down irrespective of their application to us, we are not subject to another man’s will and are therefore free . . . This, however, is true only if by “law” we mean the general rules that apply equally to everybody. This generality is probably the most important aspect of that attribute of law which we have called its “abstractness.” As a true law should not name any particulars, so it should especially not single out any specific person or group of persons.22

If we associate legality with integrity in this way, then we will favor a conception of the former that reflects and enhances the association. We prefer an account of what law is, and of how it is to be identified, that incorporates the value—integrity—whose pertinence and importance we recognize. If one way of deciding Mrs. Sorenson’s case will treat her as equal before the law, in the sense that integrity assumes, and another will not, then we prefer a conception of legality that encourages the first and discourages the second decision. I tried to construct such a conception of law in Law’s

178

Hart’s Postscript and the Point of Political Philosophy

Empire; I described it briefly earlier in this essay and I will not expand that description now. I want to emphasize, instead, that Law’s Empire reports only one way in which integrity and legality can be understood in each other’s terms, and readers who are dissatisfied with my own construction should not reject the general project for that reason. I suppose I should, however, anticipate a different objection that someone might wish to make at this point. He might object that the correct decision in Mrs. Sorenson’s case depends on what the law actually is, not on what we would like the law to be because we are attracted to some other ideal, like integrity. But, as I have been trying to argue for many pages, we cannot identify the correct tests for deciding what the law really is without deploying and defending a conception of legality, and we cannot do that without deciding what, if anything, is really good about legality. Jurisprudence is an exercise in substantive political morality. Of course, we cannot successfully propose an analysis of legality that bears no relation to legal practice: a successful account of any value must be able to be seen as an account of that value as it exists and functions in a scheme of values we share. Just as a claim about Mrs. Sorenson’s legal rights must fit the legal practice of the jurisdiction in which the case arises, so any claim about what legality is must fit legal practice more generally. But more than one conception of legality will fit well enough; that is why we have different judicial philosophies represented even on the same bench. The cutting edge of a jurisprudential argument is its moral edge. interpretive positivism

The difficulties I have been describing in Hart’s self-professed methodology, which insists that theories of law are descriptive and neutral, can all be cured by recasting his arguments in the interpretive mode I have been suggesting. We strive to understand legality by understanding what is distinctly important and valuable in it, and we are tempted, initially, by the idea that legality is important because it provides authority in circumstances when authority is needed. But that claim invites a further conceptual question. Authority, too, is a contested concept: we need an account of authority that shows what the value is in it. The key to that further question lies in the mix of other values that legal positivists have traditionally

Hart’s Postscript and the Point of Political Philosophy

179

celebrated, and, particularly, in the efficiency that authority brings. As positivists from Hobbes to Hart have pointed out, and as history has amply confirmed, political authority makes policy and coordination possible, and though policy and coordination may not work to everyone’s benefit, they often, perhaps even usually, do. We are guided by this larger matrix of ideas in settling on conceptions of the discrete concepts it engages: the concepts of legality, efficiency, and authority. We must settle on conceptions of each that allows it to play its part in the larger story. So we adopt an “exclusive” positivist conception of legality, which insists that morality plays no role in identifying true claims of law, and we also adopt what Joseph Raz calls a “service” conception of authority, which insists that there is no exercise of authority except when what has been directed can be identified without recourse to reasons the directive is meant to resolve and replace.23 We no longer suppose that these conceptual claims are neutral, Archimedean excavations of rules buried in concepts that everyone with a full grasp of the concept or a full knowledge of the language will recognize. We may still say, as positivists have, that we have identified the salient aspects of our concepts that help us best to understand ourselves or our practice or our world. But now we make explicit what is obscure in those unhelpful claims: we understand ourselves and our practices better in one particular way—by designing conceptions of our values that show what, on reflection, we find most valuable in them, each and in the whole. We make no pretense that our conclusions are uncontroversial or disengaged from concrete political decision. If our constructions show that most of what most people think about law is a mistake—if they show that the claims of law that both sides make in Sorenson’s case are wrong because none of them respects the sources thesis—then that is not an embarrassment to us, any more than it would embarrass us if our conclusions about equality showed that most people have steadily misunderstood what equality really is. That is, I think, the best we can do for the central claims of legal positivism. It sounds tinny and artificial, I know, because in fact it would not make our law more certain or predictable or our government more efficient or effective if our judges were suddenly converted to legal positivism and explicitly and rigorously enforced the sources thesis. On the con-

180

Hart’s Postscript and the Point of Political Philosophy

trary, judges would then rely much less on claims of law than they now do. If I am right, American judges would be forced to declare that there is no law in America at all, except the bare, uninterpreted words of the Constitution.24 Even if they somehow avoided that frightening conclusion, they would be forced to subvert rather than to serve legality, even on the positivist’s conception of that virtue, because they would be forced to declare much too often that either the law said nothing about the matter in controversy, or that the law was too unjust or unwise or ineffective to enforce. Judges who thought it intolerable that Sorenson should have no remedy, for example, would be forced to declare that, in spite of the fact that the law favored the defendant, they would ignore the law and hence ignore legality and award her compensation. They would announce that they had a “discretion” to change the law (or, what comes to the same thing, to fill in gaps in the law they had discovered) through the exercise of a fresh legislative power that contradicts the most basic understanding of what legality requires. So it may seem perverse, or at least ungenerous, for me to attribute to positivists such a self-defeating argument for their position. But we should now notice that when positivism was first proposed, and when it was an actual force among lawyers and judges rather than only an academic position, the political situation was very different. Bentham, for example, wrote in an age of simpler and more stable commerce and a more homogenous moral culture: he could plausibly hope, as he did, for statutory codifications that would rarely leave gaps or require controversial interpretation. In those circumstances judges wielding moral tests for law posed a distinct threat to utilitarian efficiency that could be avoided most simply by denying them any such power. Even in the earlier years of the last century, progressive lawyers shared Bentham’s views: progress, they thought, was available through administrative agencies, acting under broad parliamentary mandates, issuing detailed regulations that could be applied and enforced by technicians. Or, in the United States, through detailed uniform codes compiled by a national law institute trained by academic lawyers and proposed for adoption by the several states. Once again, in this atmosphere, judges claiming power to distil moral principles from an ancient and unsuited common law seemed archaic, conservative, and cha-

Hart’s Postscript and the Point of Political Philosophy

181

otic. The danger of such a claim was brilliantly illustrated by the Supreme Court’s 1904 Lochner decision, which held that the conception of liberty embedded in the Fourteenth Amendment made progressive legislation that limited the number of hours bakers could be asked to work each day unconstitutional.25 Legal positivism, progressives thought, saved law from such reactionary morality. Oliver Wendell Holmes’s positivism was a working legal doctrine: he cited positivism in dissenting from Supreme Court decisions in which, in his view, justices had assumed an illegitimate power to make their own law by pretending to find principles embedded in the law as a whole. “The common law is not a brooding omnipresence in the sky,” he declared in one famous dissent, “but the articulate voice of some sovereign or quasi sovereign that can be identified; although some decisions with which I have disagreed seem to me to have forgotten the fact.”26 The jurisprudential argument between positivism and older theories of law was at the center of the long controversy about whether federal judges, when they had jurisdiction only because the parties were from different states, were constitutionally obliged to enforce the common law of one of those states as that law had been declared by the state’s own courts, or whether they were permitted to decide differently by finding and applying principles of “general” law not recognized by any state court. In Erie Railroad v. Tompkins, the Supreme Court finally decided that there was no such thing as “general” law: there was only law as declared by particular states.27 Justice Brandeis, for the Court, quoted another famous Holmes passage: Law in the sense in which courts speak of it today does not exist without some definite authority behind it . . . the authority and only authority is the State, and if that be so, the voice adopted by the State as its own [whether it be of its Legislature or of its Supreme Court] should utter the last word.

Brandeis made plain the practical importance of this view of law: the contrary view, long followed by the federal courts, destroyed uniformity because it produced different results on the same issue in state and federal courts, encouraging out-of-state plaintiffs to bring suits in federal courts when that was to their advantage. Of course, the Court could have reached the same result—for those practical reasons—without embracing positiv-

182

Hart’s Postscript and the Point of Political Philosophy

ism, but the muscular rhetoric of that legal doctrine had great appeal because it allowed Holmes, Brandeis, Learned Hand,28 and other “progressives” to paint their more conservative opponents as victims of an incoherent metaphysics. Changes in society’s expectations of law and judges were well under way, however, even in the 1930s when they wrote, and with accelerating velocity in the decades that followed, that made positivism’s general conception of legality steadily more implausible and selfdefeating. Elaborate statutory schemes became increasingly important sources of law, but these schemes were not—could not be—detailed codes. They were more and more constructed of general statements of principle and policy that needed to be elaborated in concrete administrative and judicial decisions; if judges had continued to say that law stopped where explicit sovereign direction ran out, they would have had constantly to declare, as I said, that legality was either irrelevant to or compromised in their judgments. In the 1950s, moreover, several Supreme Court justices began a new turn in American constitutional law that made jurisprudence a riveting issue of national politics. They began to interpret the abstract clauses of the Constitution, including the due process and equal protection clauses, as stating general moral principles that give individual citizens important rights against national and state governments, rights whose existence presupposed that law was not limited to deliberate enactment, and whose contours could only be identified through substantive moral and political judgment. That initiative suddenly reversed the political valence of the jurisprudential argument: conservatives became positivists who argued that the Court was making up new constitutional rights of racial equality and freedom in reproductive decisions, for example, and therefore subverting legality. Some of the liberals who approved the Court’s direction then moved from positivism toward a different conception of legality that stressed the principled integrity of the American constitutional settlement. In the last decades, the most conservative Supreme Court justices have engineered a further change in valence: their initiatives increasingly require them to ignore much Supreme Court precedent, and they therefore find a better justification in conservative political principle than in any orthodox version of legal positivism.

Hart’s Postscript and the Point of Political Philosophy

183

When Hart wrote The Concept of Law he could no longer rely, as Bentham and Holmes could, on the contemporary appeal of the positivist conception of legality. Hart’s account of positivism’s efficiency is a Just-So story from an imagined ancient past: a supposed pre-historical transition from the chaos of primary-rule tribal inefficiency to the crisp authority of secondary rules embraced in a liberating and near-uniform explosion of consensus. Those who followed his lead have continued to write about authority, efficiency, and coordination. But they cannot confirm their claims in actual political practice either, and that may explain why they fall back, as Hart did, on self-descriptions that seem to isolate their theories from such practice. They say that they are probing the very concept or nature of law, which remains the same in spite of shifting features of political practice or structure, or that, in any rate, they offer only descriptive accounts of what that practice is, withholding any judgment about what it should be or become. That is the methodological camouflage that I have challenged in this essay. If, as I have argued, the self-description cannot be made both intelligible and defensible, then we must concentrate on the more comprehensible justification I tried to substitute—the substantive, positivist account of the value of legality that I have now described. It is a virtue of that description, I think, that it brings to the surface the appeal positivism had for lawyers and judges, and for scholars in substantive fields of law, in times past, when its conception of legality seemed more plausible than it now does. concluding thoughts

I have been emphasizing similarities between the concept of legality—as a foundation for legal philosophy—and other political concepts, and I shall close by noting an important difference. Legality is sensitive in its application, to a far greater degree than is liberty, equality, or democracy, to the history and standing practices of the community that aims to respect the value, because a political community displays legality, among other requirements, by keeping faith in certain ways with its past. It is central to legality that a government’s executive decisions be guided and justified by standards already in place, rather than by new ones made up ex post facto, and these standards must include not only substantive laws but also the

184

Hart’s Postscript and the Point of Political Philosophy

institutional standards that give authority to various officials to create, enforce, and adjudicate such standards for the future. Revolution may be consistent with liberty, equality, and democracy. It may, and often has been, necessary in order to achieve even a decent level of those values. But revolution, even when it promises to improve legality in the future, almost always involves an immediate assault on it. So any even moderately detailed account of what legality requires in concrete terms in some particular jurisdiction must attend very carefully to the special institutional practices and history of that jurisdiction, and even a moderately detailed account of what it requires in one place will be different, and perhaps very different, from a parallel account of what it requires elsewhere. (Arguing and deciding about these concrete requirements in a particular community is the quotidian work of that community’s practicing lawyers, at one level, and of its academic lawyers at another.) That is also true, to some more limited extent, about other political virtues: the concrete institutional arrangements that count as improving democracy or advancing equality or better protecting liberty in a country with one political demography and history may well be different from those that count in that way in another. But though legality is evidently even more sensitive, in detail, to special features of political practice and history than these other virtues, it does not follow, for legality any more than for the others, that nothing of importance can or should be done to explore the value at a philosophical level that transcends most details of place. For just as we can explore the general concept of democracy by developing an attractive abstract conception of that concept, so we can also aim at a conception of legality of similar abstraction, and then attempt to see what follows, by way of concrete propositions of law, more locally. There is no bright-line conceptual or logical difference of the kind the Archimedeans want between jurisprudence so conceived and the more ordinary, day to day, concerns of lawyers and legal scholars I just mentioned. But there is nevertheless a sufficient difference in level of abstraction and in relevant skills to explain why the philosophical issues seem different, and are ordinarily in the hands of people with somewhat different training, from the more concrete ones. Any attempt at an ecumenical conception of legality faces pressure from

Hart’s Postscript and the Point of Political Philosophy

185

two directions. It must aim at sufficient content to avoid vacuity but also at sufficient abstraction to avoid parochialism.29 I tried to steer the needed course between these dangers in Law’s Empire: I said that legality is best served through a process of constructive interpretation along the lines, and responding to the two dimensions, mentioned above. My views have been sufficiently controversial to suggest that I escaped vacuity, but it is unclear how far I escaped parochialism. It is a frequent objection among British critics that my project is either parochial in inspiration—that it aims at no more than explaining the legal practice of my own country—or obviously parochial in result because we can somehow see, without much thought or research, that it fits only that one legal practice.30 In fact, my account aims at very great generality, and how far it succeeds in that aim can only be assessed by a much more painstaking exercise in comparative legal interpretation than these critics have undertaken. I said, earlier, in discussing other political values, that we cannot tell in advance how far we might succeed in finding plausible conceptions of these that reconcile them with one another rather than leaving them, as they are so often declared to be, in conflict. We must do our best, and then see how far we have succeeded. We must take the same view of the different question of how much abstraction an informative account of legality can achieve. We must wait and see. That leads me to a final story. A while ago, talking to Professor John Gardner of Oxford University, I said that I thought that legal philosophy should be interesting. He jumped on me. “Don’t you see?” he replied. “That’s your trouble.” I am guilty of his charge. But let me say what I mean by “interesting.” I believe that legal philosophy should be of interest to disciplines both more and less abstract than itself. It should be of interest to other departments of philosophy—political philosophy, of course, but other departments as well—and it should be of interest to lawyers and judges. Much legal philosophy has indeed proved of great interest to lawyers and judges. There is, just now, an explosion of interest in legal philosophy, not just in the United States, but in Europe, South Africa, and China, for example, as well. But this explosion is taking place not within courses called “jurisprudence,” which I fear remain rather dreary, but within substantive areas of law: constitutional law, of course, which has been theorydriven for a long time, but torts, contracts, conflicts of law, federal juris-

186

Hart’s Postscript and the Point of Political Philosophy

diction, and even, most recently, tax law as well. I don’t just mean that these courses engage theoretical as well as practical issues: they engage exactly the issues I have been discussing: about the content of legality and its implications for the content of law. But legal philosophers who regard their work as descriptive or conceptual as distinct from normative have, in my view, lost an opportunity to join these discussions and debates, and in some universities the dominion of jurisprudence has shrunk in consequence. On occasions like this one it is hard to resist speaking directly to young scholars who have not yet joined a doctrinal army. So I close with this appeal to those of you who plan to take up legal philosophy. When you do, take up philosophy’s rightful burdens, and abandon the cloak of neutrality. Speak for Mrs. Sorenson and for all the others whose fate depends on novel claims about what the law already is. Or, if you can’t speak for them, at least speak to them, and explain why they have no right to what they ask. Speak to the lawyers and judges who must puzzle about what to do with the new Human Rights Act. Don’t tell the judges that they should exercise their discretion as they think best. They want to know how to understand the Act as law, how to decide, and from what record, how freedom and equality have now been made not just political ideals but legal rights. If you help them, if you speak to the world in this way, then you will remain more true to Herbert Hart’s genius and passion than if you follow his narrower ideas about the character and limits of analytic jurisprudence. I warn you, however, that if you set out in this way you are in grave danger of being, well, interesting.

chapter seven

Thirty Years On

Introduction In The Practice of Principle, Professor Jules Coleman of the Yale Law School defends what he calls a version of legal positivism.1 A classic form of that theory of law holds that a community’s law consists only of what its lawmaking officials have declared to be the law, so that it is a mistake to suppose that some nonpositive force or agency—objective moral truth or God or the spirit of an age or the diffuse will of the people or the tramp of history through time, for example—can be a source of law unless lawmaking officials have declared it to be. Coleman sets his discussion in a narrow historical context. Over thirty years ago, I published a criticism of positivism.2 I argued that it is not faithful to the actual practices of citizens, lawyers, and judges in complex political communities: in practice, I said, people who argue about the content of law draw on moral considerations in a way that positivism cannot explain. Coleman treats my article as an important catalyst to the further development of the position I criticized. He says that though my challenge was “in many ways, misguided” (p. 67), and though “no one nowadays considers this argument convincing” (p. 105), it has nevertheless dominated recent jurisprudence because “two different and incompatible strategies of response [to it] have been articulated” (p. 67), and these strategies have produced two versions of positivism, and an exciting and illuminating contest between them. The first of these is “exclusive” positivism, which insists on the traditional positivist thesis that what the law requires or prohibits can never

188

Thirty Years On

depend on any moral test. Coleman names Professor Joseph Raz of Oxford as the leading contemporary proponent of exclusive positivism and discusses Raz’s views at some length. The second form of positivism is “inclusive” positivism, which allows moral criteria to figure in tests for identifying valid law, but only if the legal community has adopted a convention that so stipulates. Coleman sponsors that second form, and he devotes much of his book to arguing that his version of inclusive positivism is superior to any form of exclusive positivism, and greatly superior to my alternative, nonpositivist interpretation of law. Coleman’s book is clear, philosophically ambitious, and densely argued. It therefore provides a useful occasion to inspect the state of legal positivism three decades after the challenge he treats as catalytic. Have any of the subsequent formulations of legal positivism succeeded in reconciling that theory with actual legal practice? If so, which formulation is most successful? I shall argue that the arguments Coleman advances, and those he attributes to other positivists, are not successful. Exclusive positivism, at least in Raz’s version, is Ptolemaic dogma: it deploys artificial conceptions of law and authority whose only point seems to be to keep positivism alive at any cost. Inclusive positivism is worse: it is not positivism at all, but only an attempt to keep the name “positivism” for a conception of law and legal practice that is entirely alien to positivism. If I am right in these harsh judgments, a further question arises. Why are legal positivists so anxious to defend positivism when they can find no successful arguments for it? I shall later offer what I believe to be at least part of the answer: positivists are drawn to their conception of law not for its inherent appeal, but because it allows them to treat legal philosophy as an autonomous, analytic, and self-contained discipline.

Pickwickian Positivism Coleman claims that his version of inclusive legal positivism provides a better understanding of law than does my theory, and it is therefore puzzling that his actual account is so stunningly like my own. It is, in fact, hard to see any genuine difference.3 I said that the content of the law is not settled by any uniform behavior or conviction of lawyers and judges, but is

Thirty Years On

189

often understood to be controversial among them; that when lawyers disagree about the law they sometimes defend their different positions and try to resolve their disagreements by appealing to moral considerations; and that when the disagreement is particularly deep, these moral considerations may include claims about the best understanding of the fundamental point or purpose of the practice of law as a whole.4 Compare Coleman’s formulation. The framework of legal practice, he says, is created and sustained by the behavior of participants, but the content of the rules that constitute the practice is not. The rules are the result of ongoing negotiations. Thus there may well be disagreement about their content— disagreement, moreover, that is substantive and important, and which, given the framework, might well be settled by an appeal to substantive moral argument about how one ought to proceed, and that may invite discussion about the point of the practice. (p. 99)

It is therefore “not surprising that in resolving such disputes, the parties offer conflicting conceptions of the practice in which they jointly participate, conceptions that appeal to differing ideas of its point or function. In doing so, they may make substantive moral arguments” (pp. 157–58). This does not sound like positivism; it sounds like Hercules and his colleagues at work.5 But Coleman nevertheless insists that his theory of law is very different from mine and that his is a species of legal positivism, for a single reason on which he bases his entire case. The nerve of positivism, he says, is the thesis that the criteria for law “are and must be” matters of convention, and the difference between us is that in his view, but not mine, the legal practice that we describe in such similar ways is entirely a conventional practice (p. 100). Of course, if Coleman claimed simply that widespread and unchallenged understandings among lawyers and judges play an important role in familiar legal systems—that the American legal system would not function as effectively as it does if American lawyers did not agree that the requirements of the Constitution are part of American law, for example—his claim would be uncontroversial; neither I nor any other legal theorist would disagree. If he added only that these widespread understandings constituted conventions, his claim would be more contro-

190

Thirty Years On

versial, but hardly surprising. His claim is powerful because it insists on the much stronger thesis that the criteria of law are exhausted by conventions: that valid legal reasoning consists only of the application of special legal conventions to particular factual circumstances. If he were able to sustain this stronger claim then he would indeed have shown an important difference between our positions, and he would indeed be entitled to describe his account as positivistic. Coleman mainly follows H. L. A. Hart’s own influential version of positivism. Hart argued that every legal system necessarily depends on a master rule, or “rule of recognition,” for identifying any and all valid propositions of law. This rule exists only because it is accepted (at least by officials) as a matter of convention.6 If this thesis is correct, then positivism is vindicated, because legal conventions are formed by the complex behavior and attitudes of officials and other participants in the legal process, and by nothing more. But for that very reason, Coleman’s claim that law rests on convention seems at war with his admission that lawyers and judges often disagree about what the law is in ways that reflect moral disagreement, including disagreement about the very point of legal institutions. Convention is built on consensus, not on disagreement. Philosophers often use, as an example of a convention, the understanding that people must drive in the right lane of a two-lane road. That convention exists where it does because almost everyone does drive in the right lane, objects when anyone drives in the left lane, and supposes that it is an essential part of the reason for driving in the right lane and objecting when people do not that almost everyone does drive in the right lane and objects when others do not. When a group of people disagrees about what behavior is required or appropriate, it seems odd to say that they have a convention that decides the issue. Suppose that in a community where people drive on the right, there is no agreement about whether drivers may use the left lane for passing. Some drivers do so. But others do not: they wait until a third lane appears for that purpose, and they criticize those who do use the left lane as driving improperly. In such a community, at least according to the generally accepted view, there is no convention about passing etiquette. Coleman, however, insists that judges are following a convention about how to identify what the law requires even when

Thirty Years On

191

they disagree about what it requires. So he faces the challenge of explaining how that is possible. He uses what I believe to be two distinct strategies in trying to overcome that challenge (though he does not emphasize or even seem to recognize the difference between the two). The first depends on a distinction between two kinds of disagreement that the parties to an alleged convention might discover they have. They might discover that they disagree about the content of the supposed convention, that is, about what the rules of that convention actually are. Or they might find that though they agree about what the rules are, they disagree about whether a particular rule applies in some particular case, which means that they disagree about the convention’s application. Coleman concedes, when he is pursuing this strategy, that if judges systematically disagree about the content of the rules they should follow for identifying the law, then his claim that law necessarily rests on convention must fail, because that kind of disagreement shows that no such convention exists. He says that if we must interpret deep judicial disagreements about how to identify law as disagreements about content, “we should certainly have to grant Dworkin his conclusion that the rule of recognition must be a ‘normative,’ and not a social, rule” (p. 117). But he insists that judicial disagreement, even when it has a moral dimension, is always a matter of application, not of content. In some cases the distinction between content and application is clear enough, and we are able confidently to describe a disagreement as one about application. If the judges on a panel agree that they must end each day’s deliberation before 7:00 p.m., but disagree about what time it is, they are disagreeing about the application rather than the content of their convention. But the judicial disagreements Coleman has in mind are very different and it might seem odd to describe the most important of these as disagreements about application only. Suppose the judges on a particular court accept as a matter of convention that they must follow the past decisions of higher courts, but disagree about whether they must follow their own past decisions. It would seem natural to say that they share no convention about the deference they owe to their own previous decisions. We could avoid that conclusion, however, and classify their disagreement as one about the correct application of a convention instead, through what I

192

Thirty Years On

shall call the abstraction strategy. We could re-describe their convention in abstract moral terms; we might say, for example, that they share a convention that requires them to follow precedent in circumstances when it would be “right” or “proper” or “desirable,” all things considered, for them to do so. We might then say that they disagree only about the correct application of that more abstract convention—that they disagree only about whether it actually would be right or desirable for them always to follow their own precedents. We can use this strategy to convert any disagreement that any group might have over the standards that should govern its conduct into a supposed disagreement over the application of some more abstract moral convention they share. The drivers I discussed, for example, who disagree about passing in the opposite lane no doubt agree that drivers should drive “properly.” So we can say that their disagreement about passing in the opposite lane is only a disagreement about the correct application of a convention—the convention that requires proper driving— to the special issue of passing in the opposite lane. Coleman uses this strategy of abstraction to defend his claim that all law is based on convention. He concedes, as I said, that lawyers and judges often disagree about the law even when they agree about all pertinent facts: they disagree, for example, about whether certain laws that discriminate against homosexuals are valid. But he insists that this is only a disagreement about the application of a more abstract rule for recognizing law they all share, a rule of recognition that is framed in moral terms. He illustrates such abstract rules in this way: [A] rule of recognition that claims that no norm can be law unless it is fair appears to treat morality as a necessary condition of legality. A rule of recognition that claims that certain norms are law because they express a dimension of justice or fairness treats morality as a sufficient condition of legality. Clauses like the equal protection clause of the 14th Amendment to the United States Constitution or similar provisions in the Canadian Charter are, arguably, instances of rules of recognition that treat certain features of the morality of a norm as a necessary condition of its legality. On the other hand, the principle “No man should profit from his own wrongdoing” might be thought to be law because it expresses a dimension of justice or fairness, in which case its doing so appears to be sufficient for its legality. (p. 126)

Thirty Years On

193

In Coleman’s view, then, disputes about the correct application of the Equal Protection Clause are disputes about the application of an abstract moral convention—that unfair statutes are not valid laws—to particular cases. Coleman insists that all Supreme Court justices accept that convention, and when they disagree about whether laws criminalizing homosexuality are valid, they are merely disagreeing about the convention’s application to those laws. So even those parts of American constitutional law that most invite moral disagreement actually rest on convention. Coleman’s abstraction strategy trivializes positivism, however, in at least three ways. First, since there is no limit to how abstract a convention we may attribute to lawyers, we can use the strategy to count any legal practice as conventional no matter how much it seems to defy positivism’s traditional requirements. Imagine a community whose judges agree only that they should decide cases in the “proper” or “desirable” or “fair” way, but disagree radically about which methods of finding law result in proper or fair or desirable decisions, because they disagree radically about what is proper or desirable or fair. According to the abstraction strategy, their practice satisfies positivism’s test for a legal system, because they share a conventional rule of recognition—the abstract moral rule that whatever is proper or fair or desirable is law—and disagree only about the application of that supposedly conventional rule to particular cases. Second, positivism rescued in this way can no longer claim to show what is distinctive about law and legal reasoning, because once we accept the strategy, we can easily regard any community’s moral practices as resting on convention in the same way. Though Americans disagree about a great range of moral issues, they almost all agree that people should behave in the “right” or “proper” or “just” way, so their less abstract disagreements are all disagreements about the correct application of their more abstract shared convention. Third, the strategy eviscerates the idea of convention itself. A convention exists only when each person acts in a certain way because others are acting in that way as well; a convention makes the appropriateness of behavior dependent on the convergent behavior of others. That is why the claim that all law rests finally on convention is a positivistic claim. But it is implausible to think that any judge’s conviction that he ought to decide

194

Thirty Years On

cases in a “proper” way depends on the convergent behavior of other judges. A judge would think he should decide in a proper way whatever other judges do or think. What is the alternative? Deciding improperly? Perhaps Coleman would object that his abstraction strategy does not require him to attribute conventions to lawyers and judges as abstract as those I have just been imagining. But once he takes the first step up the ladder of abstraction—once he declares that any particular dispute about the proper standards for identifying law is actually a dispute about the correct application of a more abstract convention—he has no way of limiting the degree of abstraction the strategy might use. In any case, he must embrace a much higher level of abstraction than he appears to recognize even to show that most familiar legal systems, including the American system, actually rest solely on convention. His statement I just quoted suggests that American judges agree, as a matter of convention, that the Equal Protection Clause and other provisions of the Constitution make the validity of particular laws depend on moral tests, and disagree only about what those moral tests actually require. But that is certainly not true. On the contrary, the proposition that the Equal Protection Clause makes law depend on morality is itself deeply controversial. Many lawyers, including some Justices of the Supreme Court, insist that it is true, while others, including other Justices of the Supreme Court, heatedly deny it. Many of the latter insist that the clause makes the law depend on historical facts about what the framers thought unfair, or what Americans generally think unfair, or something of that sort. If Coleman tried to produce a master legal convention that fit the views of substantially all American lawyers and judges about how the abstract moral clauses of the Constitution should be interpreted, he could not stop much short of something as flimsy as my imagined “convention” that judges should interpret the Constitution in the “proper” manner—that is, in the manner that the best political theory justifies. Even that supposed convention would not be abstract enough to capture all American legal practice, however, because most disputes of political morality about how to identify valid American law do not turn in any substantial way on the text of the Constitution at all.7 Coleman’s second strategy, as I understand it, does not depend on any distinction between the content and the application of a convention, and

Thirty Years On

195

therefore does not require an abstraction strategy. He argues instead that the judicial enterprise, in any community that has law at all, is itself a kind of convention; law rests on convention, that is, not because officials all accept some fundamental rule of recognition, but because officials all accept that their enterprise is a pervasively cooperative one. He uses the concept of “a shared cooperative activity” (SCA), taken from Professor Michael Bratman.8 Coleman describes an SCA as “something we do together—taking a walk together, building a house together, and singing a duet together are all examples of SCA” (p. 96). Such joint activities display three characteristic features: mutual responsiveness (each participant is “responsive to the intentions and actions of the other”); commitment to the joint activity (“the participants each have an appropriate commitment (though perhaps for different reasons) to the joint activity”); and commitment to mutual support (each “is committed to supporting the efforts of the other to play her role in the joint activity”) (p. 96). Coleman claims that the activity of officials in any legal system can be seen as constituting an SCA with those features. “Officials,” he says, “can and do disagree about the content of the criteria of legality . . . and [believe] that such disputes are resolvable by substantive moral argument” (p. 158).9 Nevertheless, this disagreement “is perfectly compatible with the rule of recognition regulating a conventional social practice, and thus with the rule of recognition being a conventional rule” because “the sense in which the SCA is conventional is plain. Its existence does not depend on the arguments offered on its behalf, but rather on its being practiced—on the fact that individuals display the attitudes constitutive of shared intentions” (p. 158). Coleman believes that these two claims—that an SCA is necessarily rooted in convention and that legal practice necessarily constitutes an SCA—bring together, in an elegant way, all the different arguments that he has been making for inclusive positivism over many years (p. 99). But each of these claims is strikingly unsuccessful. First, the examples that Coleman cites of activities that display the characteristics assigned to SCAs, such as taking a walk or building a house together, need not constitute or involve conventions in any way. Two people walking together might find it useful to adopt or observe conventions—about who walks first through a narrow passage on a forest path, for example—but they need not, and if they do

196

Thirty Years On

not it would not follow that they are not engaged in the activity, or that they do not display sensitivity to each other’s intentions, or that they are not committed to the project, or to supporting each other if necessary. They do indeed need to communicate with one another, and constantly adjust their behavior in light of what the other does. But convention need play no part in this process. In any case, the joint activity may certainly proceed without being grounded in some fundamental and governing convention of the kind that would be necessary to support a conventionalist account of the act of, say, building a house together. Each party to a shared cooperative activity may judge for himself what is appropriate for him to do at each moment, given what others are doing, and with no guidance from any convention constituted by other people’s past or expected behavior. Second, it is hardly a conceptual necessity that the actors in the legal processes of any country display the mix of attitudes that Coleman lists as defining an SCA. He says that “[i]t is a conceptual truth about law that officials must coordinate their behavior with one another in various ways that are responsive to the intentions and actions of others” (p. 98). But that exceedingly abstract description fits almost every form of social life— people who take themselves to be moral also accept that they must coordinate their behavior with others “in various ways” even though they disagree about what that abstract commitment actually requires. Bratman’s conditions, as his article makes plain, require much more by way of concrete cooperation than that, and it is an empirical question whether the legal officials of any community meet these more exacting conditions— whether the judges share more concrete ambitions, for example, and how far each is committed to supporting rather than undermining the efforts of the others. Some American judges hope to reverse decades of constitutional law by increasing the power of the states against the national government, while others are committed to doing all they can to oppose that change. Some judges hope to make the law a more efficient instrument for securing national prosperity, while others, who reject that ambition entirely, hope instead to use the law to relieve great poverty. American judges are more divided than united by these ambitions. They take their walks and build their houses alone or in parties, not all together.

Thirty Years On

197

Coleman’s energetic efforts to show that legal systems must rest on some fundamental organizing convention, as a matter of conceptual necessity, all fail. I should at least mention, however, a further philosophical argument he makes about the relation between a convention and the rules it “fixes” (p. 81). This further argument might be thought to support his convention thesis, but in fact it threatens that thesis. He taxes me with the “mistaken” idea “that the convergence of behavior [that constitutes a convention] fully determines the content of the rule” that the convention fixes (p. 79 n.10). He fails, however, to distinguish two different claims, one of which I do make and the other of which is mistaken. We might think, as I do, that a community does not have a conventional rule requiring specific behavior unless most people in the community exhibit that behavior: that there is no convention fixing a rule that forbids passing in the left lane of a two-lane road, for example, when half the population thinks there is nothing wrong with passing there and does so openly, merrily, and with no compunction. This is the claim I need to support my view that if judges disagree fundamentally about the criteria for identifying valid law, then they do not share a convention that stipulates criteria for identifying valid law. Or we might think, as I do not, that it is sufficient for a convention to fix a particular rule that there is convergent behavior consistent with that rule. The argument Coleman offers to demonstrate my mistake, which he takes from Wittgenstein, shows only that the second of these claims is mistaken. Suppose the judges of a particular community have decided every civil case in favor of the oldest party in the case and have always cited as their justification the fact that the winner is the oldest party. We could not conclude that these judges are all following a rule that requires them always to decide in favor of the oldest party. Perhaps (to adapt Wittgenstein’s suggestion) half of them are following that rule and the other half a different rule that requires them to decide in favor of the oldest party provided that the case is heard before December 31, 2004, and to decide in favor of the youngest party if the case is heard thereafter. So for us to think that a group of judges shares a conventional rule about how to identify law, “[t]hey must tend to agree . . . on which future behavior will accord with the rule and which will not” (p. 80).10 Coleman adds that the judges

198

Thirty Years On

may disagree to some extent about how exactly to formulate the rule they are following, “especially if they were asked to specify it in all its details, or to project it to cover a range of difficult hypothetical cases” (p. 81). What is important, he insists, is that the judges do in fact reach the same decisions when the difficult hypothetical cases actually arise, despite the differences in rule formulation: “if the same rule is being followed, then participants must share a grasp or understanding of the rule that is reflected . . . in convergent behavior” (p. 81). “They must tend to agree, in other words, on which future behavior will accord with the rule and which will not” (p. 80). These comments seem to concede what Coleman struggles later to reject: that sharing a rule requires, at a minimum, sharing an understanding of what the rule actually and concretely requires in fully specified factual situations. To summarize: Coleman’s version of legal positivism is best described as anti-positivism. He has wholly decamped from the philosophical heritage he undertakes to defend. He covers his retreat by claiming to remain true to the cardinal tenet of positivism, which is that law is always a matter of convention. But his use of convention pursues victory through surrender. His first strategy trivializes the idea of a convention and makes it practically and theoretically useless. His second strategy, which hopes to convert cooperation into convention, fails because cooperation need not depend on convention, and because a legal system need not, as a matter of conceptual necessity, depend on full cooperation. We have made no progress in understanding the persistence of positivism’s acolytes—in understanding why Coleman, for example, is so anxious to fly the flag of positivism that he is willing to abandon every article of its faith to do so.

Ptolemaic Positivism My aim in this essay is to evaluate the arguments legal positivists have made over the last thirty years to defend their position. I must therefore discuss the defense strategy that Coleman identifies as the principal rival to his own. He calls that strategy “exclusive” positivism, and he names Joseph Raz as its avatar. Exclusive positivism holds that moral tests or considerations cannot figure in the criteria for identifying true propositions of

Thirty Years On

199

law. Raz’s argument for that bold proposition is complex; I shall try to explain it in some detail, but it might be helpful to summarize the argument in advance. He declares, first, that it is part of the very concept of law that law claims legitimate authority over some group; second, that that claim presupposes that legal directives are capable of being authoritative; and third, that no directive can be authoritative unless the content of that directive—what it requires people to do—can be ascertained without making any moral judgment. The argument, even in that skeletal form, might strike you as odd. It is important, practically and politically, to determine what judges may and must do in the exercise of their responsibility to enforce the law, and to distinguish that from other judicial acts and decisions that must rely on a different and more controversial kind of justification. It would be bizarre for such a crucial practical distinction to turn on an abstract analysis of the concept of authority. In fact, every step in the argument I summarized is highly problematic. The trouble begins with the initial personification: “I will assume,” Raz says, “that necessarily law, every legal system which is in force anywhere, has de facto authority. That entails that the law either claims that it possesses legitimate authority or is held to possess it, or both.”11 What can it mean to say that “the law” claims legitimate authority? This type of personification is often used in philosophy as a shorthand way of stating the meaning or content of a class of propositions. A philosopher might say, for example, that morality claims to impose categorical requirements, or that physics claims to reveal the deep structure of the physical universe. He means that no proposition is a true proposition of morality unless it accurately reports categorical (rather than only hypothetical) requirements or that no proposition is a true proposition of physics unless it correctly reports physical structure. If we read Raz’s personification in this familiar way, we take him to mean that no proposition of law is true unless it successfully reports an exercise of legitimate authority. But that would imply not that morality cannot be a test for law, as Raz claims, but that it must be a test for law, because, as he recognizes, no exercise of authority is legitimate “if the moral or normative conditions for one’s directives being authoritative are absent.”12 It is difficult to find a sensible alternative reading of Raz’s personificat-

200

Thirty Years On

ion. He sometimes suggests that when he says that “law” claims legitimate authority he means that legal officials claim that authority; legal officials do this when they insist that they have a “right” to impose obligations on citizens and that these citizens “owe them allegiance” and “ought to obey the law.”13 It is one thing to suppose that legal officials often make such claims; it is quite another to suppose that unless they make such claims there is necessarily no law. In fact, many officials do not. Oliver Wendell Holmes, for example, thought the very idea of moral obligation a confusion. He did not suppose that legal enactments replace the ordinary reasons people have for acting with some overriding obligation-imposing directive, but rather that these enactments add new reasons to the ordinary ones by making the cost of acting in certain ways more expensive. Whether a community has law does not depend on how many of its legal officials share Holmes’s views. So we cannot make sense of Raz’s crucial personification by supposing it to refer to the actual beliefs or attitudes of officials. True, he offers an alternative: it is enough, he says in the passage I quoted earlier, that the law “is held to possess” legitimate authority. He means, presumably, that it is enough if almost all the citizens think their law possesses that authority. But that does not seem necessary either. Suppose the citizens, like the officials, subscribe to Holmes’s view. Does law then cease, only to spring up again when a different and better jurisprudence takes hold? Is it not more sensible to say that Holmes was wrong, and that if he had converted everyone to his view then everyone would have been wrong? That American law does, in general, impose moral obligations of obedience on its citizens whether or not they think it does? So it remains mysterious how we should deconstruct Raz’s figure of speech. But suppose we assume, for the sake of argument, that he means to make the empirical assertion that every legal official believes that the laws he enacts create moral obligations, and let us further assume that this assertion is true. Now consider Raz’s second step. He points out that the officials’ claims would not be sensible unless the laws they created were capable of legitimate authority, and he concludes that nothing is law unless it is capable of legitimate authority. There are at least two flaws in that conclusion. First, it does not follow from the fact that some laws have legitimate authority—which is all we must assume to suppose that the of-

Thirty Years On

201

ficials’ claims are sensible—that nothing is a law unless it is capable of such authority. Legislators who insist that all of the laws they make impose moral obligations may not believe that all laws do, or even that all laws everywhere are capable of doing so. They may think that, just as a conceptual matter, they would make law if they enacted a statute declaring that the tides must cease to ebb and flow, though this would be a silly law that of course could not create any moral obligations. Second, even if all officials do believe that laws must necessarily be capable of imposing moral obligations and therefore capable of legitimate authority, this opinion may show simply that they are mistaken about the concept they employ. People often make sincere claims that rest on conceptual misunderstandings. Many people believe, for example, that even justified taxation necessarily compromises the liberty of taxpayers. In my view, such people are making a conceptual mistake: they do not understand the nature of their claims because they misunderstand liberty.14 Even if practically everyone made that claim, it would not follow that taxes necessarily or inherently compromise liberty. Raz says, optimistically, that officials cannot be “systematically” confused about the concept of authority because “given the centrality of legal institutions in our structures of authority, their claims and conceptions are formed by and contribute to our concept of authority.”15 But there may not be any conception of authority that counts as “our” conception. Just as different people even within a single community may hold different conceptions of liberty, so they may hold different conceptions of authority. Even large groups of them may hold mistaken ones.16 As we will soon see, Raz’s own conception of authority is eccentric. Even if he is right that it is the best conception, or the one that lawyers ought to adopt, it does not follow (and it is plainly not true) that it is the one they all have adopted already. Suppose we do grant, however, once again only for the sake of argument, that Raz is right that law must necessarily be “capable” of constituting legitimate authority. That would seem to mean, at least at first blush, that nothing is a law unless it meets all the necessary conditions of having legitimate authority. Raz believes that there are several such conditions. Some of these are moral conditions: if a legal system “lacks the moral attributes required to endow it with legitimate authority then it has none.”17

202

Thirty Years On

Raz would presumably agree, then, that a putative law has no legitimate authority if it commands what is morally wicked, or if it issues from an illegitimate power like a usurping dictator. Other conditions of legitimate authority are non-moral. Nothing can exercise legitimate authority, Raz reminds us, if, like a tree, it cannot communicate with others. Raz believes that it is another non-moral condition of a law’s achieving legitimate authority that its content can be identified without moral reasoning or judgment. (That is the crucial third step of the argument I summarized earlier, and we shall soon turn to it.) The moral conditions Raz recognizes pose a serious problem for him. If he accepted what seems to follow from his prior assumptions—that a genuine law must meet all the necessary conditions of legitimate authority, including moral ones—he could not be an exclusive positivist. Whether a law is too wicked to be legitimate is a moral question, and an exclusive positivist cannot allow the existence of law to turn on the right answer to a moral question. Raz understands this difficulty, because he is careful to declare that being “capable” of legitimate authority requires meeting all of the nonmoral conditions of that status but does not require meeting any of the moral ones. He says that this distinction is “natural,” though he does not explain why.18 It would save his argument from collapse into natural law, but it seems to have no independent merit. How can we say that a law is “capable” of achieving legitimate authority if it lacks even a single necessary condition of legitimate authority—if it could not possibly have legitimate authority no matter what other conditions were met or no matter what other circumstances held? Nor does it help to say that the non-moral conditions are conceptual in some way that the moral conditions are not. That distinction is both mistaken and irrelevant. It is mistaken because it makes perfect sense to frame an answer to the pertinent moral question as a conceptual claim: we can sensibly say that it follows from the very concept of legitimate authority that a wicked law cannot have legitimate authority. The distinction is in any case irrelevant because Raz’s crucial claim at this stage of his argument is about capacity, not concept. He supposes that it is a conceptual truth that law must be capable of legitimate authority, and the question now is only what is necessary to that capability. We can divide this into two questions: What is necessary for authority? What

Thirty Years On

203

is necessary for legitimacy? Raz argues that we must use our own concept of authority to determine the necessary conditions of authority.19 But then why should we not use our own ideas about legitimacy to decide what the necessary conditions of legitimacy are? If we accept Raz’s claim that law must necessarily be capable of legitimate authority, and we believe that law can never achieve legitimate authority if it is intrinsically wicked, then we must conclude that law cannot be inherently wicked, which means that positivism is false. Now we must turn to the third—and most important—step that I identified in my initial summary of Raz’s argument. Once again, we should grant his prior claims to test this third step independently, so we now assume, as a conceptual truth, that law claims and must be capable of legitimate authority, and that this means only that it must satisfy the nonmoral conditions of legitimate authority. Raz summarizes what he takes to be the two most important of these non-moral conditions in the following statement: First, a directive can be authoritatively binding only if it is, or is at least presented as, someone’s view of how its subjects ought to behave. Second, it must be possible to identify the directive as being issued by the alleged authority without relying on reasons or considerations on which [the] directive purports to adjudicate.20

The first of these two conditions is puzzling. If we take it literally, it means that very little of the legislation or common law of the United States can be authoritative. An ordinary statute is a compromise of the views of many different legislators and other influential actors in the political process, such as industries, lobbyists, and citizens’ groups. It rarely represents, or is even presented as, the views of any single legislator as to how citizens “ought to behave.” A common law doctrine is an accretion of many precedent decisions. It is unlikely to represent, and is rarely presented as, any single judge’s view of what citizens should do. Raz must mean that nothing is binding as statutory or common law unless it can be presented as the view of the legislature or the common law as institutions rather than as collections of individuals. But neither Congress nor the common law has a mind or views, so we have another troublesome

204

Thirty Years On

personification on our hands. We might try to unpack the personification in an innocuous way, by taking it to mean only that we must be able to summarize the content of any particular congressional enactment or common law doctrine in language of the form: “It is the view of Congress (or the common law) that people must behave in the following way: . . .” Nothing is easier; we simply stick that introduction before our favored interpretation of the statute or favored reading of the common law. But that innocuous reading of Raz’s personification would not capture his meaning. He roundly declares that my own account of law violates his first condition.21 He explains that in my view identifying law is often a matter of finding the best justification for past legislative decisions, and then adds that “[t]he best justification, or some aspects of it, may never have been thought of, let alone endorsed, by anyone.”22 But I can certainly claim that the statute endorses a principle that none of the individual legislators had in mind if I mean only that we can sensibly attribute that principle, by way of justification, to the statute itself. So Raz must have in mind something less trivial in setting out his first non-moral condition. But it remains mysterious what that could be. Lawyers sometimes talk as if a legislature, as an institution, had a collective mind composed of the mental states of the actual legislators, combined in some never-specified way. But Raz does not hold such a naïve view. Of course, if exclusive positivism were right, then I would be wrong to suppose that moral judgment is sometimes pertinent in deciding what a statute really says or what the common law’s “view” really is.23 But Raz offers the first non-moral condition of legitimate authority as part of his argument for exclusive positivism, not as a ukase that presupposes that doctrine’s truth. The second of Raz’s non-moral conditions of authority—that the content of an authoritative directive must be identifiable without recourse to moral judgment—is, of course, the heart of his exclusive positivism. This condition encapsulates his distinctive view of the point of authority: Authority, he says, occupies “a mediating role between the precepts of morality and their application by people in their behaviour.”24 Before the exercise of some authority’s authority, people are in direct touch, as it were, with a variety of moral and other reasons for and against actions they might consider. Authority interposes itself between people and their rea-

Thirty Years On

205

sons by weighing and balancing those reasons itself and then issuing a new, consolidating directive that replaces those multitudinous moral and other reasons with a single, exclusionary instruction. Those who accept the authority will henceforth exclude the reasons the authority has weighed for them from their own calculations as reasons for action and will rely only on the new, authoritative, direction.25 Before any law is adopted regulating the matter, for instance, people may have a variety of reasons for and against parading a lion in Piccadilly. When an authority enacts a law forbidding the practice, the authority has ruled that the reasons anyone has against the practice are stronger, in combination, than the reasons anyone has for it. Accepting the ruling as authoritative means not reassessing these various reasons or balancing the supposedly authoritative directive against those reasons, but simply taking the directive as the reason not to parade a lion in Piccadilly. The ruling would not be authoritative if people still had to weigh the reasons they always had for or against parading lions in order to decide what the statute said or meant. The law would not have replaced those reasons, but left them alive and kicking. This account of the nature and point of authority insists on a certain attitude toward authority. People must decide whether they accept a particular institution as authoritative. They might make that decision by asking themselves whether, in general, the institution is better positioned to weigh reasons on their behalf than they are to weigh those reasons for themselves. If they think it is better positioned, then they should accept the institution as an exclusionary authority. Of course, they cannot ask that question and make that decision retrospectively, by asking in a particular case whether the authority’s actual decision shows that it is better at weighing reasons for and against that very decision than they are. Doing that would subvert the point of authority altogether, because people would have to consider and weigh the background moral reasons to decide whether they should accept a particular decision as a replacement for those same background reasons. They must decide in general and in advance. So the option to accept an institution’s directives as authoritative would not even be open, Raz supposes, if the institution did not accept his second non-moral condition—that the content of an authoritative directive must be identifiable without recourse to moral judgment.

206

Thirty Years On

This is a coherent account of the point of authority. It presupposes, however, a degree of deference toward legal authority that almost no one shows in modern democracies. We do not treat even those laws we regard as perfectly valid and legitimate as excluding and replacing the background reasons the framers of that law rightly considered in adopting it. We rather regard those laws as creating rights and duties that normally trump those other reasons. The reasons remain, and we sometimes need to consult them to decide whether, in particular circumstances, they are so extraordinarily powerful or important that the law’s trump should not prevail. The American Constitution (at least in most scholars’ opinion) allows only Congress, not the President acting alone, to suspend the writ of habeas corpus, and the framers of that clause plainly took into account the reasons a president might have for suspending the writ on his own. Most of us treat the Constitution as both legitimate and authoritative. But many commentators nevertheless think both that Abraham Lincoln was morally right to suspend habeas corpus during the Civil War and that he acted illegally. Raz says that people who accept authority may nevertheless ignore it “if new evidence of great importance unexpectedly comes up.”26 But the urgent requirements of war were hardly new evidence: the framers, after all, fought a war themselves. Lincoln did not deny the Constitution’s authority in making his decision; he simply weighed that authority against competing reasons of the kind the framers had also taken into account that retained their vitality. Lincoln found that the latter were, under the circumstances, strong enough to outweigh the former. We must now inspect Raz’s account of authority in a different way, however, because he presents that account, not as a recommendation of deference to constituted authority that people are free to accept or reject, but as conceptual truth. It is part of the very concept or essence of authority, he insists, that nothing can count as an authority of those putatively subject to it must engage in moral reflection to decide whether to obey it or what it has said. It follows from that conceptual truth, given the conclusions of the earlier steps of Raz’s argument, that nothing can count as law if citizens must use moral judgment to identify its content. Consider the following extreme example. Suppose a nation’s legislature adopts a law declaring that henceforth, on pain of severe criminal punishment, subjects

Thirty Years On

207

must never act immorally in any aspect of their lives. That is an exceptionally silly statute, and life in that nation will thereafter be repulsive as well as dangerous. According to Raz, however, it would be a conceptual mistake to describe the statute as law as all. Even in this extreme example, his claim seems too strong. The statute, after all, has normative consequences for those disposed to accept its authority. They now have an additional reason to reflect carefully on the moral quality of everything they do and to act punctiliously, not only because they are now subject to official sanction, but also because their community has declared, through its criminal law, the cardinal importance of moral diligence. They would not be making a conceptual mistake if they said they were behaving differently out of deference to the authority of the new law. They would not say, however, that the statute had merely empowered officials to judge their conduct according to the officials’ own moral standards. If they were jailed for an act they thought scrupulously moral, they would insist they had been jailed contrary to law. As Coleman recognizes—it is the nerve of his “inclusive” positivism— nothing in the ordinary concept of authority prevents us from treating as authoritative a rule or principle that incorporates a moral standard. Suppose that a businessman in a trade where “caveat emptor” prevails converts to a religion whose sacred text enjoins its adherents to deal “honestly and fairly” in commerce. He will behave differently, and he will sensibly say that in doing so, he is deferring to the authority of his new religion— even though he must ponder the same reasons he always had to decide what that authority commands. Suppose he wonders one day whether it would be unfair not to disclose an evident defect to a buyer who has not noticed it. If he decides that it would be unfair and discloses the defect, he can sensibly say that he has deferred to religious authority. Sacred text forbids what is unfair; non-disclosure is unfair; therefore sacred text requires disclosure. It would be inaccurate to say that the sacred text has not directed him to disclose, but only to consider whether non-disclosure is unfair. His religion tells him to avoid what is unfair, not to avoid what he judges to be unfair. If he decides after careful reflection that non-disclosure is perfectly fair but years later changes his mind, he will then think that he once disobeyed a religious command.

208

Thirty Years On

We should not dwell on these extreme cases in which a directive simply incorporates some thin moral concept by reference, however, because the use that contemporary legal practice actually makes of morality is much more complex and selective. A statute may stipulate that no contract that “unreasonably” restrains trade is enforceable, for example, or a constitution may rule out any process of criminal law that denies “due” process. Citizens, lawyers, and officials deciding what these provisions require in practice must indeed reflect on some of the same issues that they would consider if they were concerned only to act morally—but only some of them and in a different way. They must ask what force should be assigned to these moral considerations in the context of the regulation at issue, and against the general background of other laws and regulations. They must, in short, take up what I have called the constructive interpretation of discrete regulations in a discrete context. Though (as critics of positivism have long insisted) that constructive interpretation does have a moral dimension, it does not recapitulate any process of reasoning that would be required but for the statute in question and its legal context.27 Morality plays the same complex and subtle role in common law adjudication. Consider a judge who must decide as a matter of first impression whether a patient who has been damaged by taking a negligently manufactured drug over many years, but who cannot show which of the several manufacturers of the drug made the pills she took, is entitled to recover compensation from all the manufacturers in proportion to their market shares in the years in question.28 The judge will naturally consider and balance two questions of fairness: whether it is fair to a patient in that position to deny recovery altogether, and whether it is fair to hold a manufacturer liable for damages for an injury it has not been shown to have caused. Nevertheless, the judge will defer to the authority of settled law in several ways in his overall judgment, and his judgment might well differ from the one he would make if called upon to vote, as a legislator, for a law settling the issue one way or another. He might think, for example, that given various precedents and other aspects of the legal background, he should not consider the impact of market-share liability on the economic health of the pharmaceutical industry or on medical research. I do not mean that he would necessarily be right to exclude such considerations in deciding what the law requires, or that a legislator writing on a clean slate

Thirty Years On

209

would necessarily be right to include them. I mean only that the example confirms the authority of law over judicial decisions in spite of the role that morality plays in deciding exactly what law, as an authority, requires or permits. Raz must have in mind, therefore, a special and eccentric conception of authority when he insists that law could have no authority if moral issues entered into fixing what the law requires even in this diffuse way. He calls his conception of authority a “service” conception, and recognizes that other, perhaps more familiar, conceptions have no such implication.29 Why, then, does he insist that only his “service” conception can elucidate the nature or concept of law? It would, of course, be a different matter if there were some other compelling reason to prefer exclusive positivism to any other theory of law. But the argument is supposed to go the other way around; we are supposed to be persuaded of exclusive positivism because we already accept the service conception. We need an independent case for that conception, and I find none in Raz’s arguments. The heroic artificiality of Raz’s view of authority is underscored, moreover, when we notice how much it contradicts common sense. We saw earlier, in discussing Coleman’s inclusive positivism, that the abstract provisions of the American Constitution, like the Equal Protection and Due Process Clauses and the provisions that protect free speech and freedom of religion, present evident problems for positivism in any form. Coleman supposes, as we saw, that these clauses incorporate moral standards and therefore make the validity of any other law depend on the right answer to a moral question. But because Raz is an exclusive positivist who denies that moral judgment is ever relevant to the identification of law, he cannot take that view. What view then can he take of the legal force of these abstract constitutional clauses? Coleman reports it as Raz’s opinion that these clauses do not themselves invalidate any other law, but only direct judges to determine whether particular statutes ought not be enforced in spite of the fact that they are perfectly valid (p. 110). That thesis stands ordinary opinion on its head: most lawyers and laymen think not that school segregation laws were perfectly valid until the Supreme Court decided they should not be enforced, but rather that the Court struck these laws down because it rightly found them constitutionally invalid. When the Court does strike down a statute on constitutional grounds, moreover, it

210

Thirty Years On

almost always treats that statute as if it were already invalid. It denies that the statute had any legal force even before the Court acted. So if Coleman correctly represents Raz’s account of the abstract constitutional clauses, it would indeed be counterintuitive. It is not clear, however, that Raz can consistently hold the view Coleman attributes to him. Raz says that no reading of the constitutional clauses is valid as a matter of law unless that reading is authorized by a legal rule of interpretation that is itself valid on exclusive positivist grounds—that is, an interpretation itself settled by some near-uniform legal practice.30 But as we saw earlier, no pertinent rule of constitutional interpretation has been settled in the United States in that way. Some lawyers read the abstract clauses as outlawing legislation that contradicts certain moral principles; other lawyers read those clauses as outlawing legislation that contradicts the framers’ understanding of what those principles forbid. Raz says that given such disagreement, the legal force of the abstract clauses must be considered “unsettled.”31 Presumably he thinks that the words of the Equal Protection Clause, for example, are part of our law because everyone agrees that they are. But supposing that those words authorized the Supreme Court to strike down some particular legislation would require accepting one controversial reading of the Clause and rejecting others. So presumably Raz sides with the most savage critics who say that for almost two centuries the Supreme Court has been exercising a power that no legal authority has given it. That conclusion seems troubling, but no alternative open to him is less so. Suppose he argues, for example, that even though the Supreme Court’s early exercises of judicial review were not authorized by law because no settled rule of constitutional interpretation justified reading the Constitution as bestowing such a power, the Court’s own powers of creating law through precedent mean that those first few exercises gave it the legal authority it now claims. But what then are the limits of that authority? When would the Court make a legal mistake in overruling a statute? Does the law now authorize the Court, if for some reason it thinks it wise, to treat the traffic laws of some state as if they were not valid laws? It might seem natural to say at least this: the Due Process and Equal Protection Clauses give the Court no power to strike down statutes that no reasonable person could think deeply unjust. But Raz cannot accept even that limitation on

Thirty Years On

211

the Court’s authority to declare legislation invalid once he has recognized that authority at all, because it is a moral question whether anarchists who think that traffic laws are subversive of a fundamental liberty are reasonable, even if the answer to that question seems clear. Raz must therefore choose between effectively denying that the Constitution is law and denying that anything but the Constitution is law. Why does he hold to a theory of law that has such a distressing consequence?

Positivism and Parochialism Jeremy Bentham, the first philosopher to present a systematic version of legal positivism, had politics very much in mind.32 He hoped to undermine the political power of judges who claimed to have discovered law in natural rights or in ancient traditions beyond what Parliament, as the congress of the people, had explicitly declared. Positivism had a democratic flavor then, and as democracy became supposedly more progressive, positivism became part of the anthem of that progress. Oliver Wendell Holmes and Learned Hand appealed to positivism to support progressive economic and social legislation against conservative Supreme Court justices who invoked supposed natural rights protecting established property to justify holding such legislation unconstitutional. In 1938, at the zenith of positivism’s practical importance, the Supreme Court embraced it to help justify one of the most consequential decisions in its history.33 The Court held that when federal courts have jurisdiction over a matter only because the parties are from different states, they cannot appeal to some independent, overarching body of legal tradition, but must enforce the law as declared by the authorities of one of the states involved. Justice Brandeis cited one of Holmes’s earlier statements of the positivist creed: [L]aw in the sense in which courts speak of it today does not exist without some definite authority behind it . . . [T]he authority and only authority is the State, and if that be so, the voice adopted by the State as its own [whether it be of its Legislature or of its Supreme Court] should utter the last word.34

The political influence of legal positivism has sharply declined in the last several decades, however, and it is no longer an important force either in legal practice or in legal education. Government has become too com-

212

Thirty Years On

plex to suit positivism’s austerity. The thesis that a community’s law consists only of the explicit commands of legislative bodies seems natural and convenient when explicit legislative codes can purport to supply all the law that a community needs. When technological change and commercial innovation outdistance the supply of positive law, however—as they increasingly did in the years following the Second World War—judges and other legal officials must turn to more general principles of strategy and fairness to adapt and develop law in response. It then seems artificial and pointless to deny that these principles, too, figure in determining what law requires. Following the war, moreover, the idea steadily gained in popularity and in constitutional practice that the moral rights people have against lawmaking institutions have legal force, so that if the legislature condemns a class of citizens to second-class status, its act is not simply wrong but also void. Once again, it seemed increasingly pointless to declare that these moral constraints on government were not themselves part of the community’s law. The political appeal of positivism correspondingly drained away. It was associated no longer with democratic progress, but with conservative majoritarianism; it was liberal judges who appealed to morality in justifying greater legal protection for individual rights. Most academic lawyers assumed that if a general theory about the nature of law was needed at all, it had to be more subtle than legal positivism. Academic defenders of legal positivism, like Coleman and Raz, remain. But their arguments, as I have tried to show, have the artificiality and strain of theories that defenders of a sacred faith construct in the face of embarrassing evidence. What is that sacred faith? The remaining positivists are not political conservatives who hope to resist the spread of individual and minority rights by insisting on a majoritarian conception of law. On the contrary, they argue that legal positivism imposes no constraints on what judges and legal officials actually do: they have traveled far from Bentham and from Erie. They celebrate positivism as an accurate description of the very concept of law or as the most illuminating theoretical description of legal phenomena over time. Sometimes they treat these as pretty much the same thing. But they offer no serious explanation of the kind of conceptual analysis they have in mind and no empirical evidence that might support large generalizations about the forms and histories of

Thirty Years On

213

legal institutions. They make little attempt to connect their philosophy of law either to political philosophy generally or to substantive legal practice, scholarship, or theory. They teach courses limited to “legal philosophy” or analytic jurisprudence in which they distinguish and compare different contemporary versions of positivism, they attend conferences dedicated to those subjects, and they comment on each other’s orthodoxies and heresies in the most minute detail in their own dedicated journals. Why? Part of the answer, I believe, lies in the continuing influence of two talented philosophers who long ago published influential accounts of positivism: H. L. A. Hart in the Anglo-Saxon world and Hans Kelsen in the rest of the world. But that partial answer also deepens the puzzle: why should these philosophers—who, like Minerva’s owl, flew in the dusk of their traditions—still command reverence and encourage intellectual insularity? I do not know, but I now suspect that at least part of the explanation is to be found in the appeal not of positivism as a theory of law, but of legal philosophy as an independent, self-contained subject and profession. Positivists since Hart (including Hart in the Postscript to his book, The Concept of Law,35 which was published after his death) have defended with great fervor a guild-claim: that their work is conceptual and descriptive in a way that distinguishes it from a variety of other crafts and professions. On their understanding, legal philosophy is distinct not only from the actual practice of law, but also from the academic study of substantive and procedural fields of law because both practice and academic study are about the laws of some particular jurisdiction, whereas legal philosophy is about law in general. It is also distinct from and independent of normative political philosophy because it is conceptual and descriptive rather than substantive and normative. It is different from the sociology of law or legal anthropology because those are empirical disciplines, whereas legal philosophy is conceptual. It is, in short, a discipline that can be pursued on its own with neither background experience nor training in or even familiarity with any literature or research beyond its own narrow world and few disciples. The analogy to scholastic theology is, once again, tempting. In their pursuit of guild-exclusivity, positivists must fight a war on several fronts. In what sense is legal philosophy conceptual? In what sense, if

214

Thirty Years On

it is conceptual, can it also be descriptive? John Austin’s nineteenth-century defense of positivism aimed to be both conceptual and descriptive in an old-fashioned but easily understood sense—he offered a definition of the word “law” as it figures in legal argument and discourse.36 Fifteen years ago, I took Hart’s aim in The Concept of Law to be a more sophisticated form of the same ambition: he aimed, I said, to explicate the concept of law by bringing to the surface the shared criteria that lawyers and laymen follow, often unawares, in assessing propositions of law.37 Hart himself disclaimed this understanding of his methods in his posthumously published Postscript,38 as have several of his followers (p. 200).39 I continue to think, however, that it makes more sense of his work than any description of his goals that he or they have suggested. It is true that Hart said that he regarded his book as an exercise in “descriptive sociology.”40 But that remark is more obscuring than clarifying. What kind of sociology is conceptual? What kind makes no use of empirical evidence? What kind defines itself as studying not just legal practices and institutions here and there, but the very concept of law everywhere? Positivists have apparently settled, at least temporarily, on a response to these challenges, which, they say, shows how legal philosophy can be conceptual, descriptive, and global at once. Coleman duly reports this solution (p. 10 n.13): instances of law, wherever they appear and in whatever shape, share a common “structure” (p. 10 n.13) that belongs to the very essence of law, and it is the role of conceptual, descriptive jurisprudence to reveal that common structure. Jurisprudence is therefore in one sense normative—it aims to capture, in an elegant and revealing way, what is truly important and fundamental about law’s structure rather than what is merely contingent or peripheral. But it is not normative in the manner of substantive political philosophy because it does not evaluate the structure it discloses or choose among structures in such a way as to improve legal practice. “[W]hereas a Dworkinian interpretation is committed to displaying law in its ‘best light,’” Coleman says, “the method employed here is not; it is committed only to identifying what principles, if any, reveal the actual structure and content of practical inference in the law” (p. 10 n.13). Jurisprudence is in one sense conceptual—it teaches us what belongs to the very essence of law’s fundamental structure. But it is not a priori, and

Thirty Years On

215

it says nothing about proper or improper usage of the word “law.” Jurisprudence is in one sense empirical—it is continuous with empirical studies of law, which provide the raw material for its judgments about law’s structure. But it goes beyond the empirical as normally understood because it also organizes that raw material into a systematic and revealing account of law’s essence. “Descriptive sociology,” Coleman says, “enters not at the stage of providing the theory of the concept, but at the preliminary stage of providing the raw materials about which one is to theorize” (p. 200). Each of these claims makes trouble for the others, however. Coleman celebrates the contribution that the social sciences can make to legal philosophy: “By attending to these inquiries outside of or beyond philosophy, we can obtain a rich and valuable picture of the forms of governance and organization that have been characterized as constituting law in different times and places, and under very different circumstances” (p. 201). But his argument neither uses nor even mentions any of the “rich and valuable” materials of legal anthropology, for example—neither does the work of Hart or of any of the other positivists Coleman discusses—and it is hard to see what use he could make of them. How could induction from a thousand very different cases of legal institutions, and from the varying motives and assumptions of thousands of actors in different times and places, reveal the “essence” or “very nature” of law’s structure? We might just as well declare that heterosexual union is the “essence” of marriage because marriage has always been heterosexual (if that is indeed true), so that homosexual marriage is ruled out on “conceptual grounds.” There is an even deeper problem, however. Though positivists now write, as Coleman does, about laying bare law’s essential structure by showing what is constant to that institution over time, they have said nothing to defend the mysterious idea on what this is all premised, which is that law has an essential structure that can be exposed purely through description. Atoms and animal DNA have inherent physical structures, and it makes sense to suppose that these structures dictate the “essence” of hydrogen or of a lion. But there is nothing comparable about a complex social practice: where should we look for its “essence” or nature? What physical or historical or social fact could demonstrate, for example, that it

216

Thirty Years On

is part of law’s “structure” or “essence” or “nature” that it must be conventional, as Coleman claims, or that it must be authoritative, as stipulated by Raz’s “service” conception of authority? What in the physical or historical or social world forces these conceptual “truths” upon us, quite independently of our political and moral purposes? Positivists must either answer these questions or give up their suddenly popular claim that they are probing some deep historical, social, anthropological, or psychological reality. Until someone redeems that claim through an intelligible account of the “nature” or “essence” of law, it will remain only a comforting mantra: positivism’s phlogiston. I must add, finally, that Coleman rubs salt in these self-inflicted wounds by claiming, in the subtitle of and throughout his book, that his account of law is a “pragmatist” one. His explanation of pragmatism, however, consists of a list of semantic and methodological assumptions (pp. 3–12), few of which do any real work in his argument, and all of which would be accepted by philosophers who are very far from the pragmatist tradition. In fact, nothing could be further from the spirit of genuine American pragmatism than his conviction that we can discover “conceptual” truths about a political practice without considering which understanding of that practice best advances our practical, moral, and political goals.

Appendix: Points of Personal Privilege I have already reported that Coleman’s book discusses my own work. I should add that it does so extensively, and in a particularly troubling way. He makes large but undefended claims about my “philosophical confusions” (p. 155), “deeply embedded” mistakes (p. 181), and “Disney-like” arguments (p. 185). He attributes important views to me that I have never, so far as I can recall, either held or expressed. In several instances (I have already mentioned some of these) he uses my own arguments against me. These errors are compounded by an almost complete failure of citation: my alleged beliefs and confusions are often reported with no reference whatsoever to where they are supposedly stated or committed. I have undoubtedly been guilty of failures of citation myself, but never, I think, on this scale. I have left my complaints to this separate section because they

Thirty Years On

217

are of little general interest. But I have found, in the past, that uncorrected misreports and unanswered criticism take on lives of their own in the literature of jurisprudence. I shall therefore try to correct what seem to me Coleman’s most serious errors of attribution and criticism.41 1. I begin with his description of my summary of legal positivism, not because his mistakes about this are particularly important to his own argument, but because it is a common complaint among positivists, always made without documentation, that I have misrepresented their views and created a straw man to attack. Coleman, for example, refers to one of my arguments “as yet another instance of Dworkin’s mischaracterizing [the positivists’] project in an effort ultimately to discredit it” (p. 155). That is a serious charge, but the few examples he offers, sprinkled throughout his book, conspicuously fail to support it. He says that I attribute to positivists “the claim that every norm that falls within the category ‘law’ must be a rule” (p. 104), and then adds that “no legal positivist has ever actually held that all legal standards are rules,” and that Hart “denied the claim long before Dworkin attributed it to him” (p. 107). But I defined “rule” as a term of art, and it is not likely that Hart was able to deny any claim I made about rules “long before” I made it. I argued not that positivists claimed that all legal standards were rules on my stipulated definition—I do not know how many have embraced that definition—but that the tenets of positivism I described would fit only rules so defined.42 He says that I attribute to positivism “two theses, neither of which any positivist holds or must hold” (p. 155). The first is my claim that Hart was engaged in a project of semantic analysis. It is true, as I said earlier, that Hart rejected this description in his posthumously published Postscript. But, as I also said, other writers have supported my interpretation, agreeing with me that it offers the best understanding of what Hart had actually written in The Concept of Law. Coleman makes no attempt to answer their arguments or mine. The second of my allegedly plain mischaracterizations is “that the semantics to which the positivist is committed is ‘criterial’” (p. 155). In fact, in spite of Coleman’s use of quotation marks, I did not use that word in describing positivism’s mistakes. I did describe certain assumptions that I said some positivists hold about the conditions under which people may be said to share a concept. But I did not say that positivism was committed

218

Thirty Years On

to those assumptions; on the contrary, I described a different set of assumptions that would allow positivism to be restated in a stronger form.43 When Coleman explains my supposed mischaracterization in more detail, moreover, he does not deny that some positivists have held the concept sharing assumptions that I attributed to them. He says only, echoing me, that these assumptions are not necessary to positivism, and then describes an alternate semantic base for positivism that, as we shall soon see, is exactly the base that I recommended to them (pp. 156–57). He does nothing to clarify or defend the familiar claim among legal positivists that I have distorted their position. I hope the next positivist who makes the accusation will be more forthcoming. 2. Coleman claims that I am guilty of a primitive confusion between two very different claims: that morality is not necessarily a condition of law, and that morality is necessarily not a condition of law (pp. 151–52). He offers no citation to prove my guilt, however, and I have in fact insisted on the distinction. I accepted the first of these claims and have only argued against philosophers who defend the second.44 3. He announces that one of my arguments, which he refers to as the “semantic sting” argument, is “riddled with philosophical confusions” (p. 155). But he identifies only one alleged confusion: “The problem with the semantic sting argument is that it mischaracterizes what individuals must share in order to have the same concept” (pp. 156–157). According to Coleman, I suggest that people can share a concept only when they agree on criteria for its application, whereas he insists that “[i]ndividuals can share the same concept if they agree, not on criteria of application, but on a set of paradigm cases or instances of the concept. These paradigm cases, in turn, are each in principle revisable, though they cannot all or nearly all be revised at the same time” (p. 157). People who share a concept in this “paradigm” way, he continues, may disagree fundamentally about the best understanding of why the paradigms are indeed paradigms, and may therefore disagree about whether the concept applies to fresh or controversial cases. The purpose of my “semantic sting” argument was to make exactly that point: sharing a concept does not necessarily mean sharing criteria for its application, but might instead mean sharing paradigms as the basis for interpretive claims.45 Coleman’s account of concept sharing

Thirty Years On

219

through paradigms is, in fact, almost identical to my own, using many of the same terms, including “paradigm.”46 He adds that this account of concept sharing is “a standard pragmatist view” (p. 157), but again offers no citation to any writing in which this “standard” view of concept sharing is actually described. I do not claim originality for my account, of course, and the idea of a paradigm is in wide use among philosophers. But I am unaware of any other exposition of concept sharing through paradigms that is as close to my own as Coleman’s, and I would have been grateful for even one citation. 4. He declares that I see law only from the perspective of litigants and judges and that as a result I ignore the important “guidance” function that law serves for people generally (p. 166). In fact, I emphasized the importance of law as a quotidian guide for people generally, and argued for what I called a “protestant” understanding of law for that reason.47 Law serves the guidance function much better, in a complex and changing community, on that protestant understanding than on the assumption of most positivists that when fresh or controversial issues arise there is no law but instead an occasion for a judicial discretion that can be exercised only in litigation. Coleman develops his odd thought that I do not account for law’s guidance, however, through a more elaborate argument that he calls “devastating” (p. 167 n.23): Suppose one holds that it is a conceptual truth about law that something is law only if it is capable of guiding conduct, and a norm, decision, or rule is capable of guiding conduct only if those to whom the law is addressed can know in advance what it requires of them. (p. 167)

On my view, he says, it can never be established with absolute certainty in advance of argument that any proposition of law—no matter how widely accepted and unchallenged—is true, because it is always possible that someone will show that proposition to be false through a better interpretation of the settled law. So, on my view, law cannot be capable of guiding conduct. But the quoted assumption is ambiguous. It might mean that no proposition of law can be true unless it is possible for people to form the opinion, on the basis of arguments available to them, that it is true. That proposition is innocuous enough: I would have no reason to dissent

220

Thirty Years On

from it (though I would cavil at describing it as a “conceptual” truth). Or the quotation might mean that no proposition of law is true unless it is possible for people to know, in the strong sense of being able to demonstrate conclusively to everyone, without fear of contradiction, that it is true. Why should anyone suppose anything so silly as that? How could Coleman, who believes as strongly as I do that true propositions of law can be controversial, suppose it? 5. The most baffling of Coleman’s misattributions spreads over many pages and is the predicate for some of his most ungenerous accusations (p. 185). The trouble begins with the following account of “elements” of my “theory of legal content” (p. 163): One such element is Dworkin’s view that the officials whose task it is to determine the content of the law claim that the authority they exercise is legitimate. This amounts to the claim that their exercise of the coercive authority of the state is justified. A second subsidiary element of the account is Dworkin’s view that the principle of charity requires that we treat the majority of these claims to legitimacy as true. (p. 163)

Once again, there is no citation for either of these surprising “elements.” The first is not serious, but it is inaccurate. I am unaware of making any claim about the beliefs of officials in their own legitimacy. I do not know, for example, how many South African judges thought that the power they exercised in the apartheid period was legitimate, though I do know that several thought it was not. The second “element” is even more puzzling. I have made no reference to any principle of charity that requires us to assume that the actions of most officials are legitimate. The obscurity deepens a few pages later, where Coleman refers to “the way in which the principle of charity—which Dworkin takes from Donald Davidson—is thought to apply holistically over the range of authoritative legal pronouncements” (p. 168). Davidson’s principle of charity, which plays a central role in his own account of meaning and truth, has nothing to do with my account of law. Davidson himself figures in my arguments in Law’s Empire only once: as the source of an example that has nothing to do with charity.48 Coleman has invented my reliance on a principle and then berates me, again and again, for having relied on it. Perhaps he has confused

Thirty Years On

221

my appeal to the political virtue of integrity49 or my account of constructive interpretation in law50 with an appeal to Davidsonian charity about truth. If so, he has seriously misunderstood either my arguments or Davidson’s, or both. Once again, a citation to my supposed invocation of Davidson or his principle would have been enormously helpful. 6. His next set of criticisms returns to the “semantic sting” argument. His introduction is, once again, severe: “[W]e must first attend to a fundamental confusion at the heart of [Dworkin’s] argument . . . This is the confusion between the content of the concept of law and the content of the law of a particular community” (p. 180). In fact, there is an obvious and important connection between these two ideas. When a lawyer declares, for example, that in his jurisdiction “it is the law” that price-fixing contracts are illegal, he is using the concept of law in stating the content of the law of a particular community. I relied on that connection in posing this question: if two lawyers appear to disagree substantially about the content of the law, as lawyers often do, do they therefore have different concepts of law—in which case the disagreement is illusory? I pointed out that they cannot share the same concept of law if we accept what Coleman calls the “criterial” view of concept sharing, which is that people share the same concept only when they agree, at least substantially, on the criteria for the successful application of that concept to particular cases. I concluded that because lawyers obviously do genuinely disagree about the content of the law of their jurisdiction, and therefore do share a concept of law, the criterial view of how they share that concept must be wrong. I proposed the different view I refer to above. Lawyers share the concept of law as what I call an interpretive (or essentially contested) concept. They do not agree on criteria for applying the claim “it is the law that,” but rather offer rival interpretations of paradigm propositions of that form that they both accept, and then extract from these paradigm propositions different criteria for applying the judgment “it is the law that” to fresh cases about which they might disagree. Coleman, as I said in the last paragraph, also endorses this view of how lawyers employ the concept of law. So what is my “fundamental”—or, as he also says, “deeply embedded” (p. 181)—confusion? His explanation comes to this: if I am right that lawyers share the concept of law as an interpretive concept, then, contrary to what I say, they

222

Thirty Years On

actually do agree about criteria for applying that concept. They agree that it should be applied in the interpretive manner that I described (p. 182).51 But this explanation misunderstands what I meant when I said that lawyers do not share criteria; I clearly meant that they do not agree on a single set of tests for deciding which propositions of law are true. Since Coleman agrees, it remains mysterious why he supposes that I am confused. 7. Coleman ends his discussion of my work with a series of remarks about the character of legal philosophy. In my view, just as lawyers must engage in constructive interpretation of aspects of settled law and legal practice—which will often have a moral dimension—to decide what the law is on particular issues, so must legal and political philosophers also use constructive interpretation of legal practice as a whole—which will also have a moral dimension—to defend particular conceptions of the concept of law itself. That is the suggestion he calls “Disney-like” (p. 185). “This is an argument for normative jurisprudence all right,” he adds, “but how seriously can we take it? At each crucial point the inferences seem to come out of thin air” (p. 184). He ignores the long argument I actually made for that view, which ranges over most of Law’s Empire and draws on theories of interpretation, integrity, community, and equality. He repeats his bizarre claim that my argument depends on a misunderstanding of Davidson’s principle of charity, and I must therefore repeat that I make no use of that principle. But he does not address any of my arguments. Readers of his book who may be impressed at my supposed legerdemain in plucking inferences from the air will therefore be disappointed, if they care to check, by the decidedly plodding and terrestrial character of what I actually said.

chapter eight

The Concepts of Law

The Semantic Sting In this chapter, newly written for this book, I explore in greater detail certain philosophical issues raised earlier. I have emphasized throughout the importance of distinguishing among the different concepts people use to talk about law: the doctrinal concept that we use in stating what the law of some jurisdiction requires or forbids or permits (“Ignorance is no defense under the law”); the sociological concept we use to describe a particular form of political organization (“The Romans developed a complex and sophisticated form of law”); the taxonomic concept we use to classify a particular rule or principle as a legal principle rather than a principle of some other kind (“Though the rule that seven and five makes twelve figures in some legal arguments, it is not itself a rule of law”); and the aspirational concept we use to describe a distinct political virtue (“The Nuremberg tribunal was preoccupied with the nature of legality”). Each of these concepts might be said to be a concept of law, and they are plainly interconnected in various ways. But they are nevertheless different concepts from one another, and though they all generate questions about the relation between law and justice, the questions they generate are very different. I noted in the Introduction that concepts serve different kinds of functions in the thought and discourse of people who share and use them. Some concepts characteristically function as criterial, like the concept of bachelor; others as natural kind concepts, like the concept of tiger; and still others as interpretive concepts, like the concepts of justice and democ-

224

The Concepts of Law

racy. The differences among these functions are crucial in considering what form an illuminating analysis of a concept or of the nature of the objects or phenomena the concept embraces might take. If we treat a particular concept, like bachelorhood, as filling a criterial function, our analysis consists in a statement of the correct criteria for that concept’s use either in the shape of a classical definition or of a formulation of the rules those who share the concept follow, perhaps unawares, in applying it. If the concept is imprecise it would be a mistake to press the analysis into the area of imprecision by claiming, for instance, that necessarily an eighteen-yearold boy cannot be a bachelor. Our analysis should simply report that the concept’s shared criteria underdetermine the question whether he can be. In my view, as I said, the sociological concept of law is an imprecise criterial concept; and the argument, once much beloved of legal philosophers, whether the Nazis had a legal system is bootless for that reason. If we can treat a concept as a natural kind concept, however, then a much different kind of analysis of that concept is available. We can assume that the objects it collects have an intrinsic nature—an essential structure without which they would not be the kind of thing they are—even though we may not know what that is. So a helpful analysis of the concept might well consist not in a statement of the criteria people use to identify examples but in a physical or biological account of that essential nature. Concepts that function as interpretive are also different from criterial concepts, but in a different way. I described interpretive concepts in the Introduction and in Chapter 6. Sharing an interpretive concept does not require any underlying agreement or convergence on either criteria or instances. Liberals and conservatives share the concept of justice, but they do not agree either about the criteria for judgments about justice or about which institutions are just and which unjust. They share the concept because they participate in a social practice of judging acts and institutions just and unjust and because each has opinions, articulate or inarticulate, about what the most basic assumptions of that practice, its point and purpose, should be taken to be. They draw from those assumptions more concrete opinions about the right way to continue the practice on particular occasions: the right judgments to make and the right behavior in response to those judgments. An illuminating analysis of the concept of justice

The Concepts of Law

225

must be an interpretive theory of just that kind. The analyst must deploy his own sense of the values that the practice should be taken to serve and of the conceptions of the concepts in play that best serve those values. An analysis of a criterial or natural kind concept must be neutral among the various normative controversies in which the concept figures: whether bachelors should be encouraged to marry or whether tigers should be protected as an endangered species. But a useful analysis of an interpretive concept—beyond the bare statement that it is interpretive and a very general account of the practices in which it figures—cannot be neutral. It must join issue in the controversies it hopes to illuminate. I have argued for many years that in many circumstances moral facts figure among the basic truth conditions of propositions of law. In Chapters 6 and 7 I discuss a rival claim: analytic doctrinal positivism, which holds that as a conceptual matter moral facts cannot figure among these basic truth conditions. In an early book, Taking Rights Seriously, I argued that analytic positivism distorts the actual practices of lawyers and judges in contemporary legal systems and therefore provides an inadequate understanding of those practices.1 In Law’s Empire I offered an explanation of why analytic positivists make that mistake.2 I hypothesized that they assume that all concepts, including the doctrinal concept of law, are criterial concepts and that proper analysis of the doctrinal concept must therefore consist in elucidating the tests that lawyers share, except in borderline cases, for judging whether propositions of law are true. I called that assumption—that all concepts are criterial—the “semantic sting.” I then offered an analysis of law as an interpretive rather than a criterial concept to show how lawyers’ disagreements can be genuine even if they do not agree on criteria for identifying true propositions of law. I tailored that account of the semantic sting, in 1986, to fit the arguments for legal positivism then current. But a fresh generation of legal philosophers has become more sophisticated in the philosophy of language, which has itself moved on since I wrote, and my description has proved too narrow.3 I should now characterize the sting more broadly: it lies in the assumption that all concepts depend on a convergent linguistic practice of the kind I described in the Introduction: a practice that marks out the concept’s extension either through shared criteria of application or

226

The Concepts of Law

by attaching the concept to a distinct natural kind. The infection of the semantic sting, I shall now say, is the assumption that all concepts of law, including the doctrinal, depend on a convergent practice in one of those two ways. The pathology of the semantic sting remains the same. Lawyers who are stung will suppose that an analysis of the concept of law must fit—and only fit—what lawyers mainly agree is law.

Dworkin’s Fallacy In “Dworkin’s Fallacy, or What the Philosophy of Language Can’t Teach Us about the Law,” Michael Stephen Green defines a fallacy to which he gives my name. “Dworkin’s fallacy,” he writes, “is using the interpretive theory of meaning to justify an interpretive theory of law.”4 My own version of that fallacy, he believes, consists in supposing that because the doctrinal concept of law is an interpretive concept the best account of the truth conditions of propositions of law must itself be interpretive, that is, that propositions of law are true when they follow from the best interpretation of the bulk of pertinent standing law. Of course that is a fallacy. But it needs a new name because I am not guilty of it and in fact took pains to warn against it. I do believe that the doctrinal concept of law is an interpretive concept, and I also believe that the truth conditions of propositions of law are interpretive in that way. But I do not believe that the second of these beliefs follows from the first. On the contrary, I spent three chapters of my book Law’s Empire considering other, very different theories about the truth conditions of propositions of law that I said are also consistent with treating the doctrinal concept as interpretive. One of these is a form of legal pragmatism, the theory I discuss at greater length earlier in this book. Another I called conventionalism, which is a version of legal positivism offered in the spirit I described in Chapter 6, that is, as itself an interpretation of contemporary practice. I argued that my own interpretive answer to the question posed at the doctrinal stage of legal theory is best because it provides the best such interpretation, not because the doctrinal concept is itself an interpretive concept. Green’s surprising misreading is best explained as a symptom of a lin-

The Concepts of Law

227

gering semantic-sting infection. He discusses at some length a philosophical dispute between what he calls a “traditionalist” and a “realist” view of the concepts of gold, water, and tiger. The former view treats these concepts as criterial, and the latter insists that they are not criterial but function rather as natural kind concepts. Green apparently assumes that this is a dispute about how all concepts should be understood. So he says that I make “arguments about meaning similar to those of the realists in order to attack the jurisprudential position that law is exhausted by agreement or convention.”5 That is not correct, and the misunderstanding is important. I did reject a criterial account of the doctrinal concept of law. But I did not adopt a natural kind semantics for that concept; in Chapter 6 of this book I explicitly rejected a natural kind interpretation of any of the concepts of law. Instead I said that the doctrinal concept, like other important political concepts, is interpretive.

Raz on the Concepts of Law In Chapter 7 I identified Joseph Raz as among the leading contemporary analytic doctrinal positivists and explained what I took to be the errors in his attempt to defend positivism. Since I wrote that essay, Raz has published a complex exploration of his philosophical methodology with the intriguing title “Can There Be a Theory of Law?,” in which he raises a host of interesting conceptual questions. He says that legal theory is “an exploration of the nature of law” that takes the form of explaining the concept of law and therefore must satisfy a series of conditions, one of which is “setting the conditions of the knowledge involved in complete mastery of the concept, which is the knowledge of all the essential features of the thing it is a concept of, that is to say, the essential features of law.”6 Raz does not distinguish, either in this description of his project or in its execution, between the sociological and doctrinal concepts of law. On the contrary, though he recognizes the distinction he explicitly yolks the two concepts together by supposing that his account of explaining a concept holds for both of them. “Here and in what follows,” he says, “I will use ‘law,’ as it is often used, to refer sometimes to a legal system, and sometimes to a rule of law, or a statement of how the law is on a particular

228

The Concepts of Law

point. Sometimes I will use the word ambiguously to refer to one or the other of these, as it does not matter for the purposes of the discussion of this chapter which way it is understood.”7 Raz thinks both concepts are of the same kind and must be explored philosophically in the same way. In consequence his methodology fails for both. It fails for the sociological concept because that concept is not sufficiently precise to yield philosophically interesting “essential features.” As I said, the sociological concept, like the concepts of marriage, meritocracy, boxing, and the other criterial concepts we use to describe social arrangements, has too much leeway for that: its boundaries are too malleable to support an essential-feature philosophical investigation. Suppose our anthropologists report a social structure in which officials decide who has what they describe as the better moral case in personal controversies and other officials then use force to coerce any unwilling party to obey those decisions. Our shared linguistic practices do not require us to say either that that system constitutes a legal system or that it does not. Perhaps social scientists would do well for their own purposes of prediction or explanation to stipulate a more precise definition of legal system that decides that issue. But there is no reason for either our lawyers or general citizens to do that: nothing turns for us on how we make the classification. Raz’s methodology is inappropriate to the doctrinal concept for a different reason: not because it does not matter whether our officials decide that Mrs. Sorenson has a legal right to market-share damages—obviously that matters very much—but because the doctrinal concept is interpretive. Two lawyers who disagree sharply about what the law requires in various circumstances could nevertheless both have equally well mastered the concept of law. One of them, at least, is wrong about the law, but he is wrong because his legal arguments fail, not because he understands the concept of “how the law is on a particular point” less well than his rival. So we cannot say that legal theory should identify the essential nature of what lawyers, exhibiting their mastery, converge on identifying as law. Analytic doctrinal positivism points out that lawyers’ opinions normally do overlap to a considerable degree: there is generally a large area of source-based law that is for a time uncontroversial among them. It then hails that area of overlap as exhausting the extension of the doctrinal concept of law and

The Concepts of Law

229

announces that the essential nature of law so identified is source-based. This is a contrived and circular invention, not the discovery of an essential nature. Raz apparently assumes that the sociological and doctrinal concepts of law function in our thought as something in the nature of natural kind concepts. This assumption explains why he thinks it unimportant to distinguish between those two concepts of law in his analysis and why, in his general discussion of theories of concepts, he takes natural kind concepts—in particular the concept of water that philosophers exploring natural kind concepts have used as a standard case—as his examples. That is the mark of the broader semantic sting. Raz thinks that conceptual analysis is important because, as he puts it, “In large measure what we study when we study the nature of law is the nature of our own self-understanding . . . That consciousness is part of what we study when we enquire into the nature of law.”8 I am skeptical about that (I believe overheated) claim for analytical jurisprudence: if we want to study our own self-consciousness we would do much better to turn to fiction, politics, biography, depth psychology, and social science. We reflect on the character of law to know what we must do, not who we are. But so far as Raz is right in thinking that jurisprudence is a lens into our minds, the semantic sting produces an ugly distortion. Raz takes up a number of other interesting issues about the concepts of law in his methodological essay. He speaks, for example, of “our” concept of law and supposes that this is different from other concepts of law: the medieval concept, for instance, or the concept of law of some alien contemporary culture. He does not mean just that we have different attitudes or expectations of the law or different general beliefs about its importance or value or origin: these important differences, he says, are for the cultural sociologist or intellectual historian to study, not the legal philosopher. Once again, however, his failure to distinguish the sociological from other concepts of law mars his discussion. Because the sociological concept is an imprecise criterial concept, it is clear enough how the medieval sociological concept might have been somewhat different from ours. People then might have drawn the often arbitrary lines between types of social organization somewhat differently from the way most of us—or perhaps most of

230

The Concepts of Law

our social scientists—do. Perhaps they counted the customary practices of merchants in a particular market town as falling into the same category as the law of a political state, while most of us would put these into different categories and call only the latter “law.” There would be much historical but little philosophical interest in that fact. But what could it mean to say that our doctrinal concept of law is different from the medieval concept? If our doctrinal concept functioned as a natural kind concept, as Raz sometimes suggests it does, it would be difficult to see what that could mean. If English speakers of that historical period used a homonym of “water” to refer to any potable colorless liquid, including water and vodka, and used no word to refer to water distinctly, they would not have had a different concept of water from ours. They would have had no concept of water at all. If they used “water” to refer to water, but did not know (as most of us do not know) the chemical composition of water, they would still have had “our” concept (that is, the concept) of water, though they would have known less about water than some of us do. If, as I believe, the doctrinal concept is interpretive, then it would also be misleading to say that the medieval period had a different concept of law. Perhaps they had very different ideas about what sorts of material figure in the best understanding of what the law requires, just as they surely had very different ideas about what counts in deciding what justice requires. But they would no more have had a different concept of law than they had a different concept of justice. Raz discusses the further question whether a doctrinal theory of law— an account of the truth conditions of propositions of law—must be parochial, tailored only to one legal system or a group of such systems that are very like one another, or whether it can be universal. If a doctrinal theory is interpretive of legal practice, there can be no flat answer to that latter question. We can make our interpretive explication of our own practices very detailed, in which case it would of course be sensitive to features of our own practice—our rules and practices of precedent, for instance— that are special to us. Or we can make our explication much more abstract, in which case it would have much wider application. Raz says that “Dworkin’s theory of law was from the start parochial.”9 He has in mind my observation in Law’s Empire that a theory of law for us is a theory of

The Concepts of Law

231

our law; I meant that a theory carried into the detail of that book is tailored to our local practices and might not fit those of other political communities.10 So whether Raz’s characterization of my views is correct depends on what he understands as “my” theory of law: the more abstract he takes my theory to be, the less accurate is his observation. How universal can an interpretive theory of legal doctrine be? Suppose we set out to construct an interpretation of legal practice that would fit everything we took to fall under our sociological concept of law. How much detail could that highly abstract interpretation contain? Perhaps very little: it may be that once we begin the process any interpretive steps we take automatically make our interpretive account more parochial. We may find, for instance, that the only justifications that fit what we take to be legal practice in some remote place yield, at the doctrinal stage, a positivist account of the truth functions of law. I mentioned that possibility in Law’s Empire: indeed I imagined that the best interpretation even of our own legal practices might yield what I called “conventionalism.”11 Nothing in the doctrinal concept of law rules that out or, of course, guarantees it either. So I suppose the best answer to the question whether my theory of law is meant to be universal or parochial is: both. Raz raises yet another interesting question. He asks whether there can be law in the sociological sense in places that lack the concept of law. There can certainly be law in places where the sociological concept is unknown. It is not necessary for anyone in a community to understand that a legal system is a distinct form of social organization, let alone have any idea about what features might be thought to mark it off from other forms, for that community to have a legal system. But can there be law where no one has the doctrinal concept, that is, when no one understands the idea that something is required or forbidden or permitted in virtue of a practice to that effect? That is a question about the boundaries of the sociological concept. Imagine (if you can) a political community where people in black robes exercise a monopoly of coercive power, other people who are elected by the public lay down what they offer as wise counsels for the public to follow, and the black-robed people make it a practice, out of concern for efficiency, to follow those declared wise counsels in their verdicts. But no one thinks that the black-robed people are required to do

232

The Concepts of Law

this. They just do, and people plan accordingly, and though sometimes the black-robed people surprise the citizens, these events are treated as strokes of bad luck, like hurricanes. The members of this community lack the doctrinal concept of law: we cannot say of them, as I said of us in Law’s Empire, that they take up an interpretive attitude to the question of what their practices require of them and their officials. But we have the doctrinal concept and we can sensibly put that concept to work on their circumstances. I said, in the Introduction, that though our sociological concept of law is criterial and imprecise it does have boundaries, and one of these requires that it must make sense to suppose that rights, duties, and other normative relationships can be attributed to what we take to be a legal system. The people in the community I imagined cannot do that for themselves, but we can do it for them. We can say that they actually do have legal rights and duties, that their officials would not only surprise people but deny them what they are entitled to have were the officials suddenly to stop enforcing the wise counsels formally announced by the wise-counsel body. We can say that they have a legal system in spite of themselves and make our own declarations about what their law requires in particular circumstances. All these claims would of course be academic and, apart from their philosophical interest, pointless: they could have no effect on their behavior. Still, the claims would not be nonsensical. Could they also be right? That is a difficult question of moral theory. It requires us to investigate complex connections between attitude and entitlement, questions we face, for instance, when we consider whether women can have rights to equal treatment in a community in which no one, including women, even contemplates such rights or in an even more primitive community in which rights themselves are unknown. I do not mean to take up those complex issues now but only to illustrate the complexity of jurisprudential issues that are revealed when we take care to separate the distinct concepts of law.

Doctrinal and Taxonomic Concepts of Law I turn now to the different problems that arise when legal philosophers do not sufficiently distinguish the doctrinal from the taxonomic concept of

The Concepts of Law

233

law. A variety of scholars have recently suggested that I constructed two radically different arguments in criticism of analytic legal positivism: the first in a series of articles that became my 1977 book, Taking Rights Seriously, and the second in my 1986 book, Law’s Empire. In a recent article Scott Shapiro, who is drawn to analytic positivism, says that in his view legal positivists “succeeded in blunting the force” of my earlier arguments but that no one has yet provided any effective response to my later arguments so that my overall criticisms of positivism remain unanswered.12 Jules Coleman conceded in a public lecture, which he has circulated but not yet published, that the kind of arguments I made in Chapter 7 of this book have shown that his own conventionalist account of law is not a satisfactory response to my later arguments, but he says that Chapter 7 is nevertheless “embarrassing” because I refuse to admit there that he and others effectively answered my earlier arguments.13 In her recent biography of H. L. A. Hart, Nicola Lacey suggests that Hart also thought that I had “upped the ante” of my debate with him in my 1986 book and that he was uncertain until his death about how to answer the new arguments I made there.14 In a review of Lacey’s biography, another legal positivist, John Gardner, says that Hart had confident replies to my earlier arguments but that I “wrong-footed” him in Law’s Empire by introducing arguments of “first philosophy” that Hart was not intellectually equipped to handle.15 This recognition by prominent positivist philosophers that fundamental criticism of their position has not yet been answered is welcome and should improve discussion. But I do not believe that my later arguments are different from my earlier ones in any important respect. Shapiro distinguishes the two sets of arguments in the following way. My later arguments challenge positivism to explain legal disagreement: they point out that lawyers often disagree about what the law is on some matter, even though they agree about all the historical facts that positivists cite as exhausting the truth conditions of propositions of law. My earlier arguments, by contrast, make an essentially taxonomic point: that the moral principles that judges often cite to justify their legal decisions (like the principle that no one should profit from his own wrong, which figured in the Riggs v. Palmer case that I discussed in Taking Rights Seriously) are also legal principles and that taxonomic positivists are therefore wrong to separate legal from moral principles in the way they do. Shapiro says that my

234

The Concepts of Law

earlier argument so understood has been answered in the two ways I described in Chapter 7: by “inclusive” positivists like Coleman, who argue that moral principles become legal principles only when more basic legal principles that do not include moral principles designate them as legal; and, alternatively, by “exclusive” positivists like Raz, who deny that every principle that plays a role in legal argument is just for that reason a legal principle. If Shapiro were right in construing my pre–Law’s Empire arguments as taxonomic in that way, then these would indeed be pertinent responses. After all, as I said in the Introduction and earlier in this chapter, judges must often use arithmetic in deciding what legal obligations people have, even though, at least in the way most of us speak, the rules of arithmetic are not legal rules. I fear that much of what I said in an article published in 1967 encouraged Shapiro’s understanding of my argument.16 But his understanding is nevertheless incorrect, as I pointed out in an article published in 1972. Speaking of my 1967 article, I said, “My point was not that ‘the law’ contains a fixed number of standards some of which are rules and others principles. Indeed, I want to oppose the idea that ‘the law’ is a fixed set of standards of any sort. My point was rather that an accurate summary of the considerations lawyers must take into account, in deciding a particular issue of legal rights and duties, would include propositions having the form and force of principles, and that judges and lawyers themselves, when justifying their conclusions, often use propositions which must be understood in this way.”17 My target, in other words, was doctrinal, not taxonomic, positivism. I made the doctrinal argument that we cannot understand legal argument and controversy except on the assumption that the truth conditions of propositions of law include moral considerations. I did not mean to make, I said, the fallacious taxonomic argument that everything that figures among those truth conditions should be counted as belonging to a distinct set of rules or principles called legal. I have, I think, been clear about my target in my writings about positivism since that time. I therefore do not believe that the replies that took me to make a taxonomic claim actually provided any answer at all to the arguments I actually made. Perhaps a great deal of time has been wasted.

The Concepts of Law

235

The failure to distinguish between the taxonomic and doctrinal concepts of law may have caused even more damage to legal theory in recent decades. Raz has clarified his position in another important article not yet published when I wrote Chapter 7.18 He argues that since judges are human beings they are always subject to the demands of morality as every human being is in everything he does, including deciding cases. One of a judge’s responsibilities, in normal circumstances, is to enforce the law that competent authorities have made, and that responsibility normally eclipses other moral responsibilities he would have in the absence of pertinent law. But the eclipse may be partial: moral light may continue to shine on the problem through or around the law these authorities have created. Raz uses the American Constitution to illustrate one way in which law’s eclipse may be only partial. The First Amendment, he says, allows the moral requirements generated by a moral right to free speech to remain binding on judges in spite of legislative enactments that purport to exclude that moral right. We should not say, he insists, that the amendment incorporates a moral principle protecting free speech and makes that a legal principle. Rather we should say that the amendment directs judges to reason morally on the question whether they should refuse to enforce a particular statute because it offends that particular moral right. He summarizes this point in a distinction and an analogy. We should distinguish, he says, between reasoning about the law and reasoning in accordance with the law. When judges conclude that the First Amendment requires them to decide moral questions about free speech, they are reasoning about the law. When they tackle those moral questions, they are no longer reasoning about what the law is but reasoning as the law directs them to reason. If an accident occurs in Greece but a lawsuit is brought in Poland, Polish law may direct that Polish judges reason about Greek law in reaching their decision. But it would be misleading to say that Greek law has then become part of Polish law. Raz’s strategy differs from Coleman’s in one striking way: for Coleman judges are bound by morality only so far as law incorporates morality, while for Raz judges are bound by morality except insofar as law excludes it. But if we take Raz to be defending doctrinal positivism, his strategy is nevertheless defeated by the same facts that embarrass Coleman. He can

236

The Concepts of Law

protect doctrinal positivism only by supposing that law’s impact on a judge’s moral obligations—the degree of eclipse that law manages—can itself be determined without regard to morality. That is presumably why he says that “it is generally assumed” that the First Amendment refers to “a moral right of free speech.”19 He supposes that this interpretation is established by canons of constitutional interpretation that substantially all American lawyers accept so that we need appeal to no moral principle to establish that the interpretation is correct. But the opposite is true. Many lawyers, including myself, believe that the First Amendment makes the validity of statutes depend on a moral right.20 But a great many other lawyers reject that view. They believe that the First Amendment requires judges to enforce free speech not as a moral principle but as a historical fact: that judges are legally and morally bound to enforce the conception of free speech that was recognized at the time the First Amendment was enacted.21 So a judge faced with a First Amendment challenge must choose between these rival understandings of that amendment’s force and, as I explain in Chapter 6, he must make that choice on grounds of political morality. Judges who take different views about the nature and virtues of democracy will understand the role of a Constitution in a democratic society rather differently, and their views on that moral question will determine whether they interpret the First Amendment in one of those ways rather than another. The argument about how to interpret constitutional provisions does not bottom out in any source-based rule; it is an argument of political morality all the way down. So Raz’s strategy, like Coleman’s, fails to save doctrinal positivism from a fact that both concede: that judges often reason about morality in reaching their judgments. It is not plain, however, that Raz wants to save doctrinal positivism. His favorite analogy shows why. It would indeed seem odd to say that Greek tort law is part of Polish law. But Greek law certainly figures in the truth conditions of the proposition that under Polish law a particular Polish defendant is legally liable to a particular Greek plaintiff for damages arising out of an accident in Piraeus. Polish judges could not reach a correct conclusion about the state of Polish law on that discrete question of liability without taking Greek law into account. It is the same with morality. Let us suppose, contrary to fact, that Raz is

The Concepts of Law

237

right in thinking that lawyers all agree that the First Amendment makes the validity of statutes turn on the best understanding of the moral right of free speech. Suppose further that Congress has enacted a statute forbidding candidates for federal office to spend more than stipulated sums on their campaigns. Now consider the proposition of law that John Kerry was legally forbidden to spend more than the stipulated sum on his presidential campaign.22 On Raz’s account the truth or falsity of that proposition depends on the soundness of the moral judgment that Congress invaded freedom of speech when it enacted that regulation. It does not matter how we put that point: whether we say that in virtue of the First Amendment the law does not prevent morality from having that impact, or that the First Amendment directs that morality should have that impact. Either the negative or the positive formulation of Raz’s position has the consequence that the truth of that proposition of law—and, of course, battalions of other propositions of law—depend on the correct resolution of a moral issue. We can imagine arguments that Raz’s position is consistent with a doctrinal positivism, but none of these is successful. It would not help to say that the First Amendment only directs judges to reason about morality in some fashion or other and not to reason well, as if a botched and false moral analysis would acquit the judge’s duty just as well as a sophisticated and correct one. That is not, on any interpretation, what the First Amendment says. Nor is there any room here for a distinction between a litigant’s right to the best moral reading and his right to a particular reading because it is the best one. We cannot say that he is legally entitled to the best reading in the abstract but that because ascertaining the best reading requires moral reasoning he is not legally entitled to any particular concrete reading. Rights and obligations, unlike propositional attitudes, provide a context that is transparent to substitution: if a politician is legally entitled to the best understanding of freedom of speech, and if the best understanding condemns the campaign expenditure limitation law, then he is legally entitled to unlimited expenditures.23 Nor will it help to say that the proposition I cited, regarding John Kerry’s lack of entitlement to unlimited expenditures, is not really a proposition of law of the kind doctrinal positivists have in mind: that it is only a singular proposition, not some-

238

The Concepts of Law

thing in the nature of a rule of law. We can make the proposition I cited progressively more abstract and general—that under American law political candidates are entitled to unlimited expenditures, for example—until it has all the generality of a familiar rule of law. So Raz’s “exclusive” positivism may not after all be doctrinal positivism. He may not wish to defend the thesis that as a conceptual matter legal rights and duties may always be ascertained without moral reflection just by consulting “sources.” He may rather be a taxonomic positivist. An argument leading to a particular conclusion of law typically involves a great variety of propositions: reports about what lawmaking institutions have done, information about parallel lawmaking institutions in other jurisdictions, claims about economic or social or historical facts or about personal circumstances, arithmetical assumptions, moral principles and observations, and so forth. A taxonomic positivist insists on a distinction among all those propositions: a distinction between those that do and those that do not describe “the law” of the jurisdiction in question. An “inclusive” taxonomic positivist says that sometimes—when they have been properly incorporated—moral principles of different sorts count as part of “the law,” and an “exclusive” taxonomic positivist denies this. Is this an argument worth having? It is of course important what we take to be relevant to deciding what legal rights and duties people and officials have. But nothing important turns on which part of what is relevant we describe as “the law.” Why should we not say that we have considerable leeway in making that linguistic choice so that both the “inclusive” and the “exclusive” diction are acceptable? It would indeed be odd to say that arithmetical principles are part of Massachusetts law; a judge who calculates damages supposing that five and seven add up to eleven makes, we want to say, a mathematical, not a legal, mistake. We can explain this linguistic preference in a variety of ways: arithmetic is in no way special to law and certainly not to the law of any particular jurisdiction, so it would be at least highly misleading, even though not unintelligible, to say that it belongs to Massachusetts law. We can also explain our reluctance to say that Greek law becomes part of Polish law when a Greek accident is tried in a Polish court. We want our diction to acknowledge that Greek law has a place in Polish courts only in very limited circumstances and that the con-

The Concepts of Law

239

tent of even that part of Greek law that is pertinent is to be ascertained by interpreting Greek, not Polish, legislation and practice in tort cases. Some terminological choices seem more evenly balanced than these examples, however. Suppose legal practice in some jurisdiction gives legal effect to the traditional practices and expectations of merchant custom so that what a merchant’s legal rights are often depends on what the custom is. Shall we say that this custom is part of the law or not? Either choice would be defensible, and it would not matter which choice we make. When particular moral principles—like my early example of the principle that no one should profit from his own wrong—have been cited and relied on often in a particular legal jurisdiction, and particularly when they are ignored or less often cited in other jurisdictions, then it is tempting to say that these principles have become part of the law of that jurisdiction. But that locution risks the mistake I called attention to in the 1972 remarks I quoted earlier: the mistake of supposing that the “law” of a community consists of a finite body of rules, principles, and other standards that might in theory all be listed and counted. So I suppose that if I had to choose I would opt for exclusive taxonomic positivism, though my heart isn’t in it. Either choice will do if we take care to avoid the mistakes that each might encourage. Raz disagrees, and much of his recent essay is devoted to trying to show that the taxonomic debate is important. He does not claim that the phrase “part of the law” describes a natural kind: that it would be the same kind of mistake to suppose that moral principles are part of the law as it would be to suppose that some dogs are part of tigerhood. On the contrary, he concedes that the line between standards that have become part of the law and those that are not part of the law, but that the law nevertheless requires be “followed,” is “particularly vague.”24 He nevertheless thinks it necessary to insist that even when moral principles are among the truth conditions of claims about legal rights and duties they nevertheless not be counted as part of the law. He offers two reasons. First, he says, law is contingent—it can fail to exist—but morality cannot fail to exist. Second, in a community that has law there are (or at least can be) moral rights and duties that are not legal rights and duties. So there must be a boundary between law and morality.

240

The Concepts of Law

Neither of these reasons is impressive. If we decided to say that the principle that people should not profit from their wrongs is part of New York law, we would not be denying either that law is contingent or that morality is not contingent. We might perfectly consistently add that the principle would still be true even if it were not part of New York law and, indeed, even if there were no New York law. Nor if we decide to say that do we deny that there are other moral principles that are not part of New York law—or that there is a difference between law and morality. When we say that John Donne made certain words part of his poetry we do not deny that there are words he did not make part of his poetry or that there is a difference between the concept of Donne’s poetry and the concept of a word. Matthew Kramer, who counts himself an inclusive taxonomic positivist, says that though the inclusive–exclusive controversy started as a dispute about the best way to respond to my criticisms of positivism, it has taken on a life of its own.25 That is in one sense a welcome development because neither version of taxonomic positivism offers any response at all to my criticisms of doctrinal positivism. As Shapiro, Coleman, and other positivists have said, that response remains to be made if indeed it can be made. But the controversy is depressing on its own. It is sad that the important jurisprudential tradition of Bentham, Austin, Holmes, and Hart has now settled into a debate over something so scholastic. There is, however, a silver lining to that cloud. It may signal that there is no longer much interest in defending the proposition that Hart, for example, was once so anxious to establish: that as a conceptual matter the validity of claims of law depends only on social, not moral, facts. Doctrinal positivism thrives, but in its political, not its analytic, form.

chapter nine

Rawls and the Law

Rawls as Legal Philosopher It is a mark of John Rawls’s greatness as a political philosopher that my topic, “Rawls and the Law,” can be approached in so many ways. Politicians around the world cite his ideas and American and other judges appeal to his work, so we might talk about the impact that he has already had on the law in different countries. Or we might consider the impact that he might have: We might ask what changes in American tax or tort law his famous difference principle would recommend, for example. Those are indeed among the issues that have been explored. We might also examine his impact in the other direction. We might speculate about how important it was to the shape of Rawls’s theory of justice that he lived and worked in a political community as dominated by law as ours is, and in which certain crucial political issues—matters of basic liberty and constitutional essentials, if you will—are removed from ordinary politics and made the special concern of courts. I plan to talk about Rawls and the law in a different way: about Rawls as himself a legal philosopher and, indeed, lawyer. He did not suppose himself a legal philosopher, and though there are several important discussions of law in his work, some of which I will mention, he made his main contributions to legal theory through his political philosophy, because legal theory is a department of political philosophy and Rawls wrote abstractly about the whole discipline. Here I want briefly to identify but also to evaluate those aspects of Rawls’s theory of justice that speak directly to traditional issues of jurisprudence.

242

Rawls and the Law

I will begin with a short list of those traditional issues. Any general legal theory must answer the ancient question of what law is. But that ancient question in fact poses two different issues. The first is methodological: What kind of theory counts as an answer to that question? Are general theories of law descriptive theories? If so, what do they describe? Are they pieces of conceptual analysis? If so, what makes one analysis of the concept of law better than another? Are they normative political theories? If so, then how does a theory of what law is differ from a theory about what law should be? Legal philosophers disagree about these methodological matters. What view, if any, follows from Rawls’s philosophy? The second question is then obvious. Once a legal theory has taken a position about the methodological issue, it must try to answer the substantive one. Given what a theory of law should be, which theory of law is the most successful? For some time, theories of law have been divided, by both their authors and commentators, into roughly two groups: positivist theories of law, which insist that what the law of any jurisdiction requires or permits is only a matter of social fact, and anti-positivist theories, which claim that what the law requires sometimes depends not on social facts alone but also on controversial normative issues including moral issues. Rawls did not, so far as I know, explicitly choose one or another of these general positions as his own. But do his theories support one choice rather than the other? That second question inevitably presents a third one. On either a positivist or an anti-positivist theory, a judge will often face “hard” cases in which what lawyers regard as settled law does not decide the immediate issue. Positivists say that in such a case a judge must exercise discretion to make new law. Anti-positivists describe the same necessity in different ways: A lawyer who thinks, as I do, that judges must aim at integrity in their decisions will agree that what integrity requires will often, perhaps usually, be controversial, so that a fresh judgment is needed. Both sides, or rather all versions of each side, must confront the question of what sorts or sources of argument are appropriate to that judicial responsibility. What kinds of reasons may or should judges offer to defend their fresh judgments? May they appeal to religious convictions? To their personal moral convictions? To philosophical systems of morality or eschatology?

Rawls and the Law

243

To macroeconomic phenomena—would it be a good argument for a judge to say that he is deciding in a particular way because that will help the dollar on the international currency markets? These crucial questions are relatively neglected in legal theory. But Rawls developed a doctrine, which he called the doctrine of public reason, about the arguments public officials may properly use to justify their decisions, and he said emphatically that the doctrine of public reason applies with particular stringency to judges. We must examine this doctrine. If we find it unsatisfactory, as I suspect it is, then we must ask whether any other part of Rawls’s general theory is more helpful in defining the character of proper judicial reasoning. The fourth issue is particularly pressing for legal theorists in America and in other mature democracies where constitutional courts have the power to invalidate laws adopted by legislators elected by and accountable to the people. Is that power consistent with democratic principles? If not, is it unjust for that reason? Rawls spoke directly to that issue, on different occasions, and we must notice what he said. But he recently spoke to what is in many ways an even livelier and more important issue, a matter we might say, of constitutional strategy rather than legitimacy. Should a constitutional court decline to decide certain issues—for example, about abortion or assisted suicide—because its nation is not ready for a judicial resolution of the issue? Should the Court stay its hand to allow ordinary politics to reduce the issue’s divisiveness and perhaps reach a compromise more acceptable to the whole community? Several prominent legal scholars have endorsed that suggestion, and Rawls has said that he thinks their argument a “good” one. We should consider why he thinks their argument good. Finally, I would like to touch on what might seem a much more abstract issue. Can controversial claims about what the law requires be objectively rather than merely subjectively true? This is not an issue that troubles lawyers and judges in their day-to-day practice. It nevertheless has considerable practical importance, because many issues of legal and civic policy turn on it, including whether the rule of law really is different from rule by men and women with power, whether it makes sense to suppose that we have a general moral obligation to obey the law, and whether judicial review of legislative enactments really is legitimate. Some legal theories are

244

Rawls and the Law

in fact built around the supposition that legal practice is essentially subjective: for example, the influential movement called American legal realism, which in our time morphed into the incandescent, though brief, fireworks of critical legal studies. Rawls said a good deal about truth and objectivity, some of it inconclusive and even obscure, but much of it helpful when lawyers turn to these more explicitly philosophical issues.

The Nature of Legal Philosophy Assume, for the moment, as almost all lawyers do, that a proposition about legal rights and duties can be true. If so, then a theory of law should tell us under what circumstances such a proposition is true. What in the world can make it true, for example, that the speed limit around here is 55, or that Microsoft violated the antitrust laws, or that affirmative action is unconstitutional? Legal philosophers defend general theories of law that attempt answers to that question. Legal positivists claim that a proposition of law can be true only in virtue of social facts: facts, for example, about what a legislature has declared or a judge decided in some prior case. I shall say something about the merits of that view in a moment, but we must first consider an antecedent question. What kind of a claim are they, the positivists, making? What could make their claim about the truth conditions of law itself true? Many legal philosophers believe that their theories of law are descriptive theories about the social practices or conventions that the bulk of lawyers follow in making, defending, and judging propositions of law. Of course, lawyers often disagree about which legal propositions are true and which false. They disagree, for example, about the legal position of a woman who has suffered side effects from a drug that her mother had taken many years prior, but who cannot identify the manufacturer of the particular pharmaceuticals her mother had taken at any particular time because the pill was manufactured by several companies and she does not know whose pill she took when.1 Is she legally entitled to recover damages from all of the companies that manufacture the drug in proportion to their market share? But these legal philosophers assume that if this kind of disagreement is genuine, then the lawyers must agree about a more basic issue. They must agree on the right tests to use to decide if a proposition of law

Rawls and the Law

245

is true; otherwise—if different lawyers used different tests—they would simply be talking past one another. If that assumption is correct, then a philosophical theory of law should aim to describe that background agreement. It should tell us what law is by telling us what tests lawyers actually use to identify true or sound propositions of law. On this view, legal philosophy is best understood as a descriptive exercise: it is an exercise in legal sociology. But, in fact, it is extremely difficult to account for any general theory of law if we take it to be descriptive in that way. Consider the version of legal positivism developed by H. L. A. Hart.2 He argued for what he called the “sources” thesis which holds (in substance) that propositions of law are true when, but only when, they can be inferred from explicit decisions taken by legal institutions, like legislatures, that are authorized by convention to make such decisions.3 If a lawyer can show that it follows from something that the pertinent legislature has said that the woman in our example is legally entitled to market-share damages, then he has shown that she is so entitled. But if that proposition does not follow from anything any authorized institution has said or done, then it is not true. In a posthumously published Postscript to his book, The Concept of Law, Hart insisted that this sources thesis is purely descriptive.4 But it is mysterious in what sense it could be thought descriptive. Hart did not intend it, he insisted, as a description of how lawyers talk, how they use the word “law.” For it is plainly not part of the very meaning of “law” that law can only be valid in virtue of positive enactment. Nor could he have intended it as a description of what all lawyers accept as belonging to the very concept of law, as we accept that it belongs to the concept of bachelor that a bachelor is unmarried. For lawyers disagree about whether the sources thesis is right: The lawyers who think that the woman in our example has a legal right to market-share damages plainly reject the sources thesis, because no institution had declared such liability before imaginative lawyers5 argued for it. If these lawyers are mistaken, their mistake is a legal not a conceptual one. Nor, for the same reason, could Hart have meant his sources thesis as the sociological hypothesis that lawyers everywhere actually claim law on their side only when the sources thesis has been satisfied. That hypothesis, too, is plainly false. How, then, should we understand a theory of law like the sources thesis?

246

Rawls and the Law

Rawls spoke directly to that issue by example—through his analysis of the concept of justice. He did not suppose that everyone who shares and uses the concept of justice shares some substantial background understanding about what makes an institution just or unjust. On the contrary, he insisted that people have radically different conceptions of justice. They do, he allowed, share some very abstract understanding that makes these all conceptions of justice rather than of some other virtue. But this shared understanding is exceedingly thin, all but empty of real content. What makes disagreement about justice possible is that people sufficiently agree on certain specific instances or examples—everyone agrees that slavery is unjust, that wage exploitation is unjust, and so forth. So Rawls recommended that philosophers of justice engage in the interpretive enterprise he called seeking reflective equilibrium. We try to generate principles of some general scope and to match those general principles to the concrete judgments about what is just and unjust with which we begin, shifting our views about either principles or concrete judgments, or both, as becomes necessary to achieve an interpretive fit. We can restate this interpretive exercise as a method for legal philosophy. We can identify what apparently goes without saying is part of our law—the speed limit, the tax code, the ordinary everyday rules of property, contract, and so forth, that we are all familiar with. These are, we might say, paradigm instances of law. And then we can construct the other pole of an interpretive equilibrium because we share an abstract ideal that can play the same role in legal theory as the concept of justice played for Rawls. This is the concept of law—though sometimes, when we are emphasizing its political character, we describe it in another way, as the concept of legality or the concept of the rule of law. We can then try to provide a suitable conception of legality, that is, a conception of legality that brings our various preanalytic assumptions about concrete propositions of law into equilibrium with the general principles of political morality that seem best to explain the character and value of legality. In that way we can embed a theory about the truth conditions of propositions of law in a larger conception of value that we find convincing. A positivist theory of law will offer a thesis about the truth conditions of such propositions, like the sources thesis, that is supported by a positivist conception of legality

Rawls and the Law

247

that is in turn supported by an appropriate more general theory of justice. That interpretive design provides the best way of understanding the arguments that leading legal philosophers have actually made. Legal philosophy so conceived is in a way descriptive because it begins with some understanding about what is taken for granted within the community to which it is addressed, but it is in other ways substantive and normative because the equilibrium it seeks is with principles judged for independent appeal. So Rawls’s work is, right from the start, a major contribution to legal philosophy’s self-understanding.

What Is Law? Now let us turn to the substantive side of the ancient question. Which understanding of law—a positivist understanding or some other—is most successful as a conception of the concept of legality? To bring out the implications of Rawls’s ideas for that further question, we might embed the question in his imaginary construction of justice as fairness. Suppose that the representatives in the “original position”6 he described are asked to choose, in addition to general principles of justice, a conception of legality as well. They are offered, to make the point simpler, a menu of only two choices. They may choose a simplified positivist account of legality, which specifies that judges use a particular test for true propositions of law, or a simplified interpretivist non-positivist account. On the stipulated simple positivist account, judges are to enforce rules laid down by the legislature so far as these rules are unambiguous or can be made unambiguous by consulting legislative history and other standard sources of legislative intent. But when, as will often happen, rules laid down and interpreted only in that way are insufficient to decide the case, then judges should announce that the law provides no answer, and then legislate themselves to fill the gap so created. They should legislate, however, modestly and marginally, and as they believe that the legislature now in power would legislate if seized of the issue. Judges should do, that is, what they think the pertinent parliament would have done. On the rival simple interpretivist account, judges should enforce rules laid down by the legislature, interpreted in the same way, but when confronted by a so-

248

Rawls and the Law

called gap, should not attempt to legislate as the legislature would, but should instead try to identify the principles of fairness or justice that best justify the law of the community as a whole and apply those principles to the new case. Now suppose that, contrary to Rawls’s supposition, the representatives have settled on a general, all-embracing utilitarian conception of justice. Then they will think that they have a strong case for choosing the simplified positivist conception of law over the simplified non-positivist one. For there is a strong affinity between a utilitarian conception of justice and a positivist conception of legality: It is no accident that the two founders of modern legal positivism, Bentham and Austin, were arch utilitarians. As Bentham pointed out, sound utilitarian legislation must be organized and directed from a single source: The best program for maximizing utility is an integrated program in which different laws and policies can be tweaked and coordinated so as to yield a maximum utility bang.7 The legislature is the best institution for achieving that maximum bang, because it can survey the whole architecture of law and policy, and because its composition and procedures of election are conducive to providing information about the mix of preferences in the community that is indispensable to sound calculations of the trade-offs necessary to achieve maximum aggregate utility. Judges are essential to the retail enforcement of rules designed to maximize utility over time, but they should be as little as possible architects of policy, because that would be multiply inefficient. So they should be told that when the legislature’s dictum has run out without issuing a decision, they should declare there is no law controlling their decisions that comes from any other source. They should announce a gap, which they then fill as modestly as possible, as lieutenants of their political masters, in the spirit of what those masters would themselves do, going, as the arch-positivist and -utilitarian Oliver Wendell Holmes put it, from the molar to the molecular.8 That is the positive case for positivism from the utilitarian perspective. There is a corresponding negative case from that perspective against interpretivism: that it is irrational. For utilitarians, moral and political principles are simply rules of thumb for achieving maximum utility in the long run, and there can be no independent value, and much mischief, in

Rawls and the Law

249

pursuing a coherence of principle for its own sake. Utility is surely better served by concentrating on the future alone, with no backward glance at integrity with the past, except so far as this is in itself strategically wise. Now suppose, however, that representatives in the original position do choose as Rawls supposes they would. They reject utilitarianism in favor of the two principles of justice, one of which gives priority to certain basic liberties and the other of which seeks to protect the position of the worstoff group in society. Then it would seem natural for them also to choose interpretivism over positivism, because interpretivism would then be a better bet to achieve justice, at retail as well as wholesale, in the long run. The two principles require implementation at successive levels of detail. They require, first, a constitutional stage at which institutions are designed so as most likely to produce the outcomes that the two basic principles demand. Then they require those institutions to make, at what Rawls calls a legislative stage, more specific decisions about laws and policies guided by more specific principles of justice in service of the basic principles. People who place a lexical priority on equal liberty, and then a further priority on protecting the position of the worse-off group, will be particularly sensitive to the possibility of slippage at this legislative stage. They will worry that a legislature dependent on majority approval will be under great pressure to advance the interests of some groups at the expense of others. They will, therefore, be attracted to the idea of a judiciary with independent powers and responsibilities. They will be drawn to the idea of judicial review of a written constitution, and I will discuss the implications of Rawls’s arguments on that score later.9 But they will also be drawn to the idea that judges should also exercise a less potent, but still important, supervision over the application and development of the more quotidian law made by legislatures. And to the further idea that they should exercise that power in the direction of equality before the law, that is, in the direction of insisting that, so far as a reasonable doctrine of legislative supremacy permits, whatever principles are presupposed by what the legislature has done for some groups be available generally to all. They will have that strong reason for favoring an interpretivist conception of law that deems people to have legal rights not only to what legislative institutions have specifically directed, but also to the principled elaboration of those directions.

250

Rawls and the Law

Coherence is the best protection against discrimination. That is, after all, the premise of the Equal Protection Clause of our own Constitution’s Fourteenth Amendment. Rawls did not make this argument for interpretivism; indeed, so far as I am aware, he made no explicit argument for any conception of law. But he did explicitly endorse the principle that I said supports interpretivism, and he endorsed that principle in the course of a discussion of legality or the rule of law. Let me quote: [T]he precept that like decisions be given in like cases significantly limits the discretion of judges and others in authority. The precept forces them to justify the distinctions that they make between persons by reference to the relevant legal rules and principles. In any particular case, if the rules are at all complicated and call for interpretation, it may be easy to justify an arbitrary decision. But as the number of cases increases, plausible justifications for biased judgments become more difficult to construct. The requirement of consistency holds of course for the interpretation of all rules and for justifications at all levels.10

Note Rawls’s emphasis on complexity as itself a constraint, and his insistence that consistency hold, as he put it, for “all rules . . . at all levels.”11 Citizens are best protected from arbitrariness and discrimination when judges interpreting the law and elaborating it in hard cases are responsible for coherence, not simply with particular doctrines here and there, but, as best as it can be achieved, principled coherence with the whole structure of law. Now it might be objected that in spite of the historical affinity I cited a positivist need not be a utilitarian. We can suppose, instead, a positivist judge who is not a utilitarian, and who stands ready, in cases in which he supposes he has discretion, to adopt the rule that he believes best comports with justice on some other understanding. Why wouldn’t people who have chosen Rawls’s two principles of justice in the original position also choose that characterization of a judge’s role? Why isn’t that a better choice on grounds of what Rawls called imperfect procedural justice? But this suggestion neglects the fact that judges, even if they aim only at justice, will nevertheless often disagree about what justice is, and that judges may themselves be influenced by preconception, prejudice, or the other

Rawls and the Law

251

enemies of impartial justice. People choosing a conception of law have no reason to think that a decision in their own case will better reflect justice, on any conception of what that is, if judges are free to disregard principled coherence with what other officials and judges have done than if they are asked to respect principled coherence. They might well think that they safeguard themselves better against arbitrariness or discrimination if they do not instruct judges to do justice as they see it, but seek to discipline judges by insisting that they do their best to respect principled consistency as they see it. That is, as I said, the assumption of our Equal Protection Clause.12

The Constraints of Legal Reasoning Now I turn to a more specific question that must be faced by all conceptions of law, but which is particularly difficult for some. How should judges reason in hard cases? Under the simple package of positivism and utilitarianism that I described, judges must make fresh judgments to fill gaps in the law, but the package dictates the character of that judicial reasoning. It holds that judges should try to do what the legislature would have done. Interpretivism, as well as other legal theories, also supposes that judges must make fresh judgments of political morality in hard cases: It instructs them to seek an interpretive equilibrium between the legal structure as a whole and the general principles that are best understood as justifying that structure. That, as I have argued elsewhere, is in fact the traditional common law method.13 But are there any constraints on what kind of principles judges can cite in constructing this interpretive equilibrium, that is, in justifying the law’s record as a whole? It would certainly seem wrong for them to deploy certain kinds of arguments. They must not appeal to their personal interests or to the interests of some group to which they are connected. That obvious constraint seems part of the very idea of a justification. But may they appeal to their religious convictions, if they have any, or to the doctrines of their church, if they have one? After all, some judges think, as a matter of their deepest conviction, that religion provides the most compelling or perhaps the only true justification of political morality and, therefore, the most compelling or only true justification of past legal decisions. In the United States, reli-

252

Rawls and the Law

gious argument might be thought excluded from judicial reasoning by the First Amendment. But what about elsewhere? In a country, for example, like the United Kingdom or Israel, in which there is an established religion? If religion is also an impermissible ground for adjudication even there, is this constraint ad hoc to religion? Or does it follow as only one case from some more general principle of political morality? What about arguments of moral philosophy, for example? May a judge properly appeal in his opinions to the philosophical doctrines of Immanuel Kant or John Stuart Mill? May he appeal, as a number of American judges have in fact done, to the philosophical writings of John Rawls?14 May a judge appeal to macroeconomic policy? May he decide that one principle better justifies the legal structure as a whole because following that principle will help control inflation or promote savings? Rawls’s doctrine of public reason is devoted exactly to defining the kinds of arguments that are permissible for officials in a politically liberal community, and he insists that the doctrine applies with particular stringency to judges. I find the doctrine of public reason difficult to define and defend, however. I will try to summarize my difficulties here. There are two ways of stating what the doctrine requires. The first, and more basic, appeals to the important idea of reciprocity. The doctrine permits only those justifications that all reasonable members of the political community can reasonably accept. The second is presumably the upshot of that more basic test. Public reason requires officials to offer justifications that are based on the political values of the community and not on comprehensive religious or moral or philosophical doctrines. The doctrine, therefore, requires judges searching for a justification of the law’s structure to avoid controversial religious, moral, or philosophical doctrines. I do not see, however, what the doctrine of reciprocity excludes. If I believe that a particular controversial moral position is plainly right—for example, that individuals ought to take charge of their own lives and bear the financial responsibility for any mistakes they make themselves—then how can I not believe that other people in my community can reasonably accept the same view, whether or not it is likely that they will accept it? Perhaps Rawls means that judges should not appeal to ideas that some reasonable citizens could not accept without abandoning their convictions of a certain sort—their X convictions. But we seem to have no basis for

Rawls and the Law

253

stipulating what these X convictions are. I accept that religious convictions are special for several reasons. Certainly someone who believes that the religious truth is only available through divine grace, or some other privileged access, cannot hold that all reasonable citizens could reasonably embrace his own religious convictions. But Rawls offers no reason to think that the test of reciprocity excludes any reasonable convictions beyond religious convictions. I have equally great difficulties with the distinction between political values on the one hand and comprehensive moral convictions on the other. Rawls’s own conception of justice as fairness depends critically on what seem to be controversial moral positions. The difference principle, for example, is generated and defended in reflective equilibrium by a set of assumptions, including assumptions about the fundamental moral irrelevance of effort or responsibility: If the arrangement that best maximizes the position of the worst-off group turns out to reward slackers, that is no objection. Rawls defends this conclusion by supposing that effort is influenced by endowment.15 So it is, but it is not exhausted by endowment, and the question of how the interaction between the two is to figure seems a mixed question of psychology and morality of just the kind that divides different comprehensive moral views about personal responsibility. Rawls’s position is certainly controversial in our community, and some people reject it in favor of a theory of distributive justice that depends more on personal responsibility. These difficulties are confirmed, I think, by Rawls’s examples of the idea of public reason in operation. He discusses the abortion controversy on several occasions, though in each case only very briefly. His discussion assumes that the question whether an early fetus has rights and interests of its own, including a right to life, is a question for a comprehensive moral or religious or philosophical position and is not settled by any political value of a liberal community. But how can we take a position about whether American women have a constitutional right to abortion—how could the Supreme Court decide Roe v. Wade16 or Planned Parenthood of Southeastern Pennsylvania v. Casey17—without taking some position on that comprehensive issue? There seems no default position here. The view that a fetus does not have interests and rights of its own is as much drawn from a comprehensive position as the view that it does, and we cannot

254

Rawls and the Law

reach a decision about abortion without adopting one of these two views. The Equal Protection Clause applies to all persons, and any argument that a woman has a constitutional right to abortion in the first trimester of pregnancy must deny that a fetus is a “person” within the meaning of that clause.18 I therefore doubt that Rawls’s doctrine of public reason can help us much in filling out a conception of legality and adjudication. We must look elsewhere. In my view, we can find the necessary constraints on judicial argument in the conception of law that I said Rawls’s general arguments suggest: interpretivism. If we accept an interpretivist conception, we do not need a separate doctrine like the doctrine of public reason. Judges may not appeal to religious convictions or goals in liberal societies because such convictions cannot figure in an overall comprehensive justification of the legal structure of a liberal and tolerant pluralistic community. This interpretive constraint cannot, however, exclude moral as distinct from religious convictions. Judges interpreting a string of cases in tort law can appeal to Rawls’s theory of justice as a ground for rejecting a utilitarian interpretation of past decisions and doctrines in favor of an interpretation more firmly grounded in a conception of equality.19 One more point: In his discussion of public reason, Rawls says that in any case judges may not appeal to their personal moral convictions. If that means that a judge cannot argue that one justification of past law is superior because he happens to think so, then it is obviously correct. A judge’s intellectual biography is not a legal argument. But if it means that a judge may not give any place to controversial moral opinions in his judgment, because he would then be citing the moral opinions that he but not others think right, then it states an impossible demand. On no conception of law—positivist or interpretivist—can judges in complex pluralistic communities acquit their institutional responsibilities without relying on controversial moral convictions.20

Constitutionalism The institution of judicial review, under which appointed judges have the power to declare enactments of legislatures and other representative insti-

Rawls and the Law

255

tutions void because they offend constitutional guarantees of individual rights, is often said to be anti-democratic because it allows a few unelected and virtually unsackable lawyers to override the considered verdicts of elected representatives. Rawls addressed himself to that classic complaint on several occasions. He made plain, first, that according to his favored conception of justice as fairness, the various institutions that a community constructs at what he calls the constitutional level, in the light of principles of justice chosen behind the veil of ignorance, are chosen in the spirit of perfect rather than pure procedural justice. They are chosen, that is, with an eye to outcomes. The principles of justice establish the basic liberties and their priority, and the question to be decided at the constitutional stage is an instrumental one: Which scheme of institutions is best suited to protect those liberties? Of course, among the equal liberties that institutions must be designed to protect are the political liberties, which include the right to vote and to participate in politics. But, as Rawls says in Justice as Fairness: A Restatement, these and the other basic liberties are themselves to be seen as quasiinstrumental.21 They are justified as essential for the development and exercise of the two fundamental moral powers, that is, the power to form and act on a sense of justice and to form and act on a conception of the good. This means, as I understand it, that though people have a basic right to broadly democratic procedures, because extensive rights to vote and participate in politics are plainly and inescapably necessary to the development of these moral powers, people have no basic right to any particular form of democracy, and therefore no basic right that democratic institutions follow any particular design or have any particular jurisdiction. The question is rather which parliamentary structure and jurisdiction has the best prospects of securing the other mandated or desired outcomes. So nothing in Rawls’s general conception of justice as fairness supports the so-called “majoritarian” objection to judicial review in its most comprehensive form. But his theory leaves room, at the constitutional level of construction, for the more limited objection that in fact the American structure of constitutionalism and judicial review cannot be justified in that instrumental way, that the basic liberties including the political liberties would be better served by some other arrangement, which might be

256

Rawls and the Law

pure parliamentary sovereignty or a mixed case like that of the United Kingdom after the enactment of the Human Rights Act, which permits Parliament to legislate in violation of the rights the Act specifies if Parliament clearly states its intention to do so. Though Rawls did not attempt anything like a thorough outcome-based case for the American model against such more majoritarian rivals, he makes several arguments that seem to support roughly the American model. He distinguishes, for example, between parliamentary and popular sovereignty, and says that the American model is consistent with popular sovereignty. That model promotes people’s basic moral powers, he says, because the people in general not only endorsed the original Constitution, but have prompted and overseen its cardinal developments since—in the Reconstruction period, for example, and in the New Deal. (In that view he follows, as he says, the arguments of Bruce Ackerman.)22 Second, he points out a further way in which constitutionalism and judicial review help rather than constrain the development of the two moral powers. He says that the fact that the Supreme Court acts as a forum of principle encourages and focuses public political discussion of central moral issues.23 I can now turn to the different issue I mentioned earlier: not the legitimacy of judicial review, but its proper strategy. The Supreme Court is often pressed to recognize a concrete constitutional right that it has not recognized before, and whose standing as a right is very much in dispute among thoughtful people in the nation. If it recognizes and enforces that new right, its decision will be massively resented, and its own standing and legitimacy may be called into question. The Court faced that situation in Brown v. Board of Education24 and the other early racial discrimination cases of the 1950s. It faced it in the school prayer cases, in the abortion cases beginning with Roe v. Wade,25 and in the more recent cases about assisted suicide for terminally ill patients. It is widely argued that in such situations the Court ought to refuse to recognize the new right in order to allow the political process more time to consider the merits of the issue through local politics and decision, which might vary across states, and which might therefore provide a kind of experimentation in what Justice Brandeis called the laboratories of the several states.26 It might do that in some cases through its certiorari policy;

Rawls and the Law

257

it might decline to take a case that required it to decide basic issues of individual rights because it thought it wiser to let those issues percolate in politics further. In most such cases, however, one or more lower courts will have spoken to the issue in a way that requires the Supreme Court to decide whether the Constitution grants the claimed right. In that case the passive or cautionary strategy I described would require the Court to hold that the asserted controversial right did not exist, as it did hold, for example, in the assisted suicide cases.27 Rawls had himself urged the Supreme Court to recognize a limited right to assisted suicide: he, along with several others, signed a brief urging that decision as an amicus curiae.28 But Rawls later said that the cautionary argument is what he called a “good” argument for the decision the Court made against his advice, and he also called it a “good” argument against the Court’s 1973 decision to recognize a limited right to abortion in Roe v. Wade,29 which of course doesn’t mean that he thought it finally a persuasive argument.30 However, there seems to me a straightforward and powerful—even knock-down—Rawlsian argument against the cautious view. In these contentious cases a plaintiff or group of plaintiffs claims that some law or practice denies their basic liberties and therefore offends the first principle of justice, which in justice as fairness is given priority over everything else, presumably including civil peace and quiet. Of course, any particular justice might not believe that the targeted law or practice does deny a basic liberty. But we must assume that Justice Rawls would be convinced by the argument he himself put forward, for example, in the Philosophers’ Brief. We can easily see how a utilitarian who calls himself a pragmatist might be attracted to the cautious argument. But why should Rawls be? Why should he think that the cautious argument is even a “good” one? One answer might be epistemic. Perhaps Rawls thought it arguable that a Supreme Court justice, recognizing the fearsome Burdens of Reason, should accept that his own judgment might well be flawed, that the political process might over some years work out a different compromise that would be very widely accepted, and that this compromise, if it is ever achieved, would be a more accurate or reasonable account of the basic liberty in question than a majority of justices could devise for themselves in

258

Rawls and the Law

advance. There are several obvious difficulties in this answer, however, and we may expose these using the abortion issue once again as an example. First, it seems unlikely that a non-divisive compromise would soon have been reached in the politics of this country. The Europeans have by and large settled on a position that, with cosmetic formalities, allows abortion on demand.31 This has not generated continuous controversy there, but that is because Europe is not plagued with fundamentalist religious movements or any serious fundamentalist sensibility. But we are so plagued, as has once again been demonstrated by the recently legislated ban on socalled partial birth abortion.32 The only solution that would still militant objection from fundamentalists here would be a harsh anti-abortion regime, and that would not be tolerable to women’s movements that can be almost equally militant. Second, even if a compromise were reached politically that proved reasonably acceptable to all, there would be no reason to think that this compromise would have more accurately or reasonably identified the basic liberties in question. On the contrary, whatever view one takes of those basic liberties, it seems likely that a compromise would mean injustice to some. Suppose, for example, that most people ceased much to object to anti-abortion laws in their own states, but only because women who wanted an abortion could conveniently travel to a nearby state in which abortion was legal. That would deny the equal value of liberty to people too poor to afford the various expenses of the travel. Could Rawls have thought that it is indeterminate whether there is a basic liberty to an abortion, or to assisted suicide, or to a prayer-free school? If so, he might then have thought that there is a good case for leaving such issues to politics, because politics is superior to adjudication when quasipure-procedural justice is all that is anyway on offer. But it is extremely implausible that Rawls thought that issues like these are matters of indeterminacy, because he himself took positions on several of these issues. Nor could he consistently suppose (as several scholars in effect have supposed) that state-by-state politics is a better vehicle for developing the various public virtues he recognized than is adjudication. His argument that Supreme Court adjudication stimulates the development of the two moral powers seems to apply as thoroughly to divisive cases as to less dramatic ones; indeed, more so.

Rawls and the Law

259

Might he have appealed to the virtue of civility, arguing that it is better not to take decisions that will seem deeply offensive to some citizens? But these decisions will seem equally offensive to the losing side if legislatures rather than courts impose them. In any case, this kind of civility is aimed at a mere modus vivendi, which Rawls rejected as inadequate, rather than at anything one might defend in principle. Of course, if the authority of the Supreme Court or of the constitutional arrangement as a whole were actually at stake, that would be different. We could understand the wisdom of a cautious counsel in that case: Better to ignore the rights of a few people than to sacrifice the system that protects everyone’s rights in the long run. But, of course, that is not the situation. Contrary to Justice Frankfurter’s worries, the Court’s authority survived Brown v. Board of Education33 and the miscegenation cases;34 it has also survived Roe v. Wade35 and the school prayer decisions.36 It could have survived deciding for a limited right to assisted suicide. Indeed, I’m tempted to think that since it has apparently survived the shame of Bush v. Gore,37 it can survive almost anything.

Truth and Objectivity I want finally to notice, though I must do so very briefly, the last aspect of Rawls’s views that I cited as particularly important for legal theory. It is a frequent objection to celebrations of the rule of law that legal judgments, particularly in hard cases, cannot be reports of any objective truth, but simply express the speaker’s psychological state of approval or disapproval. This is a familiar skeptical view about morality and other departments of value, but it is of particular practical importance in law, because it is thought to provide a substantive argument in various controversies: for example, about whether people have a moral duty to obey the law, or whether judicial review of majoritarian legislation is defensible. In Political Liberalism, Rawls identified a conception of objectivity that he believed suitable to political claims, and much of what he said holds for controversial claims of law as well. He insisted that objectivity, in the sense he defined, does not depend on any assumption that political or legal reasoning is a case of perception, that is, that a political or legal claim can be

260

Rawls and the Law

objectively true only when the belief that it is true is caused by the situation it reports. Legal facts are not in any causal relationship with lawyers’ central nervous system. But why should it follow that a controversial proposition of law—that the manufacturers of a dangerous medicine are legally responsible for injuries in proportion to their market share, for example—cannot be objectively true? Whether a proposition claims objective truth depends on its content. It claims objective truth if it claims that its truth is independent of anyone’s belief or preference: that manufacturers would be liable, on the present state of the law, even if lawyers didn’t think so. That is all the claim of objectivity means. Whether that claim is successful depends on the legal arguments we can offer for it, that is, on our reasons for thinking that manufacturers would still be liable even if lawyers didn’t think so. If we think that our reasons for thinking that are good reasons, then we must also think that the proposition that the manufacturers are liable is objectively true. Objectivity so understood doesn’t depend on a metaphysical assumption that seems popular among some so-called moral realists. They think that a proposition can be objectively true only if, in addition to the substantive reasons we can offer for embracing the proposition, the proposition also has a ground in some kind of reality that goes beyond these reasons. They are wrong: Substantive reasons are enough. But they must not be isolated reasons. Our arguments for objectivity are sufficient only if they are sufficiently systematic and mutually and reciprocally examined. Rawls puts that crucial point this way: Political constructivism does not look for something for the reasonableness of the statement that slavery is unjust to consist in, as if the reasonableness of it needed some kind of grounding. We may accept provisionally, though with confidence, certain considered judgments as fixed points, as what we take as basic facts, such as slavery is unjust. But we have a fully philosophical political conception only when such facts are coherently connected together by concepts and principles acceptable to us on due reflection.38

I do not agree with all of Rawls’s discussion of objectivity. Indeed, I think some of it is shown to be unnecessary and unjustified by the rest— his view that we cannot properly claim objectivity for a domain unless we

Rawls and the Law

261

can explain what we take to be error in that domain in a non-questionbegging way, for example. But I commend his general discussion of objectivity to lawyers who hope to understand what their arguments are really about.

Confession Some of you will have noticed a certain congruence between the positions in legal theory I say Rawls’s arguments support and those I have myself tried to defend, and you may think this no accident. So I offer you a confession, but with no apology. The work of philosophical icons is rich enough to allow appropriation through interpretation. Each of us has his or her own Immanuel Kant, and from now on we will struggle, each of us, for the benediction of John Rawls. And with very good reason. After all the books, all the footnotes, all the wonderful discussions, we are only just beginning to grasp how much we have to learn from that man.

Notes

Introduction 1. It may be objected that the distinctions I make do not establish that there are different concepts of law, but only that the single concept of law can be used in different ways. Even if that were so, the important differences among jurisprudential questions that I emphasize would remain: in particular it would remain important to distinguish the question of truth conditions of propositions of law from the sociological and taxonomic questions with which it has often been confused. But it is not so: the concepts are different, though as I said densely interrelated, because they collect different instances. The doctrinal concept collects valid normative claims or propositions, and the sociological concept collects institutions or patterns of behavior. We must take care not to fall into the trap set by the careless use of personification by some legal philosophers that I discuss in Chapters 7 and 8. Suppose we say both that “the law provides a good living for lawyers” and that “the law provides that wills with only one witness are invalid.” We might be tempted to think that these two propositions offer two reports about the same entity so that they use only one concept of “the law”: the concept of the entity that does the providing in both instances. But that would be a serious mistake. These are only personifications, and when we collapse them no single entity remains. The second proposition is only a metaphorical way of stating a proposition of law; it is not a report of what some entity has actually said or demanded. See also Chapter 8, note 23. 2. See Max Weber on Law in Economy and Society, ed. Max Rheinstein (Cambridge, Mass.: Harvard University Press, 1954), 13.

264

Notes to Pages 3–22

3. Lon L. Fuller, The Morality of Law (New Haven: Yale University Press, 1965). 4. See Dworkin, Law’s Empire (Cambridge, Mass.: Harvard University Press, 1986), 102–8. 5. See Joseph Raz, The Concept of a Legal System, 2d ed. (Oxford: Oxford University Press, 1980), 34. 6. I may have contributed to the mistake. In an early essay I suggested that “the law” contains not just rules but certain principles as well. See Taking Rights Seriously (Cambridge, Mass.: Harvard University Press, 1978), Chapter 2. I quickly corrected myself, however. See ibid., Chapter 3, 76. See also Chapter 8 of this book. 7. Perhaps some or all interpretive concepts began their conceptual lives as criterial: perhaps people had the concept of justice when they all understood “unjust” to refer only to acts condemned by convention, for instance. But if so these concepts ceased to function as criterial a very long time ago. The reverse process is, however, a common occurrence. An imprecise criterial concept becomes interpretive when it is embedded in a rule or direction or principle on whose correct interpretation something important turns, for example. If a legislature was silly enough to adopt a special tax allowance for bachelors, a judge would one day have to decide whether an unmarried eighteen-year-old male qualified. He would decide not by stipulating a more precise definition of “bachelor” but by puzzling about which decision better served what he took the point of the tax allowance to be. 8. See Thomas Nagel, “The Psychophysical Nexus,” in his collection Concealment and Exposure: and Other Essays (Oxford: Oxford University Press, 2002), 194. 9. Of course I don’t mean that lawyers make these judgments self-consciously. Education, training, and experience provide them with a sense that is best explained as consisting in an intuitive answer to these questions. 10. Harvard University Press, 1986. 11. See my discussion of the two dimensions of integrity in Dworkin and His Critics, ed. Justine Burley (Malden, Mass.: Blackwell, 2004), 381–82, and in Law’s Empire, 410–11. 12. See Dworkin, A Matter of Principle (Cambridge, Mass.: Harvard University Press, 1985), Chapters 12 and 13. See also Law’s Empire, Chapter 8.

Notes to Pages 25–38

265

13. See Cass R. Sunstein, “From Theory to Practice,” 29 Arizona State Law Journal 389 (1997). 14. See Law’s Empire, 250–54. 15. See Cass R. Sunstein, Daniel Kahneman, David Schkade, and Ilana Ritov, “Predictably Incoherent Judgments,” 54 Stanford Law Review 1153, 1200– 1201 (2002). 16. See Taking Rights Seriously, Chapter 2. 17. H. L. A. Hart, The Concept of Law, 2d ed. (Oxford: Oxford University Press, 1994), 269. 18. See Liam Murphy, “The Political Question of the Concept of Law,” in Jules Coleman, ed., Hart’s Postscript (Oxford: Oxford University Press, 2001), 371. See also, e.g., Tom Campbell, The Legal Theory of Ethical Positivism (Aldershot: Dartmouth Publishing, 1996); Abner S. Greene, “Symposium: Theories of Taking the Constitution Seriously outside the Courts: Can We Be Legal Positivists without Being Constitutional Positivists?” 73 Fordham L. Rev. 1401 (2005). 19. See H. L. A. Hart, “Positivism and the Separation of Law and Morals,” 71 Harvard Law Review (1958), reprinted in his Essays in Jurisprudence and Philosophy (Oxford: Clarendon Press, 1983), 49. 20. See Dworkin, Freedom’s Law (Cambridge, Mass.: Harvard University Press, 1996). 21. See Hart, The Concept of Law, Chapter 5. 22. Law’s Empire, 418 n.29. 23. See, e.g., Stephen Perry, “Hart’s Methodological Positivism,” in Coleman, ed., Hart’s Postscript, 311. 24. Nicola Lacey, A Life of H. L. A. Hart (Oxford: Oxford University Press, 2004). Lacey describes Hart’s fascination with linguistic philosophy on pp. 144–46. Her account of his reflections as a visitor to Harvard Law School in 1956, as revealed in his contemporary letters and notebooks, show how thoroughly he saw himself as engaged in what he referred to as “linguistics.” See ibid., Chapter 8. 25. See Taking Rights Seriously, Chapter 2.

1. Pragmatism and Law 1. Bernard Williams, London Review of Books (January 1991). 2. Richard Rorty, “The Banality of Pragmatism and the Poetry of Justice,” in

266

3.

4.

5.

6. 7. 8. 9. 10.

11.

Notes to Pages 38–43

Michael Brint and William Weaver, eds., Pragmatism in Law and Society (Boulder, Colo.: Westview Press, 1991). Some critics, including Brian Barry and Joseph Raz, suggest that I have changed my mind about the character and importance of the one-rightanswer claim. For better or for worse, I have not. See Ronald Dworkin, Taking Rights Seriously (Cambridge, Mass.: Harvard University Press, 1977), Chapters 4 and, particularly, 13. See also my somewhat earlier article “Is There Really No Right Answer in Hard Cases?” which was reprinted as Chapter 5 in Ronald Dworkin, A Matter of Principle (Cambridge, Mass.: Harvard University Press, 1985); Chapter 7 of that collection; and Chapter 7 of Ronald Dworkin, Law’s Empire (Cambridge, Mass.: Harvard University Press, 1986). That is what makes it so difficult to state the issue, if any, that divides “realists” from “antirealists” in metaphysics, and, more generally, to formulate any philosophically very deep form of skepticism. Notice that I do not claim that lawyers all agree about which side is favored by the best arguments. (I could hardly claim that because a hard case is one in which lawyers do disagree.) Nor do I claim that some algorithmic decision procedure is available that dictates what the right answer is. I have elsewhere described how I think lawyers should think about hard cases, and my description emphasizes how dense with individual judgment that process is. Ronald Dworkin, “The Right to Death: The Great Abortion Case,” New York Review of Books (January 31, 1991): 14–17. See Dworkin, A Matter of Principle, Chapter 5. Dworkin, Law’s Empire, Chapter 7. See Marshall Cohen, ed., Ronald Dworkin and Contemporary Jurisprudence (London: Duckworth, 1984), 271–75 and Chapter 7. These are now collected in Stanley Fish, Doing What Comes Naturally (Durham, N.C.: Duke University Press, 1990). See “Working the Chain Gang: Interpretation in Law and Literature,” Chapter 1; “Wrong Again,” Chapter 2; and “Still Wrong After All These Years,” Chapter 16. See also pp. 384–92. Judge Posner’s book is characteristic of that phenomenally prolific author’s virtues and defects. It is clear, erudite, punchy, knock-about, witty, and relentlessly superficial. He sets out, as his main theoretical aim, to attack what he describes as the right-answer thesis. He has in mind the thesis I described and tried to clarify earlier. He says he means, in claiming

Notes to Pages 44–47

12. 13. 14.

15. 16.

17.

267

that there is no “objective” answer in hard cases, that the experts do not agree in such cases. As it is exactly that feature—disagreement—that makes such cases hard, he wins a quite total victory, for it is undeniable that the experts do not agree in cases in which they disagree. But, of course, that is not what the “right-answer” argument is about. As I said earlier, it is about a legal question of jurisprudential size and philosophical dimension. I described, earlier, some of the facets of that question that other writers have taken up. Posner fastidiously avoids them all, sticking to his trivial claim about disagreement. There is a good deal of interest and fun in his book, and he does range over a wide variety of issues, discussing, for example, statutory and constitutional interpretation. Fish, Doing What Comes Naturally, 342. Fish, “Still Wrong,” in Doing What Comes Naturally, 111–12. It might seem that Fish is appealing here to what I have elsewhere called the demonstrability thesis: that nothing can count as a good argument for any view unless it is demonstrably persuasive, that is, unless no one who is rational can or will resist it. So nothing can count, within practice, as showing that a particular argument is an example of inventing unless everyone agrees that it is. If that is Fish’s point, this is just another example of importing standards of good argument that are foreign to a practice into it from some external level of skepticism. Fish has said, however, that he agrees with me that this kind of external skepticism is pointless. See “Still Wrong,” in Doing What Comes Naturally, 370–71. See Law’s Empire, Chapter 2. Fish recently offered a more dynamic account of the epistemic structure of interpretation; indeed, one that seems congenial to the account I offered in reply to Knapp, about how interpreters are constrained. Fish says, for example, “Even though the mind is informed by assumptions that limit what it can even notice, among these is the assumption that one’s assumptions are subject to challenge and possible revision under certain circumstances and according to certain procedures. . . .” That seems to recognize that interpretation can be internally critical, at the level of practice, which Fish denied in the passage quoted earlier. But he does not suggest that he has changed his mind, perhaps because he presents this dynamic account not as an explanation of the reflective character of interpretive practice but more passively as an explanation of how interpretive styles can come to change. See Fish, Doing What Comes Naturally, 146. Fish says that it would be impossible for judges to be pragmatists in the

268

Notes to Pages 47–48

sense I described earlier because judges cannot help but be influenced by their legal training. That is a non sequitur; the fact that judges are still recognizably acting as judges when they ignore precedent doesn’t mean that is not what they are doing. He says that because interpretation is necessary in reading any statute, the style of adjudication I called conventionalism is impossible. But I defined conventionalism as holding that law is a matter of uncontroversial interpretation, not no interpretation. He says that because even a judge who rejects any responsibility of continuity with the past nevertheless will make a principled decision in some sense, it follows that some sort of integrity is inevitable. But not, of course, the demanding kind of integrity I described as essential to law as integrity. 18. See Fish, Doing What Comes Naturally, 386–87. 19. Part of Professor Fish’s response to my essay (Section V, Chapter 3 of Brint and Weaver, eds., Pragmatism in Law and Society) is welcome: He not only recognizes that theory is very much part of some practices, but he also acknowledges a key component of the best explanation of how theory works within interpretive practices like law. He says that in such practices “competent practitioners operate within a strong understanding of what the practice they are engaged in is for.” He might have added that this fact accounts for the argumentative and dynamic character of such practices. Lawyers often puzzle and disagree about what the law, properly understood, really requires in some situation because, though they share a sense that law is for something—that the various rules and practices that form law’s history have a point—they have different, rival and controversial, accounts of what that point is, either in general or with respect to particular departments or doctrines or rules of law. So legal reasoning is best understood as interpretive in the following way: Lawyers reason about what the law in new or controversial cases is by constructing what they take to be the best justification for past rules and practices, and then trying to extrapolate that justification forward into those new cases. In that way they interpret and re-interpret their institution’s past, formulating, re-formulating, testing and probing rival justifications. They disagree with one another when and because they adopt somewhat different justifications for the same history, or extrapolate much the same justification differently. This process is not self-conscious or explicit in every case: “Easy” cases are those in which any plausible interpretation of the past would dictate the same decision now, and the new decision therefore

Note to Page 48

269

seems unreflective and near automatic. But every appellate judge, at least, faces hard cases, in which the process of justification and extrapolation becomes more self-conscious and explicit, closer to the fully reflective and explicit form it takes in, say, classroom argument, which is only another, differently structured and motivated, forum in which the same practices unfolds. (In Law’s Empire, I try to defend the views about adjudication summarized in this paragraph.) If Fish had continued his account in that way, he would have given an intelligible and accurate account of how legal theory “folds into” legal practice, and also of how academic lawyers and legal philosophers can try to help that enterprise. He is not yet ready, however, to give theory so prominent a place in interpretive practices. So he continued in a very different way more congenial to his former anti-theoretical stance. He says that though lawyers understand that law serves a point, their understanding is “not theoretical in any interestingly meaningful way” because it “generates without the addition of further reflection a sense of what is and is not appropriate, useful, or effective in particular situations.” He still wants, in other words, to picture lawyers and judges as like natural, unreflective athletes: instinctive craftsmen who react unthinkingly to legal problems, deciding as they have been trained to do, as no one trained in that way can help but do, obeying the ancient practices of their profession because it would be unthinkable to do otherwise, supplying justifications for these rules only if asked, and then just by repeating empty phrases they memorized in law school, idle justifications that have nothing to do with their actual practice, except to impress like hydraulics textbooks on a plumber’s shelf. That is an exceptionally poor description of actual legal practice. Fish’s account leaves no room for puzzle or progress or controversy or revolution: It cannot explain how lawyers can worry or disagree or change their minds about what the law is. As I said, his account of interpretive practices leaves them flat and passive. He insists, for example, that judges are simply incapable of challenging settled procedures of adjudication: He thinks it would be as “unthinkable” for them to reconsider conventional principles of court hierarchy and precedent as it would be for them to decide cases by randomized Shakespearian citation. But legal history is choked with examples of judges who called procedural orthodoxy into question. Some challenges ended in failure—those federal court judges

270

Note to Page 48

who claimed the right not to follow past Supreme Court decisions when they thought the Court was about to change its mind, for example, have so far persuaded no one else and have been overruled. In other cases the challenge was dramatic and successful; a few decades ago, for example, the House of Lords, Britain’s highest court, suddenly announced that contrary to settled practice it would no longer be bound by its own past decisions, and though the new practice was thought shocking by some British lawyers, few of them question it now. These are only random examples: Legal history or legal process could provide hundreds of others. In almost all such cases the challenge to orthodoxy and convention was wrapped in an argument of the same underlying structure: that the purposes of adjudication, precedent, hierarchy, and the rest—at least as seen by those proposing the change—would be better served by some more or less radical departure from what had seemed beyond question. Fish makes parallel claims about substantive legal principles. He says that lawyers would be dumbfounded if asked for a justification of the “tools” they use in considering contract cases, like the doctrines of offer and acceptance, mistake, impossibility, frustration, breach, and so forth; he says lawyers no more rely on theories or justifications in using these doctrines than carpenters rely on theories to use nails. But any standard history of how contract law developed in the centuries after Slade’s Case shows how bad an analogy that is, how thoroughly it misstates the role theoretical argument and disagreement played in that process. Each of the doctrines Fish mentions changed in content from period to period, and they still differ from jurisdiction to jurisdiction in the common law world; the changes and differences reflect, among other things, different emphases on the relative importance of freedom of contract, efficiency in commerce, imposing fairness on commercial practice, and protecting people with inadequate bargaining power, just to name four of a large number of theoretical claims lawyers have made or rejected about the point and justification of contract law. Any contemporary contracts case book, moreover, shows how vivid these controversies remain. For the doctrines Fish calls second-nature tools are intensely controversial. It is controversial not only what should count as an offer or an acceptance or a mistake, for example, but also how central these ideas should be to the law’s enforcement of consensual transactions, as the development of doctrines of quasi-contract and contracts of adhesion, and the limited

Notes to Pages 48–49

271

replacement of contract by status, among other trends, seems plainly to show. Once again, the heart of these controversies is the kind of theoretical argument—urging and challenging different justifications—that Fish wants to treat as merely decorative. As I said earlier in this article, legal skeptics challenge one ordinary assumption of legal argument—the assumption that legal questions have right answers. But these skeptics insist, as much as anyone, that legal argument is nevertheless theoretical in just the way Fish denies, because they describe it as an attempt by each side self-consciously to advance its own vision of private law. I therefore believe that though Fish’s views are now less radical and shocking than once they seemed, he still seriously misunderstands the role of theoretical argument in interpretive practices like law and literary criticism. But I must point out one passage, near the end of his response, that might suggest a different conclusion. If I am right, he says, that a lawyer or judge must engage in theoretical reflection just to carry out his role competently, then “there seems little reason to call it theory,” since it is simply the quality of being skilled at one’s job. But if, on the other hand, “theory is used in a more exalted sense . . . we are back in the realm of meta-commentary and high abstraction.” The first of these claims must have surprised every reader. It is surely part of a philosopher’s or a cosmologist’s or a welfare economist’s being “skilled” at his job that he be able to engage in very complex theoretical argument, and we have, in that fact, not “little” but compelling reasons to call what they do theory. Does Fish really mean that theory plays no “meaningful” role in the work of any profession? Or does he mean only that he, for his part, will not use the word “theory” to describe any form of thinking, no matter how selfconscious, that goes with skill at a job, but will reserve that word to describe mental processes that somehow float free of practice in the never-never land of “meta-commentary and high abstraction”? If so, we would finally have nothing left to disagree about, except that since I don’t believe in the never-never world I use “theory” in the normal way.

2. In Praise of Theory 1. Sindell v. Abbott Labs., 607 P.2d 924, 935–38 (1980). 2. The Supreme Court has since decided these cases. See Washington v. Glucksberg, 117 S. Ct. 2258 (1997).

272

Notes to Pages 51–66

3. Richard A. Posner, Overcoming Law (Cambridge, Mass.: Harvard University Press, 1995). 4. Cass R. Sunstein, Legal Reasoning and Political Conflict (New York: Oxford University Press, 1996) (hereinafter Sunstein, Legal Reasoning). 5. I do not mean to revisit here the question of what counts as a justification, and how the interpretive dimensions of fit and morality interact in producing one. See Ronald Dworkin, Law’s Empire (Cambridge, Mass.: Harvard University Press, 1986), 44–86. 6. Ibid., 250–54. 7. 111 N.E. 1050 (N.Y. 1916). 8. Posner, Overcoming Law, 8–10. 9. Ronald Dworkin, “Objectivity and Truth: You’d Better Believe It,” 25 Phil. & Pub. Aff. 87, 89–94 (1996). 10. Posner, Overcoming Law, 11. 11. Dworkin, Law’s Empire, 176–224. 12. See Ronald Dworkin, A Matter of Principle (Cambridge, Mass.: Harvard University Press, 1985), 235–89 (1986) (discussing Posner and utilitarianism). 13. See note 7 above. 14. H. L. A. Hart, The Concept of Law (New York: Oxford University Press, 1961). 15. Edward H. Levi, An Introduction to Legal Reasoning (Chicago: University of Chicago Press, 1949) (first published in 15 U. Chi. L. Rev. 501 (1948)). 16. In 1993, Sunstein proposed a grand, Herculean project for the First Amendment. “I suggest,” he said, “that the First Amendment should be taken to set out a general principle of free expression, and that the contours of that principle should not be limited to the particular understandings of those who wrote and ratified it.” In a footnote attached to that sentence he explained that he had in mind an interpretive exercise of the kind defended, he said, by me. Cass R. Sunstein, Democracy and the Problem of Free Speech (New York: Free Press, 1993), xv, 253. In 1996, however, in a review of my book Freedom’s Law, he attacked my ambition to produce a general principle of free expression for the First Amendment, and he wholly rejected the interpretive method he had only recently endorsed, noting that “the Supreme Court has not made a decision about ‘the point’ of that amendment . . . [T]he complex body of free speech law is not united by a single overarching theory.” Cass R. Sunstein, Book Review,

Notes to Pages 66–75

17. 18. 19.

20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.

273

New Republic, May 13, 1996, 35, reviewing Ronald Dworkin, Freedom’s Law: The Moral Reading of the American Constitution (Cambridge, Mass.: Harvard University Press, 1996) (hereinafter Sunstein, “Review”). Youth’s wisdom had all too short a season. I should add that Sunstein’s review contains a surprising number of misdescriptions of my book. Sunstein, Legal Reasoning, 38–41 (distinguishing three claims for “incomplete” theory). Ibid., 50 (citations omitted). For a more detailed study of how far, if at all, Sunstein’s account ends in an account different from the one I have defended, see Alexander Kaufman, “Incompletely Theorized Agreement: A Plausible Ideal for Legal Reasoning,” 85 Geo. L.J. 395 (1996). Dworkin, Law’s Empire, 265. Ibid., 250–54. See Cass R. Sunstein, “Incompletely Theorized Agreements,” 108 Harv. L. Rev. 1733, 1760–62 (1995); see also Sunstein, Legal Reasoning, 44–46. Sunstein, Legal Reasoning, 53. For evidence that Sunstein does not distinguish the two, see generally Sunstein, “Review.” Sunstein, Legal Reasoning, 54. Cf. ibid., 54–61, with Dworkin, Law’s Empire, 176–90. Sunstein, Legal Reasoning, 55–56. Ibid., 56–57. Ibid., 57. See Laurence H. Tribe, “Comment,” in Antonin Scalia, A Matter of Interpretation: Federal Courts and the Law (Princeton, N.J.: Princeton University Press, 1997), 65, 72–73.

3. Darwin’s New Bulldog 1. See Richard A. Posner, “Against Constitutional Theory,” 73 N.Y.U. L. Rev. (1998); Richard A. Posner, “Conceptions of Legal Theory: A Reply to Ronald Dworkin,” 29 Ariz. St. L.J. 377 (1997) (hereinafter Posner, “Conceptions of Legal Theory”); Richard A. Posner, “The Problematics of Moral and Legal Theory,” 111 Harv. L. Rev. 1637, 1640 (1998) (hereinafter Posner, “Problematics”). 2. Posner, “Conceptions of Legal Theory,” 379.

274

3. 4. 5. 6.

7. 8. 9. 10. 11.

12.

Notes to Pages 75–78

Ronald Dworkin, “In Praise of Theory,” 29 Ariz. St. L.J. 353 (1997). See Ronald Dworkin, “Reply,” 29 Ariz. St. L.J. 431 (1997). Posner, “Problematics,” 1640. Posner suggests that “consensus is the only basis on which truth claims can or should be accepted, because consensus makes ‘truth’ rather than truth forcing consensus.” Ibid., 1657. This “postmodernist” view of science is presented hypothetically, but Posner’s repeated (though sometimes contradicted) claims that diversity of opinion demonstrates lack of objective truth presuppose the view. For further discussion of what I have called Posner’s postmodernist “flirtation,” see Dworkin, “Reply,” 439–440. For a critique of such views, in the context of a recent celebrated exposure of their shallowness, see Paul Boghossian, “What the Sokal Hoax Ought to Teach Us,” Times Literary Supplement, Dec. 13, 1996, 14. See Posner, “Problematics,” 1647. Ibid., 1655. Ibid., 1640. Ronald Dworkin, “Objectivity and Truth: You’d Better Believe It,” 25 Phil. & Pub. Aff. 87 (1996). The most revealing and important mistake is his report that “[Dworkin] runs moral relativism, moral subjectivism, and moral skepticism together, treating them as different names for what he calls ‘external [moral] skepticism.’” Posner, “Problematics,” 1642 n.6 (second alteration in original). The central point of my article was exactly to the contrary: each of these positions makes sense only as a form of what I called “internal” skepticism. Nor does it capture the point of the distinction between external and internal skepticism to say, as Posner does, that the latter is only partial. Ibid. The difference is rather that even though internal skepticism may be global, it is itself rooted in (possibly counterfactual) substantive normative judgments. This might be an appropriate place for me to add for the record that I am unable to understand much of what Posner says throughout his essay about “metaphysics,” “moral realism,” and “right answers.” He is apparently using those terms in an idiosyncratic way. I am asking for Posner’s own views: I will stipulate that he can find articles to cite that are critical of this and other aspects of my work. He is fond of argument by bare citation: see, for example, his citation of a particularly weak critique of Rawls as dispositive of that eminent philoso-

Notes to Pages 78–81

13. 14. 15. 16. 17. 18. 19. 20.

21. 22.

275

pher’s contribution to moral and political philosophy, discussed in Martha Nussbaum’s Response. See Martha Nussbaum, “Still Worthy of Praise,” 111 Harv. L. Rev. 1776, 1778 and n.11 (1998). Such citations can backfire, since the writer is normally taken to endorse the opinions he cites and should therefore check them. In note 98, for example, Posner reports, apparently with approval, Duncan Kennedy’s opinion that I have recommended, as “the legally ‘right answer,’” a variety of positions that include “civil disobedience, nonprosecution of draft card burners, [and] the explicit consideration of distributive consequences rather than reliance on efficiency.” Posner, “Problematics,” 1686 n.98, quoting Duncan Kennedy, A Critique of Adjudication (Cambridge, Mass.: Harvard University Press, 1997), 127–28 (internal quotation marks omitted). I do not see how it would aid Posner’s argument if this were all true. But it is false. I have been careful to say that my views about civil disobedience, including draft-card burning, are not judgments of law, see, e.g., Ronald Dworkin, Taking Rights Seriously (Cambridge, Mass.: Harvard University Press, 1977), 206; and I have been careful to deny legal standing to my opinions about distributive justice, see, e.g., Ronald Dworkin, Freedom’s Law (Cambridge, Mass.: Harvard University Press, 1996), 36. Posner, “Problematics,” 1639. Ibid., 1656–57. See Ronald Dworkin, Life’s Dominion (New York: Knopf, 1993), 28–29; Dworkin, “In Praise of Theory,” 358. Dworkin, “In Praise of Theory,” 356. Ibid., 356–57. See John Rawls, Political Liberalism (New York: Columbia University Press, 1993). See Dworkin, Freedom’s Law. See T. Scanlon, “A Theory of Freedom of Expression,” in Ronald Dworkin, ed., The Philosophy of Law (New York: Oxford University Press, 1977), 153. See H. L. A. Hart, Punishment and Responsibility (Oxford: Clarendon Press, 1968). See Planned Parenthood v. Casey, 505 U.S. 833, 869 (1992) (joint opinion of O’Connor, Kennedy, and Souter, JJ.). Another opinion in that case, Justice Stevens’s, made explicit reference to philosophical literature that had

276

23. 24. 25.

26. 27. 28. 29. 30. 31.

32.

Notes to Pages 81–84

proposed a ground of decision for the case similar to one of the grounds that figured in the three-judge opinion. See ibid., 913 n.2 (Stevens, J., concurring in part and dissenting in part). See sources cited in Dworkin, “Reply,” 435–56. Posner, “Problematics,” 1639. However, it may be that nihilism is in fact the substantive position that flows most naturally from his various contradictory moral opinions described below, 89–90. I do not mean to exclude that possibility by considering the alternative interpretation I later explore. Posner, “Problematics,” 1642. Ibid. See 89ff. See Rawls, Political Liberalism, 3–11. See Dworkin, Freedom’s Law, 1–38. Posner could deny this only if he insisted that the kind of moral code he had in mind was a laundry list of discrete moral judgments about very particular matters, and contained no more general principles the interpretation of which could be a matter of dispute. But that would leave no room at all for the “normative reasoning” that he says he does not want to criticize and that a relativist could recognize. It would also mean that no actual community has a moral code, which would make his relativism alarmingly irrelevant. He says that British judges ignore morality, but his comments on British legal practice are ill-informed and contradict what he has himself recently written. Posner says, “Law in England is an autonomous discipline. Novel issues are resolved largely by interpretation of authoritative texts, consisting of statutes, regulations, and judicial decisions, none of which incorporates a controversial moral theory. Sometimes English judges have to make policy choices, but so rarely that when they do so they have the feeling that they’re ‘step[ping] outside the law.’” Posner, “Problematics,” 1693, quoting H. L. A. Hart, The Concept of Law (Oxford: Clarendon Press, 1994), 272 (alteration in original). In his own book, Posner rejects the descriptive accuracy of Hart’s statement, as much for British as for American practice. See Richard A. Posner, Law and Legal Theory in England and America (Oxford: Clarendon Press, 1996), 15. Posner says, as a criticism of the “stepping outside the law” remark, that “[a] further objection to the positivist conception is that judges and lawyers are not aware of a division

Notes to Pages 85–86

33. 34. 35.

36. 37. 38. 39. 40.

41.

42. 43.

277

between the judge as applier and as maker of law,” and he adds that Hart “is right to point out that the cases in this area are frequently indeterminate and that in deciding such cases the judge is making a value choice . . . rather than engaging solely in analysis, reflection, or some special mode of inquiry called ‘legal reasoning.’” Ibid., 18. Posner’s new description of British practice was never true, and it is a culpable misdescription now. (For an account of British judicial use of political theory in administrative law, for example, see Jeffrey Jowell, “Restraining the State: Politics, Principle and Judicial Review,” 50 Current Legal Problems 189 (M. D. A. Freeman and A. D. E. Lewis, eds., 1997).) It might be instructive for Posner to compare, for example, the Supreme Court’s assisted suicide decisions, which he discusses, see Posner, “Problematics,” 1700–1702, with the British Court of Appeals decision in a parallel case, which expressly relied on the writings of contemporary moral philosophers, see Airedale NHS Trust v. Bland, [1993] 2 W.L.R. 316, 351 (C.A.). See Posner, “Conceptions of Legal Theory,” 388. See Dworkin, “Reply,” 435–36. James A. Henderson, Jr., “Judicial Reliance on Public Policy: An Empirical Analysis of Products Liability Decisions,” 59 Geo. Wash. L. Rev. 1570, 1595 n.131 (1991). Posner, “Problematics,” 1639. Ibid., 1697. Ibid., 1695. Ibid. Ibid., 1700, discussing Brief for Ronald Dworkin, Thomas Nagel, Robert Nozick, John Rawls, Thomas Scanlon, and Judith Jarvis Thompson as Amici Curiae in Support of Respondents, Vacco v. Quill, 117 S. Ct. 2293 (1997) (No. 95-1858), Washington v. Glucksberg, 117 S. Ct. 2258 (1997) (No. 96-110), reprinted in “Assisted Suicide: The Philosophers’ Brief,” N.Y. Rev. Books, Mar. 27, 1997, 41. This is a compressed account of the decision. For a fuller account, see my article “Assisted Suicide: What the Court Really Said,” N.Y. Rev. Books, Sept. 25, 1997, 40. Posner, “Problematics,” 1703. Responding to the suggestion that it would have been wiser for the Supreme Court to have postponed recognizing a constitutional right to abortion, I said that if the Court had done so, but had ultimately decided

278

Notes to Pages 86–87

in favor of abortion rights, it would have incurred the acknowledged “moral cost” of the ruined lives of many young women. Dworkin, “Reply,” 437. Posner replies that the Court, in deciding when it did, also incurred the moral cost of the death of those fetuses who would not have been aborted if it had delayed its decision. See Posner, “Problematics,” 1703. But the Court’s ultimate decision means that, in its view, an early abortion does not involve any violation of rights, and therefore that deciding earlier rather than later does not involve any moral cost of that kind. Posner misses that point because he fails to recognize that the Court’s decision in Roe v. Wade necessarily adjudicated moral issues: the Court did not “balance” moral costs, but rather, at least so far as was necessary to its decision, defined them. 44. See note 22. 45. Posner, “Problematics,” 1701. 46. Most of these statements have a strikingly defensive tone. In an odd footnote, for example, evidently anticipating his surprising distinction, he suggests that my own arguments that judges need moral theory would fall flat if I meant only that they need political theory. See ibid., 1639 n.1. That is baffling: most of my examples of the kind of moral theory judges need, particularly in my work in constitutional law, are principles that, on Posner’s apparent suggestion, would be political rather than personal. Elsewhere he anticipates the objection that he himself relies on ethical arguments, but he offers, in reply, only opaque reassurances. “Ethics and practical reason are not interchangeable with moral theory,” he says, “unless the term is to be used unhelpfully to denote all normative reasoning on social questions.” Ibid., 1697. Of course, moral theory of the kind under discussion doesn’t include strategic or instrumental “reasoning on social questions.” But why doesn’t it include reasoning on social questions that is normative not in these senses, but in the categorical sense of moral reasoning? What sense would any definition of moral theory make if it left moral issues about politics out? On another occasion, Posner says of his argument that on a proper understanding of democracy, euthanasia should be left to the political process: “That is not a moral point unless morality is a synonym for policy.” Ibid., 1701. “Morality” is indeed not a synonym for “policy” if the latter term is used to refer to instrumental or strategic considerations. But Posner’s argument about democracy is not strategic or instrumental; it states a controversial position about how de-

Notes to Pages 87–89

47.

48. 49. 50. 51.

52. 53. 54. 55.

56. 57.

279

mocracy is best understood and administered, and that is—what else could it be?—an argument of political morality. Still elsewhere, he says that judges testing the constitutionality of affirmative action plans need not rely on moral judgments, though in some such cases, judges will have to decide “politically.” See ibid., 1706–7. He doesn’t mean according to party affiliation, I assume, but according to their best judgment of the soundness of the rival factions’ claims in political morality, so the statement completes the self-contradiction. Perhaps Posner intends some distinction he does not elaborate between personal and political morality. That distinction is hardly pellucid. (Is the fairness of judges’ imposing market-share liability a matter of personal or political morality?) In any case, Posner’s ukase would be unmotivated if it barred only personal morality from judicial reasoning. Theories of political morality, including his own theory of democracy, share all the supposed defects—they are certainly controversial and undemonstrable— that he says render any moral judgment unfit for judicial use. See Posner, “Problematics,” 1702–3. Ibid., 1704. See ibid., 1705. Posner briefly criticizes my past use of another case, Riggs v. Palmer, 22 N.E. 188 (N.Y. 1889). See Posner, “Problematics,” 1707. He says there was no moral issue in the case, because everyone agreed that what the murdering heir had done was wrong. I used the case, however, as an example of a decision that was difficult not because the moral judgment involved was controversial, but because it was controversial how great a role that independent moral principle should be allowed to play in statutory interpretation. See Ronald Dworkin, Law’s Empire (Cambridge, Mass.: Harvard University Press, 1986), 15–20. Posner, “Problematics,” 1704. See ibid., 1705–6. Ibid. (internal quotation marks omitted). Romer v. Evans, 517 U.S. 620 (1996). For an elaboration of this argument, see Ronald Dworkin, “Sex, Death, and the Courts,” New York Review of Books, August 8, 1996, 44. See Posner, “Problematics,” 1646. See William James, Pragmatism, ed. Bruce Kuklick (Indianapolis: Hackett, 1981 [1907]), 7–21.

280

Notes to Pages 89–93

58. Posner, “Problematics,” 1642. 59. See ibid. Posner is misled by a discussion in Bernard Williams’s book. See Bernard Williams, Morality: An Introduction to Ethics (New York: Harper and Row, 1972), 20–21. Williams is discussing a “functionalist” version of relativism that is crucially different from the one Posner claims as his own. 60. Posner, “Problematics,” 1643. 61. See ibid. 62. Ibid., 1642. 63. See ibid., 1644. 64. Posner, “Conceptions of Legal Theory,” 382. 65. Posner, “Problematics,” 1704–5. 66. Ibid., 1642. 67. Ibid., 1641. 68. See Richard A. Posner, “Utilitarianism, Economics, and Legal Theory,” 8 J. Legal Stud. 103, 119–27 (1979). 69. See Posner, “Problematics,” 1670 and n.62. 70. See Chapter 1. 71. Posner, “Problematics,” 1642. 72. See ibid., 1708. In a revealing remark in the same vein, Posner suggests that just as teachers may do a better job without studying the theory of education, so judges can do well without studying moral theory, and he calls my claim to the contrary “empty.” See ibid., 1697–98. There are two relevant differences between the two professions. First, teachers are not often called upon to justify what they have done by trying to explain in writing why it is right; they sometimes are, and then they are indeed engaged in a kind of education theory. Second, it is sometimes reasonably clear and uncontroversial what counts as success in teaching—an improvement in students’ test scores, for example—and we can then test a teacher’s trial-and-error efforts instrumentally. It is a signal failure of Posner’s jurisprudence, as this analogy reveals, that he takes the same view of a judge’s job. 73. Ibid., 1643 (internal quotation marks omitted). 74. Ibid., 1704. 75. See Oliver Wendell Holmes, “The Path of the Law,” 10 Harv. L. Rev. 457, 459–60 (1897). 76. Until Posner or someone else explicitly embraces Darwinian pragmatism,

Notes to Pages 94–101

77.

78.

79. 80.

81. 82. 83.

84. 85.

281

I will not say much about it as a normative theory. (It may fall into the category of theories whose careful statement is a sufficient refutation.) It is a weakness of the new evolutionary model of moral development that much of it seems pseudo-science, capable of explaining anything that has actually happened. We might easily, for example, construct an evolutionary account of the very practices Posner condemns. Moral reflection, through various layers of justificatory ascent including the philosophical, is as much a part of human nature as anything else. The capacity and taste for such reflection presumably did not develop much in the ancestral environment that mainly concerns evolutionary theory, but here, as in the case of theoretical science, our later history may have built on capacities present and useful very early in the story. See Richard Posner, “Bush v. Gore as Pragmatic Adjudication,” in A Badly Flawed Election: Debating Bush v. Gore, the Supreme Court, and American Democracy, ed. Ronald Dworkin (New York: New Press, 2002). Richard Posner, Breaking the Deadlock: The 2000 Election, the Constitution, and the Courts (Princeton, N.J.: Princeton University Press, 2001), 171. Ibid., 185–86. I do not agree with Posner’s analysis of the history or content of philosophical pragmatism, but since that analysis does not bear on his argument about Bush v. Gore I will not defend my disagreement here. His account of non-philosophical judicial pragmatism, however, seems to match my own description of judicial pragmatism in Law’s Empire. Posner, “Bush v. Gore as Pragmatic Adjudication,” 201. Ibid., 192, 201. It would make no difference to this point whether the imagined pragmatist judge intended to decide on the merits, not on the equal protection grounds Posner thinks fallacious, but on the Article II grounds he thinks somewhat more respectable. The pragmatic judge would still have had to compare on December 9 the results of refusing any further intervention then with the results of declaring somewhat later that the Florida court had violated Article II, and he would have had no reason to assume on December 9 that the recounters would in any case lay down tools then and, for some inexplicable reason, only take them up again on December 13. Posner, Breaking the Deadlock, 180. Suppose we assume, as Posner imagines some justice believing, that a

282

Notes to Pages 101–122

Gore presidency would be “a national calamity.” (See Posner, “Bush v. Gore as Pragmatic Adjudication,” 207.) Suppose we also assume, as Posner supposes, that the “systemic consequences” of allowing such an opinion to influence a Supreme Court decision would be very bad. If the badness of those systemic consequences outweighs the calamity of a Gore presidency, in the consequentialist scales, in the long run, then a pragmatic judge would not be tempted to follow his political opinion. But if, on the contrary, the calamity outweighs the systemic consequences, even in the long run, then why should he hesitate to do what is, all things considered, the best? Posner’s apparent nervousness at this point suggests that he is, at best, a halfhearted pragmatist. 86. Ibid. 87. We might even add to that proposition, if we think it relevant and true, that it will in fact have the best consequences in the long run if judges decide such cases only on principle. 88. See Chapter 7 of this book.

5. Originalism and Fidelity 1. Supreme Court Chief Justice John Roberts made exactly that point in his Senate confirmation hearings. See my article “Judge Roberts on Trial” in the New York Review of Books, October 20, 2005. 2. See Dworkin, Law’s Empire (Cambridge, Mass.: Harvard University Press, 1986), Chapter 9. 3. There is another possibility pointed out to me by Ori Simchen: that we should read the Eighth Amendment’s reference to cruelty as subjective but not dated, so that it forbids punishments that are widely regarded as cruel at the time such punishments are imposed. As we shall see, Justice Scalia supposes that anyone who rejects the dated subjective reading must adopt an undated subjective one: that is why he insists that his opponents think that the force of the Constitution depends on popular opinion from time to time. But the plausible alternative to a dated subjective reading is not an undated subjective one, but a principled one: one that translates the Eighth Amendment as referring to punishments that really are cruel. 4. Those who say that “substantive due process” is an oxymoronic phrase, because substance and process are opposites, overlook the crucial fact that a demand for coherence of principle, which has evident substantive consequences, is part of what makes a process of decision making a legal pro-

Notes to Pages 124–149

5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

16.

283

cess. See my arguments for integrity as a distinctly legal ideal in Law’s Empire. Antonin Scalia, A Matter of Interpretation (Princeton, N.J.: Princeton University Press, 1977). Ibid., 145. Tribe, in Scalia, Interpretation, 75, 78, 70. Emphasis added. That is a pervasive theme in my books Law’s Empire and Freedom’s Law (Cambridge, Mass.: Harvard University Press, 1996). See Law’s Empire, Chapter 9. Scalia, Interpretation, 77. Ibid., 64. Ibid., 69. Ibid. Cambridge, Mass.: Harvard University Press, 1996. I have chosen this name, on this occasion, just for convenience. I do not mean to endorse any suggestion about the connections between legal pragmatism and pragmatism as a more general movement or school in philosophy. See my article “Sex and Death in the Courts,” New York Review of Books, August 8, 1996, 44.

6. Hart’s Postscript and the Point of Political Philosophy 1. See my book Law’s Empire (Cambridge, Mass.: Harvard University Press, 1986). 2. See my article “Objectivity and Truth: You’d Better Believe It,” 25 Phil. & Pub. Aff. (1996) (hereafter referred to as “Objectivity and Truth”). 3. My example is invented. For real cases involving market-share liability, see, e.g., Sindell v. Abbott Labs., 607 P.2d 924, 935–38 (1980), and cases cited therein. 4. See Ira S. Bushey & Sons Inc. v. United States, 398 F 2nd 167 (1968). 5. H. L. A. Hart, The Concept of Law (Oxford: Oxford University Press, 1994), 269. 6. Ibid., 240. 7. See my book Freedom’s Law: The Moral Reading of the American Constitution (Cambridge, Mass.: Harvard University Press, 1996), particularly the Introduction. 8. Someone might well say, pointing to what he takes to be a clear case—

284

9. 10.

11. 12. 13. 14.

15. 16. 17. 18.

19. 20. 21. 22. 23. 24.

25.

Notes to Pages 151–181

China for example—“You wouldn’t call that a democracy, would you?” But this is a tactical move, and the response—“Yes, I would, and so would most people”—would be disappointing but not itself, even if true, a refutation. See my discussion of the “semantic sting” in Law’s Empire. I do not mean to rule out a poetic claim along this line: if April is the cruelest month, so seven might be called, in an appropriate context, the most unjust number. See “Objectivity and Truth.” See Law’s Empire. See my book Sovereign Virtue (Cambridge, Mass.: Harvard University Press, 2001), Chapter 6. “Justice for Hedgehogs,” the unpublished Dewey Lectures at Columbia University that I mentioned in the introduction to Sovereign Virtue, is more explicitly an attempt to illustrate this kind of philosophy. See Chapter 4 of this volume. Others agree. See, e.g., N. Stavropoulos, “Hart’s Semantics,” in J. Coleman, ed., Hart’s Postscript (Oxford: Oxford University Press, 2001), 59. See Freedom’s Law. See the debate between Justice Scalia and myself in Antonin Scalia, A Matter of Interpretation: Federal Courts and the Law (Princeton, N.J.: Princeton University Press, 1997), 117. See also my article “The Arduous Virtue of Fidelity: Originalism, Scalia, Tribe, and Nerve,” 65 Fordham L Rev 1249 (1997). See L. Murphy, “The Political Question of the Concept of Law,” in Coleman, ed., Hart’s Postscript. I elaborate and criticize this argument from democracy to positivism in Freedom’s Law. A. V. Dicey, Introduction to the Study of the Law of the Constitution, 8th ed. (London: Macmillan, 1915), 114. F. A. Hayek, The Constitution of Liberty (London: Routledge, 1960), 153. See J. Raz, Ethics in the Public Domain: Essays in the Morality of Law and Politics (Oxford: Oxford University Press, 1994). See Chapter 7 of this book. That essay, written some time after the lecture published here was given, briefly summarizes some of the material in the next several paragraphs of this text. Lochner v. New York, 198 U.S. 45 (1905).

Notes to Pages 181–189

285

26. 27. 28. 29.

Southern Pacific Co. v. Jensen, 244 U.S. 205, 222, Holmes dissenting. 304 U.S. 64 (1938). See Freedom’s Law, Chapter 17. There are further conditions of success. Any successful conception of legality must preserve the distinctness of that concept from other political values, including procedural fairness and substantive justice, no matter how closely related and interdependent our theories declare these various concepts to be. If we believe that even quite unjust political arrangements may nevertheless display the virtue of legality, as most of us do, then our account of legality must permit and explain that judgment. How this is to be done is the nerve of an old jurisprudential chestnut: can very wicked places have law? I argued, again in Law’s Empire, that we can answer this question in different ways provided that we surround our answer with enough else by way of an account of legality to capture the necessary distinctions and discriminations. Hart said, in his Postscript, that my remarks on this score concede everything at issue to legal positivism. But he misunderstood. 30. The criticism is not confined to British critics: it appealed to Judge Richard Posner in his Clarendon Lecture at Oxford, though perhaps more as an observation than a criticism, because he added that Hart’s jurisprudence is equally parochial. See Posner, Law and Legal Theory in England and America (Oxford: Clarendon Press, 1997).

7. Thirty Years On 1. Jules Coleman, The Practice of Principle: In Defense of a Pragmatist Approach to Legal Theory (Oxford: Oxford University Press, 2001). Unless otherwise indicated, all quotations from Coleman are from this work. 2. Ronald Dworkin, “The Model of Rules,” 35 U. Chi. L. Rev. 14 (1967), reprinted as Ronald Dworkin, “The Model of Rules I,” in Taking Rights Seriously (Cambridge, Mass.: Harvard University Press, 1978), 14. 3. Coleman is sensitive to this difficulty. In two long footnotes (p. 4 n.3 and p. 10 n.13), he reports and attempts to rebut the view of unnamed readers that his methods and conclusions are very much like mine. 4. I set out to defend this account of legal practice in Ronald Dworkin, Law’s Empire (Cambridge, Mass.: Harvard University Press, 1986). 5. In Law’s Empire, I imagine an ideal judge, whom I call Hercules, who de-

286

6. 7.

8. 9. 10.

11. 12. 13. 14.

Notes to Pages 190–201

velops an overall account of the point of legal practice as a whole and of the best moral justification for the settled law of his community, and uses his conclusions as the basis for deciding fresh cases that come before him. Ibid., 239–40. See H. L. A. Hart, The Concept of Law, 2d ed. (Oxford: Clarendon Press, 1994), 94. See Benjamin C. Zipursky, “The Model of Social Facts,” in Jules Coleman, ed., Hart’s Postscript (Oxford: Oxford University Press, 2001), 219, 251–53. See Michael E. Bratman, “Shared Cooperative Activity,” 101 Phil. Rev. 327 (1992). Emphasis has been added. See Ludwig Wittgenstein, Philosophical Investigations, trans. G. E. M. Anscombe (Oxford: Blackwell, 1953), 202. Wittgenstein argues for a more radical conclusion than Coleman seems to accept: even an explicit agreement among judges about how to decide all future cases would not show that the judges were following the same rule. That agreement would have to be expressed in a proposition, and no proposition could identify the infinite number of cases that might arise. Some commentators read Wittgenstein’s argument as having the skeptical consequence that there is no such thing as following a rule. See, e.g., Saul Kripke, Wittgenstein on Rules and Private Language: An Elementary Exposition (Oxford: Blackwell, 1982), 55. (There may be no more misleading title in the literature of philosophy; Kripke’s argument is far from elementary.) Coleman cites Kripke (p. 81), but he cannot accept Kripke’s skeptical claims about the upshot of Wittgenstein’s argument, because Coleman wants to establish not that there are no conventional rules for law, but that law is entirely a matter of conventional rules. Joseph Raz, Ethics in the Public Domain: Essays in the Morality of Law and Politics (Oxford: Clarendon Press, 1994), 199. Ibid., 199–200. Ibid., 201. In my view, liberty means the freedom to use what is properly or morally your property as you wish provided you respect the rights of others. So liberty is not infringed by just taxation. See Ronald Dworkin, Sovereign Virtue (Cambridge, Mass.: Harvard University Press, 2000), 120–83. But of course others think it is I who misunderstand the concept of liberty.

Notes to Pages 201–211

15. 16. 17. 18. 19. 20. 21. 22. 23.

24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34.

287

My present point is that the correct understanding of certain concepts we share is often controversial. We agree that there is a correct understanding or conception of that concept, but disagree about what that correct understanding or conception is. See Dworkin, Law’s Empire, 45–86. So even if the officials of some community all adopted one understanding of the concept of law, that would not show that their understanding was correct. Raz, Ethics in the Public Domain, 201. See Dworkin, Sovereign Virtue. Raz, Ethics in the Public Domain, 199. See ibid., 202. Ibid., 204. Ibid., 202. Ibid., 208. Ibid. In one passage Raz notes that a community may have adopted, as a matter of law consistent with exclusive positivism, a purely factual test for determining the legislature’s “view” on some matter. Ibid., 217. In fact, however, almost no standards of statutory or constitutional interpretation command the general acceptance in the United States that an exclusive positivist would require to regard these as law. It would apparently follow, for Raz, that American legislatures have no “view” at all as to how citizens should behave. Ibid., 209–10. Ibid., 196–97. Ibid., 197. For an account of constructive interpretation in law, see Dworkin, Law’s Empire, 62–86. See, e.g., Sindell v. Abbott Labs., 607 P.2d 924, 936–38 (Cal. 1980). Raz, Ethics in the Public Domain, 204. See ibid., 214–17. Ibid., 217. See Jeremy Bentham, An Introduction to the Principles of Morals and Legislation (New York: Hafner, 1948 [1823]). Erie R.R. Co. v. Tompkins, 304 U.S. 64 (1938). Ibid., 79 (third alteration in original) (quoting Black & White Taxicab & Transfer Co. v. Brown & Yellow Taxicab & Transfer Co., 276 U.S. 518, 533, 535 (1928) (Holmes, J., dissenting)).

288

Notes to Pages 213–222

35. Hart, The Concept of Law, 2d ed. 36. John Austin, The Province of Jurisprudence Determined, ed. Wilfrid E. Rumble (Cambridge: Cambridge University Press, 1995 [1832]), 18–37. 37. Dworkin, Law’s Empire, 34–35. 38. Hart, The Concept of Law, 2d ed., 246. 39. With characteristic firmness, Coleman says there is “no doubt” that my understanding was wrong (p. 200 n.25). But other writers disagree. See, e.g., Nicos Stavropoulos, “Hart’s Semantics,” in Coleman, ed., Hart’s Postscript, 59, 98. Even some writers who believe that I misinterpreted Hart concede that there is evidence for my interpretation in his book and that some of his statements require “charitable” interpretation to be seen as not confirming my interpretation. See, e.g., Timothy A. O. Endicott, “Herbert Hart and the Semantic Sting,” in Coleman, ed., Hart’s Postscript, 39, 41–47. 40. Hart, The Concept of Law, 2d ed., v. 41. I should say here that I do not mean to accept Coleman’s other attributions, many more of which are inaccurate, or his other criticisms. 42. Dworkin, “The Model of Rules I,” 7–13. 43. Dworkin, Law’s Empire, 130–50. 44. In Law’s Empire, I assume this distinction throughout: see, for example, the discussions of natural law theories at pages 35–36, and of the relation between law and morals on pages 96–98 and 101–4. Also, see the discussion of legal “mistakes” in Taking Rights Seriously, 118–30. 45. See Dworkin, Law’s Empire, 55–73. 46. Timothy Endicott argues that the claim that any paradigm is in principle revisable—the claim that Coleman accepts—is the nerve of my interpretive approach to law, and defends positivism against my “semantic sting” argument by contesting that claim. See Endicott, “Herbert Hart and the Semantic Sting.” 47. Dworkin, Law’s Empire, 413. 48. Ibid., 202. 49. Ibid., 95–96, passim. 50. Ibid., 52. 51. I quote the pertinent sentences: “Suppose, for example, that we share the view that law is a contestable concept in the sense that wherever there is law, what the law is is always a matter of potential dispute, and requires an interpretive practice . . . [S]uch disagreement is part of what we take law to be—part of our shared understanding of the kind of thing it is. Thus,

Notes to Pages 225–228

289

not only is disagreement about the criteria of legality in our community compatible with our sharing the same criteria for applying the concept of law, in this case our disagreement about the criteria of legality in our community is intelligible to us just because we share the same criteria for applying the concept” (p. 182).

8. The Concepts of Law 1. Cambridge, Mass.: Harvard University Press, 1977. 2. Cambridge, Mass.: Harvard University Press, 1986. 3. For a subtle and illuminating account of the bearing of philosophy of language on legal theory, see Nicos Stavropoulos, Objectivity in Law (Oxford: Clarendon Press, 1996). I am grateful to Mr. Stavropoulos for very helpful comments on a draft of the Introduction to this book and of this chapter. 4. 89 Va. L. Rev. 1897 (2003); quotation on 1918. 5. Ibid., 1908. Green elsewhere suggests that perhaps my account of interpretive concepts is not a realist account after all, but is rather “similar” to Nelson Goodman’s starkly nominalist theory of logic. Goodman held that our shared sense of the validity of certain rules of deductive inference is the upshot not of our perceiving Platonic forms of valid inference but of our together achieving an equilibrium between the inferences we are disposed to accept and the rules of inference we are disposed to embrace. Goodman’s account supposes that the equilibrium we have achieved after a (mythic) process of critical adjustment is both wholly contingent—“we” might have settled on a very different equilibrium and then had a very different logic—and wholly a social construction: rules of logic could not have the force they do for “us” unless we all unquestionably now accepted that if A, and if A then B, then B. In that way Goodman’s nominalism about deductive and inductive logic is like what Green calls traditionalism. The meaning of terms is settled by a uniform practice achieved after some period of semantic negotiation. Nothing could be more remote from my account of interpretive concepts than this extreme nominalism. The nerve of my account is that interpretive concepts are contested, not settled, by practice, and I hold that the value questions posed by the use of such concepts have, at least in principle, right answers. 6. Joseph Raz, in The Blackwell Guide to the Philosophy of Law and Legal Theory (Malden, Mass.: Blackwell, 2005), 324, 326. 7. Ibid., 341 n.6.

290

Notes to Pages 229–233

8. Ibid., 331. 9. Ibid., 332. 10. Raz suggests that I believe that a theory of law must be parochial but he offers, as my supposed argument for that conclusion, a bad argument that I do not recognize. He adds, however, that this is a point on which I am not clear. I believe these to be further symptoms of the semantic sting. 11. Law’s Empire, 114. 12. Scott Shapiro, “The ‘Hart-Dworkin’ Debate: A Short Guide for the Perplexed,” forthcoming in Arthur Ripstein, ed., The Cambridge Companion to Dworkin (Cambridge: Cambridge University Press). Shapiro suggests in this article a type of response that he believes positivists could make to my “later” arguments. 13. Coleman has generously given me permission to describe his view in this way. 14. Lacey, A Life of H. L. A. Hart (Oxford: Oxford University Press, 2004). 15. Gardner’s review is at 121 Law Quarterly Review 329 (2005). He says that I discomforted Hart by introducing alien issues into legal philosophy: “Dworkin was arguing that the classic debates in the philosophy of law were mainly to be resolved at the level of first philosophy, or metaphysics. They were not debates merely about the nature of law, legal rules, and so on, but about the nature of human understanding of such things as law and legal rules. By driving him up towards these high planes of first philosophy, Dworkin deprived Hart of his unselfconscious (but fully justified) confidence in his work as a philosopher of law, and made his final replies to Dworkin seem frail and defensive. To switch metaphors, Dworkin tempted Hart to worry about how he was staying upright, so that he wobbled dramatically. Many think that it would have been better for the philosophy of law had he ignored the distraction, and kept his eye firmly on his original destination.” Elsewhere Gardner expands on that metaphor, saying that “most of Hart’s work is philosophically unselfconscious. He works on the problems he works on, not on the further problem of how those problems are to be worked on or what kind of problems they are. He is the metaphysical monocyclist who, as soon as he begins to wonder how he stays upright, wobbles and risks falling off. His forays into the investigation of his own philosophical outlook, with the possible exception of his rejection of a crudely lexicographical approach in his inaugural lecture, were not notably successful.” This idiot-savant judgment of

Notes to Pages 234–237

16. 17. 18. 19. 20. 21. 22.

23.

291

Hart’s philosophical abilities is undeserved. Hart had difficulty not because he had a wobbly grasp of “first philosophy” but because he had sufficient philosophical grasp to see problems that some of his successors in analytic positivism do not. Gardner’s account of the issues that I supposedly introduced is in any case strikingly wrong. I wrote nothing, in Law’s Empire or elsewhere, about “the nature of human understanding of such things as law and legal rules.” My concerns were not epistemological, but conceptual. I said that though Hart aimed to explicate the doctrinal concept of law he had misunderstood the character of that concept and that his claims about the nature of law were bound to be mistaken for that reason. See Taking Rights Seriously (Cambridge, Mass.: Harvard University Press, 1977), Chapter 2. Ibid., 76. Joseph Raz, “Incorporation by Law,” 10 Legal Theory 1–17 (2004). Ibid., 10. See my book Freedom’s Law (Cambridge, Mass.: Harvard University Press, 1996). See Antonin Scalia, A Matter of Interpretation (Princeton, N.J.: Princeton University Press, 1997). The Supreme Court held the proposition false some time ago. See my book Sovereign Virtue (Cambridge, Mass.: Harvard University Press, 2000), Chapter 10. This mistake might be encouraged by the undisciplined personifications to which some legal theorists are addicted and that I criticized in Chapter 7. Propositional attitudes are opaque to substitution: it does not follow from the fact that Jephta directed that the first person he saw after his victory be sacrificed to the gods in gratitude, and the fact that the first person he saw was his daughter, that Jephta directed that his own daughter be sacrificed. He did not know that the first person he would see would be his daughter. So if we fall into the habit of saying that the law or the Constitution “directs” or “commands” that no law be adopted that violates free speech, for example, we might be tempted to think that it does not follow from the further (moral) fact that campaign finance restrictions violate free speech that the Constitution has directed or commanded that no campaign finance restrictions be adopted. We might think: the Constitution might not know that campaign restrictions violate free speech. But

292

Notes to Pages 239–248

that would be a serious mistake: a residue of John Austin’s now-abandoned “command” version of analytic positivism. “The law commands that” can only sensibly be understood as a shorthand way of stating propositions about legal rights, duties, powers, and so forth, and these propositions are transparent to substitution. The fact that Congress has “directed” something is of no pertinence to any proposition of law unless it follows from that direction—as it might not follow—that people have the legal rights, duties, and powers that Congress has directed they have. 24. Raz, “Incorporation by Law,” 12. 25. Matthew H. Kramer, “On Morality as a Necessary or Sufficient Condition for Legality,” 48 Am. J. Juris. 53 (2003). Kramer’s essay shows the perils of any attempt to take the distinction between the two forms of taxonomic positivism as deep or important. He argues that law includes all standards that are both binding on judges and “free-floating,” by which he means that the standard has not been adopted by any authority. When moral principles are binding on judges, they pass that test because they have not been adopted by any authority. But the principles of mathematics also pass the test.

9. Rawls and the Law 1. See Sindell v. Abbott Labs., 607 P.2d 924, 936–38 (Cal. 1980). 2. See H. L. A. Hart, The Concept of Law, 2d ed. (Oxford: Clarendon Press, 1994), vii. 3. See ibid. 4. See ibid. 5. These lawyers were themselves inspired with the idea for market-share liability by a student Comment published by the Fordham Law Review. See Naomi Sheiner, Comment, “DES and a Proposed Theory of Enterprise Liability,” 46 Fordham L. Rev. 963 (1978). 6. John Rawls, A Theory of Justice, rev. ed. (Cambridge, Mass.: Harvard University Press, 1999), 15–19. 7. See generally Jeremy Bentham, An Introduction to the Principles of Morals and Legislation (New York: Hafner, 1948 [1823]). 8. See So. Pac. Co. v. Jensen, 244 U.S. 205, 221 (1917) (Holmes, J., dissenting) (“I recognize without hesitation that judges do and must legislate, but

Notes to Pages 249–254

9. 10. 11. 12.

13. 14.

15. 16. 17. 18.

293

they can do so only interstitially; they are confined from molar to molecular motions”). See the section “Constitutionalism” below. Rawls, A Theory of Justice, 209. Ibid. I do not mean that the argument for interpretivism that I have drawn from Rawls match my own arguments for such a conception. I mean only to show the bearing of Rawls’s work on this central question of jurisprudence. But, at least on one interpretation of the basic structure of Rawls’s argument from the original position, his arguments are in fact not very far from my own. I believe that integrity expresses the right view of equal citizenship: Principles applied to one person must be applied to others unless clearly directed otherwise by competent institutions. In my view, some idea of equality of that sort provides at least part of the set of ideas that the heuristic device of the original position is best understood to model and enforce. However, in footnote 19 to his article “Justice as Fairness: Political Not Metaphysical,” Rawls considers and rejects my interpretation. See John Rawls, “Justice as Fairness: Political Not Metaphysical,” in Collected Papers, ed. Samuel Freeman (Cambridge, Mass.: Harvard University Press, 1999), 388, 400 n.19. See generally Ronald Dworkin, Law’s Empire (Cambridge, Mass.: Harvard University Press, 1986), 276–312. See, e.g., Uhl v. Thoroughbred Tech. and Telecomms., Inc., 309 F.3d 978, 985 (7th Cir. 2002) (referring to Rawls’s “veil of ignorance” from A Theory of Justice); Goetz v. Crosson, 967 F.2d 29, 39 (2d Cir. 1992) (quoting A Theory of Justice); Memphis Dev. Found. v. Factors Etc., Inc., 616 F.2d 956, 959 (6th Cir. 1980) (same); W. Addition Cmty. Org. v. NLRB, 485 F.2d 917, 938 (D.C. Cir. 1973) (same). See Rawls, A Theory of Justice, 274. 410 U.S. 113 (1973). 505 U.S. 833 (1992). Some philosophers believe that a moral right to abortion can be defended even if we assume that a fetus is a person, because even on that assumption a woman has no moral responsibility to continue the burdens of pregnancy. For a discussion of that suggestion, see my book Life’s Dominion (New York: Vintage Books, 1994 [1993]), 102–17. But even if we ac-

294

19. 20.

21. 22. 23. 24. 25. 26.

27. 28.

29. 30. 31.

32. 33.

Notes to Pages 254–259

cept that view, it does not follow that a constitutional right to abortion can be defended in that way. If a state may properly treat a fetus as a person, it may constitutionally treat it as a person toward whom a mother has a special responsibility that excludes elective abortion. See Dworkin, Law’s Empire, 276–312. See Dworkin, “The Secular Papacy,” in Robert Badinter and Stephen Breyer, eds., Judges in Compemporary Democracy: An International Conversation (New York: NYU Press, 2003), 67. John Rawls, Justice as Fairness: A Restatement, ed. Erin Kelly (Cambridge, Mass.: Harvard University Press, 2001), 112. See Bruce Ackerman, We the People (Cambridge, Mass.: Harvard University Press, 1991). John Rawls, Political Liberalism (New York: Columbia University Press, 1996). 347 U.S. 483 (1954). 410 U.S. 113 (1973). See New State Ice Co. v. Liebmann, 285 U.S. 262, 311 (1932) (Brandeis, J., dissenting) (“It is one of the happy incidents of the federal system that a single courageous State may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country”). Vacco v. Quill, 521 U.S. 793 (1997); Washington v. Glucksberg, 521 U.S. 702 (1997). Brief of Amici Curiae Ronald Dworkin et al., Glucksberg (No. 95-1858, 96-110), available at 1996 WL 708956. In addition to John Rawls, the brief was signed by Thomas Nagel, Thomas Scanlon, myself, Robert Nozick, and Judith Jarvis Thomson. 410 U.S. 113 (1973). See John Rawls, Commonweal interview with John Rawls, in Rawls, Collected Papers, 616, 618. For an exhaustive discussion of the status of abortion law in Europe, see Inter-Departmental Working Group on Abortion, Gov’t of Ir., Green Paper on Abortion ¶3.02, available at http://www.taoiseach.gov.ie/index.asp?docID=238 (last visited Apr. 14, 2004). Pub. L. No. 108-105, §3(a), 117 Stat. 1206 (2003). 347 U.S. 483 (1954).

Notes to Pages 259–260

295

34. Loving v. Virginia, 388 U.S. 1 (1967); McLaughlin v. Florida, 379 U.S. 184 (1964). 35. 410 U.S. 113 (1973). 36. Sch. Dist of Abington Township, Pa. v. Schempp, 374 U.S. 203 (1963); Engel v. Vitale, 370 U.S. 421 (1962). 37. 531 U.S. 98 (2000). 38. Rawls, Political Liberalism, 124.

Sources

Chapter 1 was originally published as a portion of “Pragmatism, Right Answers, and True Banality” in Pragmatism and Law and Society, ed. Michael Brint and William Weaver (Boulder, Colo.: Westview Press, 1991). Chapter 2 was originally published in 29 Arizona Law Review (Summer 1997). Chapter 3, minus the appendix, was originally published in 111 Harvard Law Review (1998). The appendix was originally published as part of the Introduction to A Badly Flawed Election: Debating Bush v. Gore, the Supreme Court, and American Democracy, ed. Ronald Dworkin (New York: New Press, 2002). Chapter 4 was originally published as “Do Liberal Values Conflict?” in The Legacy of Isaiah Berlin, ed. Mark Lilla, Ronald Dworkin, and Robert Silvers (New York: New York Review Books, 2001). Chapter 5 is a compressed and edited version of my article “The Arduous Virtue of Fidelity: Originalism, Scalia, Tribe, and Nerve,” 65 Fordham L. Rev. 1249 (1997). Chapter 6 was originally published as “Hart’s Postscript and the Character of Political Philosophy” in Oxford Journal of Legal Studies, vol. 24, no. 1 (2004). Chapter 7 was originally published in 115 Harvard Law Review (2002). Chapter 9 was originally published in 72 Fordham Law Review (2004).

Index

Abortion, 6, 17, 64, 68–69, 79, 81, 86, 89, 91, 103–104, 123, 132, 137, 141–142, 253–254, 256–259, 277n43, 293n18 Abraham, 110–111 Abstraction strategy, 192–195 Abstract language/concepts, 122, 128, 132– 133, 135–136, 138, 148, 177, 182, 184– 185, 210 Academic moralism, 77–78, 83–84, 89, 93 Accuracy, 172–174 Ackerman, Bruce, 256 Adaptationist concept of morality, 90– 91, 93 Adjudication, 71–72, 74, 88, 94–95, 103– 104, 119, 136, 208, 252, 254, 258, 268n17, 269n19 Adjudicative stage, 18–21, 26–27, 30 Affirmative action, 82, 136–137, 244, 279n46 Afghanistan, 85, 106 African Americans, 87–88, 110, 117, 132, 137 Amorality, 93 Analogies, 69–70, 73, 137 Analytic doctrinal positivism, 30–33, 225, 227–228, 240, 291n15 Anarchism, 211 Anthropology, 3, 78, 107, 216, 228; moral, 76–77; political, 153; legal, 166, 213, 215 Anti-theory approach, 50–51, 57–58, 72–73,

76, 78, 91; pragmatism, 23–25, 60–65; metaphysics, 58–60; professionalism, 65–72 Antitrust laws, 244 Application of convention, 191–194 Archimedeans, 47, 141–143, 146–149, 153, 158–159, 164–165, 167, 170, 179, 184 Arithmetic, 234, 238, 292n25 Arrest, 169 Art, 141, 157 Aspirational concept of law, 5, 13, 223 Assisted suicide, 49–50, 79, 85–86, 122, 132, 256–259 Assumptions, 267n16 Austin, J. L., 31, 65, 214, 240, 248, 292n23 Authority, 33, 110, 175, 178–179, 181, 183, 199–211, 216, 292n25; legitimate, 199– 203. See also Power Autonomy, 86 Balkans, 170 Barry, Brian, 266n3 Baseball, 48, 81 Bentham, Jeremy, 65, 174–175, 180, 183, 211–212, 240, 248 Berlin, Isaiah, 26, 105–116, 145–146, 153, 158–159, 161–162 Bill of Rights, U.S., 125, 130. See also specific amendments Biology, 153; moral, 91–93

300

Index

Blackstone, William, 173 Blake, William, 10 Bork, Robert, 168 Bowers v. Hardwick, 138 Brandeis, Louis, 181–182, 211, 256 Bratman, Michael, 195–196 Britain, 6, 108, 145, 163, 177, 180, 185, 211, 230, 252, 256, 270n19, 276n32 Brown v. Board of Education, 87–88, 123, 256, 259 Burke, Edmund, 173 Bush, George W., 24, 94–104, 117–118, 132 Bush, Jeb, 96 Bush v. Gore, 24, 94–104, 259, 281n80 California, 49–50 Campaign expenditures, 237–238, 291n23 Canada, 192 Capital punishment, 121, 125–126, 130, 167–168 Cardozo, Benjamin, 55, 63 Casey decision, 81, 86, 253 Charity, 220–222 Charter, Canadian, 192 Chicago Law School, 51, 57–65, 135 China, 89, 185, 284n8 Churchill, Winston, 108 Civil disobedience, 275n12 Civil War, 28, 94, 99, 103, 117, 132, 206 Claims of law, 162–166, 169–170, 179–180, 240 Clitoridectomy, 76–77 Codes, uniform, 180, 182 Coercive power, 3, 18, 169, 172, 174, 176, 231 Coleman, Jules, 32–33, 187–198, 207, 209– 210, 212, 214–222, 233–236, 240, 285n3, 286n10, 288nn39,46; The Practice of Principle, 187 Colorado, 138 Commerce, 207–208, 230, 239, 243 Commitment, 195–196 Common law, 22, 153, 180–181, 203–204, 208, 251, 270n19 Community, 142, 151, 184, 187, 207, 222, 247; conventions of, 163–164; and shared

activities, 195–197; laws of, 212, 221, 239, 287n23; legal systems of, 231–232; political values of, 252–253; liberal, 253–254 Compensatory justice, 59, 170 “Comprehensive” theory approach, 67 Concepts, 9; criterial, 9–12, 224–225, 229, 264n7; natural kind, 10–12, 152, 154– 155, 166, 223, 225, 227, 229–230; interpretive, 10–12, 22, 29, 31–32, 158, 169, 221–226, 264n7, 289n5 Concepts of law, 1, 214, 221, 263n1; doctrinal, 2, 4–5, 223, 225, 227–240; sociological, 3–5, 223–224, 227–229, 231–232; taxonomic, 4–5, 223, 232–240; aspirational, 5, 223; and semantic sting, 223–226; and Dworkin’s fallacy, 226–227; Raz on, 227–232 Conceptual investigations, 142, 145–147, 150, 154–155, 166, 170, 186, 213–216, 242 Confederacy, 99 Congress, U.S., 96, 98–99, 203–204, 206, 237, 292n23 Conscience, freedom of, 49–50, 52 Consequentialism, 21, 24, 61, 73, 95–104 Conservatism, 173, 212, 224 Constitution, British, 177 Constitution, Colorado, 138 Constitution, U.S., 5–7, 28, 30, 42, 49–50, 57, 71, 94, 103, 117–123, 125–135, 137– 138, 167–168, 180, 182, 189, 192, 194, 206, 209–211, 235–237, 249–250, 256– 257, 282n3, 291n23; and textual fidelity, 118–133, 135–136, 138; abstractions in, 122, 128, 132–133, 135–136, 138, 182, 210 Constitutionalism, 254–259 Constitutional law, 2, 5, 7, 16–17, 28–30, 49, 52–54, 56–57, 62, 68, 71–72, 86, 88, 93, 95, 104, 117–123, 125–138, 167–168, 175, 180–182, 185, 189, 192–194, 196, 209– 212, 235–237, 243, 249–250, 278n46, 294n18 Constitutions: and morality, 6, 16–17, 28– 29, 138–139; and political values, 107– 108, 146–148

Index Constraints, 21–22 Content of convention, 191–194 Contracts, 163, 185, 208, 221, 270n19 Conventionalism, 226, 231, 233, 268n17 Conventions, 163–164, 188–198; application/content of, 191–194 Convergent linguistic practice, 11–12 Coordination, 179, 183 Coordination responsibilities, 66–67, 72 Coventry, bombing of, 108 Criminal law, 146, 151, 206–208 Criterial concepts of law, 9–12, 224–225, 229, 264n7 Critical legal studies, 43, 57, 244 Critical race theory, 34, 57 Cruel punishment, 28–30, 120–122, 125– 126, 130, 167–168, 282n3 Cruzan decision, 42 Cultural sociology, 229–230 Damages, 6, 8, 14, 16–17, 22–23, 49, 53, 64, 69, 163, 170, 208, 244–245 Darwinism, 91–93, 154, 158, 167, 280n76 Davidson, Donald, 36, 220–222 Deconstruction, 36, 57, 200 Democracy, 70, 73, 81, 86, 106, 135, 155– 156, 173, 206, 278n46, 284n8; liberty in, 50, 52, 71, 147, 161, 255; equality in, 58, 161; as political value, 114, 116, 142, 145, 149–152, 154, 158, 168, 170, 183–184, 223–224; majority rule in, 133, 146–148, 160, 176; partnership, 134, 139; judicial review in, 151, 153; and positivism, 211– 212; and law, 223–224, 243 Democratic Party, 98 Demonstrability thesis, 267n14 Deontology, 61 Descriptive concepts, 142, 145, 147, 149– 150, 152, 154, 164–166, 170, 178, 183, 186, 213–214, 242, 244–245, 247 Desire, satisfaction of, 21 Detached values, 154–160, 179 Dewey, John, 36 Dicey, A. V., 177 DNA, 10, 152–153, 166, 215 Doctrinal concepts of law, 2, 4–5, 9, 12–13,

301

19, 22–23, 30–32, 223, 225, 227–240, 263n1 Doctrinal positivism, 235–238, 240; political, 26–30, 240; analytic, 30–33, 225, 227–228, 240 Doctrinal stage, 13–18, 22, 26–27, 226 Dogmatism, 61 Donne, John, 240 Draft card burning, 275n12 Dred Scott decision, 117 Drug manufacturers, 7–9, 14, 16–17, 49–52, 64, 68–69, 73, 143–144, 163, 208, 244– 245, 260 Due Process Clause, 28, 68, 121–123, 134, 182, 208–210 Dworkin, Ronald, 45, 61, 67, 85, 126–127, 191, 214, 217, 220–221, 226, 290n15; Freedom’s Law, 133–134; Law’s Empire, 12, 31, 47, 61, 68, 166, 171, 177–178, 185, 220, 222, 225–226, 230–234, 269n19, 285n29, 291n5; Sovereign Virtue, 162; Taking Rights Seriously, 225, 233 “Dworkin’s fallacy,” 226–227 Economic efficiency, 21, 24 Economics, 50, 136, 153–154, 167, 238, 243, 252; in law, 22, 34 Education, 87–88, 123, 256, 280n72 Efficiency, 172, 174–176, 179–180, 183, 231 Eighth Amendment, 20, 120–121, 125–126, 130, 167–168, 282n3 Embedded approach. See Theory-embedded approach Emotive concepts, 149 Empirical concepts, 145, 164, 166–167, 213–215 Endicott, Timothy, 288n46 Engaged concepts, 142, 154–155, 169 England. See Britain Environmental protection, 24, 81 Equality, 62, 88, 115, 123, 156, 177, 222, 249, 254, 258; as political value, 26, 105, 111–114, 116, 143, 145–146, 149–154, 158–159, 161–162, 165, 169–170, 183– 184, 186; of citizenship, 53, 58, 120–122, 133–134, 137, 139; and political integrity,

302

Index

Equality (continued) 73, 176; conflict with liberty, 106, 108– 109, 142 Equal Protection Clause, 6, 17, 28, 87–88, 120–121, 123, 134, 182, 193–194, 209– 210, 250–251, 254, 281n83 Equilibrium, interpretive, 246–247, 251, 253, 289n5 Erie Railroad v. Tompkins, 181, 212 Essential features of law, 227–229 Ethics, 160–161, 168, 278n46 Euthanasia, 79, 86 Evaluative investigations, 145, 150, 164–165 Evolutionary theory, 281n76 Exclusive positivism, 187–188, 198–211, 234, 238–240, 287n23 Expectation originalism, 29–30 Expediency, 169 Experimental aspect of pragmatism, 63–65 Exposition responsibilities, 66–67, 72 External level of practice, 45 External level of thought/speech, 38– 39, 44 Fairness, 30, 33, 68, 85, 136, 171, 175, 192– 193, 207, 212; and justice, 136, 171, 192– 193, 285n29; justice as, 247–248, 253, 255, 257 Feminism, 34, 258 Fidelity, textual, 117–133, 135–136, 138 Fifth Amendment, 5, 121, 125 First Amendment, 49–50, 71, 121, 127, 129– 131, 134, 235–237, 252, 272n16 Fish, Stanley, 23, 43–48, 267nn14,16,17, 268n19 Flag burning, 49–50, 52, 69, 74 Florida, 96–100, 281n83 Forster, E. M., 162 Foundationalism, 44 Fourteenth Amendment, 5, 28, 50, 120–121, 123, 181, 192, 250 Frankfurter, Felix, 259 Freedom. See Liberty Friendship, 141–142, 157–158, 160, 162 Fugitive Slave Act, 132 Fuller, Lon, 3

Functionalist relativism, 280n59 Fundamentalists, 258 Gardner, John, 185, 233, 290n15 Gay marriage, 10, 153–154, 215 Gender discrimination, 73 Gender equality, 24 Generalizations, empirical, 167 General legal theory. See Legal theory Genocide, 58–60 Georgia, 138 Germany, 108, 169–170 God, 110, 173–174, 187 “Go Fish” game, 4 Good life, 156–158, 160 Goodman, Nelson, 289n5 Gore, Al, 24, 94–104, 282n85 Green, Michael Stephen, 33, 226–227, 289n5 Habeas corpus, suspension of, 94, 103, 206 Hand, Learned, 1, 27, 182, 211 Happiness, 21 Hart, H. L. A., 28, 30–32, 65, 81, 140–145, 164–168, 170, 175, 178–179, 183, 186, 190, 213–215, 217, 233, 240, 245, 265n24, 276n32, 285nn29–30, 288n39, 290n15; The Concept of Law, 26, 30, 65, 140, 163, 165–167, 183, 213–214, 217, 245 Harvard Law School, 265n24 Harvard University, 97 Hate speech, 110–111 Hayek, F. A., 177 Health care, 106, 109, 122–123 Hercules, 51, 55–56, 67–68, 70, 189 History, 238; social, 153; intellectual, 229– 230 Hitler, Adolf, 110 Hobbes, Thomas, 175, 179 Holmes, Oliver Wendell, Jr., 1, 27, 58, 135, 181–183, 200, 211, 240, 248; “The Path of Law,” 93 Holocaust, 67, 110 Homosexuality, 119, 121, 130, 138, 153– 154, 192–193, 215 Housing, 122, 138

Index Humanism, 161 Human rights, 81 Human Rights Act, 186, 256 Immorality, 93, 104, 207 Inclusive positivism, 32, 188–198, 207, 234, 238, 240 “Incomplete” theory approach, 66–72 Independent of purposes, 39–40 Integrated values, 156–160, 162 Integrity, 158, 160; political, 73, 172, 176– 178, 221–222, 242, 268n17, 293n12 Intellectual history, 229–230 Internal level of practice, 45 Internal level of thought/speech, 38 Internal skepticism, 274n11 Interpretive concepts, 10–12, 22, 29, 31–32, 158, 169, 221–226, 264n7, 289n5 Interpretive equilibrium, 246–247, 251, 253, 289n5 Interpretive positivism, 178–183 Interpretive practices, 43–48 Interpretivism, 249–251, 254, 293n12 Ireland, 89 Isaac, 110 Islam, 85, 106 Israel, 252 James, William, 36, 89 Japan, 99 Japanese Americans, 94 Jephta, 291n23 Jews, 67, 110 Jim Crow laws, 137 Joint activities, 195–197 Judges, 70, 163, 185–186, 189, 196, 199, 203, 220, 241, 244, 252, 269n19, 286n10, 292n25; reasoning ability of, 75; and moral issues, 84–86, 88–89, 129, 131, 175, 180, 190, 235–237, 254, 276n32, 278n46, 280n72; and fairness, 85, 208; and pragmatism, 94–95, 102–103, 267n17, 281n83; and textual fidelity, 117–118, 121–124, 132–134, 138; and constitutional principles, 137–139, 141; jurisdiction of, 181–182; positivism of, 187–188,

303

211–212, 225, 242, 276n32; disagreements about law, 190–194; and conventions, 197–198; use of arithmetic, 234, 238; difficult decisions by, 242–243; determining what is law, 247–251 Judgment responsibilities, 66–67, 72 Judicial decisions, 5–7, 164 Judicial review, 147, 149, 151, 153, 210, 243, 254–256, 259 Jurisdictions, 71, 166, 181–182, 184–186, 211, 213, 221, 238–239, 242, 255, 270n19 Jurisprudence, 169, 171–172, 181–182, 184– 187, 213–215, 217, 222, 229, 232, 240– 241, 280n72, 293n12; accuracy, 172–174; efficiency, 172, 174–176; integrity, 172, 176–178 Jurisprudential stage, 12–13, 22, 227 Justice, 114, 142, 151, 158; and law, 1, 35, 223; racial, 24, 59, 87–88, 132, 182; social, 36, 76, 137; compensatory, 59, 170; Rawls’s view of, 81, 161, 241, 246–251, 253–255, 257; and textual fidelity, 132; and fairness, 136, 171, 192–193, 285n29; as abstraction, 148; as concept, 148–149, 224, 264n7; political, 149; value of, 155– 156; versus expediency, 169; as fairness, 247–248, 253, 255, 257; utilitarian view of, 248–250 Justification, 199, 268n19 Justificatory ascent, 53–57, 80–81 Kant, Immanuel, 69, 80, 88, 137, 252, 261 Kelsen, Hans, 213 Kennedy, Duncan, 275n12 Kerry, John, 237 Kramer, Matthew, 240, 292n25 Kripke, Saul, 286n10 Lacey, Nicola, 31, 233, 265n24 Language games, 58–59 Law, 33–35, 141, 143, 145, 213, 223; concepts of, 1–5, 214, 221, 223–240, 263n1; and justice, 1, 35, 223; and morality, 1– 35, 144, 168, 175, 181, 187–188, 233–240, 242; constitutional, 2, 5, 7, 16–17, 28–30, 49, 52–54, 56–57, 62, 68, 71–72, 86, 88,

304

Index

Law (continued) 93, 95, 104, 117–123, 125–138, 167–168, 175, 180–182, 185, 189, 192–194, 196, 209–212, 235–237, 243, 249–250, 278n46, 294n18; propositions of, 2–6, 13–15, 18– 20, 23–27, 30–31, 166, 184, 190, 198–199, 214, 219–220, 225–226, 230, 233–234, 237, 244–246, 260, 263n1; and judicial decisions, 5–7; general theory of, 9–21; and pragmatism, 21–25, 36–48; common, 22, 153, 180–181, 203–204, 208, 251, 270n19; and moral pluralism, 25– 26; and doctrinal positivism, 26–33; criminal, 146, 151, 206–208; claims of, 162–166, 169–170, 179–180, 240; Hart’s defense of, 162–168; rule of, 169, 177, 227, 243, 246; essential features of, 227– 229; defined, 247–251 Law and economics movement, 22, 34 Lawrence v. Texas, 138 Legal anthropology, 166, 213, 215 Legal conventions, 163–164, 188–198 Legality, 142–143, 150, 152, 158, 183–186, 246, 254, 285n29, 289n51; value of, 168– 171; and jurisprudence, 171–178; and interpretive positivism, 178–183 Legal philosophy, 28, 31, 33–34, 140–143, 163, 183, 185–186, 188, 213, 215, 222, 224–225, 229, 232, 241–244, 261, 269n19, 290n15; nature of, 244–247; determining what is law, 247–251; constraints of legal reasoning, 251–254; constitutionalism, 254–259; truth/objectivity, 259–261 Legal positivism, 26–33, 43, 172, 174–176, 178–183, 187–189, 198, 211–212, 217– 218, 225–226, 233, 245, 285n29 Legal pragmatism, 21–25, 36–48, 135–136, 226 Legal realism, 244 Legal reasoning, 242–243, 251–254, 268n19, 277n32 Legal sociology, 34, 166–167, 213–215, 245 Legal systems, 3–4 Legal theory, 2, 9–21, 34–35, 42, 49–51, 72– 74, 140–141, 145, 171, 187, 189, 212–213, 227, 230–231, 235, 244–245, 269n19,

289n10, 291n23; semantic stage, 9–12, 22–23, 29–30; jurisprudential stage, 12– 13, 22, 227; doctrinal stage, 13–18, 22, 26–27, 226; adjudicative stage, 18–21, 26–27, 30; embedded view, 51–53; Hercules/Minerva, 53–57; Chicago Law School, 57–65; professionalism, 65–72; traditional issues, 242–244 Legal values, 168–172 Legislative interpretation, 17, 247–248 Legitimate authority, 199–203 Levi, Edward, 69; Introduction to Legal Reasoning, 66 Liability, 8, 14, 49–53, 64, 68–69, 73, 81, 85, 143–145, 163, 208, 244–245, 260 Liberalism, 16, 212, 224, 252–254 Libertarianism, 146 Liberty, 68, 123, 155, 166, 168, 177, 181, 201, 241, 249, 257–258; as political value, 26, 105, 111–114, 116, 143, 145–146, 149–154, 158–159, 161–162, 165, 169– 170, 183–184, 186; of expression, 49–50, 52, 58–59, 69–71, 73–74, 81, 110–111, 121, 129, 147, 235–237, 272n16, 291n23; conflict with equality, 106, 108–109, 142; religious, 134; sexual, 136–138, 182; political, 255; defined, 286n14 Lincoln, Abraham, 94, 103, 206 Linguistic philosophy, 265n24 Local priority, 25, 70 Lochner v. New York, 70, 181 Logic, 289n5 Love, 160 Machiavelli, Niccolò, 94 Mackie, John, 43 MacPherson v. Buick Motor Company, 55, 63 Majoritarianism, 133–134, 139, 146–148, 160, 176, 212, 249, 255–256, 259 Marriage, 9–10, 153–154, 215 Martinez, Denny, 48 Massachusetts, 238 Matter of Interpretation, A (Scalia), 124 Metaethics, 79, 141–142 Metaphysics, 58–60, 80, 182, 266n4, 290n15 Microsoft Corporation, 244

Index Military service, 6, 28 Mill, John Stuart, 145, 252 Milton, John: Paradise Lost, 120–121, 130 Minerva, 55–56, 213 Miscegenation, 259 Missouri, 42 Modesty, 72–73, 158, 160 Money, 146 Moore, G. E., 157 Moral anthropology, 76–77 Moral biology, 91–93 Morality, 24, 33–35, 75–76, 141, 161, 168, 253, 276n32; and law, 1–35, 144, 168, 175, 181, 187–188, 233–240, 242; political, 5, 13–21, 27–29, 31–32, 50, 56–58, 67, 86–87, 128, 131, 133, 139, 162, 174, 178, 246, 251–252, 279nn46,47; and judicial decisions, 5–7; and constitutions, 6, 16– 17, 28–29, 138–139; and legal pragmatism, 21–25; and doctrinal positivism, 26–33; and pragmatism, 37, 43; independence of, 76–78; adaptationist concept of, 90–91, 93; and inclusive positivism, 189–196; and exclusive positivism, 198– 211 Moral nihilism, 82 Moral philosophy, 6, 76–77, 80, 83, 113, 142, 252 Moral pluralism, 25–26, 105–116 Moral psychology, 76–77 Moral realism, 260 Moral reasoning, 278n46 Moral reflection, 281n76 Moral relativism, 75, 82–83, 89–90, 92–94, 274n11 Moral skepticism, 274n11 Moral sociology, 76, 82 Moral subjectivism, 90, 274n11 Moral theory, 61–62, 65, 73, 75–76, 278n46; defined, 78–81; “strong” thesis, 81–84; “weak” thesis, 84–88; and new pragmatism, 88–94 Moral values, 106, 110–111, 122–123, 129, 135–136, 253–254 Murphy, Liam, 27–28, 175

305

Nagel, Thomas, 107 Natural kind concepts, 10–12, 152, 154– 155, 166, 223, 225, 227, 229–230 Nazis, 4, 67, 108, 169–170, 224 Negligence, 6–8, 14, 16, 81 Neutrality, 142–143, 146–149, 154–155, 165, 178–179, 186, 225 New Deal, 256 New York, 240 Nihilism, 36, 82, 276n25 No-right-answer thesis, 42–43 Normative concepts, 143, 146–147, 149– 150, 154–155, 164–165, 170, 186, 213, 222, 225, 232, 242, 247, 276n31, 278n46, 281n76 Norms, 192 Nuremberg war crimes trials, 169–170, 223 Objective truth, 58–60, 244, 259–261, 274n6 Originalism, 28–30, 118, 125–126, 130 Overlapping consensus, 66–67 Paradigms, 218–219, 221 Parliament, British, 163, 180, 211, 256, 270n19 Parochialism, 185, 211–216, 230–231, 290n10 Partnership democracy, 134, 139 Patriotism, 141–142, 151, 162 Peirce, Charles Sanders, 36 Persian Gulf War, 37 Personification, 263n1 Philosophy: legal, 28, 31, 33–34, 140–143, 163, 183, 185–186, 188, 213, 215, 222, 224–225, 229, 232, 241–261, 269n19, 290n15; political, 50, 105, 113, 140, 142, 146, 153, 155, 161–162, 168, 185, 213– 214, 222, 241; moral, 76–77, 80, 83, 113, 142, 252; linguistic, 265n24 Planned Parenthood of Southeastern Pennsylvania v. Casey, 81, 86, 253 Plato, 168, 173, 289n5 Pluralism, 254; moral, 25–26, 105–116; political, 161; value, 168 Political anthropology, 153

306

Index

Political doctrinal positivism, 26–30, 240 Political integrity, 73, 172, 176–178, 221– 222, 242, 268n17, 293n12 Political justice, 149 Political liberties, 255 Political morality, 5, 13–21, 27–29, 31–32, 50, 56–58, 67, 86–87, 128, 131, 133, 139, 162, 174, 178, 246, 251–252, 279nn46,47 Political philosophy, 50, 105, 113, 140, 142, 146, 153, 155, 161–162, 168, 185, 213– 214, 222, 241 Political pluralism, 161 Political power, 23, 172 Political theory, 278n46 Political values, 106, 108–109, 111, 113–114, 145–162, 164–165, 168–169, 171, 176, 183–186, 252–253 Politics, 36, 79, 84, 86, 97, 111–113, 139, 141, 161, 241, 255–258, 278n46 Popular sovereignty, 256 Pornography, 74 Positivism, 216–222; doctrinal, 26–33, 43, 225, 227–228, 235–238, 240; legal, 26–33, 43, 172, 174–176, 178–183, 187–189, 198, 211–212, 217–218, 225–226, 233, 245, 285n29; political, 26–30, 240; sociological, 26–27; taxonomic, 26–27, 30, 292n25; analytic, 30–33, 225, 227–228, 233, 240, 291n15; inclusive, 32, 188–198, 207, 234, 238, 240; interpretive, 178–183; exclusive, 187–188, 198–211, 234, 238– 240, 287n23; and parochialism, 211–216; and legal theory, 242, 244, 246–249, 251 Posner, Richard, 24–25, 43, 51, 56–57, 60– 63, 72–73, 75–104, 119, 266n11, 274nn6,11,12, 276nn31,32, 278nn43,46, 279nn47,51, 280nn72,76, 281nn80,83,85, 285n30; Breaking the Deadlock, 97; Overcoming Law, 57–58 Postmodernism, 57, 107, 274n6 Poverty, 106, 109, 114, 196 Power: coercive, 3, 18, 169, 172, 174, 176, 231; political, 23, 172. See also Authority Practical approach. See Anti-theory approach Practices, interpretive, 43–48

Pragmatism, 36, 216, 257, 267n17, 281nn80,83; legal, 21–25, 36–41, 135– 136, 226; new, 36–41, 88–94; and right answers, 41–43; and interpretive practices, 43–48; as anti-theory approach, 60– 65; and moral theory, 88–94; Darwinian, 91–94; and Bush v. Gore, 94–104 Prayer, school, 256, 258–259 Precedents, 70 President, U.S., 2, 94, 120–121, 206, 237 Pre-structuralism, 57 Princeton University, 124 Professionalism, 65–72 Progressives, 181–182, 211 Progressive taxation, 154 Propositions of law, 2–6, 15, 18–20, 26, 166, 184, 190, 198–199, 214, 219–220; truth conditions for, 13–14, 23–25, 27, 30–31, 225–226, 230, 233–234, 237, 244–246, 260, 263n1 Psychology, 82, 216, 253; moral, 76–77 Ptolemy, 188 Punishment, 28–30, 81, 120–122, 125–126, 130, 167–169, 206, 282n3 Putnam, Hilary, 37 Quine, W. V. O., 36 Racial discrimination, 58–60, 73, 82, 87–88, 123, 136, 141–142, 256 Racial justice, 24, 59, 87–88, 132, 182 Racism, 73, 110–111 Rawls, John, 34, 66–67, 81, 83, 161, 241– 244, 246–261, 274n12, 293n12; Justice as Fairness, 255; Political Liberalism, 259 Raz, Joseph, 32–33, 175, 179, 188, 198–212, 216, 227–231, 234–239, 266n3, 287n23, 290n10; “Can There Be a Theory of Law?” 227 “Realist” pragmatism, 23 Reasoning, legal, 242–243, 251–254, 268n19, 277n32 Reciprocity, 252–253 Reconstruction, 256 Relativism, moral, 75, 82–83, 89–90, 92–94, 274n11

Index Religion, 156–157, 207, 258; freedom of, 134; and legal reasoning, 242, 251–254 Republican Party, 96, 98 Revolution, 184 Rhode Island, 2 Ricardo, David, 167 Riggs v. Palmer, 233 Right answers, 41–43, 266nn3,11, 275n12, 289n5 Roberts, John, 282n1 Roe v. Wade, 86, 117, 123, 253, 256–257, 259, 278n43 Rome, ancient, 223 Romer v. Evans, 88, 138 Rorty, Richard, 23, 36–40, 42, 44, 59–60, 91 Rule consequentialism, 101–102 Rule of law, 169, 177, 227, 243, 246 Rule of recognition, 32–33, 164, 190, 192– 193, 195 Safety, 24 Scalia, Antonin, 29–30, 118, 124–131, 168, 282n3 Scanlon, Thomas, 81 Science, 141, 157 Scotland, 145 Security, 146 Segregation, 87–88, 123 Semantic originalism, 29–30, 125–126, 130 Semantic stage, 9, 22–23, 29–30; criterial concepts, 9–10; natural kind concepts, 10; interpretive concepts, 10–12 “Semantic sting” argument, 218–219, 221, 223–227, 229, 288n46 Sexism, 73 Sexual liberty, 136–138, 182 Shakespeare, William: Hamlet, 120–121 Shapiro, Scott, 33, 233–234, 240 Shared activities, 195–197 Shared cooperative activity, 195–196 Shared criteria, 151–152, 166, 214, 221–222 Simchen, Ori, 282n3 Skepticism, 40, 42–43, 60, 93–94, 148, 175, 266n4, 267n14, 274n11, 286n10 Slavery, 117, 132 Social Darwinism, 158

307

Social history, 153 Social justice, 36, 76, 137 Sociological concept of law, 3–5, 9, 27, 223– 224, 227–229, 231–232, 263n1 Sociological positivism, 26–27 Sociology, 3, 78, 165, 216, 238; legal, 34, 166–167, 213–215, 245; moral, 76, 82; cultural, 229–230 Sodomy, 119, 138 Sources thesis, 245 South Africa, 185, 220 Speech, freedom of, 49–50, 52, 58–59, 69– 71, 73–74, 81, 110–111, 121, 129, 147, 235–237, 272n16, 291n23 Stalin, Joseph, 107 Statutes, 6, 193, 201, 203–205, 207, 209–210 Stavropoulos, Nicos, 289n3 Stevens, John Paul, 275n22 “Strong” thesis, in moral theory, 81–84 Style, 158, 160 Substantive concepts, 142–143, 147–149, 183, 213–214, 247, 260, 282n4 Substitution, 291n23 Suicide. See Assisted suicide Summers, Lawrence, 97, 99 Sunstein, Cass, 24–25, 51, 57, 66–73, 272n16 Supreme Court, California, 49–50 Supreme Court, Florida, 96–98 Supreme Court, Missouri, 42 Supreme Court, U.S., 1, 24, 29, 42, 85–88, 94–104, 117–119, 122–123, 125, 137–138, 168, 181–182, 193–194, 209–212, 243, 253, 256–259, 270n19, 272n16, 277nn32,43, 282n85. See also specific cases Taliban, 85, 106 Tanner Foundation Lectures, 124–125, 128 Taxation, 5, 106, 109, 112, 114, 146, 148, 151, 154–155, 159, 163, 186, 201, 241 Taxonomic concept of law, 4–5, 9, 27, 223, 232–240, 263n1 Taxonomic positivism, 26–27, 30, 292n25 Texas, 138 Textual fidelity, 117–133, 135–136, 138 Theoretical assent, 25

308

Index

Theory. See Legal theory Theory-embedded approach, 50–53, 56–65, 67, 69–71, 104, 181 Tribe, Laurence, 30, 72, 118, 125–131 Truth, 56, 149, 204, 216; conditions, 13–14, 23–25, 27, 30–31, 225–226, 230, 233–234, 237, 244–246, 260, 263n1; objective, 58– 60, 244, 259–261, 274n6; of claims of law, 163–164, 170 Uniform codes, 180, 182 Utilitarianism, 21, 24, 62, 73, 80–81, 180, 248–251, 254, 257 Value judgments, 141–146, 164 Value pluralism, 168 Values, 26, 105–116; moral, 106, 110–111, 122–123, 129, 135–136, 253–254; politi-

cal, 106, 108–109, 111, 113–114, 145–162, 164–165, 168–169, 171, 176, 183–186, 252–253; detached/integrated, 154–160, 162; legal, 168–172 Violence, 146 Voting, 255 “Weak” thesis, in moral theory, 84–88 Wealth, 21, 149, 196 Weber, Max, 3 Welfare, 21, 61–62 Williams, Bernard, 37, 280n59 Wittgenstein, Ludwig, 36, 58, 197, 286n10 Women’s movement, 34, 258 World War II, 94, 99, 108, 169–170 Yugoslavia, 170