The Problems of Jurisprudence

of

» c e

Posner n

Richard A.

n this book, one of our country's most distinguished scholarjudges shares with us his vision of the law. For the past two thousand years, the philosophy of law has been dominated by two rival doctrines. One contends that law is more than politics and yields, in the hands of skillful judges, correct answers to even the most difficult legal questions; the other maintains that law is politics through and through and that judges wield essentially arbitrary powers. Rejecting these doctrines as too metaphysical in the first instance and too nihilistic in the second, Richard Posner argues for a pragmatic jurisprudence.

I

"Richard Posner is one of the nation's most influential thinkers ... He spearheaded the law and economics movement that has, in the eyes of many, revolutionized the entire legal enterprise ... The bulk of his book consists of close analyses ... and critical discussions ... all of which are bolstered by an impressive array of references to the rich literature of contemporary law, philosophy, science and jurisprudence." —Calvin Woodward, New York Times Book Review "IPosner'sJ book will challenge everyone. All sides will likely draw upon the work, even as they differ with it and among themselves ... (It is an] extraordinarily wide-ranging, intellectually stimulating book." —Robert A. Katzmann, Christian Science Monitor "As one reads through this remarkable book, Posner's vision of law slowly builds to a climax that... is awesome ... [He] tackles all the basic problems of legal thought and of moral and political philosophy since the time of Plato and Aristotle (and| enables us to see the richly textured fabric of our legal system in a more coherent manner than has anyone else." —Stephen B. Presser, Chicago Tribune Richard A. Posner is a judge of the U.S. Court of Appeals for the Seventh Circuit, a senior lecturer at the University of Chicago Law School, and the author of many books, including Sex and Reason and Law and Literature (both from Harvard). Harvard University Press Cambridge, Massachusetts London, England

A L*

THE PROBLEMS of JURISPRUDENCE

Richard A. Posner

W158003 Harvard University Press Cambridge, Massachusetts London, England

C o p y r i g h t © 1990 b y t h e President and F e l l o w s of H a r v a r d C o l l e g e All r i g h t s reserved P r i n t e d in t h e U n i t e d States o f A m e r i c a F O U R T H P R I N T I N G , 1994

First H a r v a r d U n i v e r s i t y Press p a p e r b a c k e d i t i o n , 1993 Library of Congress Cataloging-in-Publication

Data

Posner, R i c h a r d A . T h e p r o b l e m s o f j u r i s p r u d e n c e / Richard A. P o s n e r . p. cm. Includes b i b l i o g r a p h i c a l references. I S B N 0 - 6 7 4 - 7 0 8 7 5 - X (alk. paper) (cloth) I S B N 0 - 6 7 4 - 7 0 8 7 6 - 8 (paper) 1. J u r i s p r u d e n c e . I. Title. K230. P 6 6 5 P 7 6 1990 89-26781 340—dc20 CIP

In M e m o r i a m Paul M . Bator

Contents

Preface I n t r o d u c t i o n : T h e B i r t h o f L a w a n d t h e Rise of Jurisprudence The Origins of Law and Jurisprudence 4 A Short History of Jurisprudence 9 A Preview of the Book 24

PART I. T H E E P I S T E M O L O G Y OF LAW 1. L a w as Logic, R u l e s , a n d Science The Syllogism and Other Methods of Logic Rules, Standards, and Discretion 42 Scientific Observation 61

38

2. Legal R e a s o n i n g as Practical R e a s o n i n g What Is Practical Reason? 71 Authority 19 Reasoning by Analogy 86 A Note on Legal Education 98 3. O t h e r Illustrations o f Practical R e a s o n i n g in L a w Interpretation 101 Means-End Rationality 105 Tacit Knowing 108 Submitting to the Test of Time 112 4. L e g i t i m a c y in A d j u d i c a t i o n The Problem of Rational Prejudgment 124 Consensus 125 Policy versus Pedigree as Warrants for Judicial Action How Are Judges' Visions Changed? 148 Critical Legal Studies 153

130

PART II. T H E O N T O L O G Y OF LAW 5. O n t o l o g y , t h e M i n d , a n d B e h a v i o r i s m

161

Ontological Skepticism 161 Mental and Other Metaphysical Entities in Law Behaviorism and the Judicial Perspective 186

167

6. A r e T h e r e R i g h t A n s w e r s to Legal Q u e s t i o n s ? Questions of Law 197 Questions of Fact 203

197

7. W h a t Is Law, a n d W h y A s k ? Is It a Body of Rules or Principles, an Activity, Holmes, Nietzsche, and Pragmatism 239

220 or Both?

220

PART III. I N T E R P R E T A T I O N R E V I S I T E D 8. C o m m o n L a w v e r s u s S t a t u t e L a w

247

9. O b j e c t i v i t y in S t a t u t o r y I n t e r p r e t a t i o n

262

The Plain-Meaning Fallacy 262 The Quest for Interpretive Theory 269 Indeterminate Statutory Cases 278 10. H o w to D e c i d e S t a t u t o r y a n d C o n s t i t u t i o n a l Cases Is Communication Ever Possible? 293 Beyond Interpretation 299 A Case Study of Politics and Pragmatism 302

286

PART IV. S U B S T A N T I V E J U S T I C E 11. C o r r e c t i v e , R e t r i b u t i v e , P r o c e d u r a l , and Distributive Justice

313

Corrective Justice and the Rule of Law 313 A Note on Retributive Justice—and on Rights 330 Formal Justice 332 Distributive Justice 334 What Has Moral Philosophy to Offer Law? 348 12. T h e E c o n o m i c A p p r o a c h t o L a w The Approach 353 Criticisms of the Positive Theory 362 Criticisms of the Normative Theory 374 Common Law Revisited 387

353

13. Literary, Feminist, and C o m m u n i t a r i a n P e r s p e c t i v e s o n Jurisprudence Law and Literature 393 Natural Law and Feminist Jurisprudence 404 Communitarianism 414

393

PART V. J U R I S P R U D E N C E W I T H O U T FOUNDATIONS 14. N e o t r a d i t i o n a l i s m The Decline of Law as an Autonomous Discipline The Neotraditionalist Response 433 15. A P r a g m a t i s t M a n i f e s t o Index

423 424 454 471

Preface

By "jurisprudence" I mean the most fundamental, general, and theoretical plane of analysis of the social phenomenon called law. For the most part it deals with problems, and uses perspectives, remote f r o m the daily concerns of legal practitioners: problems that cannot be solved by reference to or by reasoning f r o m conventional legal materials; perspectives that cannot be reduced to legal doctrines or to legal reasoning. Many of the problems of jurisprudence cross doctrinal, temporal, and national boundaries. "Philosophy" is the name we give to the analysis of fundamental questions; thus the traditional definition of jurisprudence as the philosophy of law, or as the application of philosophy to law, is prima facie appropriate. Problems of jurisprudence include whether and in what sense law is objective (determinate, impersonal) and autonomous rather than political or personal; the meaning of legal justice; the appropriate and the actual role of the judge; the role of discretion in judging; the origins of law; the place of social science and moral philosophy in law; the role of tradition in law; the possibility of making law a science; whether law progresses; and the problematics of interpreting legal texts. All these are examples of what may be termed "wholesale" problems of jurisprudence, to distinguish them f r o m such "retail" problems as the philosophical pros and cons of forbidding abortion or capital punishment, or of imposing tort liability for failing to rescue a stranger. T h e retail problems are no less w o r t h y than the wholesale ones, nor even less fundamental in a useful sense; they are only narrower. T h e line is as indistinct as it is unimportant, and it will be crossed frequently in this book—for example, to discuss in Chapter 5 both the state of mind required for criminal liability and the distinction between voluntary and involuntary confessions; in Chapter 6 the problem of legal proof; in

Chapter 9 affirmative action; in C h a p t e r 11 abortion; in C h a p t e r 15 free speech. Yet I shall suggest that p h i l o s o p h y has less to contribute to the retail problems of j u r i s p r u d e n c e than to the wholesale ones (see last section of C h a p t e r 11). T h e principal tools that I use in dealing w i t h b o t h the wholesale and the retail p r o b l e m s of j u r i s p r u d e n c e are those of analytical philosophy, the study of " n o r m a t i v e p r o b l e m s about reasons and reasoning" (L. J o n athan C o h e n , The Dialogue of Reason: An Analysis of Analytical Philosophy 11, 1986). A l t h o u g h moral and political philosophy figure as well in m y analysis, the emphasis o n the analytical approach places the question of law's objectivity at center stage. I w a n t to m a k e clear at the outset that I a m a c o n s u m e r rather than a producer of philosophy. I seek neither to c o m p e t e w i t h professional philosophers n o r to take sides in philosophical debates, but only to mine philosophy for insights useful to law. At times, t h o u g h , this requires m e to indicate which a m o n g competing philosophical positions on a question I find m o r e persuasive. T h e spirit in w h i c h I approach the p r o b l e m s of j u r i s p r u d e n c e is that of the epitaph that William Butler Yeats c o m p o s e d for his t o m b s t o n e : Cast a cold eye O n life, on death. Horseman, pass by!

" C o l d " for Yeats is n o t the coldness of indifference or hostility, for w e read in " T h e Wild Swans at C o o l e " of swans that "lover by lover, / T h e y paddle in the cold / C o m p a n i o n a b l e streams or climb the air." It is the cold of d e t a c h m e n t — a clear-eyed, no-nonsense, but passionate detachment, as in these lines f r o m " T h e Fisherman": 'Before I am old I shall have written him one Poem maybe as cold A n d passionate as the dawn.'

I try to examine the p r o b l e m s of j u r i s p r u d e n c e with the coldness reco m m e n d e d by Yeats, and hence w i t h o u t the piety and cant that are de rigueur in m a n y discussions of law. I defend a j u r i s p r u d e n c e that is critical of f o r m a l i s m (less pejoratively, of traditional legalism) and that has affinities w i t h legal realism, shorn h o w e v e r of the left-of-center politics characteristic of that m o v e m e n t and its offspring. It is a jurisprudence that, like legal realism, d r a w s o n the p h i l o s o p h y of p r a g m a t i s m (though not only on that philosophy) but that, unlike s o m e versions of legal realism, seeks to d e m y t h o l o g i z e law w i t h o u t either denigrating or diabolizing it.

I have tried to keep the presentation as simple as possible, w i t h a m i n i m u m of legal and philosophical j a r g o n . T h e r e is n o reason w h y the fascinating subject o f j u r i s p r u d e n c e should be the exclusive preserve of the handful of academic lawyers w h o specialize in it. It should be of interest to all lawyers and law students—and also to the increasing n u m ber of philosophers, economists, political scientists, sociologists, psychologists, and historians interested in law and justice. Because this is a book not about t h e scholarship of j u r i s p r u d e n c e but about the problems which that scholarship addresses, the reader should n o t expect a c o m prehensive exegesis of the classics of jurisprudence. N o r do I discuss all the problems of jurisprudence, although I touch on e n o u g h of t h e m to warrant titling the b o o k as I have. At times the touch is light, h o w e v e r , and perhaps the implicit title is The Problems of Jurisprudence as I See Them. M y interest in j u r i s p r u d e n c e goes back a n u m b e r of years. In the 1970s I began writing about B e n t h a m ' s attack on Blackstone, about the n o r mative foundations of e c o n o m i c analysis of law, about law's roots in revenge, and about corrective justice. Since b e c o m i n g a j u d g e in 1981, I have, naturally, b e c o m e fascinated b y the issue of objectivity in a d j u dication, and it occupies center stage in this b o o k . T h e approach I take to the issue is sketched in " T h e Jurisprudence of Skepticism," 86 Michigan Law Review 827 (1988), and I reprint portions of that article in this book along w i t h portions of several other articles: " T h e C o n c e p t of Corrective Justice in Recent Theories of Tort Law," 10 Journal of Legal Studies 187 (1981), © 1981 T h e U n i v e r s i t y of Chicago; "Lawyers and Philosophers: A c k e r m a n and O t h e r s , " 1981 American Bar Foundation Research Journal 231, © 1981 American Bar Foundation; "Wealth M a x imization Revisited," 2 Notre Dame Journal of Law, Ethics and Public Policy 85 (1985), © 1985 T h o m a s J. White Center on Law & G o v e r n m e n t , University of N o t r e D a m e ; " T h e Decline of L a w as an A u t o n o m o u s Discipline: 1962-1987," 100 Harvard Law Review 761 (1987), copyright © 1987 T h e H a r v a r d L a w Review Association; "Legal Formalism, Legal Realism, and the Interpretation of Statutes and the C o n s t i t u t i o n , " 37 Case Western Reserve Law Review 179 (1987), c o p y r i g h t © 1987 Case Western Reserve University; " T h e Law and E c o n o m i c s M o v e m e n t , " 77 American Economic Review Papers and Proceedings 1 (May 1987), copyright American E c o n o m i c Association 1987; " C o n v e n t i o n a l i s m : T h e Key to Law as an A u t o n o m o u s Discipline?" 38 University of Toronto Law Journal 333 (1988), © University of T o r o n t o Press 1988; " C o n s e r v a t i v e F e m inism," 1989 University o f Chicago Legal Forum 191, © 1989 T h e U n i versity of Chicago. T h e b o o k is largely new, h o w e v e r , and its previously published c o m p o n e n t s appear here heavily revised and rearranged.

Selections f r o m " A C o a t , " " T h e Fisherman," and " U n d e r Ben Bulben" are reprinted w i t h permission of Macmillan Publishing C o m p a n y f r o m Collected Poems ofW. B. Yeats, c o p y r i g h t 1916, 1919 by Macmillan Publishing C o m p a n y , renewed 1944, 1947, b y Bertha Georgie Yeats and copyright 1940 by Georgie Yeats, renewed 1968 by Bertha Georgie Yeats, Michael Butler Yeats, and A n n e Yeats. I a m m o r e than usually indebted for help in bringing this b o o k to its present f o r m . Ricardo Barrera, J a m e s Taggart, Catherine Van H o r n , Darren Fortunato, Laura Neebling, A d a m Pritchard, Barbara Smith, Steven Hetcher, and Philip Clark provided excellent research assistance. Ronald Allen, Dennis Black, Steven B u r t o n , Frank Easterbrook, William Eskridge, R o b e r t Ferguson, D a v i d Friedman, Robert Fullinwider, Linda H i r s c h m a n , Stephen H o l m e s , David Luban, Stephen McAllister, Eric Posner, Margaret Radin, Eva Saks, George Stigler, Stephen Stigler, David Strauss, Lloyd Weinreb, R o b i n West, and an a n o n y m o u s reader for the H a r v a r d University Press m a d e valuable c o m m e n t s on the m a n uscript; so did participants in the University of Chicago's Seminar for C o n t e m p o r a r y Social and Political T h e o r y and in the University of Maryland's Legal T h e o r y W o r k s h o p . T h e extensive and provocative c o m m e n t s of Lawrence Lessig, Frank Michelman, and Martha M i n o w require a special a c k n o w l e d g m e n t , as d o Albert Alschuler's valiant efforts, t h r o u g h c o m m e n t s and correspondence, to save m e f r o m error; the c o m m e n t s , support, guidance, and wise counseling of Michael A r o n s o n , General Editor at the H a r v a r d University Press; and the sensitive and meticulous copyediting of Elizabeth Gretz. T h e philosophers Daniel Brudney, J o n a t h a n C o h e n , Jules C o l e m a n , Michael C o r r a d o , and Russell H a r d i n w e r e kind e n o u g h to read parts or all of the m a n u script and point out errors in it; Brudney's c o m m e n t s resulted in substantial changes in several chapters. To Ronald Allen and Steven B u r t o n I o w e an additional debt for helpful conversations on the subject matter of the b o o k , while to Cass Sunstein I o w e a particularly large debt for extensive c o m m e n t s , and n u m e r o u s conversations, that resulted in m a j o r revisions. M y last a c k n o w l e d g m e n t is to the late Paul Bator, w h o in a review of an earlier b o o k of mine called m e "a captive of a thin and unsatisfactory e p i s t e m o l o g y " ( " T h e Judicial Universe of J u d g e Richard Posner," 52 University of Chicago Law Review 1146, 1161 [1985]). I f o u n d this an arresting accusation and one w i t h considerable merit, and it stimulated m e to e x a m i n e the p r o b l e m s of j u r i s p r u d e n c e in greater depth than I had ever expected to. In addition, conversations w i t h Professor Bator on the subject matter of the b o o k helped to shape its themes and avert m a n y pitfalls.

THE P R O B L E M S OF J U R I S P R U D E N C E

Introduction The Birth of Law and the Rise of Jurisprudence

Jurisprudence addresses the questions about law that an intelligent layperson of speculative b e n t — n o t a l a w y e r — m i g h t think particularly interesting. W h a t is law? A system of rules? O f rules plus judicial discretion? O f principles? O r is it j u s t organized public opinion? Is it a thing, entity, or concept at all—and if not, does this m a k e the original question ("What is law?") meaningless? W h e r e does law c o m e f r o m ? Can there be law w i t h o u t lawyers and judges? Is there progress in law? H o w do w e k n o w w h e n a legal question has been answered correctly? What is " k n o w l e d g e " of l a w k n o w l e d g e of? W h a t conditions are necessary or sufficient to m a k e law objective? Should law even try to be objective? Can legal findings of fact ever be verified? Is there really a distinct f o r m of legal reasoning or is it identical to s o m e other f o r m , such as moral or e c o n o m i c reasoning? Is l a w an a u t o n o m o u s discipline? H o w , if at all, does it differ f r o m politics? D o e s the criminal law presuppose the existence of minds? O f free will? D o e s the case for p r o tecting free speech depend o n the existence of truth? Is objective interpretation of statutes and constitutions a chimera? W h a t is the p u r p o s e of law? H o w d o w e identify a w e l l - f u n c t i o n i n g legal system? Is law a science, a humanity, or neither? If a law is sufficiently vicious does it cease to be law? Does " l a w " m e a n the same thing in all these questions and, if not, w h a t is the range of meanings of the w o r d ? A practicing lawyer or a j u d g e is apt to think questions of this sort at best irrelevant to w h a t he does, at w o r s t naive, impractical, even childlike ( h o w high is up?). 1

'

1. C o m p a r e w i t h the q u e s t i o n s listed a b o v e such l a w y e r s ' q u e s t i o n s as: C a n the d o c t r i n e of pendent j u r i s d i c t i o n be used to add a p a r t y to a case or j u s t to add a claim? D o e s the replacement of c o n t r i b u t o r y by c o m p a r a t i v e n e g l i g e n c e i m p l y the d e m i s e of the d o c t r i n e of last clear chancc? C a n a p e r s o n w h o r e t u r n s a lost article t o its o w n e r claim the r e w a r d posted for its return if he d i d n ' t k n o w a b o u t the r e w a r d ? D o e s a l l o w i n g a federal m a g i s t r a t e

To question the point of asking such questions is valid; I hope to show that there is a point. When I entered Harvard Law School as a student thirty years ago, the emphasis in legal education was heavily practical, in the sense of antitheoretical. (It is only slightly less so today.) I recall vividly the questions with which m y first-year courses began. In torts it was whether an assault, to be actionable at law, requires a touching of the victim. The vehicle for discussing this question was a fourteenth-century case in which the defendant had s w u n g at a tavernkeeper with an axe and missed. 2 In property the opening question was the meaning of "adverseness" in the doctrine of adverse possession. A person w h o occupies land in the honest but mistaken belief that it is his can by passage of time acquire a good title, but only if his possession is adverse to the "true" owner: a tenant, whose possession is not adverse to his landlord, does not acquire ownership of the premises he rents, no matter h o w long he has rented them. In contracts w e examined the difference between liquidated damages—an estimate made in the contract itself of h o w much the promisee is likely to lose in the event of a breach—and penalties. Liquidated-damages clauses are enforceable, penalty clauses not. And in civil procedure we examined the difference between "substance" and "procedure." T h e federal rules of civil procedure authorize a litigant to require his adversary to undergo a medical examination if the adversary's physical condition is pertinent to the litigation, as it often will be in a personal-injury case. The question in the case we read was whether the rule is substantive, and consequently outside the congressional authorization for rules of procedure, or procedural, and hence within it. 3 This way of studying law, which involves beginning the study in

to preside o v e r a j u r y trial in a case w h e r e federal j u r i s d i c t i o n is based o n diversity o f citizenship violate Article III of t h e C o n s t i t u t i o n ? T h e s e q u e s t i o n s are quite general b y l a w yers' s t a n d a r d s , b u t t h e y are n o t j u r i s p r u d e n t i a l q u e s t i o n s in m y sense. T h e y are not general e n o u g h , t h e y are n o t f u n d a m e n t a l , and t h e y are n o t — a n d p e r h a p s c a n n o t usefully be— a p p r o a c h e d f r o m the s t a n d p o i n t o f p h i l o s o p h y . I d o n o t s u g g e s t that the w o r d " j u r i s p r u d e n c e " has an established m e a n i n g ; it does n o t . See, for e x a m p l e , R. H . S. Tur, " W h a t Is J u r i s p r u d e n c e ? " 28 Philosophical Quarterly 149 (1978). I use the w o r d to m e a n the set o f issues in o r a b o u t law that p h i l o s o p h y can illuminate. 2. I. de S. & Wife v. W. de S., Y. B. Liber A s s i s a r u m , 22 E d w . 3, f. 99, pi. 60 (1348 or 1349); for text and discussion, see m y b o o k Tort Law: Cases and Economic Analysis 13—16 (1982). 3. Sibbach v. Wilson & C o . , 312 U . S . 1 (1941).

medias res, taking the structure of the legal system and the principal rules of law for granted, and treating cases decided in different eras as if all had been decided yesterday (rather than treating old cases as historical curiosities), remains an invaluable m e t h o d of professional instruction. It conveys to the student the texture of legal practice, and it does so m o r e coherently and efficiently than w o u l d an apprenticeship system of legal training. It inducts the students into the f o l k w a y s of the profession. It drills t h e m in the critical use of logic—an essential t h o u g h sometimes o v e r w o r k e d technique of legal analysis. It teaches t h e m to think o n their feet. It instills—again m o r e efficiently than w o u l d alternative systems of training—essential lessons, such as that legal language often differs f r o m lay language (the legal meanings of "assault," "adverse," "substance," "procedure," and "penalty" are all counterintuitive), that—a related point—legal meanings are heavily dependent on context ("procedure," notoriously, has m a n y different legal meanings), that m a n y rules of law must be inferred f r o m — t h e y are not stated clearly, or usably, i n — j u d i cial decisions, and that at its margins law is far m o r e plastic than the lay public believes. A layperson is apt to think that the answer to every legal question is contained in a b o o k s o m e w h e r e , and that all one has to k n o w is where to look. T h e law student soon learns better. These lessons, b a c k g r o u n d , and competences, and the "Socratic" m e t h o d by w h i c h they are imparted, continue to supply essential p r e p aration for the practice of l a w — b r o a d l y conceived to include j u d g i n g and other law functions as well as practice in a law f i r m — a t the high level of skill that m a n y American practitioners display. But the concern of this b o o k is not w i t h — m o r e precisely, n o t at—the practice level of law; and it should be apparent that the pedagogic approach j u s t described is not oriented to asking and answering the kind of questions with which I began, unless the teacher is Socrates himself. T h e relationship between a conventional legal education and doing j u r i s p r u d e n c e is m u c h like that between learning a language and d o i n g linguistics. T h e general, f u n d a m e n t a l , and deeply problematic character of the "questions w i t h w h i c h I began m a r k t h e m as philosophical. To practical people, however, including j u d g e s and lawyers and even m a n y law p r o fessors, philosophy is an exasperating subject. Philosophers seem preoccupied with questions that n o one w i t h a m o d i c u m of c o m m o n sense and a living to earn w o u l d waste a m i n u t e on, such as: H o w do w e k n o w that any other person has a m i n d , w h e n w e can never observe it? H o w can w e prove that the sun will rise t o m o r r o w , or that n o zebras wear overcoats in the wilds—or can we? H o w even in principle can ideas, w h i c h have n o spatio-temporal locus, affect bodies? If y o u are in

a room that is locked f r o m the outside but do not know it, and while it is locked you decide not to leave, are you exercising free will in making this decision or are you "acting" under compulsion? H o w do you know you are not a brain in a vat, being fed impressions of an external world by a mad scientist? I ask nonphilosophical readers to suspend their disbelief, for I hope to show that questions such as these, remote as they may seem not only from c o m m o n sense but from c o m m o n law (and every other sort of law), can shed light on the perennial problems of jurisprudence. Meanwhile, Bertrand Russell's eloquent defense of philosophy is worth pondering: T h e man w h o has no tincture of philosophy goes through life imprisoned in the prejudices derived f r o m c o m m o n sense, f r o m the habitual beliefs of his age or his nation, and f r o m convictions which have g r o w n up in his mind without the co-operation or consent o f h i s deliberate reason. To such a man the world tends to become definite, finite, obvious; c o m m o n objects rouse no questions, and unfamiliar possibilities are contemptuously rejected. As soon as w e begin to philosophize, on the contrary, we find . . . that even the most everyday things lead to problems to which only very incomplete answers can be given. Philosophy, though unable to tell us with certainty what is the true answer to the doubts which it raises, is able to suggest m a n y possibilities which enlarge our thoughts and free them f r o m the tyranny of custom. Thus, while diminishing our feeling of certainty as to what things are, it greatly increases our knowledge as to what they may be; it removes the s o m e w h a t arrogant dogmatism of those w h o have never travelled into the region of liberating doubt, and it keeps alive our sense of w o n d e r by showing familiar things in an unfamiliar aspect. 4

The Origins of Law and

Jurisprudence

To understand the problems of jurisprudence—even to understand why there are such problems—it is necessary to k n o w a little about the origins of law, itself a jurisprudential question when pursued in a speculative fashion. 5 The ultimate source of these problems lies in the divi4. The Problems of Philosophy 156—157 (1912). " P h i l o s o p h y , b e g i n n i n g in w o n d e r , as Plato and Aristotle said, is able to fancy e v e r y t h i n g different f r o m w h a t it is. It sees the familiar as if it w e r e s t r a n g e , and the s t r a n g e as if it w e r e familiar. It can take t h i n g s u p and lay t h e m d o w n again. Its m i n d is full of air that plays r o u n d every subject. It rouses us f r o m o u r native d o g m a t i c s l u m b e r and breaks u p o u r caked p r e j u d i c e s . " William J a m e s , Some Problems oj Philosophy: A Beginning of on Introduction to Philosophy 7 (1911). 5. Such q u e s t i o n s are the d o m a i n of historical j u r i s p r u d e n c e , the field m a d e f a m o u s by Savigny, Maine, and H o l m e s . See the interesting discussion in E d w i n W. Patterson, " H i s -

sion o f l a b o r . Even the simplest society has n o r m s , tacit or explicit, that evolve f r o m the needs of the society before there are j u d g e s or other officials/' T h e fact that n o r m s precede f o r m a l legal systems is, n o doubt, one cause of belief in "natural law." In its strongest f o r m , this term, which w e will encounter m a n y times, denotes the idea that there is a b o d y of suprapolitical principles that u n d e r w r i t e "positive law," m e a n ing law laid d o w n by courts, legislatures, or other state organs. W h e n a customary n o r m is violated to someone's i n j u r y in a simple, "prelegal" society, the instinct of the victim or his family to take revenge is activated. Tacit n o r m s enforced by the threat of revenge arc the r u d i m e n tary f o r m of a legal system; or, if one prefers, a f o r e r u n n e r to it, for it is u n i m p o r t a n t f r o m a practical standpoint w h e t h e r o n e calls a system of enforcing customary n o r m s t h r o u g h revenge " l a w " or "prelaw." What is i m p o r t a n t is that the grave drawbacks of a revenge system 7 make it intolerable except in the smallest or m o s t primitive societies. There are h u g e advantages to having specialists in the creation and enforcement of n o r m s , and as soon as society can afford t h e m these specialists emerge. T h e history of law in f o r m s recognizable to us (that is, of publicly declared and enforced norms) is to a significant degree one of increasing specialization in the p e r f o r m a n c e of legal tasks. In the first stage after the pure system of private revenge, a chief or king, or perhaps even a popular assembly, will legislate and adjudicate as undifferentiated aspects of governing. Examples are A g a m e m n o n in Euripides' Hecuba, C r e o n in Antigone, and the j u r y that tried and c o n d e m n e d Socrates. 8 Gradually these functions are hived off to specialists, but even before

torical and E v o l u t i o n a r y T h e o r i e s of Law," 51 Columbia Law Review 681 (1951). For recent c o n t r i b u t i o n s see Peter Stein, " T h e Tasks o f Historical J u r i s p r u d e n c e , " in The Legal Mind: Essays for Tony Houore 293 (Neil M a c C o r m i c k and Peter Birks eds. 1986); Stein, Legal Involution: The Story of an Idea (3980); Stein, " A d a m S m i t h ' s J u r i s p r u d e n c e — B e t w e e n M o r a l i t y and E c o n o m i c s , " 64 Cornell Law Review 621 (1979); E. D o n a l d Elliott, " T h e E v o lutionary Tradition in J u r i s p r u d e n c e , " 85 Columbia Law Review 38 (1985); H e r b e r t H o v e n k a m p , " E v o l u t i o n a r y M o d e l s in J u r i s p r u d e n c e , " 64 Texas Law Review 645 (1985). 6. For an interesting discussion, f r o m a philosophical perspective, of the e m e r g e n c e of n o r m s , see Gilbert H a r m a n , The Nature of Morality: An Introduction to Ethics 110-111 (1977). 7. Discussed in m y b o o k s The Economics of Justice, pt. 2 (1981), and Law and Literature: A Misunderstood Relation, ch. 1 (1988). 8. T h e trial and c o n d e m n a t i o n of Socrates (sec T h o m a s C . B r i c k h o u s e and N i c h o l a s D . Smith, Socrates on Trial 24—37 [1989]) by a j u r y of h u n d r e d s , w i t h no professional j u d g e s , no deliberations, and n o possibility of appeal, illustrates the limitations of p o p u l a r j u s t i c e . So does the inability of even a m o d e r n j u r y to a d m i n i s t e r any b u t the simplest remedies. Injunctions and o t h e r equitable r e m e d i e s had to a w a i t the a p p e a r a n c e of professional j u d g e s .

that happens the system of social control will stand in marked contrast to that of a prelegal culture, where the victim is both the adjudicator and the enforcer. As soon as law enforcement is handed over to officials, however, an "agency problem" arises: What is to guarantee that officials will act in the interest of victims? What is to prevent legal specialists f r o m becoming a professional caste with goals that diverge sharply f r o m social need and popular understanding? H o w , in other words, is law to be kept human but not too human, too personal, too subjective, capricious? H o w does an official even find out what happened in the case presented for decision, when he will not have been a participant in the events giving rise to the dispute? T h e underlying problem is the general one, central to political theory, of securing competent and responsible rulers. The traditional focus of political theory has been on the overtly political branches of government—the legislature and the executive. T h e focus of Anglo-American legal theory has been on the individuals w h o resolve disputes over the application of legal n o r m s and w h o often in the course of doing so m o d ify, refine, or elaborate those norms—the judges. The position of the judge, as a person called on to resolve a dispute in a manner almost certain to harm one party and benefit the other, is inherently precarious. We can see this by comparing his position with that of the doctor. The doctor is not usually helping one person at the expense of another (an exception is triage), and therefore no one has much incentive to interfere with his w o r k or to challenge his competence. But it is natural for a powerful person w h o finds himself in a dispute with a weak one to try to use his power to influence the outcome; and it is equally natural for judges, at least the strong-minded ones, to want to prevent interference that puts them in the sway of the powerful. There have been times when and places where such interference was c o m m o n and made a mockery of law. In a m o d e r n society the threat is more likely to come f r o m an interest group or f r o m public opinion than f r o m an individual, but in any case it is not trivial and continues to affect the way in which judges and their defenders in the legal community think about law. The issue of judicial independence has a private as well as a social aspect. Independence is something judges want the way academics want tenure. Judges do not want to be handmaidens of the powerful. But if independence means only that judges decide cases as they like without pressure f r o m other officials, it is not obvious that an independent judiciary is in the public interest; the people may be exchanging one set of tyrants for another. Where will judges look for guidance once they have

achieved independence f r o m rank political interference? Will they merely act as politicians exempt f r o m the usual political constraints, or will they be hemmed in s o m e h o w by professional norms? Is there a body of objective norms (either "positive law" or "natural law") or a set of analytic methods ("legal reasoning") that can be used to ensure that judicial decisions will be objective, determinate, impersonal? If not, are judges reduced to ruling by fiat made impressive by the hieratic stagecraft of judging—the raised bench, the robes, the oaths, the j a r g o n and rhetoric? The answers may depend on what is meant by "objective." If we stress the first t w o syllables and insist that a judicial decision, to be more than fiat, must correspond to something that is really "out there," we shall be forced to choose between natural law in a strong sense and the legal nihilist's view that law is simply the will of the judges. We may be forced to a similar pass even if, rejecting ontological objectivity (objectivity as correspondence to an external reality), we use "objective" in a weaker sense, which is the scientific sense, of replicable. A finding is objective in this sense if different investigators, not sharing the same ideological or other preconceptions (except—maybe a big "except"— that they would have to share the scientific worldview), would be bound to agree with it. (Hence the association of objectivity with both impersonality and determinacy—and sometimes I use all three terms synonymously.) Although I call this the scientific sense of objectivity, its applications are not limited to science; there are many nonscientific propositions that persons of diverse ideology can agree on. T h e problem for law is that there is little tendency for inquirers w h o hold different ideologies to converge on the answers to difficult legal questions. N o t all legal questions are difficult, of course; and one of the points I shall be emphasizing is that there really are easy legal questions—many of them. But naturally the focus of professional attention is on the difficult ones. And only if w e are content to define "objective" in a third sense (which I shall sometimes call the "conversational" sense), as merely reasonable—that is, as not willful, not personal, not (narrowly) political, not utterly indeterminate though not determinate in the ontological or scientific sense, but as amenable to and accompanied by persuasive though not necessarily convincing explanation—will we be able to locate, with respect to difficult legal questions, a middle ground between the natural lawyer's view and the legal nihilist's view. Although m y focus is on judges, judicial independence is desired almost as much by lawyers. To demonstrate professionalism and justify judicial independence, judges have evolved (with m u c h help f r o m lawyers, of course) technical concepts and methods that in turn support the

lawyers' claim to a m o n o p o l y of representing persons involved in legal disputes. T h e proposition that law is strongly objective and therefore impersonal, and concomitantly that law is autonomous and professional, provides the most easily understood raison d'etre of the lawyer and the law professor as well as of the j u d g e and helps explain the frequency with which practicing lawyers and academic lawyers, as well as judges, make exaggerated claims for the proposition. The division of labor is not the only historical movement that has made the issue of objectivity central to jurisprudence. Another is the growing efficacy of law in society. There have been—to overschematize—three stages in that growth. In the first, symbolized in modern times by the Soviet Constitution of 1936, law is largely aspirational—a set of n o r m s that look good on paper but are not enforced, because the courts are co-opted or overawed by other forces in the society. In the second stage, which in this country spans the approximate period 1800 to 1960, judges achieve genuine independence and their decrees are obeyed, but the profession pays little attention to the actual system for "delivery" of legal services, litigation is costly and cumbersome, and as a result much of the law on the books is as dead a letter as in the first stage. In the third stage, which is the stage the United States is in now, the focus of professional concern shifts to the law in action, and great efforts are made—as by subsidizing the legal expenses of indigent criminal defendants, excluding illegally obtained evidence f r o m criminal trials, relaxing the limitations on standing to sue, facilitating class actions, requiring prisons to have law libraries, and making losing defendants pay the plaintiffs' attorney's fees—to ensure that legal rights can actually, dependably, be enforced. 9 In this stage, which is also associated with a g r o w t h in "claims consciousness," 10 the legal system becomes immensely costly, intrusive, and politically controversial. The stakes in the age-old struggle for judicial independence seem greater

9. See, for e x a m p l e , Innovations in the Legal Services ( E r h a r d B l a n k c n b u r g ed. 1980). T h e g r o w i n g p e n e t r a t i o n of law in A m e r i c a n life is the t h e m e of L a w r e n c e M . Friedman, Total Justice (1985). 10. See the interesting discussion in William L. F. Felstiner, Richard L. Abel, and Austin Sarat, " T h e E m e r g e n c e and T r a n s f o r m a t i o n of D i s p u t e s : N a m i n g , Blaming, C l a i m i n g . . . ," 15 Law and Society Review 631 (1980—81). T h e three stages can coexist. T h a t is, at a given m o m e n t o n e field of law m i g h t be in stage 1, a n o t h e r in stage 2, and a n o t h e r in stage 3. For e x a m p l e , antitrust l a w reached stage 3 long before civil rights law did; e m p l o y m e n t d i s c r i m i n a t i o n law m a y n o t yet have reached stage 3; public international law m a y n o t h a v e reached stage 1. R e c e n t years m a y have seen a slight recession f r o m stage 3 in s o m e areas of A m e r i c a n law.

than ever, and the problems of j u r i s p r u d e n c e are correspondingly m a g nified.

A Short History

of Jurisprudence

T h e question where the j u d g e s get the n o r m s they apply arises early. Consider its presentation in Antigone, w r i t t e n in the fifth century B.C. Creon, the ruler of Thebes, decrees that Polynices, w h o died in a traitorous attack on the city, shall as p u n i s h m e n t be denied the h o n o r a b l e burial that had been given to Eteocles, Polynices' brother, w h o died defending the city against the traitor. Creon's reasoning is civic and instrumental; he wants to safeguard the state and, to that end, to deter treason by making the p u n i s h m e n t for it as horrible as possible. A n t i gone, the sister of Eteocles and Polynices, defends a rival concept, law as conformity to the dictates of n a t u r e , " w h i c h requires her to arrange a proper burial for her brother, traitor t h o u g h he was. These opposed concepts of law receive parallel f o r m u l a t i o n s in the Republic. T h e civic or political concept is expressed b y T h r a s y m a c h u s : "if one reasons correctly, one will c o m e to agree that the j u s t [dikaion] is the same thing e v e r y w h e r e — w h a t e v e r is to the profit or advantage [sumpheron] of the mightier [kreittonos]."n If all T h r a s y m a c h u s had meant was that law bends to the w h i m s of p o w e r f u l individuals, he w o u l d have been describing not a concept of law but the fact of lawlessness. But w h a t he actually meant, I think, was that law is, and b y implication should be, the expression of the values of the d o m i n a n t g r o u p s in society. 13 H e was stating the case for the u n c o m p r o m i s i n g f o r m of legal positivism that teaches that (political) m i g h t makes (legal) right, and that nothing else can d o so. Against this Socrates proposed a conception

11. T h i s is " n a t u r a l l a w " in its literal sense—a c o n c e p t also p r o m i n e n t in Oedipus TyrO n the natural law tradition, and particularly its m o d e r n shift a w a y f r o m nature, see Lloyd L. Weinreb, Natural Law and Justice (1987); Weinreb, " T h e N a t u r a l Law T r a d i t i o n : C o m m e n t s on Finnis," 36 Journal of Legal Education 501 (1986); cf. R o b e r t P. George, "Recent Criticism of N a t u r a l L a w T h e o r y , " 55 University of Chicago Law Review 1371 (1988); Russell H i t t i n g e r , A Critique of the New Natural Law Theory (1987). A still serviceable i n t r o d u c t i o n to natural law t h i n k i n g is A. P. d ' E n t r e v e s , Natural Law: An Introduction to Legal Philosophy (1951). AIIIIUS.

12. Platonis Opera, vol. 4, p. 339a (Ioannes B u r n e t ed. 1902) (Politeia, Bk. I, 11. 2 - 4 ) . T h e translation is m y o w n . 13. T h e r e is m u c h d i s a g r e e m e n t , t h o u g h , o v e r the precise m e a n i n g of T h r a s y m a c h u s ' s concept of justice. See K i m o n Lycos, Plato on Justice and Power: Reading Book I of Plato's "Republic" (1987), and references cited there.

o f justice as s o m e t h i n g that m a n discovers rather than creates—that exists literally apart f r o m h i m , in the Heaven of Forms. M u c h of j u r i s p r u d e n c e simply rings changes on the disagreement between C r e o n and Antigone, T h r a s y m a c h u s and Socrates; 14 it will be apparent o n w h i c h side w e can expect to find m o s t lawyers. In the seventeenth century w e find J a m e s I asking E d w a r d C o k e , the Chief Justice of England, w h y J a m e s is n o t c o m p e t e n t to deliver legal j u d g m e n t s : law is f o u n d e d o n reason, and his reasoning p o w e r s are as g o o d as the j u d g e s ' . C o k e could n o t challenge either premise, b u t replied that the reason o n w h i c h law was f o u n d e d was a species o f "artificial reason" that only a person trained and experienced in law could exercise. 15 C o k e ' s reply encapsulated three elements of the o r t h o d o x lawyer's view of law: that law is reason, n o t fiat; that it is a special sort of reason and n o t j u s t c o m m o n sense, applied moral philosophy, or policy analysis; and, a related point, that only lawyers k n o w it. L a w so conceived resembles Socrates' F o r m o f Justice in that it exists apart f r o m the political organs of g o v e r n m e n t , as the preserve of a professional caste, and m o r e f u n d a m e n t a l l y because of its " f o r m a l i s m , " its independence f r o m the w o r l d o f fact. But C o k e ' s is n o t a natural law concept—at least n o t explicitly. To C o k e the law is n o t a b o d y of ethical or religious principles that subtends, inspires, and is available to criticize specific j u d g m e n t s . If it were, King James's competence to render j u d g m e n t s could not be q u e s t i o n e d — n o t b y C o k e , anyway. Law is the b o d y o f doctrines developed b y j u d g e s and expressed or implied in their judicial opinions. T h i s concept of law is not only distinct f r o m but in tension with that of Socrates, w h o m I have placed first in Coke's ancestral line. Socrates' belief in natural law carried with it a disdain f o r sophistry—that is, for legal reasoning, criticized by Socrates in Plato's Gorgias and exemplified b y c o m m o n l a w j u d g e s such as C o k e . Lawyers' tricks d o n o t sit c o m fortably w i t h natural law in the strong f o r m urged b y Socrates. Further m a r r i n g the s y m m e t r y of m y classification, J a m e s I, w h o m I have g r o u p e d w i t h A n t i g o n e and Socrates against C r e o n and Thrasymachus, could be t h o u g h t a sort o f C r e o n , with a natural l a w twist (the divine right o f kings). We begin to see an association o f legal positivism with 14. Throughout, "Socrates" refers to the character in Plato's dialogues, not the historical Socrates, w h o s e v i e w o f law and justice may well have differed from that attributed to him by Plato in the late dialogues, including the Republic and (possibly) Gorgias. 15. Prohibitions del Roy, 12 C o . Rep. 63, 65, 77 Eng. Rep. 1342, 1343 (1608). See GeraldJ. Postema, Bentham and the Common Law Tradition, ch. 1 (1986). C o k e was speaking o f the c o m m o n l a w — l a w made by j u d g e s as a by-product o f deciding cases—rather than o f statute law.

legal formalism, as in Coke, and natural law with legal realism (conceived as antiformalism), as in James I and perhaps in Socrates; these associations blur the more familiar association of positivism with realism (Thrasymachus, Creon) and natural law with formalism (Blackstone and also to a degree Coke, as we are about to see). This terminology will become clearer as we proceed, although one of the things I hope to show in this book is that w e can live quite happily without such terminology. The following matrix presents the basic permutations:

Legal formalism Legal realism

Natural law

Positive law

Blackstone James I

Coke Thrasymachus

Coke provided a foil for Hobbes, w h o argued that law is the c o m mand of the sovereign. 16 Although this concept of law resembles Thrasymachus's (not to mention James I's), the dispute between Hobbes and Coke might appear to be merely over the locus of sovereignty in resolving legal disputes: is it in the king or in the judges? T h e appearance is misleading. Coke would not be likely to question the sovereignty of James I so frontally. His point was merely that the judges, by virtue of their training and experience, have privileged access to a body of authoritative norms. Yet it is a subversive point after all, for it implies that the judges are expositors of natural law, in the sense of law that is authoritative yet does not emanate f r o m "the sovereign." Although this does not quite make them competing sovereigns with the king, it does c o m promise the king's sovereignty. That is what troubled Hobbes and set him against Coke. In the next century, Blackstone made the natural law implications of Coke's legalism explicit in the course of painting an admiring picture of the artificial reason of the English c o m m o n law in all its doctrinal and institutional intricacies. 17 Although conscious both of the utilitarian function of the c o m m o n law and of its most conspicuous imperfections, 16. See Leviathan, pt. II, ch. 26 and a Student of the Common Laws Cairns, Legal Philosophy from Plato Hale played C o k e to H o b b e s (that still been alive), see J. H . H e x t e r , 471 (1980).

(1651); see also H o b b e s , A Dialogue between a Philosopher of England (Joseph C r o p s e y ed. 1971), and H u n t i n g t o n to Hegel, ch. 7 (1949). For a discussion of h o w M a t t h e w is, a n s w e r e d H o b b e s as C o k e w o u l d h a v e d o n e if he had " T h o m a s H o b b e s and the L a w , " 65 Cornell Law Review

17. William B l a c k s t o n e , Commentaries

on the Laws of England (4 vols., 1765—1769).

including the excessive severity of many criminal punishments and the arbitrary limitations on trial procedure (for example, the parties to a suit could not testify as witnesses, or criminal defendants be represented by counsel), Blackstone tried to give the c o m m o n law a transcendental aura. H e argued that the c o m m o n law was rooted in Saxon customary law, which was itself natural law—specifically, the law of God as it had been perceived by h u m a n reason in clearer-sighted times. Blackstone gave the c o m m o n law, in other words, a natural law pedigree. As a concomitant, he downplayed the creative role of judges, calling them the "oracles" of the law. By depicting them as unmediating transmitters of divinely originating commands, this metaphor placed the judges at a far remove f r o m politicians and statesmen. But judges are not unmediated transmitters. O n l y lawyers can understand them; lawyers translate the oracular discourse for the laity. Blackstone provided a natural target for Jeremy Bentham. Acutely conscious of the many dysfunctional features of eighteenth-century law, Bentham wanted to refashion the law—and quickly, too—into an instrument for maximizing the greatest happiness of the greatest n u m ber. Coke's professionalism (the "artificial reason") and Blackstone's incrementalism, traditionalism, and transcendentalism were obstacles to this program. In combination, these "isms" implied that the c o m m o n law should not or even could not (without impiety) be changed to conf o r m to the greatest-happiness principle—not soon, anyway—or, at the very least, that any change must be agreeable to the legal profession. The profession might not be sympathetic to the greatest-happiness principle, and in any event had a vested interest in the status quo—a fusty, barnacled, superstitious, reactionary status quo, it seemed to Bentham. I f i Bentham not only made penetrating criticisms of specific legal rules and institutions (while willfully failing to appreciate the strengths of the c o m m o n law system) but deployed a version of legal positivism most damaging to Blackstone's transcendental pretensions. He showed that much of Blackstone's conceptual apparatus consisted of transparent and u n w o r t h y fictions. 19 There was nothing "natural," universal, or tran-

18. O n the conflict b e t w e e n B e n t h a m and B l a c k s t o n e , see The Economics of Justice, note 7 above, ch. 1; P o s t e m a , n o t e 15 a b o v e , p t . 2. T h e character of B e n t h a m ' s j u r i s p r u d e n c e is m o s t fully displayed in his p o s t h u m o u s treatise Of Laws in General (H. L. A. H a r t ed. 1970). 19. Such as the oneness of m a n and w o m a n in m a r r i a g e : as explained b y one of Bent h a m ' s successors in the activity of d e b u n k i n g legal fictions, in an essay originally published in 1870, " A t law h u s b a n d and w i f e lie u n d e r certain incapacities in respect to one another. T h e y are n o t subject to these incapacities because they are one p e r s o n , but as if they were one person. T h e s e incapacities aside, the l a w regards t h e m as t w a i n , because they are

sccndental about law. It was the creation of distinctly earthbound political authorities—legislators and, at the time, especially judges. Erroneously believing that he had found in rule by representative legislatures the answer to the practical question of h o w to get government to dedicate itself to the greatest-happiness principle, Bentham wanted to shift the center of legal rule making f r o m the courts to a suitably reformed Parliament, which he hoped would draft detailed codes covering every area of law and would thereby minimize judicial discretion. The judges would be minor bureaucrats rather than politicians parading in phony oracular colors—minor because the codes would be drafted in ordinary language and so require no professional training to decipher and because they would be so comprehensive that judges would have no scope for legislating in the guise of filling gaps left by legislators. Bentham's is the most comprehensive scheme for making law a true and complete expression of popular sovereignty. This may seem to make him a Creon, Thrasymachus, James I, or Hobbes, albeit with a democratic twist—and the twist is an accident of Bentham's mistaken belief that a popular sovereign would adopt the greatest-happiness principle; if an enlightened despot could have been trusted to do the same, Bentham would have favored, as indeed was his original impulse, vesting all sovereign powers in the despot. Yet with the contingent, almost fortuitous, charactcr of Bentham's attachment to democratic politics acknowledged, it still is the case that he differs importantly f r o m Creon, Thrasymachus, and (less clearly) James I. For at the root of his jurisprudence is not a merely civic or political concept, such as "what's good for Thebes," but the greatest-happiness principle—as sovereign and imperative a n o r m to a utilitarian as nature is to a natural lawyer. Even if Parliament had taken up Bentham's invitation to codify the law, the essential elements of the orthodox view would have survived. The judges would have exchanged a divine for a secular overseer but would still have had to possess substantial professional skills, as C o n t i nental judges do notwithstanding the extensive codification of law on the Continent. Bentham did not understand this because he thought language purely referential and transparent to reality. 20 This view led

t w a i n . " Nicholas St. J o h n Green, " P r o x i m a t e and R e m o t e C a u s e , " in Green, Essays anil Notes on the Law of Tori and Crime 31, 32 (1933) (emphasis in original—as t h r o u g h o u t this book unless o t h e r w i s e indicated). 20. See J a m e s Steintragcr, Benlliam 2 2 - 2 8 (1977). T h e idea is c a p t u r e d in one of H o l m e s ' s aphorisms: "We m u s t think t h i n g s n o t w o r d s . " " L a w in Science and Science in L a w , " 12 Harvard Law Review 443, 460 (1899). In o t h e r w o r d s , all w o r d s are n a m e s .

him to suppose that a legal code could be drafted that would establish an undistorted path of communication between the legislators on the one hand and the judges and citizens on the other and would make the interpretation and application of law mechanical—a task for low-level civil servants. Bentham's idea of the law as utilitarian has a significance that transcends the debates between natural law and positivism and between formalism and realism. Rejecting on the one hand every moralism (except utilitarianism itself) as a foundation for law, and on the other hand every merely emotive, traditional, rhetorical, or (in short) unreflective and unsystematic approach to law, Bentham sought to place law on a scientific footing—to make it a practical h u m a n instrument for the achievement of definite social ends. 21 In this limited but important respect Bentham is the originator of the pragmatic concept of law advocated in this book. Despite Bentham's promptings, neither the English Parliament nor American legislatures showed much taste for codification. Throughout the nineteenth century and well into this one, therefore, most AngloAmerican law was still c o m m o n law. Bentham's onslaught, however, together with the steady diminution in religious belief a m o n g the educated classes, had left the intellectual foundations of the orthodox view in disarray. If the c o m m o n law did not come f r o m God, where did it come from? It was one thing to speak of "natural law" when nature was conceived to be the expression of divine love or order, and quite another to find universal legal n o r m s in Darwinian nature, red in tooth and claw. T h e natural law project has never recovered f r o m what Nietzsche called the death of God (at the hands of Darwin). If not f r o m God or nature, where could the c o m m o n law have come but f r o m the judges themselves? That would make them legislators—unelected ones, to boot. Many of our states confronted this possibility head on, by making their judiciaries elective. But this experiment merely undermined judicial independence and encouraged the perception (at times self-perception) of judges as nothing more than legislators in robes (if that: until late in the nineteenth century, most American judges did not wear robes). Despite its persistence, the concept of an elective judiciary is generally and correctly regarded as a failure. Into the breach stepped the American legal formalists, of w h o m the best k n o w n today is Dean Christopher C o l u m b u s Langdell of the H a r vard Law School. O n one level the formalists were Platonists, believing 21. See N a n c y L. R o s e n b l u m , Bentham's

Theory of the Modern State, ch. 3 (1978).

that there existed a handful of permanent, unchanging, indispensable principles of law imperfectly embodied in the many thousands of p u b lished judicial opinions, and that the goal of legal reasoning was to penetrate the opinions to the principles. That is the flavor of Langdell's o w n scholarship. 22 O n another—more interesting but not sharply distinct— level the formalists had reconceived law as an inductive science. (In this they had been anticipated, however, by Francis Bacon.) 23 T h e reports of appellate decisions were the data f r o m which the principles of the c o m mon law could be inferred—principles such as that a promise is not legally enforceable unless supported by consideration, or that liability in tort requires proof of blameworthy conduct. O n c e these principles were brought to light the correct outcome of a case could be deduced. T h u s the principles could be used both to show that outliers in the sample of appellate cases f r o m which the principles had been derived had been decided incorrectly and to guide the decision of n e w cases. Although man-made, the principles could minimize judicial discretion, just as Blackstone's divinely originated c o m m o n law had been thought to. And although these principles did not have divine backing they had something almost as good—the p o w e r of scientific induction and the verdict of time. The orthodox view had been reinstated in secular, scientific dress. But the challenge remained of explaining just h o w the creation of legal principles as by-products of judicial decision making could confer on them a political legitimacy equivalent to that of legislative rules. T h e use of induction to derive principles f r o m cases left a gap between "is" and "ought," and the verdict of time could easily be questioned in a period when social conditions were changing rapidly. As Bentham had taken on Blackstone, so it fell to Holmes to take on the heirs of Blackstone. Three features of formalism repelled Holmes. The first was its conceptualism and scientism. Formalism considered law a set of principles and legal reasoning a branch of exact inquiry, a

22. See, for e x a m p l e , the preface to C . C . Langdell, A Selection of Cases on the Law of Contracts (1871). 23. See Paul H . Kocher, "Francis B a c o n o n t h e Science o f J u r i s p r u d e n c e , " 18 Journal of the History of Ideas 3, 7 (1957). T h e f o r m a l i s t s a c k n o w l e d g e d B a c o n ' s priority. See P e r r y Miller, The Life of the Mind in America: From the Revolution to the Civil War 159 (1965). As Miller's b o o k m a k e s clear, Langdell's a p p r o a c h w a s n o t original even in A m e r i c a . F o r m a l ism had b e c o m e the A m e r i c a n legal p r o f e s s i o n ' s o r t h o d o x y b e f o r e t h e C i v i l War. See id., ch. 3 ( " T h e Science of the L a w " ) , especially p p . 159-163; also M . H . H o e f l i c h , " L a w and G e o m e t r y : Legal Science f r o m Leibniz to L a n g d e l l , " 30 American Journal of Legal History 95, 112-121 (1986).

source of apodictic certainty. Second and closely related was the static quality of formalism, its penchant f o r treating the cases f r o m which the principles of law are inferred as a set of data having n o chronological dimension and the principles themselves as timeless, like the propositions of Euclidean geometry. 2 4 T h e third objectionable feature of formalism, c o m b i n i n g the first t w o , was the separation of law f r o m life. A formal system like g e o m e t r y is a system of relations a m o n g ideas, not between ideas and physical reality; the formalists' induction was f r o m judicial opinions, n o t f r o m the problems and practices of social life. (And most of the opinions in w h i c h the American formalists f o u n d clues to the principles of law were not even American, b u t English.) Holmes's o u t l o o k was severely t h o u g h imperfectly pragmatic. 2 5 There are n o conceptual entities; the m e a n i n g of an idea lies n o t in its definition, its Form, its relation to other ideas, b u t rather in its consequences in the w o r l d of fact. B y the same token there are n o legal principles in an interesting sense. T h e r e is merely the fact that if you do thus and so, eventually the sheriff will s h o w u p at y o u r d o o r and take away s o m e of y o u r property to satisfy a legal j u d g m e n t against you, and if you resist h i m the police or, if necessary, the a r m y will be s u m m o n e d to crush y o u r resistance. As it is unlikely that any of this r o u g h stuff w o u l d ensue f r o m violating a precept discovered b y a law professor in his library, it m u s t be that the force of the state is deployed in the n a m e of law for concrete political reasons. T h e rules of law are the systematized and coercive e m b o d i m e n t of the salient opinions of the m o s t p o w e r f u l groups in society and therefore change as those opinions and groups change. But H o l m e s , unlike T h r a s y m a c h u s , did not believe that legal doctrines are merely decrees, reflecting the w h i m s of a tyrant or the caprice o f public opinion. H o l m e s was writing at a time w h e n there was a professional judiciary. Legal doctrines have an intellectual structure and to s o m e extent a life of their o w n (or at least j u d g e s have lives of their o w n ) , independent of the political forces that give the doctrines their w e i g h t and direction. To paraphrase Hamlet, the political forces rough h e w the law b u t the j u d g e s shape it. A m a n of the century of D a r w i n , Spencer, Hegel, M a r x , and Nietzsche, H o l m e s t h o u g h t that the best

24. See T h o m a s C . Grey, "Langdell's O r t h o d o x y , " 45 University of Pittsburgh Law Review 1 (1983). 25. As w e shall see in C h a p t e r 7, H o l m e s ' s o u t l o o k resembles that o f his c o n t e m p o r a r y and a c q u a i n t a n c e C h a r l e s Sanders Peirce, the f o u n d e r of p r a g m a t i s m , but is b y no means " o r t h o d o x l y p r a g m a t i s t , " a s s u m i n g there could be such a t h i n g . Also, it is unclear to w h a t extent H o l m e s w a s i n f l u e n c e d by Peirce and to w h a t e x t e n t b y their c o m m o n Zeitgeist. See T h o m a s C . Grey, " A p p e n d i x : H o l m e s and the P r a g m a t i s t s , " in Grey, " H o l m e s and Legal P r a g m a t i s m , " 41 Stanford Law Review 787, 864 (1989).

j u d g e was the one w h o shaped the law to c o n f o r m m o s t closely to, even to anticipate, the desires of the d o m i n a n t g r o u p s in the c o m m u n i t y , which is to say, w h o j u m p e d o n history's b a n d w a g o n . Holmes's The Common Law (1881) is an extended paean to j u d g e s ' skill in adapting c o m m o n law doctrines to changes in durable public opinion, which reflects social need and power. T h e b o o k points out, for example, that w h e n law was merely a substitute for revenge, legal doctrines were designed to give victims of the kinds of w r o n g that engender a thirst for revenge a r e m e d y a p p r o x i m a t i n g w h a t w o u l d have been c o n sidered necessary to slake that thirst before there were legal remedies. When revenge fell o u t of favor and other goals, such as deterrence and compensation, came to the fore, the old doctrines were ingeniously adapted to the n e w goals. Consider the law of deodands. A deodand is an inanimate object that causes a death. In ancient law, deodands w e r e treated as criminals, and destroyed. This behavior was consistent w i t h law's roots in revenge. If you are hurt, you will w a n t to lash out at w h o e v e r — o r , if y o u cannot locate the precise line between animate and inanimate objects, w h a t e v e r — h u r t you. T h u s a tree that fell o n a person and killed h i m was a deodand and had to be destroyed. All this eventually became dysfunctional but, as H o l m e s explained, the old law of deodands turned out to have a n e w use. T h e victim of a m a r i t i m e accident m a y encounter great difficulty in trying to obtain legal redress directly against the o w n e r of the ship that did the h a r m , because the o w n e r m a y live far away, quite possibly in a foreign jurisdiction. B y analogy to the law of deodands, w h y n o t treat the ship itself as the wrongdoer, and hence let the victim sue the ship in the first port in which it berths after the accident? U n t i l the suit is resolved, the ship will not be allowed to leave the port unless the ship's o w n e r or master posts a bond. In this w a y the victim can obtain and collect a j u d g m e n t w i t h o u t traveling to the o w n e r ' s domicile. A d m i r a l t y law adopted this approach and still follows it. H o l m e s did not believe that the evolution of the law w a s complete. Indeed, a great believer in evolution in all spheres, he did n o t consider h u m a n biological evolution complete, and looked w i t h apparent f a v o r on eugenic breeding of h u m a n beings. 2 6 H e t h o u g h t it likely that political p o w e r w o u l d shift to the w o r k i n g class and that crime m i g h t c o m e

26. See, for e x a m p l e . B u c k v. Bell, 274 U . S . 200 (1927); Letter to H a r o l d Laski o f M a y 12, 1927, in Holmes-Laski Letters, vol. 2, p. 942 ( M a r k D e W o l f e H o w e ed. 1953) ("I w r o t e and delivered a decision [Buck v. Bell] u p h o l d i n g the c o n s t i t u t i o n a l i t y of a state l a w for sterilizing imbeciles the o t h e r d a y — a n d felt that I w a s g e t t i n g near t o t h e first principle of real r e f o r m " ) .

to be considered a disease rather than b l a m e w o r t h y conduct. H e t h o u g h t it inevitable and on the w h o l e desirable that law w o u l d change to a c c o m m o d a t e these shifts in p o w e r and opinion—that it w o u l d , for example, b e c o m e m o r e favorable to labor unions—and he did not think the shift should be i m p e d e d by adherence to principles derived f r o m judicial opinions. 2 7 His tolerant attitude t o w a r d federal and state legislation challenged u n d e r the C o n s t i t u t i o n reflected his view that law should not be an obstacle to social change. T h e exception that he m a d e for g o v e r n m e n t a l action that restricts f r e e d o m of speech was consistent, because it was the o p p o n e n t s of social change w h o w a n t e d to stifle free speech. Holmes's jurisprudential views are scattered a m o n g thousands o f j u d i cial opinions, letters, academic writings, occasional pieces, and even anecdotes. It w o u l d be remarkable if views so presented were consistent, and they are not. 2 8 B u t they are m o r e consistent than they m a y seem to be, as a brace of examples will help show. T h e first is Holmes's belief in an "objective" t h e o r y of contractual liability (indeed, of liability generally, including criminal liability): a party to a contract m a y be b o u n d to a contractual u n d e r t a k i n g that neither party intended but that seems implied b y the w o r d s of the contract. T h e second example is Holmes's rejection of the doctrine of unconstitutional conditions: he t h o u g h t that since a person has n o right to be a police officer (or any other type of public employee), an officer cannot complain if the police department has a rule d e n y i n g its employees free speech. T h e greater p o w e r — t h e p o w e r n o t to hire—includes the lesser p o w e r to condition e m p l o y m e n t on the employee's agreeing to surrender his constitutional rights. T h e objective t h e o r y of contract followed directly for H o l m e s f r o m his rejection of occult, immaterial entities; another person's m i n d is closed to us (and w h a t is the m i n d m a d e of, anyway?), so intentions should not figure in the determination of legal duties. T h e rejection of the doctrine of unconstitutional conditions followed f r o m Holmes's belief that law registers the balance of political power, a belief that led H o l m e s always to look for the source of p o w e r and to m a k e law f o l l o w it. Power—closely related to physical force as it is (for Holmes, p o w e r meant the ability to

27. See, for e x a m p l e , G i t l o w v. N e w York, 268 U . S . 652, 673 (1925) (dissenting o p i n ion): "If, in the long r u n , the beliefs expressed in proletarian d i c t a t o r s h i p are destined to be accepted b y the d o m i n a n t forces of the c o m m u n i t y , the o n l y m e a n i n g of free speech is that t h e y s h o u l d be g i v e n their chance and have their w a y . " 28. As p o w e r f u l l y a r g u e d , a l t h o u g h w i t h s o m e e x a g g e r a t i o n ( o w i n g I t h i n k to the a u t h o r ' s lack of s y m p a t h y f o r his subject), in K o b e r t W. G o r d o n , " H o l m e s ' Common Law as Legal and Social Science," 10 Hofslra Law Review 719 (1982).

mobilize force in support of one's desires)—is an attractive concept to someone w h o wants a material, n o n m o r a l g r o u n d for legal obligations. 2 '' Holmes's idea of law as a superstructure is m o r e i m p o r t a n t than the specific base that he f o u n d w h e n he b u r r o w e d beneath it. Whether the base is p o w e r or, w h a t is closely related in a democratic society, d o m i nant public o p i n i o n — o r even s o m e social or ethical objective, such as maximizing social wealth or utility (and H o l m e s flirted w i t h that possibility, too)—Holmes's approach invites the student of law to seek an explanation for its rules and outcomes, its doctrines and institutions, outside the law itself, at least as " l a w " is conventionally defined. This search continues to be an i m p o r t a n t part of legal theory. I have dwelt on H o l m e s at such length because his jurisprudential ideas are so fascinating and fruitful; he remains the leading figure in American jurisprudence. But of course the history of j u r i s p r u d e n c e did not end with H o l m e s . B e n j a m i n C a r d o z o in the next generation consolidated m a n y of Holmes's insights in an i m p o r t a n t b o o k , The Nature of the Judicial Process (1921)—a clear exposition, as w e shall see, of a mature pragmatic jurisprudence. H o l m e s and C a r d o z o , along w i t h J o h n C h i p m a n Gray and Roscoe P o u n d — b u t particularly H o l m e s — h a d laid the foundations for the "legal realist" m o v e m e n t , w h i c h flourished in the 1920s and 1930s. 30 I shall pass over the realists. Given the similarity 29. "I believe that force, m i t i g a t e d so far as m a y be b y g o o d m a n n e r s , is the ultima ratio, and b e t w e e n t w o g r o u p s that w a n t to m a k e inconsistent k i n d s of w o r l d I see n o r e m e d y except force." Letter to Sir Frederick Pollock of Feb. 1, 1920, in Holmes-Pollock Letters 36 (Mark D e W o l f e H o w e ed. 1961). 30. O n legal realism, sec R o b e r t S a m u e l S u m m e r s , Instrumentalism and American Legal Theory (1982); Alan H u n t , The Sociological Movement in Law, ch. 3 (1978); A n d r e w A l t m a n , "Legal Realism, Critical Legal Studies, and D w o r k i n , " 15 Philosophy and Public Affairs 2 0 5 214 (1986); Charles M . Yablon, " L a w and M e t a p h y s i c s , " 96 Yale Law Journal 613, 6 1 5 - 6 2 4 (1987); E d w a r d A. Purcell, J r . , The Crisis of Democratic Theory: Scientific Naturalism and the Problem of Value, ch. 9 (1973). For C a r d o z o ' s gentle b u t p e n e t r a t i n g criticism of the excesses of legal realism, see " J u r i s p r u d e n c e , " in Selected Writings of Benjamin Nathan Cardozo: The Choice ofTycho Brahe 7 ( M a r g a r e t E. Hall ed. 1947). T h e t e r m "realist" is u n f o r t u n a t e . T h e legal realists were not scientific realists ( w h o are "idealists" in the lay sense, because they believe in the existence of a w o r l d b e h i n d sense data—an invisible w o r l d ) . As e m p h a s i z e d by S u m m e r s , they w e r e p r a g m a t i s t s , in all the varied hues of that t e r m . Legal realism is largely an A m e r i c a n p h e n o m e n o n . T h e E n g l i s h legal tradition is h i g h l y formalist, the C o n t i n e n t a l heavily i m b u e d w i t h natural law t h i n k i n g well s u m m a r i z e d in Peter G o o d r i c h , Reading the Law: A Critical Introduction to Legal Method and Techniques 6 - 8 (1986). T h e reason for the C o n t i n e n t a l affection for natural l a w m a y be the greater i m p o r tance of legislatures vis-a-vis c o u r t s in the C o n t i n e n t a l tradition than in the A n g l o - A m e r ican. In the C o n t i n e n t a l system the j u d g e s ' docility can usually be a s s u m e d (for an i m p o r tant e x c e p t i o n — G e r m a n j u d g e s in the W e i m a r republic—see C h a p t e r 7), and the focus

of their views to Holmes's, and the emphasis I place in this book on pragmatic jurisprudence, this m a y seem a surprising omission. But it is m o r e p o s t p o n e m e n t than omission; I refer to realism, and m o r e frequently to its current epigone, critical legal studies, in subsequent chapters. If I seem nevertheless to slight legal realism, this may be because I have difficulty u n d e r s t a n d i n g w h a t is original in it. At least about the large questions of j u r i s p r u d e n c e that are the focus of this b o o k , the legal realists had little to say that H o l m e s and C a r d o z o had n o t said earlier. Holmes's relationship to m o d e r n legal positivism 3 1 deserves c o m m e n t . In the influential positivist view expounded by H . L. A. H a r t in his b o o k The Concept of Law (1961), the law is the set of rules laid d o w n by legislators, j u d g e s , and other authorized lawgivers. W h e n the rules run out or fail to fit (as they often do), the j u d g e s have discretion to modify, trim, or extend t h e m as m a y be necessary to m a k e t h e m cover the case at hand. Alternatively—but it comes to the same thing—one of the rules that c o m p o s e the law is a jurisdictional rule authorizing j u d g e s to exercise discretion w h e n e v e r there is a gap or ambiguity in the substantive rules. In this v i e w the j u d g e is, as H o l m e s put it, an interstitial legislator. H o l m e s argued that law is a prediction of what the j u d g e s will d o

shifts t o the principles that o u g h t to g u i d e an e n l i g h t e n e d legislature. T h e differences in actual b e h a v i o r b e t w e e n A n g l o - A m e r i c a n j u d g e s o n the one h a n d and C o n t i n e n t a l j u d g e s o n the o t h e r s h o u l d n o t be e x a g g e r a t e d , h o w e v e r . See K o n r a d Z w e i g a r t and H e i n Kotz, Introduction to Comparative Law, vol. 1, pp. 2 7 8 - 2 8 0 (2d rev. ed., T o n y Weir trans., 1987). 31. A n o t h e r u n f o r t u n a t e choice of w o r d s , since positivism in law is n o t at all the same t h i n g as p o s i t i v i s m in p h i l o s o p h y . P o s i t i v i s m in law can m e a n either that legal obligation does n o t depend on m o r a l o b l i g a t i o n , and t h e r e f o r e the expression " i m m o r a l l a w " is not an o x y m o r o n , or (the s t r o n g e r v e r s i o n of legal positivism) that law can e m a n a t e only f r o m j u d g e s and o t h e r d u l y a u t h o r i z e d g o v e r n m e n t officials, that ethical o b l i g a t i o n alone cannot create a legal o b l i g a t i o n . See Jules L. C o l e m a n , " N e g a t i v e and P o s i t i v e P o s i t i v i s m , " in C o l e m a n , Markets, Morals, and the Law 3 (1988); see also D a v i d Lyons, " M o r a l Aspects of Legal T h e o r y , " 7 Midwest Studies in Philosophy 223 (1982). E i t h e r w a y the i m p o r t a n t thing is that positive law is n o t necessarily g r o u n d e d in or vetted b y natural law. P o s i t i v i s m in p h i l o s o p h y is the idea that o n l y perceptible t h i n g s are real: succinctly, "seeing is believing." Ian H a c k i n g , Representing and Intervening: Introductory Topics in the Philosophy of Natural Science 63 (1983). T h i s m e t a p h y s i c a l n o t i o n shares w i t h legal positivism o n l y a distrust of v a g u e entities—see D e r y c k Beyleveld and R o g e r B r o w n s w o r d , Law as a Moral Judgment 4—7 ( 1 9 8 6 ) — t h o u g h that is s o m e t h i n g ; and as G o r d o n , n o t e 28 above, at 7 2 2 - 7 2 6 , argues, H o l m e s w a s , at t i m e s a n y w a y , a scientific positivist (or, as it is o f t e n called, a logical positivist or logical empiricist) as well as a legal one. For p o w e r f u l criticism o f legal p o s i t i v i s m , sce Lon L. Fuller, The Law in Quest oj Itself (1940). As a final w r i n k l e , notice that positive analysis is f r e q u e n t l y c o n t r a s t e d w i t h n o r m a t i v e analysis, the f o r m e r b e i n g descriptive (but n o t necessarily positivistic!) and the latter prescriptive; I shall e x a m ine this usage in C h a p t e r 12.

when confronted with a specific set of facts. I discuss this view at greater length in Chapter 7 and for n o w merely note its consistency with the more conventional positivist view of the j u d g e as rule applier and, on occasion, rule modifier or creator. O n either view, law is the activity of licensed persons, the judges, rather than a body of concepts (rules, principles, whatever). Judges employ discretion to change rules, and discretion is not "principled," although it may be bounded by principles. Indeed, to speak of "employing discretion" may be too grand. Judges change rules, period. In the end the law is what the judges do with your case. The idea of judicial discretion—a blank space or black box, not the solution to the problem of deciding cases when the rules run out but merely the name of the problem—is, no matter h o w fancied up, a source of unease to the legal profession. If, much of the time, and certainly in the most interesting cases, judges are legislators, w h y are they not subject to the same political—today, democratic—controls as legislators? Two of the best-known modern answers are those of Hart and Sacks 32 and of Dworkin. 3 3 Hart and Sacks argued that the j u d g e is a legislator with a difference—his training, the procedures he follows, and the traditions of legal reasoning entitle us to repose greater confidence in his probity and competence. They also believed, however, that ordinary legislators—the members of Congress and of state legislatures—were scrupulous and reasonable, like judges, which if true made a possible convergence of the judicial and legislative powers less worrisome in any event. Hart and Sacks were writing in a period of political and legal consensus, including general satisfaction with the burst of legislative activity during the N e w Deal. In such a period their message was convincing; it is no longer. Dworkin's answer to the question of whether the j u d g e is an undemocratic legislator, offered as it was after the consensus had collapsed, was bound to seem less convincing f r o m the start. T h e considerations that judges are entitled to use in deciding whether to create or modify a rule D w o r k i n names "principles" and makes a part of law itself^—in fact the major part. Rather than use notions of public policy to change or create law, judges, in Dworkin's view, interpret principles to discover the law applicable to the case at hand. The principles take

32. See H e n r y M . H a r t , J r . , and Albert Sacks, The Legal Process: Basic Problems in the Making and Application of Law (tentative ed. 1958). 33. See Ronald D w o r k i n , Taking Rights Seriously (1977); A Matter of Principle (1985); Law's Empire (1986).

the place of the F o r m of Justice or the law of G o d . Yet as D w o r k i n acknowledges, they d o n o t c o m p o s e a consistent whole. T h e y include such frequently w a r r i n g precepts as that n o person shall be allowed to profit f r o m his o w n w r o n g d o i n g and that promises shall be obeyed, thus placing the j u d g e in a quandary as to h o w to approach a suit for breach of an illegal contract. T h e y also include vague n o r m s , such as treat like cases alike, to which both sides to a legal dispute m a y be able to appeal w i t h equal plausibility. Nevertheless D w o r k i n claims that every legal question has a right and a w r o n g answer; that is, an answer given by law and an answer contrary to law. This conclusion is an organic c o m p o n e n t of his theory rather than a baroque addition to it. If there could be t w o right answers to a legal question the j u d g e w o u l d have r o o m to exercise discretion; maybe, indeed, he w o u l d have to exercise discretion, and then the law w o u l d be determined by extralegal considerations. We w o u l d have c o m e full circle to positivism: the law w o u l d be the will of the j u d g e s or other officials. Yet defining law as broadly as D w o r k i n does provides n o escape. W h e n law is defined to include, u n d e r the rubric of "principle," the ethical and political n o r m s that j u d g e s use to decide the most difficult cases, decision according to law and decision according to political preference b e c o m e difficult, sometimes impossible, to distinguish in a society as morally heterogeneous as ours. D w o r k i n does distinguish between policy and principle, and he argues that only the latter is a fit foundation for and limitation on judicially declared rights. T h e a r g u m e n t is u n c o n vincing, the distinction arbitrary, the principles merely the policies that c o m m e n d themselves to D w o r k i n ' s brand of political liberalism. And such is the malleability of "principles" and the associated term "rights" that j u d g e s widely regarded as lawless become in D w o r k i n ' s view paragons of lawfulness if the observer shares the j u d g e s ' political preferences, while conventionally lawful j u d g e s b e c o m e exemplars of lawlessness because they disregard principles that, h o w e v e r political they m a y look, are actually part of law. T h e timid j u d g e , the j u d g e w h o hesitates to innovate, the j u d g e w h o thinks it the business of legislatures rather than of j u d g e s to legislate—to D w o r k i n he is the lawless j u d g e . T h e irony of D w o r k i n ' s project is that the m o r e broadly law is defined, the less rather than the m o r e secure the "rule of l a w " becomes. Law loses distinctness—merging first with morals, and then, w h e n it is recognized that society is morally diverse, w i t h politics and hence n o law. If law includes a broad swatch of political principles, then j u d g e s can do politics and say w i t h a g o o d conscience that they are doing law.

"Right" and " w r o n g " become epithets bestowed on the legal analyst's political friends and enemies. T h e significance of D w o r k i n as an apologist for an "activist" judiciary and the association, as in Bentham, of legal positivism with hostility to an active judiciary become transparent. 34 Dworkin is correct that judges do not automatically become lawless when they bring moral and political values into their decision making, but he is incorrect in thinking that value-laden judicial decisions can be confidently pronounced right or wrong. Without social, cultural, and political homogeneity, a legal system is not able to generate demonstrably right, or even professionally compelling, answers to difficult legal questions, whether f r o m within the legal culture or by reference to m o r al or other extralegal norms—the traditional province of natural law. For without either nature, or a political, social, and moral c o m m u n i t y so monolithic that the prevailing legal norms are "natural" in the sense of taken for granted, natural law can be but a shadow of its former self— can be but a name for the considerations that influence law even though not prescribed by a legislature or other official body. Lately Dworkin has been emphasizing the interpretive character of law. Even in a novel case, he argues, the j u d g e is not making a new rule but interpreting a legal tradition. Yet we shall see that the concept of interpretation need not, and in the most difficult cases will not, limit judicial discretion. Moreover, the American legal tradition is n o w so rich, variegated, conflicted, and ambivalent that a strand of it can easily be found to support either side in difficult cases. D w o r k i n has created a rich vocabulary for masking discretionary, political decision making by judges. 34. M a k i n g the title of D w o r k i n ' s latest b o o k — " L a w ' s Empire"—significant in a w a y perhaps not i n t e n d e d , since, as Frank M i c h e l m a n r e m i n d s me, D w o r k i n m a y s i m p l y have meant to echo H a r r i n g t o n ' s slogan, "an e m p i r e of laws and n o t of m e n . " J a m e s H a r r i n g t o n , The Commonwealth of Oceana, in The Political Works of James Harrington 155, 171 (J. G. A. Pocock ed. 1977). T h e relation b e t w e e n D w o r k i n ' s j u r i s p r u d e n c e and his defense of liberal judicial activism is discussed in Ken Kress, " T h e I n t e r p r e t i v e T u r n , " 97 Ethics 834, 8 4 4 - 8 4 6 (1987). For representative criticisms of D w o r k i n ' s enterprise, see J. L. M a c k i e , " T h e T h i r d T h e o r y of Law," in Mackie, Persons and Values: Selected Papers, vol. 2, p. 132 (Joan M a c k i c and Penelope Mackie eds. 1985); A l t m a n , n o t e 30 a b o v e , at 232-234; L a w r e n c e A. A l e x ander and Michael D . Bayles, " H e r c u l e s or Proteus? T h e M a n y T h e s e s of R o n a l d D w o r kin," 5 Social Theory and Practice 267 (1980); Stanley C . B r u b a k e r , " R e c o n s i d e r i n g D w o r kin's Case for Judicial A c t i v i s m , " 46 Journal of Politics 503 (1984); Drucilla L. C o r n e l l , "Institutionalization of M e a n i n g , Recollective I m a g i n a t i o n , and the Potential for T r a n s f o r mative Legal I n t e r p r e t a t i o n , " 136 University of Pennsylvania Law Review 1135 (1988); Weinreb, Natural Law and Justice, n o t e 11 above, at 117-122.

A Preview

of the

Book

I began the previous section b y contrasting a strong positivist (Thrasymachus) and a strong natural lawyer (Plato's Socrates). This term i n o l o g y identifies one difference b e t w e e n the t w o contending schools of jurisprudence that I have been discussing: the moral difference or, equivalently, the difference in political theory. But there is also an epistemic difference. M o s t t h o u g h n o t all legal positivists are skeptical about the claim that law is objective and a u t o n o m o u s . L a w to t h e m is too close to politics for that. So while " s t r o n g " natural lawyers tend to be the true believers in law, legal positivists tend to be legal skeptics. But these correlations are n o t necessary. A legal positivist w h o believed that the m e t h o d s of legal reasoning enabled the sovereign's c o m m a n d s to be infallibly u n d e r s t o o d and applied b y j u d g e s w o u l d be a moral but not an epistemic skeptic—that is, he w o u l d be skeptical about whether law was g o o d but not about w h e t h e r it was determinate—while a natural lawyer w h o believed that the m e t h o d s of legal reasoning often were incapable of discovering moral truth w o u l d be an epistemic but not a moral skeptic. (Plato appears to have believed that the m e t h o d s of legal reasoning obscure moral truth.) T h a t is w h y a positivist can be a legal formalist and a natural lawyer a legal realist; the f o r m e r conjunction is in fact c o m m o n . 3 5 H o l m e s flirted w i t h formalist positivism, especially in his "objective" theory of contractual (and statutory) interpretation. A n d notice that while the formalist w h o is also a positivist believes that at root law is "all politics," he believes that in application it is nonpolitical, because the j u d g e is n o t exercising discretion. This posture lets the jurist disdain politics and natural law in the same breath. For that matter, a natural lawyer can be a positivist—Coke, and m a y be D w o r k i n in his recent writings, w h i c h emphasize interpretation, fit this bill. A n d m a y b e a formalist can be a realist. E c o n o m i c analysis of law is a formalist edifice erected on a realist base, so one is not surprised to find that it has been criticized as formalist by antiformalists and as realist by antirealists. A n d to the extent that the e c o n o m i c analyst seeks to shape law to c o n f o r m to e c o n o m i c n o r m s , economic analysis of law has a natural law flavor. These conflations reveal the p o v e r t y of labels but d o not entirely obliterate the fact that, f o r m o r e than t w o millennia, the field of jurisprudence has been f o u g h t over b y t w o distinct t h o u g h variegated

35. For an e x a m p l e , see R o l f Sartorius, "Social Policy and Judicial Legislation," 8 American Philosophical Quarterly 151 (1971).

groups. O n e contends that law is m o r e than politics and in the hands of skillful j u d g e s yields—at least at certain times, in stable conditions— correct answers to even the m o s t difficult legal questions. T h e other contends that law is politics t h r o u g h and t h r o u g h and that j u d g e s exercise broad discretionary authority. O n the side of law as an objective entity and a u t o n o m o u s discipline w e have the distinguished lineage of Antigone, Socrates, C o k e , Blackstone, Langdell, Hart-Sacks, and D w o r k i n ; let us call t h e m the "Legalists." O n the side of law as politics, or law as the will of the stronger, or law as the activity of licensed professionals (judges, legislators, and so on), w e have the equally distinguished lineage of C r e o n , T h r a s y m a c h u s , J a m e s I, H o b b e s , B e n t h a m , Holmes, and H . L. A. Hart; let us call t h e m the "Skeptics." This division, which resembles William James's distinction b e t w e e n the "tenderm i n d e d " and the " t o u g h - m i n d e d , " is imperfect; Socrates is a particularly p o o r fit. A n d the lists are radically incomplete. F r o m the Skeptics I have omitted, to note j u s t a f e w examples, n o t only J o h n Austin, the great nineteenth-century expositor o f Hobbes's c o m m a n d t h e o r y o f law, and the distinguished j u r i s p r u d e A m b r o s e Bierce, w h o s e Devil's Dictionary succinctly defines " l a w f u l " as " c o m p a t i b l e w i t h the will of a j u d g e h a v ing jurisdiction," 3 6 but also the m e m b e r s o f the critical legal studies movement—heirs to the legal realists. 37 C o n t e m p o r a r y Legalists less exotic than D w o r k i n include, again to note j u s t a few, Charles Fried, Walter Berns, and Michael Oakeshott. 3 8 Legalists advance positions that are m o r e congenial to the legal profession than those of the Skeptics, but perhaps also m o r e h u m a n e and m o r e edifying; it is to Legalists such as C o k e and Blackstone that w e o w e m a n y of the distinctive A n g l o -

36. The Collected Writings of Ambrose Bierce 289 (1946). 37. See M a r k K e l m a n , A Guide to Critical Legal Studies (1987); G o o d r i c h ' s fine b o o k , cited in n o t e 30 above; The Politics of Law: A Progressive Critique ( D a v i d K a i r y s ed. 1982); M a r k Tushnet, "Critical Legal Studies: A n I n t r o d u c t i o n t o Its O r i g i n s and U n d e r p i n n i n g s , " 36 Journal of Legal Education 505 (1986). For criticism of critical legal studies, see M a r t i n Krygier, "Critical Legal Studies and Social T h e o r y — A R e s p o n s e to Alan H u n t , " 7 Oxford Journal of Legal Studies 26 (1987); William E w a l d , " U n g e r ' s P h i l o s o p h y : A Critical Legal Study," 97 Yale Law Journal 665 (1988); J o s e p h I s e n b e r g h , " W h y L a w ? " 54 University of Chicago Law Review 1117 (1987). 38. See Fried, " T h e Artificial R e a s o n o f t h e L a w o r : W h a t L a w y e r s K n o w , " 60 Texas Law Review 35 (1981); B e r n s , Taking the Constitution Seriously (1987); O a k e s h o t t , " T h e R u l e of Law," in O a k e s h o t t , On History, and Other Essays 119 (1983). Fried represents a school of t h o u g h t that I call n e o t r a d i t i o n a l i s m and discuss in C h a p t e r 14. Critical legal studies is discussed passim, b u t particularly in C h a p t e r s 1, 4, 8, and 13. O t h e r recent tendencies in legal theory that bear o n the issues I discuss in this b o o k , i n c l u d i n g f e m i n i s m (see C h a p t e r 13) and law and e c o n o m i c s ( C h a p t e r 12), are also e x a m i n e d in d u e course.

American liberties enshrined in the Bill of Rights and the Fourteenth A m e n d m e n t . T h e Skeptics, other than H . L. A. Hart, are a bit scary, although attempts to link t h e m w i t h fascism and other horrors 3 9 are misconceived and indeed flavored w i t h hysteria. Yet while the Skeptics are too skeptical and too positivist, they have, I believe, a better grip on the realities of law. T h e rest of this b o o k is devoted to exploring the basic issues debated by the t w o groups and the issues that g r o w out of those issues. I shall argue against "artificial reason," against D w o r k i n ' s " r i g h t - a n s w e r " thesis, against f o r m a l i s m , against overarching conceptions of justice such as "corrective justice," "natural law," and "wealth maximization"— t h o u g h n o t against m o d e s t versions of these n o r m a t i v e systems—but also against " s t r o n g " legal positivism (see note 31). I shall argue for an "activity" theory of l a w — t h e t h e o r y that underlies Holmes's prediction theory; for behaviorism and therefore against "rich" conceptions of mentalism, intentionality, and free will; for the critical as distinct f r o m constructive use of logic; for the idea that the j u d g e ' s proper aim in difficult cases is a reasonable result rather than a demonstrably right one; and for a concept of the j u d g e as a responsible agent rather than as a conduit of decisions m a d e elsewhere in the political system. More, I shall argue for objectivity as a cultural and political rather than epistemic attribute of legal decisions, f o r balancing rule-of-law virtues against equitable and discretionary case-specific considerations, for m a k i n g law m o r e receptive to science—but w i t h due awareness of the irremediably authoritarian character of law, w h i c h limits the scope for a scientific ethos in law—and for a consequentialist theory of interpretation. I shall argue in short for a functional, policy-saturated, nonlegalistic, naturalistic, and skeptical, 40 b u t decidedly not cynical, conception of the legal process; in a w o r d (although, I fear, an inadequate word), for a pragmatic jurisprudence. This is n o invention of mine. H o l m e s , Cardozo, and the realists were pragmatists, although H o l m e s incompletely so; and legal p r a g m a t i s m has its c o n t e m p o r a r y defenders as well, although they will disagree with m u c h of m y argument. 4 1 T h a t a r g u m e n t , I add, is b o t h positive and 39. A t t e m p t s well discussed in Purccll, n o t e 30 a b o v e , ch. 9. 40. 1 m e a n " s k e p t i c i s m " in its lay sense r a t h e r than in its philosophical sense: that is, skepticism as a t t i t u d e r a t h e r t h a n as m e t a p h y s i c s . See Michael Williams, " T h e Elimination of M e t a p h y s i c s , " in Fact, Science and Morality: Essays on A.J. Ayer's "Language, Truth, and Logic" 9, 1 4 - 1 6 ( G r a h a m M a c d o n a l d and C r i s p i n W r i g h t eds. 1986). 41. See, for e x a m p l e . S u m m e r s , n o t e 30 a b o v e ; Daniel A. Farber, "Legal P r a g m a t i s m and the C o n s t i t u t i o n , " 72 Minnesota Law Review 1331 (1988); the perspectivist approach

normative—that is, both descriptive and evaluative. I argue that A m e r ican law really is, and also should be, pragmatic, and that it can be improved by greater awareness of its pragmatic character. I use "pragmatic" in its philosophical rather than its lay sense, to refer to the theories of the American pragmatist philosophers—notably Peirce, James, Dewey, Mead, Kuhn, and Rorty—and of the European philosophers, notably Wittgenstein and Habermas, w h o have espoused similar views. 42 There is enormous diversity within this group. There

taken by M a r t h a M i n o w in her article " T h e S u p r e m e C o u r t , 1986 T e r m : F o r e w o r d , Justice E n g e n d e r e d , " 101 Harvard Law Review 10 (1987); and Grey's fine article " H o l m e s and Legal P r a g m a t i s m , " n o t e 25 a b o v e . A g o o d older discussion of p r a g m a t i c j u r i s p r u d e n c e , e m p h a sizing (as d o b o t h G r e y and I) J o h n D e w e y ' s b r a n d of p r a g m a t i s m , is E d w i n W. P a t t e r s o n , " P r a g m a t i s m as a P h i l o s o p h y o f L a w , " in The Philosophy of the Common Man: Essays in Honor of John Dewey to Celebrate His Eightieth Birthday 172 (1940); also in P a t t e r s o n , Jurisprudence: Men and Ideas of the Law, ch. 17 (1953). S u m m e r s ' s b o o k has a m o r e historical, and less philosophical, e m p h a s i s t h a n m i n e and focuses m o r e t h a n I d o on the legal realists. Farber's article contains a particularly g o o d reply to D w o r k i n ' s criticisms of legal p r a g m a t i s m , w h i c h appear in Law's Empire, n o t e 33 a b o v e , at 151—164 (see also i n d e x references in Law's Empire u n d e r " P r a g m a t i s m " ) . See 72 Minnesota Law Review at 1343-1347. D w o r kin uses the t e r m " p r a g m a t i s m " strictly in its lay, " w h a t w o r k s " sense; t h u s " t h e p r a g m a t i s t thinks j u d g e s s h o u l d always do the best t h e y can f o r the f u t u r e , in the circumstances, unchecked by any need to respect or secure consistency in principle w i t h w h a t o t h e r o f f i cials have d o n e or will d o . " Law's Empire, n o t e 33 above, at 161. T h i s is n o t an interesting or even coherent version of p r a g m a t i s m , i m p l y i n g as it does that a p r a g m a t i c j u d g e is an ad hoc decision m a k e r h a v i n g n o appreciation f o r the b e n e f i t s of consistency w i t h p r e c e d e n t and of fidelity to s t a t u t o r y and c o n s t i t u t i o n a l texts. T h i s is t a n t a m o u n t t o saying that a p r a g m a t i c j u d g e ignores all s y s t e m i c as distinct f r o m case-specific c o n s e q u e n c e s of his decisions; n o t h i n g in p r a g m a t i s m w a r r a n t s , or indeed p e r m i t s , so t r u n c a t e d a c o n c e p t i o n o f the judicial task. 42. O n the revival of interest in p r a g m a t i c p h i l o s o p h y , see Anti-Foundationalism and Practical Reasoning: Conversations between Hermenentics and Analysis (Evan S i m p s o n ed. 1987); Richard Rorty, (Contingency, Irony, and Solidarity (1989); R o r t y , Consequences of Pragmatism {Essays 1972-1980) (1982), esp. pp. 160-166; C o r n e l West, The American Evasion of Philosophy: A Genealogy of Pragmatism (1989); Hilary P u t n a m and R u t h A n n a P u t n a m , "William James's Ideas," 8 Raritan, Winter 1989, at 27; Richard H . K i n g , ' " I n O t h e r W o r d s ' : T h e Philosophical Writings of Richard R o r t y , " 19 Journal of American Studies 95 (1985); J o s e p h Margolis, Pragmatism without Foundations: Reconciling Realism and Relavitism (1986) (vol. 1 of his trilogy The Persistence of Reality); Pragmatism: Its Sources and Prospects ( R o b e r t J. M u l vancy and Philip M . Z e l t n e r eds. 1981). For b a c k g r o u n d , sec J o h n E. S m i t h , Purpose and Thought: 'The Meaning of Pragmatism (1978); D a v i d A. H o l l i n g e r , In the American Province: Studies in the History and Historiography of Ideas, ch. 2 (1985); H . S. T h a y e r , Meaning and Action: A Critical History of Pragmatism (1968). A l t h o u g h Richard R o r t y is the b e s t - k n o w n m o d e r n A m e r i c a n p r a g m a t i s t — a n d his essay " T h e P r i o r i t y of D e m o c r a c y , " in The Virginia Statute for Religious Freedom: Its Evolution and Consequences in American History 257 (Merrill D. Peterson and R o b e r t C . V a u g h a n eds. 1988), is a particularly g o o d i n t r o d u c t i o n to p r a g m a t i s m — I reject R o r t y ' s R o m a n t i c , antiscientific b r a n d of p r a g m a t i s m . See C h a p t e r 12. A n d I have very little s y m p a t h y for m o s t of H a b e r m a s ' s v i e w s .

are also close links b e t w e e n its m e m b e r s and a variety of other philosophers w h o are n o t usually called pragmatists. A m o n g these are E m e r son, Nietzsche, and P o p p e r ; utilitarians ( w h o like pragmatists are strongly instrumentalist), such as B e n t h a m ; British empiricists, such as H u m e , Mill (also a utilitarian, of course), and Russell; and such c o n t e m porary analytic philosophers as Q u i n e and Davidson. 4 3 N o t only is it difficult to tell w h e n one has strayed across the boundaries that separate p r a g m a t i s m f r o m n e i g h b o r i n g philosophical traditions, but the core of p r a g m a t i s m , if there is such a thing, is too v a r i f o r m to m a k e p r a g m a tism a single philosophy or philosophical school in a useful sense. It lacks the unity of such schools as logical posivitism—another neighbor of, and rival to, p r a g m a t i s m . To say that one is a pragmatist is to say little. T h e b r a n d of p r a g m a t i s m that I like emphasizes the scientific virtues (open-minded, n o - n o n s e n s e inquiry), elevates the process of inquiry over the results o f inquiry, prefers f e r m e n t to stasis, dislikes distinctions that m a k e n o practical difference—in other w o r d s , dislikes "metaphysics"—is d o u b t f u l of finding "objective t r u t h " in any area of inquiry, is uninterested in creating an adequate philosophical foundation f o r its t h o u g h t and action, likes experimentation, likes to kick sacred cows, a n d — w i t h i n the b o u n d s of prudence—prefers shaping the future to maintaining continuity w i t h the past. So I a m speaking of an attitude rather than a d o g m a ; an attitude w h o s e " c o m m o n d e n o m i n a t o r " is "a future-oriented instrumentalism that tries to deploy t h o u g h t as a w e a p on to enable m o r e effective action." 4 4 M o s t of this b o o k is concerned w i t h attacking the d o g m a s and letting p r a g m a t i s m emerge as the natural alternative. S o m e quotations f r o m Cardozo's The Nature of the Judicial Process will give a flavor of pragmatic j u r i s p r u d e n c e and will sound s o m e themes explored in the chapters below. " T h e final cause of law," writes C a r dozo, "is the welfare of society" (p. 66). So m u c h for the formalist idea, w h o s e scientistic provenance and pretensions are evident, of law as a b o d y of i m m u t a b l e principles. C a r d o z o does n o t mean, however, that in (for example) j u d g i n g the constitutional validity of statutes the j u d g e s "are free to substitute their o w n ideas of reason and justice for those of the m e n and w o m e n w h o m they serve. Their standard must be an objective o n e " — b u t objective in a pragmatic sense, w h i c h is not the

43. O n the relation b e t w e e n p r a g m a t i s m and British e m p i r i c i s m , see A. J. Ayer, The Origins of Pragmatism: Studies in the Philosophy of Charles Sanders Peirce and William James (1968). 44. West, n o t e 42 above, at 5.

sense of correspondence with an external reality. For, he writes, "in such matters, the thing that counts is not what I believe to be right. It is what I may reasonably believe that some other man of normal intellect and conscience might reasonably look u p o n as right" (p. 89). Legal rules are to be viewed in instrumental terms, implying contestability, revisability, mutability. "Few rules in our time are so well established that they may not be called upon any day to justify their existence as means adapted to an end. If they do not function, they are diseased. If they are diseased, they need not propagate their kind. Sometimes they are cut out and extirpated altogether. Sometimes they are left with the shadow of continued life, but sterilized, truncated, impotent for h a r m " (pp. 98-99). "I sometimes think w e w o r r y ourselves overmuch about the enduring consequences of our errors . . . In the endless process of testing and retesting, there is a constant rejection of the dross" (p. 179); "the tide rises and falls, but the sands of error crumble" (p. 177). Law is forward-looking. This is implicit in an instrumental concept of law—which is the pragmatic concept of law, law as the servant of human needs. " N o t the origin, but the goal, is the main thing. There can be no wisdom in the choice of a path unless w e k n o w where it will lead . . . The rule that functions well produces a title deed to recognition . . . The final principle of selection f o r j u d g e s . . . is one of fitness to an end" (pp. 102-103). T h e reference to "title deed" is particularly noteworthy; it is a rebuke to formalists w h o require that law to be valid be "pedigreed" by being shown to derive f r o m some authoritative source. Where does the j u d g e turn for the knowledge that is needed to weigh the social interests that shape the law? "I can only answer that he must get his knowledge . . . f r o m experience and study and reflection; in brief, f r o m life itself" (p. 113). This is vague, but points in the right direction. The j u d g e is not merely an interpreter of legal materials. He is not only a finder but also a maker of law: J o h n Marshall "gave to the constitution of the United States the impress of his o w n mind; and the form of our constitutional law is what it is, because he moulded it while it was still plastic and malleable in the fire of his o w n intense convictions" (pp. 169-170). O n e purpose of this book is to extend these insights and bring them up to date. The book is in five parts, with the first four concentrating on the major attempts to give law a satisfactory philosophical foundation and the last exploring antifoundationalist approaches. Part I focuses on the epistemology of law. T h e central questions are what is legal reasoning and to what extent can it produce knowledge and not merely opinion

or belief? I examine the role of exact inquiry (logical or scientific) in law and find it to be limited; the role of rules in law is correlatively limited. I then discuss legal reasoning as a branch of practical reasoning, conceived as the diverse set of reasoning methods (embracing intuition, authority, analogy, deliberation, interpretation, tacit knowing, subjecting propositions to the test of time, and much else) that people use when logic and science run out. Although these methods often generate determinate results in law, there are occasions—and these the cruces of legal evolution—when they do not, and then judicial decision is based perforce on policy, politics, social vision, "values," even "prejudice" (in Gadamer's sense of prejudgment). Decisions so based are not easily determined to be "right" or " w r o n g " ; the vocabulary of apodictic certainty is misplaced. Perhaps the highest aspiration of the j u d g e is reasonableness in adjudication. Part II looks at the same questions f r o m a different angle—the ontological. Epistemology and ontology are closely related, the former dealing with what we can know, the latter with what there is to know. The less there is to know, the harder it is to k n o w much. Yet it is possible to believe, for example, that there are moral "reals" (for example, Platonic Forms) but that we have no way of discovering them, and conversely that there are no moral "reals" but that objective knowledge of moral duties is possible, perhaps through consensus or through convention. Part II asks, What are the "objects" of legal inquiry? In what sense do law, and even fact, "exist"? I do not question only obviously problematic entities, such as intent. With the aid of a variety of examples, including criminal intent, coerced confessions, legal fact-finding, judicial behavior, and the debate over the question of what law is, I argue that legal inquiry is hampered by the elusive character of many of the entities that appear to be central to it. Legal thought cannot be made objective by being placed in correspondence with the "real" world. It owes whatever objectivity it has to cultural uniformity rather than to metaphysical reality or methodological rigor. Part III takes up in much greater detail than in Part I the question of interpretation, which is sometimes thought the key to objectivity in law—not only statutory and constitutional law but also c o m m o n law, that is, law overtly made by judges. I argue that the term "interpretation" is so elastic (in part because the concept or activity of interpretation serves so many different functions) that it often is a fig leaf covering judicial discretion rather than a guide to decision making. This is not to say that all legal texts are inscrutable—far f r o m it. Most such texts communicate a straightforward, readily decipherable message. But for those

that do not, neither the theory n o r the practice of "interpretation" p r o vides objective footing for a legal j u d g m e n t . We m i g h t do better to discard the term and concentrate instead o n c o m p a r i n g the practical consequences of proposed applications of a legal text in cases w h e r e its meaning is unclear. Part IV, switching the emphasis f r o m metaphysics (Parts I and II) and hermeneutics (Part III) to moral and political philosophy, examines a variety of overarching concepts of justice that could be used to channel legal inquiry. T h e concepts range f r o m corrective justice and wealth maximization to the feminist ethic of care and the r o m a n t i c yearnings of radical communitarians, w i t h intermediate stops at egalitarianism, natural rights, and civic republicanism. T h e first t w o of these approaches, at least, have significant roles to play in channeling legal inquiry, but neither one (nor b o t h together) can close the open area of judicial decision making all the way. T h e other approaches are vulnerable to disabling criticisms. B o t h in endorsing a significant role for corrective j u s tice and in urging only a limited role f o r wealth maximization, this b o o k modifies s o m e of m y previously published views. Part V examines efforts to find alternatives to systems, w h e t h e r novel or traditional, of jurisprudence. (The last chapter of Part IV paves the way, by recasting the wealth-maximization approach to law in p r a g matic terms.) T h e first chapter in this part examines the "back to basics" m o v e m e n t that I call neotraditionalism, w h i c h d r a w s on Aristotle and Burke for a j u r i s p r u d e n c e that emphasizes prudence and professional tradition rather than deduction f r o m first principles. I argue that neotraditionalism is n o m o r e directive than are the m o d e s of interdisciplinary inquiry to w h i c h it is offered as an alternative. T h e final chapter gathers together proposals of m y o w n that are scattered t h r o u g h o u t the b o o k and defends the case for a n e w pragmatist jurisprudence, one similar in some respects to neotraditionalism yet distinct f r o m and in part opposed to it. As should be evident f r o m even so bare a s u m m a r y , the questions that give structure to the b o o k are whether, in w h a t sense, and to what extent the law is a source of objective and determinate, rather than m e r e ly personal or political, answers to contentious questions. Recurring to the three senses of objectivity—the ontological, the scientific (replicable, convergent), and the conversational—my a r g u m e n t is that the first is out of the question in m o s t legal cases; the second is sometimes attainable, but given the attitudes of and the constraints o n the legal profession, and the character of the p r o b l e m s that it deals with, often not; and the third is attainable—but that isn't saying m u c h . T h e conversational

method is, by definition, inconclusive in difficult cases when society is heterogeneous. But this is to speak redundantly, for it is cultural, social, and political heterogeneity that makes cases difficult. Contemporary American law is, then, strongly objective in easy cases, weakly objective in difficult ones, but rarely either highly determinate or merely political. The only way to make it more objective is to make the courts and the legislatures more homogeneous, culturally and politically. There would be costs to doing this even if it were feasible. There would also be costs to making the courts and the legislatures more heterogeneous; for one, the law would become more political, which would damage rule-of-law virtues that are genuine, indeed precious (though limited), social goods. Extreme positions are more fun, but in jurisprudence the true as well as the good is to be found between the formalistic and "realistic" extremes depicted by Mencken: T h e average American j u d g e . . . [has] no more give and take in his mind than you will find in the mind of a terrier watching a rathole. H e converts the law into a series of rubber-stamps, and brings them d o w n upon the scalped skulls of the just and unjust alike. T h e alternative to him, as c o m m o n l y conceived, is quite as bad—an uplifter in a black robe, eagerly gulping every n e w brand of Peruna that comes out, and converting his pulpit into a sort of soap-box. 1 5

M v position may seem boringly centrist, but it will provoke both the true centrists in the profession, w h o want very much to believe that law is a u t o n o m o u s and apolitical, and the political activists w h o want to m o v e the law sharply to the left or to the right. It will be criticized by the Left as authoritarian and complacent, and by the Right as cynical and amoral. It may appear to be merely old-fashioned legal realism with a conservative twist abhorrent to the realists, but that would not be an accurate characterization. Unlike Holmes and other legal (and logical) positivists, I do not believe that legal questions can be partitioned into those that can be answered with certainty and those that cannot be answered by rational methods at all but depend on the will or caprice or politics of the judge. I am also not a "strong" legal positivist, as Holmes was; indeed, I resist the effort to dichotomize positive and natural law. M y position owes more to philosophical pragmatism than did the realist movement. If I had to choose I would range myself on the side of the Skeptics, but the pragmatic approach may enable the conflict

45. H . L. M e n c k e n , " M r . J u s t i c e H o l m e s , " in The Vintage Mencken C o o k e ed. 1955) (essay first p u b l i s h e d in 1932).

189, 195 (Alistair

between Legalists and Skeptics to be transcended. I repeat that legal positivists w h o believe that the sovereign's c o m m a n d s are readily interpretable are formalists, at least in the sense of believers in right answers to all or virtually all legal questions. I am not a formalist. Although I shall recur f r o m time to time to Holmes, D w o r k i n , H. L. A. Hart, and other jurisprudes, I shall not undertake a detailed examination of the texts or arguments of the warring schools, whose positions I have sketched—and with a broad brush—merely to give the reader a sense of the terrain that the rest of the book will explore. And while the quest, unsuccessful though I believe it must prove in the end, for grounds on which to reaffirm confidently law's objectivity and autonomy gives structure to the book, that search is indeed skeleton rather than flesh. I urge readers to attend to the particulars of m y analysis rather than merely to its direction and conclusion. If, having done so, they still conclude that the ratio of destructive to constructive criticism is too high, I ask them to ponder Voltaire's reply when he was taken to task for offering no substitute for Christianity, which he had attacked: "Je vous delivre d'une bete feroce, et vous me demandez par quoi je la remplace!"46 46. "I save y o u f r o m a ferocious beast and y o u ask m e w h a t I replace it w i t h ! "

PART I

THE EPISTEMOLOGY OF LAW

Law as Logic, Rules, and Science

This chapter begins the inquiry into whether and to what degree law is objective, impersonal, determinate: whether, in other words, it is an external (though not necessarily an effective—that is a separate question) constraint on judges; in an earlier diction that has come to seem naive, whether judges find rather than make law. I offer a moderately skeptical answer, in t w o steps. First, many important questions of law, though certainly not all and maybe only a tiny fraction (but one rich in landmark and seminal cases), cannot be answered by the use of the methods of legal reasoning, in part because those methods are not powerful when viewed as tools for construction rather than merely for criticism. Second, when legal reasoning runs out and the j u d g e has to appeal to policies, preferences, values, morals, public opinion, or whatever else may be necessary to answer a legal question in a way satisfactory to himself and his colleagues, it will often mean that the answer is indeterminate. If, to anticipate subsequent discussion, "truth is what would emerge as the result of unconstrained inquiry pursued indefinitely,"' then indeterminacy is the state where unconstrained inquiry is—interminable: where there is no light at the end of the tunnel. That is often the case in law. The first half of m y thesis denies Coke's "artificial reason" but is consistent with the possibility that legal outcomes can be made determinate by methods of analysis that o w e nothing to legal training or experience. The second half denies that every legal question has a right answer, whatever sources or procedures are appealed to. The artificial-reason thesis reflects nostalgia for a lost sense of the law's autonomy, the right1. J o s e p h Rouse, Knowledge and Power: Toward a Political Philosophy of Science 7 (1987). T h i s is n o t a satisfactory d e f i n i t i o n of t r u t h — a s w e shall see in C h a p t e r 3, there are n o satisfactory definitions of t r u t h — b u t it is a u s e f u l criterion of t r u t h .

answer thesis nostalgia for lost certitudes. I disagree with both theses, but I also disagree with the radical skepticism that equates law to politics (and hence a judge to a legislator or other politician), regards all legal analysis as bunk, all legal questions as indeterminate. Although "politics" is a word of many meanings, some of which overlap "law," it is not true that what a judge does is indistinguishable from what a legislator or an executive-branch official does. So I face two ways in this chapter, as throughout the book. I defend my middle position primarily by attacking the extremes, particularly the right-answer / artificial-reason extreme because that is the more plausible and influential one. Whether the reader w h o is persuaded by my arguments thinks I have shown that the glass is half full or half empty will depend on the reader's conception of h o w objective, impersonal, and determinate law must be to deserve the name of law—which is not to say, however, to deserve the citizen's respect. Law can be highly objective and impersonal, yet thoroughly unjust. The inquiry into justice is postponed to later chapters, but the tradeoff between formal and substantive justice should be kept in mind from the start. To decide how people, judges specifically, might acquire true beliefs about legal questions will require consideration of the various methods by which nonlawyers acquire true beliefs about nonlegal questions, as well as consideration of the use of such methods by lawyers and judges. The methods can be divided into two broad classes, exact inquiry and practical reason. The first will occupy us in this chapter. The Syllogism

and Other Methods of Logic

I begin with the apt and famous (though, as we shall have occasion to notice, nonstandard) syllogism: "All men are mortal; Socrates is a man; therefore Socrates is mortal." The validity of the argument—not the truth of the conclusion, which depends on the truth of its premises— seems utterly compelling. But that is only because the conclusion that Socrates is mortal is contained in the first premise, the definition of "man." That premise says, in effect, here is a box, labeled "men," with a bunch of things in it, every one of which is "mortal." The second premise tells us that the things in the box have name tags and that one of the tags says "Socrates." When we pluck Socrates out of the box we know he is mortal because the only things in it are mortals. We are merely taking out what we put in. We thus find the syllogism's validity compelling by virtue of a metaphor, the metaphor of the box. (It may seem odd that one's confidence

in the validity of logic should be bolstered by a metaphor—yet this is a clue to the limitations of logic and the cognitive importance of metaphor.) As we move away f r o m the simplest, most transparent examples of logical reasoning, such as the syllogism, the nature and cogency of such reasoning become less perspicuous. "We think that the proposition 2 + 2 = 4 is true by definition in much the same sense that the proposition that Socrates is mortal is true by definition. But because it is not clear in the numerical example what the container and the thing contained are, the metaphor of inclusion is no longer available to reassure us of the compelling quality of logical reasoning. And as soon as we begin to ask complicated mathematical questions, such as whether every even number is the sum of t w o odd prime numbers (as 16, for example, is the sum of 13 and 3), we enter a realm in which no mechanical or apodictic decision procedure is available and the questions are as difficult, uncertain, and inaccessible to lay understanding as the most difficult questions about the empirical as distinguished f r o m the conceptual, the definitional, world. Indeed, it has proved impossible to reduce mathematics to logic. 2 But we can set these problems to one side and stick with the simplest forms of logical reasoning, for with rare exceptions they are the only ones lawyers use. So compelling and familiar is syllogistic reasoning that lawyers and judges, ever desirous of making their activity seem as objective as possible, try hard to make legal reasoning seem as syllogistic as possible. 3 The overuse of the syllogism is the defining characteristic of the brand

2. See, for e x a m p l e , A l v m I. G o l d m a n , Epistemology and Cognition 3 4 - 3 5 (1986); G r e g o r y J. C h a i t i n , " R a n d o m n e s s in A r i t h m e t i c , " Scientific American, J u l y 1988, at 80; Ernest N a g e l and J a m e s R. N e w m a n , Godel's Proof (1958). 3. O n the use of the s y l l o g i s m in legal reasoning, see, for e x a m p l e , Steven J. B u r t o n , An Introduction to Law and Legal Reasoning, ch. 3 (1985); Neil M a c C o r m i c k , Legal Reasoning and Legal Theory 19-52 (1978); Vincent A. Wellman, "Practical R e a s o n i n g and Judicial J u s tification: T o w a r d an A d e q u a t e T h e o r y , " 57 University of Colorado Law Review 45, 6 4 - 8 0 (1985); also A. G. Guest, " L o g i c in the Law," in Oxford Essays in Jurisprudence 176 (A. G. Guest ed. 1961). O n logical reasoning in general, C a r l Wellman, Challenge and Response: Justification in Ethics, ch. 1 (1971), is particularly h e l p f u l ; and o n the subject m a t t e r of this chapter generally, see G i d o n G o t t l i e b , The Logic o f Choice: An Investigation of the Concepts of Rule and Rationality, chs. 1 and 2 (1968). T h r e e t e r m i n o l o g i c a l p o i n t s to bear in m i n d : First, b y " l o g i c " I m e a n d e d u c t i v e and mathematical logic, n o t s y s t e m a t i c t h i n k i n g in general or specialized logics such as i n d u c tive logic. I do discuss i n d u c t i o n but n o t u n d e r the r u b r i c of logic. Second, in the t e r m " s y l l o g i s m " 1 include the e n t h y m e m e , a s y l l o g i s m w i t h an implicit premise. Legal r e a s o n ing is in fact m o r e o f t e n e n t h y m e m a t i c than syllogistic in the strict sense. T h i r d , a s y l l o g i s m can have m o r e than t w o premises; and again this is characteristic of m o s t syllogistic reas o n i n g in law.

of legal formalism attacked by Holmes. But today when used pejoratively, "formalism" is more likely to refer to an exaggerated belief in the transparency of statutory or constitutional language and hence in the possibility of definitely correct answers to difficult interpretive questions than to the overuse of the syllogism. When used nonpejoratively, formalism can refer to a strong—but perhaps justifiably strong—belief in the possibility of obtaining right answers to legal questions by means of conventional methods of legal analysis, mainly the careful reading of texts to find the rules in them, followed by deduction f r o m the rules to the outcome of the particular case. 4 O r formalism can refer simply to the use of logic to reason f r o m premises to legal conclusions. T h e most useful sense of this protean term, however, derives f r o m the contrast between f o r m and substance—form referring to what is internal to law, substance to the world outside of law, as in the contrast between formal and substantive justice. T h e a u t o n o m y and objectivity of law are secured by confining legal analysis to the formal level, the level requiring only an exploration of the relations a m o n g legal ideas. A u t o n o m y and objectivity are threatened when the legal outcome depends on facts about the world, which might be the facts of a dispute or the social or ethical facts relevant to creating or interpreting a rule. Since logical reasoning is the exploration of relations between concepts, the relation between the form-substance sense of formalism and the logico-mathematical sense should be plain. Formalism comes in both natural law and legal positivist varieties. The only prerequisite to being a formalist is having supreme confidence in one's premises and in one's methods of deriving conclusions f r o m them. T h e natural law formalist is certain about the principles of justice and the p o w e r of logic to derive specific case outcomes f r o m those

4. See Frederick Schauer, " F o r m a l i s m , " 97 Yale Law Journal 509 (1988); Neil M a c C o r m i c k and O t a W e i n b e r g e r , An Institutional Theory of Law: New Approaches to Legal Positivism 1 1 5 - 1 4 4 (1986); J a m e s G. Wilson, " T h e M o r a l i t y of F o r m a l i s m , " 33 UCLA Law Review 431 (1985); R o b e r t S a m u e l S u m m e r s , Instrumentalism and American Legal Theory 137—159 (1982). Schauer a d v o c a t e s w h a t he calls " p r e s u m p t i v e f o r m a l i s m " (see also Schauer, "Is t h e C o m m o n L a w L a w ? " 77 California Law Review 455, 470 [1989])—a p r e f erence for sticking w i t h rules unless t h e y p r o d u c e p a l p a b l y a b s u r d or o u t r a g e o u s results. T h e reason for t h e p r e s u m p t i o n is n o t fully explained b u t p r e s u m a b l y is the value of certainty and predictability. T h i s invites c o n s i d e r a t i o n o f an alternative a p p r o a c h — i d e n t i f y i n g classes of rules in w h i c h certainty is particularly i m p o r t a n t and c o n f i n i n g the p r e s u m p t i o n to t h e m . T h e distinct q u e s t i o n , n o t e x a m i n e d b y Schauer, is the feasibility of f o r m a l i s m , w h e t h e r p r e s u m p t i v e or absolute; this is o n e o f the q u e s t i o n s addressed in the present c h a p ter. For e x e m p l a r y criticism of legal f o r m a l i s m , see J u d i t h N . Shklar, Legalism: Law, Morals, and Political Trials 3 3 - 3 6 (1964).

premises; the positive law formalist is certain that the law consists only of legislative or other official commands that, read carefully, yield demonstrably correct results in all cases. Either approach spares the lawyer or judge f r o m a messy encounter with empirical reality. There are not only natural law formalists and positive law formalists but also natural law realists and positive law realists. Most natural lawyers are formalists, though; and while most legal positivists are realists, there are plenty of positive law formalists. It is a safe bet that a majority of legal professionals are formalists. Formalism contains a built-in bias against legal change; this is a clue to the breadth and durability of its appeal. T h e most secure premise for deducing the outcome of a new case is the decision of the most recent case that had essentially the same facts, so that the rule expressed or implied in that case fits the new one like a glove (another comfortable metaphor). As a result, people w h o are defending the legal status quo— whether it is a liberal status quo or a conservative one—tend to assume a formalist stance. But so, paradoxically, do the people attacking the status quo. 5 Having the burden of persuasion, they want to show that their "revisionist" position is compelled by logic, is the authentic original position, that the defenders of the status quo are not merely w r o n g but demonstrably, irrefutably, and newly w r o n g — f o r example, because the precedents on which the defenders rely are invalid deductions f r o m a more authoritative premise, such as the text of the Constitution. T h e rhetorical power of formalism makes discourse about law more formalistic than the actualities of the judicial process warrant. I have been stressing the syllogism (including the enthymeme), but of course lawyers employ other elementary forms of logic as well, such as the principle that a proposition cannot be both true and false or that if two things are identical to a third they are identical to each other. Lawyers also use the language of logic where, strictly speaking, it does not belong. Most judges (myself included) would say without hesitation that it would be illogical to read into a speed-limit statute an implied exception for B M W s (as distinct f r o m one for police cars or a m b u lances). Yet this is using "illogical" as a s y n o n y m for "clearly w r o n g . " 5. A n i m p o r t a n t e x c e p t i o n , h o w e v e r , consists of t h o s e legal realists w h o attacked the very idea that there w e r e legal rules, t h o s e for w h o m every n e w case created n e w law. T h e legacy of these realists is the Federal Rules of Civil P r o c e d u r e , w h i c h do n o t require the plaintiff to plead his case in t e r m s of preexisting legal p i g e o n h o l e s . See, for e x a m p l e , J u d i t h Resnik, "Failing Faith: A d j u d i c a t o r y P r o c e d u r e in D e c l i n e , " 53 University of Chicago Law Review 494, 5 0 2 - 5 0 3 (1986); N o t e , "Plausible Pleadings: D e v e l o p i n g Standards for R u l e 11 Sanctions," 100 Harvard Law Review 630, 6 4 4 - 6 4 7 (1987).

A harmless usage unless taken literally, it misconceives the domain of logic. Logic has nothing to do with the question, easy though it may be to answer. The question is one of statutory interpretation, and we shall see that interpretation, although it sometimes yields reasonably certain answers, is not a method of logic. Given the purpose of the statute, there is no reason to exclude BMWs. But the rejection of the exception becomes a logical conclusion only after a rule has been extracted from the statute by a process that is not itself one of logic. It is similarly misleading to suggest that logic informs in a helpful way the idea of treating like things alike—an idea of great resonance for law (equal justice under law, equal protection of the laws, equality before the law, one law for rich and poor, and so forth). Although a natural extension of the logical propositions that things are identical to themselves and that two things identical to a third are identical to each other, the idea is empty without specification of the criteria for "likeness," and in law those criteria are political/' The legal principle of equal treatment distinguishes those differences between litigants that may properly be considered in the allocation of legal benefits and burdens, such as differences in skill and effort, f r o m those that, sometimes or always, may not be, such as differences in religion, race, social class, income, or relationship to the judge. The principle has nothing interesting to do with the avoidance of contradiction except insofar as it forbids completely arbitrary—"irrational"—classifications (like the exception for BMWs), as distinct f r o m vicious, invidious, or politically intolerable ones that may be perfectly rational, at least to the classifiers. But irrational classifications are rare.

Rules, Standards, and

Discretion

Despite the critical tone of the previous section, most legal questions are resolved syllogistically. A legal rule has the form of the major premise of a syllogism. For example: no contract is enforceable without consideration; the contract in suit has no consideration; therefore the contract is not enforceable. By a process not itself syllogistic or otherwise deductive, judges extract rules f r o m statutes and previous decisions and then use them as premises to decide cases syllogistically. Are wages income? Is marriage to one's sister valid? Is murder excused if the victim consents 6. See Peter Westen, " T h e E m p t y Idea of Equality," 95 Harvard Law Review 537 (1982); Westen, " T h e M e a n i n g of E q u a l i t y in Law, Science, M a t h , and M o r a l s : A Reply," 81 Michigan Law Review 604 (1983).

in advance? Is it illegal to drive sixty miles an h o u r in a forty-mile zone? Is Illinois entitled to have three U . S . senators? These and m a n y other legal questions are answered deductively by the application of clear and uncontested rules to facts determined or conceded. T h e reason such questions do not figure largely in legal debate, and indeed lie almost beneath the professional horizon, is that they are too simple to be a likely subject of litigation or even to require legal counseling. This makes it easy in doing j u r i s p r u d e n c e — t o o easy—to forget about such questions and suppose that all legal questions are indeterminate. (There is a missing link here: questions m i g h t have determinate answers w i t h o u t those answers' being logically entailed. But that is a subject taken u p in later chapters.) N o t only are m a n y legal questions determinate in the a p p r o x imate sense of logical determinacy; they are the very ones that l o o m largest in the thinking of ordinary people about law. M o s t laypeoplc demand of the law only that it tell t h e m the rules they m u s t live by, and they are therefore m o r e likely than lawyers to think of law as d e t e r m i nate, constraining. A big part of legal training, especially at the elite law schools, is exploration of the law's indeterminacies, and it creates an outlook remote f r o m that of the average person and indeed f r o m that of many lawyers. A n o t h e r reason that laypersons have a m o r e vivid sense of the law's determinacy than d o lawyers is that lawyers, s o m e t i m e s with a guilty conscience, tell the laity (I mean laypersons in general— not clients!) that the law is determinate. We must distinguish between the validity of a syllogism and its soundness—that is, its p o w e r to yield a true conclusion. Soundness depends not only o n the validity of the particular syllogism but also on the truth of the premises. 7 This is not a p r o b l e m w i t h the syllogism about Socrates, because n o one is likely to question that all m e n are mortal or that Socrates was a m a n . ( W h y n o one is likely to question the first premise is not so clear as one m i g h t think, and will be discussed later.) But with rules of law, the truth of the m a j o r and m i n o r premises is often contestable. To begin with, establishing the m i n o r premise—in other words, finding the facts—is often difficult; and finding facts is not a process of logic. H o w difficult depends in part on the rule's generality. C o m p a r e the t w o basic liability regimes for injuries caused b y accident:

7. T h e i m p o r t a n c e of the distinction b e t w e e n validity and s o u n d n e s s is s h o w n by the fact that a syllogism can have a true conclusion even t h o u g h b o t h its premises are false: "All Spartans are wise; Socrates is a Spartan; t h e r e f o r e Socrates is w i s e . " T h i s s y l l o g i s m — at once valid and u n s o u n d — i l l u s t r a t e s the l i m i t a t i o n s of s y l l o g i s m s as devices for reasoning to the t r u t h .

strict liability and negligence. U n d e r the f o r m e r the injurer is liable if he caused the accident, and usually it will be pretty straightforward to determine w h e t h e r he did cause it. U n d e r negligence the injurer is liable only if he could have avoided the accident by exercising due care. Even if " d u e care" is defined quite precisely, for example as taking precautions that cost less than the expected cost of the accidents they w o u l d prevent, there m a y be difficulty determining w h e t h e r this precondition of liability has been established. A rule, such as negligence, that requires a relatively broad factual inquiry to establish the m i n o r premise is often called a "standard" to distinguish it f r o m the simpler sort of rule in w h i c h the m i n o r premise states a single fact. This is not to say that the simpler sort of rule is always simpler to apply. T h e single fact m a y be difficult to determine or m a y not really be a fact ( " w a g e s " in m y earlier example o f an "easy" case); or the rule m a y be a m b i g u o u s . I suppress these complications for now, along w i t h m o s t of the trade-offs between rules and standards, 8 in order to focus on one point. A rule suppresses potentially relevant circumstances of the dispute (could the injurer have avoided the accident at reasonable cost?), while a standard gives the trier of fact—the j u d g e or j u r y — m o r e discretion because there are m o r e facts to find, weigh, and compare. T h e rule engenders tension w i t h the social policies that underlie it and that m a y be achieved only imperfectly w h e n the rule is applied w i t h o u t regard to the particular circumstances of the dispute. T h e standard solves that p r o b l e m — t h e p r o b l e m of achieving substantive justice rather than merely f o r m a l justice—but, b y vesting broad discretion in the officials applying it, it opens the w a y to abuse. (The solution m a y thus be illusory.) T h e m o r e flexible the criteria, the m o r e difficult it will be for an observer to determine w h e n the officials have applied t h e m intelligently and impartially. Rules create pressure for ad hoc exceptions, but standards could be t h o u g h t the very institutionalization of the ad h o c exception. In a regime of standards, the principles or policies that in a regime of rules w o u l d determine the content of the rules are used to determine the o u t c o m e of particular cases. Even if officials do not abuse the broad discretion that a regime of 8. See Isaac Ehrlich and Richard A. Posner, " A n E c o n o m i c Analysis of Legal R u l e m a k i n g , " 3 Journal of Legal Studies 257 (1974); A n t h o n y I. O g u s , " Q u a n t i t a t i v e Rules and J u d i cial Decision M a k i n g , " in The Economic Approach to Law 210 (Paul B u r r o w s and C e n t o G. Veljanovski eds. 1981); C o l i n S. D i v e r , " T h e O p t i m a l Precision of A d m i n i s t r a t i v e Rules," 93 Yale Law Journal 65 (1983); Pierre Schlag, " R u l e s and S t a n d a r d s , " 33 UCLA Law Review 379 (1985); cf. D u n c a n K e n n e d y , " F o r m and S u b s t a n c e in P r i v a t e Law A d j u d i c a t i o n , " 89 Harvard Law Review 1685 (1976).

standards gives them, standards have the disadvantage of making it harder to predict the outcome of a legal dispute—there are more variables, which makes litigation more likely and also makes it harder for people to plan and to conduct their affairs. Uncertainty is a source of cost and disutility and is increased by substituting standards for rules. A conspicuous—virtually a defining—characteristic of legal realism, critical legal studies, and feminist jurisprudence is depreciation of the benefits of rules in reducing uncertainty. Their slogan might be William Blake's dictum that "To Generalize is to be an Idiot. To Particularize is the Alone Distinction of Merit." 9 Yet no one really believes this. N o one thinks, for example, that the rule that school buses must stop, look, and listen at railroad crossings even if the warning gate is up is necessarily inferior to a standard that would require the driver merely to exercise due care in the circumstances. The rule guards against a lapse of j u d g ment by the bus driver, and this gain may well outweigh the cost in minuscule delay f r o m stopping every time. The general point illustrated by this example is that rules economize on information. T h e fact that the application of a standard requires the adjudicator to have more information than the application of a rule implies that the ratio of standards to rules will be greater, the lower the costs to courts of assimilating information. Perhaps over a long period of time there has been a general if irregular increase in the ability of courts to handle information; if so, this might explain what appears to be a historical shift in the balance between rules and standards, in favor of the latter. 10 Since syllogistic reasoning plays a smaller role in determining the outcome of a dispute governed by a standard than one governed by a rule, we may have a clue to the widely perceived decline (welcomed in some quarters, deplored in others) of legal formalism. A bigger danger to the idea that law can be made determinate by syllogistic logic impends when we switch our focus f r o m the minor premise—the facts—to the major premise—the rule itself. T h e source of 9. " A n n o t a t i o n s to Sir J o s h u a R e y n o l d s ' s D i s c o u r s e s , " in The Complete liam Blake 445, 451 ( G e o f f r e y K e y n e s ed. 1958).

Writings of Wil-

10. See m y b o o k The Economics of Justice, chs. 6 and 7 (1981). It is n o t a steady p r o g r e s sion. T h e r e was a m o v e m e n t f r o m s t a n d a r d s to rules in E n g l i s h l a w in the n i n e t e e n t h century, see O g u s , n o t e 8 above, at 216-217; there is m o r e use of strict liability in t o r t law today than there was t h i r t y years a g o ( a l t h o u g h f o r reasons largely unrelated to a desire to m a k e law simpler and m o r e certain); and the recently p r o m u l g a t e d Federal s e n t e n c i n g guidelines curtail the traditional discretion of j u d g e s in deciding h o w l o n g a sentence to give a convicted d e f e n d a n t w i t h i n the limits set b y the legislature. Indeed, there is a g r o w ing sense that the legal s y s t e m has g o n e t o o far in the direction of s u b s t i t u t i n g s t a n d a r d s for rules and is due for r e t r e n c h m e n t .

the danger is not, as one m i g h t think, that rules have exceptions. That in itself is n o t problematic, provided the exceptions are themselves rules, as n o r m a l l y they will be. T h e statute of limitations in Illinois for bringing suit o n a written contract is ten years. O n e exception is if the defendant misled the plaintiff into thinking he had m o r e than ten years to sue. T h e main rule will get us as far as concluding that if a plaintiff waits m o r e than ten years to bring suit on a written contract, the defendant will have a prima facie defense that the statute of limitations has run. We m a y then have to start over and consider w h e t h e r the plaintiff has rebutted the defense by establishing the applicability of one of the exceptions. But as each exception is itself a rule, and a rule that cleanly defeats the main rule, the plaintiff has only to prove the exception's m i n o r premise (for example, that the defendant had in fact misled him) to establish that the suit is timely. T h e problems with governance by rule lie elsewhere; they are familiar but deserve emphasis. O n e is that the d o m a i n of a rule m a y be uncertain, and then j u d g e s will continually have to decide w h e t h e r to apply the rule to situations n o t foreseen, or at least left open, w h e n the rule was laid d o w n . Rules m a k e d i c h o t o m o u s cuts in continuous p h e n o m e na. A rule that punishes breaking and entering at night m o r e severely than breaking and entering in d a y t i m e is perfectly clear at the semantic level, but by disregarding the actual relation between day and night (one of imperceptible blending) the rule fails to track the reality being described, creating a m b i g u i t y in application. T h e m o r e general point is that legal rules frequently treat as referential w o r d s (like " d a y " and "night") that do n o t have a definite referent. T h e First A m e n d m e n t forbids Congress to establish "religion" or interfere with its free exercise. Religious organizations also are e x e m p t f r o m state property taxes, and contributions to religious organizations are tax deductible. But what is "religion"? T h e w o r d does n o t denote s o m e closed o r definite set of practices or objects or entities. In the absence of a legislative definition j u d g e s have to decide w h a t the w o r d shall mean for legal purposes, not what it does mean. T h e decisions that j u d g e s are required to m a k e in fitting rules to facts can be described either as interpretation or as the m a k i n g of ad hoc exceptions and adjustments, in effect the continual remaking of the rules. Practically these are the same b u t j u d g e s prefer the first f o r m u l a tion, the interpretive, because it casts t h e m in a less creative, and therefore less usurpative-seeming, role. T h e older the rule, and the m o r e d y n a m i c the activity governed by it (that is, the greater the change over time in the activities to which the rule refers), the m o r e pressure there will be on the j u d g e s to make ad

hoc exceptions and ad h o c extensions. Indeed, the m o r e d y n a m i c the regulated activity, the less amenable it will be to governance by rules at all, as distinct f r o m standards. T h e choice in international trade theory between fixed and floating exchange rates provides an analogy. M a n y economists believe that fixed exchange rates mask rather than contain underlying e c o n o m i c changes, resulting in a substitution of a b r u p t for s m o o t h adjustments. T h e r e m a y be analogous consequences w h e n rules are chosen over standards to g o v e r n variegated and changing activities. Another p r o b l e m with rules is that inconsistent rules m a y be applicable to the same activity. H e r e logic reasserts its claims. Logically inconsistent rules cannot be applied to the same activity. T h e j u d g e has a duty to eliminate the inconsistency. But logic does n o t tell h i m w h i c h rule to discard. M a n y legal rules, moreover, are j u d g e - m a d e , and they can be j u d g e unmade. T h e c o m m o n law is a vast collection of j u d g e - m a d e rules, and much of statutory and constitutional law also consists of j u d g e - m a d e rules, loosely tethered to debatable interpretations of a m b i g u o u s enactments. T h e estoppel exception to the statute of limitations is j u d g e made, and the j u d g e s could u n m a k e it if persuaded that such an exception was unnecessary in the case of a long statute of limitations and produced too m u c h uncertainty and litigation. As l o n g as either the rule or its exceptions are contestable, the neat logical pattern of rule and exception will not resolve all cases even if the rule is b o t h clear and consistent with all related rules. Adherence to a rule is, as Wittgenstein famously explained in Part I of Philosophical Investigations, n o t a dictate of logic; the rule does n o t tell you w h e n to follow it. We thus have the paradox that a legal question m i g h t be at once determinate and indeterminate: determinate because a clear rule covers it, indeterminate because the j u d g e is not obligated to follow the rule. This makes a legal rule a little like a natural law. A n d it supports H o l m e s ' s view that the law is really j u s t a prediction of w h a t j u d g e s will do w i t h a given set of facts, because the j u d g e s are n o t bound b y the rules to d o anything. Finally, decision according to rule is n o t inherently m o r e objective than decision according to standard; the principled and the ad hoc are not, in law, as polar as the terms connote. Consider the f o l l o w i n g choices for a regime to govern First A m e n d m e n t challenges to offensive art: a rule i m m u n i z i n g all such art f r o m censorship or other regulation; a rule giving such art no protection; and a standard requiring the j u d g e s to "balance"—employing weights inevitably influenced by their p e r s o n al values—offensiveness against artistic quality and to authorize censorship w h e n the balance inclines decisively against the artist. Each of these

choices has as m u c h , or as little, g r o u n d i n g in the text and history of the First A m e n d m e n t , and in considerations of institutional c o m p e tence, as any of the others. If one of the t w o rules (which happen to be opposites) is adopted instead of the standard, decisions will appear to be strongly objective because logically deducible. But the appearance will fade as soon as the contestability of the premise is noticed, enabling the observer to see that the decisions are n o m o r e objective in an ultimate sense than those m a d e u n d e r the m o r e frankly ad hoc regime of the balancing test. Rules m a s k — t h e y d o not eliminate and m a y n o t even reduce—the role of the subjective and the political in the f o r m a t i o n of legal rights and duties. As w e b e c o m e m o r e realistic about rules, their advantages vis-a-vis standards begin to dim. T h e principal advantages are curtailing official discretion—the discretion of the officials w h o administer the rules, but not of those w h o m a k e t h e m — a n d reducing uncertainty about legal rights and duties. These advantages depend, h o w e v e r , on having rules that are relatively f e w and simple, relatively immutable, and clear in b o t h statement and application, and these are ideals that a system of legal rules ,rarely attains. E v e n if the rules are clear, moreover, they may be so n u m e r o u s that the people subject to t h e m cannot learn them; then their clarity is delusive. Standards that capture lay intuitions about right behavior (for example, the negligence standard) and that therefore are easy to learn m a y p r o d u c e greater legal certainty than a n e t w o r k of precise but technical, nonintuitive rules covering the same g r o u n d . N o t o riously, legalistic rules are traps f o r the unwary. A n d m a n y j u d g e s sense this. T h e school bus example shows, however, that it w o u l d be a mistake to d e n o u n c e rules tout court, or to expect or desire j u d g e s to subvert t h e m at every turn. Realism about rules includes, moreover, a recognition that j u d g e s consider stability valuable and therefore often refuse to reexamine rules to m a k e ad h o c exceptions to t h e m even w h e n confronted by p o w e r f u l claims of substantive justice. But j u d g e s ' reasons for generally sticking b y rules have little if anything to d o w i t h logic and often are n o t "reasons" at all. W i t h o u t g o i n g so far as to argue that a judge's decisions are influenced b y his digestion, or that j u d g e s w h o are sticklers for rules have an "authoritarian personality" and w o u l d benefit f r o m psychoanalysis," I suggest that differences in t e m p e r a m e n t are indeed

11. See J e r o m e Frank, Law and the Modern Mind (1930). T h e first part of his conjecture, a b o u t the a u t h o r i t a r i a n personality, is m o r e p e r s u a s i v e t h a n the s e c o n d — t h a t such p e r s o n alities are a m e n a b l e to p s y c h o t h e r a p y .

responsible for much of the evident difference a m o n g judges in willingness to interpret rules flexibly, recognize exceptions freely, transmute rules into standards willingly, and cede discretion to juries and to lowercourt judges comfortably. 1 2 Some judges have greater tolerance for untidiness, even disorder, than others. They see law as exploration and dialogue rather than as governance and hierarchy. Such judges may be highly sensitive to the particulars of each case, which is good, but insufficiently sensitive to the costs (mainly hidden) of legal uncertainty. O t h er judges are uncomfortable about assuming responsibility for decisions. They want to think that they are just carrying out the orders of other people (legislators, the founding fathers, higher or earlier judges). There are well-known dangers with the deferential posture 1 3 —but the P r o m e thean j u d g e is no prize either. These differences in judicial outlook and behavior reflect different j u d g m e n t s and life experiences, but they are also temperamental in the sense that they are deeply rooted in the judge's psyche and therefore resistant to experience. Let us look a little more closely at the character of legal rules, before turning again to logic and formalism. The contestability of legal rules stands in contrast to that of t w o other kinds of rules, the rules of games and the rules of language. Although the rules of a game are changed f r o m time to time, it is unthinkable (in a "serious" game) to change a rule during the game. Tennis players are not overheard saying to the referee, "It's true m y ball landed behind the baseline, but if one thought about the matter carefully one would have to conclude that the baselines are too shallow and should be extended, so m y ball ought not be deemed out of bounds." There is of course an open area in a game, an area left to the strategy and tactics of the players. T h e rules do not determine the game's outcome. But they constrain the players' play and by

12. For s o m e evidence, see Philip E. Tetlock, J a n e B e r n z w e i g , and Jack L. Gallant, " S u p r e m e C o u r t Decision M a k i n g : C o g n i t i v e Style as a P r e d i c t o r o f Ideological C o n s i s tency of V o t i n g , " 48 Journal of Personality and Social Psychology 1227 (1985); see also C h a p ters 4 and 5 of this b o o k . A d m i t t e d l y the j u r y e x a m p l e is equivocal. A f o r m a l i s t m i g h t w a n t to cede discretion to j u r i e s in o r d e r to a v o i d h a v i n g to exercise it himself. J u r y discretion is a licensed area of lawlessness in law, the existence of w h i c h p e r m i t s the j u d g e to decide all issues w i t h certainty in t h e (diminished) s p h e r e in w h i c h he is t h e d e c i s i o n m a k e r . N o t i c e h o w a c o n s e r v a t i v e f o r m a l i s t , d i s t r u s t f u l o f j u r i e s b u t eager to divest h i m s e l f o f discretion, m i g h t feel t o r n o n this issue. 13. See Stanley M i l g r a m , Obedience to Authority: An Experimental View (1974), f i n d i n g that highly obedient people, b y casting t h e m s e l v e s as t h e i n s t r u m e n t s o f another, tend to slough o f f responsibility for their actions, h o w e v e r e g r e g i o u s t h o s e actions m a y be. O n e is put in m i n d of Churchill's epithet f o r the G e r m a n soldiery in W o r l d War II: "lethal s h e e p . "

doing so make it a g a m e and not a free-for-all. A n d the rules that do this are fixed, definite, and at least in the course of the game, immutable. 1 4 Legal rules likewise are enabling as well as repressive. Just as the rules of chess enable a subtle and challenging g a m e to be played, rather than simply i m p o s i n g i r k s o m e and repressive constraints on the players, so (for example) the rules of p r o p e r t y law facilitate investment and transacting, the rules of contract law enable the m a k i n g of c o m m i t m e n t s to b u y and sell that u n f o l d over time rather than being p e r f o r m e d at the same instant, and the rules of criminal law enable people to go about their business u n a r m e d and w i t h o u t guards. But these rules do n o t so completely define the activities w h i c h they enable or facilitate that those activities lose their p u r p o s e if the rules are changed while the activities are in progress. A chess player w h o announces in the middle of the g a m e that he will n o t allow his r o o k to be captured b y his opponent's queen, because queens have too m u c h p o w e r in chess, m i g h t j u s t as well sweep all the pieces f r o m the board; and it w o u l d m a k e n o t the slightest difference w h e t h e r the u m p i r e agreed w i t h h i m , or w h e t h e r the player waited until he lost the g a m e and then asked the u m p i r e to order it replayed u n d e r rules reflecting a m o r e equitable distribution of p o w e r a m o n g the pieces. These sorts of appeals are allowed in law, however. Therefore, paradoxical as it m a y seem, the d u t y to play by the rules is m o r e imperative in games than in law; for in law the rules m a y be up for grabs. N o t all of them, of course; not even most of t h e m . Yet j u d g e s do have a limited license to " d i s o b e y " rules by changing t h e m in m i d c o u r s e . " We could eliminate the paradox by saying that every time the j u d g e s break a rule they are d o i n g so in order to comply w i t h a higherlevel rule, or that if j u d g e s are s o m e h o w licensed to break rules this s h o w s that w h a t they are doing is n o t really breaking rules. But this

14. So w e m i g h t c o n t r a s t n o r m a t i v e rules w i t h a " r u l e of t h u m b " rule that guides but does not constrain play; such a rule c o r r e s p o n d s to a recipe in c o o k i n g . For a useful t y p o l o g y o f rules, sec D a v i d K. Lewis, Convention: A Philosophical Study 100-107 (1969). C o u n t e r p a r t s in law to the principle of n o t c h a n g i n g the rules in medias res include "stare decisis" ( " s t a n d b y w h a t has been d e c i d e d " — t h a t is, f o l l o w precedent) and " l a w of the case" (in general, c o u r t s will n o t c h a n g e the applicable rules of l a w d u r i n g the course of a l a w suit); res j u d i c a t a , w h i c h (to o v e r s i m p l i f y ) p r e v e n t s the relitigation of a claim b e t w e e n the s a m e parties, p r o v i d e s a m o r e distant analogy. Stare decisis is far f r o m rigid; law of the case is flexible t o o , a l t h o u g h less so than stare decisis. M o r e o v e r , the d o c t r i n e of law of the case is n o t that the c o u r t will refrain f r o m c h a n g i n g the applicable l a w b e t w e e n the dispute giving rise to the l a w s u i t and the decision r e s o l v i n g the dispute, b u t o n l y that it will refrain f r o m c h a n g i n g that law b e t w e e n the first decision in the case and s u b s e q u e n t decisions. 15. See M o r t i m e r R. Kadish and S a n f o r d H . Kadish, Discretion to Disobey: Lawful Departures from Legal Rules 8 5 - 9 1 (1973).

A Study of

would just be w o r d play and w o u l d n o t change the fact that j u d g e s w i t h some frequency decide to discard or to change w h a t until then had seemed a valid rule. T h e y m a y have g o o d reasons to d o this, but c o m pliance with a higher-level rule is n o t a g o o d reason; it is merely a fancy way of stating the conclusion. A better analogy to rules of law m a y be "rules" of h o w to w i n a game. I mentioned strategy and tactics in chess. A related but simpler example is skiing. T h e beginner learns various rules about turning, stopping, and the like, but he is " a l l o w e d " to break t h e m w h e n skiing and often he has to, to avoid falling. T h e rules are really guides, and m a y b e this is what rules o f law are to j u d g e s . A f u r t h e r analogy is to the labor practice k n o w n as " w o r k i n g to rule." T h e w o r k e r s decide to c o m p l y fully with every j o b rule—and the result is a disastrous s l o w d o w n , because the rules were n o t meant to be obeyed to the letter. T h e j u d g e w h o takes rules too seriously may be a bureaucratic stickler. Still another analogy to rules of law is " l a w " in its scientific rather than legal or moral sense. If a n o n c o n f o r m i n g instance is f o u n d , and it cannot be ascribed to mistakes in observation or experimental procedure, the law has to be changed; a scientific law is descriptive, n o t prescriptive. Similarly, if the j u d g e s decide n o t to apply a rule b u t instead to act contrary to it, in effect they have forced a change in the rule, as within broad limits they are licensed to do. T h e principal difference, so far as is relevant here, b e t w e e n law and science on the one hand, and games on the other, is that law and science are concerned w i t h right results, while the function o f the rules of a game is to provide a f r a m e w o r k for c o m p a r i n g and appraising the contestants' p e r f o r m a n c e in the game. Law, h o w e v e r , unlike science, is c o n cerned not only with getting the result right b u t also w i t h stability, to which it will frequently sacrifice substantive justice. In short, all three activities—law, science, and g a m e playing—have different purposes, and these purposes affect the attitude of the "players" t o w a r d obedience to rules. Language resembles law in that its rules are mutable and change in use, but language is m o r e stable. Because n o o n e (except the French Academy!) is specifically authorized to change language rules, they tend to change at the glacial pace typical of customs. T h e judicial system, being hierarchical within each jurisdiction, has the p o w e r to change the law of the jurisdiction abruptly. T h a t the judiciary seldom exercises the p o w e r — a n d therefore m a y n o t really possess it, because of political constraints on j u d g e s — d o e s not make legal outcomes the result of a logical process. Those outcomes may be predictable because of the k n o w n p r o -

pensity of the j u d g e s to stick b y the existing rules—the propensity called stare decisis. But because those rules are subject to reexamination, the predictable o u t c o m e s are n o t compelled b y logic. T h e analogy is to the Socrates syllogism if spoken b y the lexicographical dictator in Nineteen Eighty-Four (O'Brien), w h o could have redefined " m a n " and " m o r t a l " at a stroke o f the pen. A n o t h e r feature c o m m o n to law and language is that b o t h are, in a sense, b e y o n d right and w r o n g . O u r language is full of errors that are n o longer subject to being corrected. For example, the collective n o u n "pease," a w o r d parallel to " w h e a t " and preserved in the nursery r h y m e "pease p o r r i d g e h o t , " was long ago misunderstood to be a plural n o u n , so " p e a " was "inferred" to be the singular and "peas" was constructed to be the "regular" plural. Hamlet's phrase " m o r e h o n o r e d in the breach," referring to a c u s t o m m o r e h o n o r a b l e to avoid than to f o l l o w — the specific referent was excessive drinking—has been misread to refer to a c u s t o m that is n o t observed, and this error, like "pea" and "peas" is a secure, an unshakable, part of our language. Because law, like language, values stability, legal mistakes, too, can b e c o m e sanctified by time and usage. All this is not to deny that there are legal rules in a useful sense. I have n o quarrel w i t h H . L. A. Hart's classic statement: We promise to visit a friend the next day. When the day comes it turns out that keeping the promise would involve neglecting someone dangerously ill. The fact that this is accepted as an adequate reason for not keeping the promise surely does not mean that there is no rule requiring promises to be kept, only a certain regularity in keeping them. It does not follow from the fact that such rules have exceptions incapable of exhaustive statement, that in every situation we are left to our discretion and are never bound to keep a promise. A rule that ends with the word 'unless . . .' is still a rule."'

A n d one m i g h t add that the fact that a rule is not eternal does not mean that it is n o t a rule while it lasts. But it is significant that Hart's example is described f r o m the standpoint of people subject to rules rather than f r o m that of the rule makers. T h e " w e " in the example are n o t the people w h o m a d e or can u n m a k e the rule that promises should be kept. F r o m the perspective of persons subject to the rules, the fact that rules have fuzzy edges or may be changed does n o t m a k e t h e m n o n b i n d i n g . B u t f r o m the standpoint of the rulers, rules m a y be so m u c h putty. T h e rulers m a y have complete,

16. The Concept of Law 136 (1961).

or at least extensive, discretion to repeal or revise or "interpret" the rules; this is one reason w h y sophisticated members of the legal profession are more aware of legal indeterminacy than are laypeople. In Hart's example, the rule about keeping promises is just there. There is n o m e n tion of its creation, and this omission obscures the fact that the people w h o create rules have a different relationship to them than do the people w h o are penalized if they disobey them and have no power to revise or repeal them. The sanctions for not enforcing a rule are not the same, and normally are not as severe, as those for not obeying a rule. We like to think of judges as being under the law, that is, as being ruled rather than ruling. But besides the point just made about the different sanctions for nonenforcement and noncompliance, judges can, within broad limits, change the rules, particularly those of the c o m m o n law and of constitutional law. This power is consistent with the fact that rules do not have to be pellucid, definite, or incapable of revision or repeal in order to bind in a meaningful sense the persons subject to them. It is merely that the judges are not subject to the rules in the same sense—which is not to deny, however, that even j u d g e - m a d e rules carry weight with judges, that is, that there really is a policy, although a flexible one, of adhering to precedent, in order both to stabilize law and to economize on judges' time.' 7 Mention of legal stability brings us face to face once again with the limitations of logic in law. T h e appearance of stability is highly sensitive to the interval over which a body of rules is examined. At any m o m e n t the law as a whole is relatively stable, because only a few rules are on the agenda of change. But expand the time period, and the picture changes. The bodies of rules that make up such otherwise disparate fields as tort law, constitutional law, and antitrust law have changed greatly over the last century, almost entirely as a result of changes in judge-made rules. T h e process by which the judges have changed these rules—or, more bluntly, have created the rules under which we n o w live—is not a logical process, not a matter of correcting errors in syllogistic reasoning made by their predecessors. To pick one of many possible examples, the vast expansion in tort liability in this century cannot be explained in terms of logic. It may have to do with a sense that accidents are more readily avoidable by potential injurers than was once the 17. For a c k n o w l e d g m e n t of this p o i n t f r o m an u n e x p e c t e d source, see D u n c a n Kennedy, " F r e e d o m and C o n s t r a i n t in A d j u d i c a t i o n : A Critical P h e n o m e n o l o g y , " 36 Journal of Legal Education 518 (1986). T h e " w e i g h t " m e t a p h o r s h o u l d n o t be taken literally, e v o k i n g as it docs an exact f o r c e — g r a v i t y — w h i c h has n o close c o u n t e r p a r t i n j u d i c i a l incentives or c o n straints.

case (and perhaps less avoidable by potential victims), or w i t h a greater capacity of courts to process cases, or w i t h a fall in the costs to accident victims of bringing suit, or w i t h a g r o w i n g enthusiasm for social insurance; it m a y have to d o above all w i t h a drive t o w a r d m a k i n g law m o r e efficacious—the "stage 3" in the g r o w t h of law discussed in the Introduction. Logic doesn't enter. It doesn't enter into the law's cycling between generous and niggardly protection of criminal defendants' rights, or between generous and niggardly protection of patent o w n e r s ' rights, or between generous and niggardly protection of e m p l o y e r s ' rights. W h e n j u d g e s and lawyers d o use logic, they use the simplest m e t h o d s . Formal logic is rarely taught in law schools or encountered in judicial opinions, briefs, or law review articles. ( N o t too m u c h should be m a d e of this, however. Formal logic is rarely f o u n d outside b o o k s and articles on logic. I a m not advocating f o r m a l notation for judicial opinions.) T h e closest thing to f o r m a l logic in law is the use by economic analysts of mathematical models of legal rules; the m o d e l of these models is Learned Hand's f o r m u l a , w h i c h defines negligence in a simple algebraic f o r m u l a . ' 8 As for the syllogism, it should be apparent by n o w that it is an unhelpful template for legal reasoning. Its function is to demonstrate the validity of a reasoning process rather than to establish the truth of the o u t c o m e of the process. It is used n o t to prove that Socrates is m o r tal—for n o one d o u b t s t h a t — b u t to d e m o n s t r a t e a logical relation, as w o u l d be clearer if w e substituted A for m a n (men), B for mortal, and C for Socrates. 1 9 Logic, like mathematics, explores relations between ideas rather than correspondence to facts. T h e legal system cannot be indifferent to issues of empirical truth. Despite appearances, it is n o t m y intention to disparage the use of logic in law. Even in difficult cases it plays a role, a critical role. Bertrand Russell put the basic point well: "Logical errors are, I think, of greater practical i m p o r t a n c e than m a n y people believe; they enable their pcr18. U n i t e d States v. C a r r o l l T o w i n g C o . , 159 F.2d 169, 173 (2d Cir. 1947) (B < PL, m e a n i n g that if the b u r d e n to the i n j u r e r of a v o i d i n g the accident was less than the loss if the accident o c c u r r e d d i s c o u n t e d [that is, m u l t i p l i e d ] b y the p r o b a b i l i t y that it w o u l d occur, the i n j u r e r is negligent). C o m p a r e A m e r i c a n H o s p i t a l S u p p l y C o r p . v. Hospital P r o d u c t s Ltd., 780 F.2d 589, 593 (7th Cir. 1986), and see generally William M . Landes and Richard A. Posner, The Economic Structure of Tort Law (1987). 19. T h e fact that A stands for t w o different w o r d s in the Socrates syllogism is a clue to its n o n s t a n d a r d status, o n w h i c h see Willard Van O r m a n Q u i n e , Methods oj Logic 107, 2 5 9 267 (4th ed. 1982). In c o n t r a s t , the s y l l o g i s m "All Greeks are d a r k - h a i r e d ; all C r e t a n s arcGreeks; t h e r e f o r e all C r e t a n s are d a r k - h a i r e d " causes n o difficulty w h e n put in the form "All A are B; all C are A; t h e r e f o r e all C are B . "

petrators to hold the c o m f o r t a b l e opinion on every subject in turn. A n y logically coherent b o d y of doctrine is sure to be in part painful and contrary to current prejudices. T h e dialectic m e t h o d — o r , m o r e generally, the habit of unfettered discussion—tends to p r o m o t e logical consistency, and is in this w a y useful. B u t it is quite unavailing w h e n the object is to discover n e w facts." 20 T h e qualification in the last sentence is, as w e shall see in the next section, highly relevant. Here is an example of h o w logic can be used critically in law. T h e Supreme C o u r t has held that the C o n s t i t u t i o n does not forbid a state to make s o d o m y between consenting h o m o s e x u a l adults a crime. 2 1 T h e C o u r t emphasized the long history of criminalizing such conduct, implying that conduct long punished criminally cannot find shelter in the concept of "privacy" that the C o u r t has claimed to have discovered in the Constitution, no matter h o w private the conduct is. But, if so, then the C o u r t ' s earlier decisions, w h i c h Bowers v. Hardwick does not purport to overrule, holding that w o m e n cannot be f o r b i d d e n to have abortions or married or u n m a r r i e d couples to practice contraception, seem unsound, for those practices t o o had l o n g been forbidden. So there is a potential inconsistency in the m a j o r i t y opinion. T h e dissenting J u s tices likewise ignored a potential inconsistency in their position. T h e y treated homosexual s o d o m y as n o different in any relevant respect f r o m heterosexual sex acts, and in particular f r o m heterosexual sodomy. If there is no socially or legally pertinent difference between h o m o s e x u a l and heterosexual activity, a state that refuses to recognize h o m o s e x u a l marriage is discriminating arbitrarily. N o n e of the dissenting Justices was willing to confront, let alone embrace, this implication of their stated position. If, as I suspect, these Justices d o not believe that the Constitution requires the recognition of h o m o s e x u a l marriage, then they do not really believe that h o m o s e x u a l s o d o m y is relevantly identical to heterosexual sodomy. Logic will not decide the m o s t difficult cases. Logic the destroyer is not logic the creator. To s h o w that an opinion is illogical is not to s h o w that the o u t c o m e is incorrect—a particularly pertinent observation w h e n both the majority and the dissenting opinions in a case are illogical. M a y b e the earlier privacy decisions should have been overruled in Bowers; or m a y b e h o m o s e x u a l marriage should be legalized. If p r o p o sitions A and B are inconsistent, one of t h e m has to go, b u t logic will not tell you which one. It can, however, kick the props out f r o m under 20. A History of Western Philosophy 93 (1945). 21. B o w e r s v. H a r d w i c k , 478 U . S . 186 (1986).

many of the arguments offered in cases and put pressure on the opponents to argue anew, and specifically to bring into the open the premises necessary to resolve the inconsistencies, for I am not denying the possibility that the inconsistencies I have identified might be dispelled in a fuller analysis. 22 The very importance of the critical use of logic in law, however, should help us to understand w h y lawyers and judges overuse logic. Law attracts and rewards people w h o are good at handling informal logic (not formal logic—the logic of lemmas and the predicate calculus and the like; that attracts a different kind of mind), and not all of them are content to use their logical skills to refute other lawyers' and judges' arguments. They want to make law as well as criticize law, and to do so with the tools they k n o w best, the tools of simple logic. All this emphasis on judges and litigation may bore or irritate lawyers w h o specialize in fields in which litigation is not the central focus. There are many such fields, and they tend to get short shrift in books of Anglo-American jurisprudence; they will in this one. The judge is a more problematic figure than the lawyer w h o negotiates a lease or other contract against a background of settled and slowly changing law or the arbitrator w h o decides a dispute arising f r o m such a contract. But there is one field in which litigation is not the focus of practice yet jurispru-

22. T h e o m i s s i o n of essential premises is of a piece w i t h the generally evasive character of the Bowers o p i n i o n s , w h i c h , w h i l e m e n t i o n i n g traditional d i s a p p r o v a l and sporadic p e r secution of h o m o s e x u a l s , disregard such potentially relevant facts as that (1) s o d o m y laws are n o l o n g e r e n f o r c e d in this c o u n t r y (there was an arrest b u t n o p r o s e c u t i o n in Bowers itself); (2) in the U n i t e d States and Western E u r o p e male h o m o s e x u a l s w e r e at the t i m e Bowers was decided the principal k n o w n carriers, as well as victims, o f the virus that causes the ghastly disease A I D S , a l t h o u g h their role as t r a n s m i t t e r s of the virus m a y be declining relative to that of d r u g addicts; (3) there is w i d e s p r e a d revulsion a m o n g heterosexuals t o w a r d specific sexual practices of h o m o s e x u a l s and t o w a r d the f l a u n t i n g of t h o s e practices and of the life style associated w i t h t h e m ; (4) a visible and significant n u m b e r of male h o m o s e x u a l s are h i g h l y p r o m i s c u o u s (this is connected w i t h p o i n t 2), a l t h o u g h the n u m b e r has d i m i n i s h e d in response t o the A I D S epidemic; (5) the biological and psychiatric q u e s tion of w h e t h e r and if so to w h a t e x t e n t sexual o r i e n t a t i o n is v o l u n t a r y is u n r e s o l v e d ; (6) h o m o s e x u a l s a p p e a r t o be a b o v e a v e r a g e in e d u c a t i o n , i n c o m e , and artistic and intellectual creativity; (7) they are increasingly assertive politically; (8) societies differ greatly in their attitudes t o w a r d h o m o s e x u a l i t y , w i t h s o m e b e i n g tolerant. For discussion and references, see Michael Ruse, Homosexuality: A Philosophical Inquiry (1988). I a m far f r o m certain of the exact b e a r i n g of all these facts, b u t I have difficulty u n d e r s t a n d i n g a responsible j u d g e ' s failing to consider t h e m in t r y i n g to decide w h e t h e r the C o n s t i t u t i o n s h o u l d be c o n s t r u e d to limit public r e g u l a t i o n of h o m o s e x u a l sex acts. O f course, the Justices m a y have t h o u g h t a b o u t these t h i n g s — m a y even have discussed t h e m a m o n g t h e m s e l v e s ( t h o u g h I d o u b t t h a t ) — b u t decided n o t to include t h e m in the o p i n i o n s . A p r u d e n t C o u r t , I believe, w o u l d have dismissed the case as m o o t in v i e w of the state's decision n o t to prosecute the plaintiff, and n o t have decided the m e r i t s .

dential problems highly pertinent to this chapter abound, and that is federal taxation. Tax legislation and regulations issue with such frequency and in such volume f r o m Congress and the Treasury Department that the judicial function of filling gaps in legislative and related texts is rarely employed. Yet the price of this approach is high, and should give pause to those inclined to agree with Bentham that the problem of judicial legitimacy could be solved by codification. Carefully drafted though they typically are, tax statutes and regulations by their very precision and exactitude create loopholes that lawyers are quick to exploit for their clients. The loopholes incite amendatory legislation or regulation. The result is a structure with fewer unintended loopholes than generally supposed, but it is a structure of immense complexity. Whether it is a net improvement on a simpler structure that would make greater use of the courts to plug loopholes is anyone's guess. The experience with taxation, an institution that conservatives do not admire, casts a sidelight on the position of Hayek and his followers that the rule of law implies governance by clear rules, k n o w n to all in advance and not subject to change through judicial interpretation, which operates retroactively. 23 A society in as much ferment as ours needs a mechanism for legal change, and it is not obvious that the combination that predominates in federal income taxation—extremely detailed statutory specifications with frequent amendments and continuous executive rule making—is always to be preferred to judicial administration of flexible standards. It is odd that conservatives in the libertarian tradition should want to expand the legislative and administrative roles in our society. Joseph Isenbergh has mounted a strong attack on the j u d g e - m a d e doctrine of tax law that authorizes the tax collector to disregard the f o r m of a transaction and penetrate to its substance in an effort to prevent the 23. See, for e x a m p l e , F. A. H a y e k , The Constitution of Liberty, eh. 10 (1960); H a y e k , Law, Legislation, and Liberty, vol. 1 (1973); N o r m a n P. Barry, Hayek's Social and Economic Philosophy, ch. 5 (1979); J a m e s M . B u c h a n a n , " G o o d E c o n o m i c s — B a d Law," 60 Virginia Law Review 483 (1974); B u c h a n a n , " C o n t r a c t a r i a n Political E c o n o m y and C o n s t i t u t i o n a l I n t e r p r e t a t i o n , " 78 American Economic Review Papers and Proceedings 135 ( M a y 1988); R a n d y E. Barnett, " F o r e w o r d : C a n Justice and the R u l e of L a w Be R e c o n c i l e d ? " 11 Harvard Journal of Law and Public Policy 597 (1988). H a y e k ' s hostility to legal s t a n d a r d s is related to the e x t r e m e skepticism of the school of " A u s t r i a n e c o n o m i c s , " of w h i c h he is t h e p r e e m i n e n t e x p o n e n t , regarding the reliability of i n f o r m a t i o n g e n e r a t e d b y g o v e r n m e n t rather t h a n b y markets. Standards, as I h a v e n o t e d , require m o r e i n f o r m a t i o n t o a d m i n i s t e r t h a n rules, and in that sense i m p l y a potentially m o r e i n t r u s i v e role f o r g o v e r n m e n t t h a n regulation by rules does. T h e " A u s t r i a n " p o s i t i o n resembles that of t h e advocates of f o r m a l , or p r o cedural, j u s t i c e (such as Lon Fuller), w h o s e p o s i t i o n I discuss in C h a p t e r s 7 and 11.

exploitation of unintended loopholes in tax statutes and regulations. 2 4 After reviewing a n u m b e r of cases in w h i c h he believes the "substance over f o r m " doctrine has been misapplied, Isenbergh concludes that " m o r e parsing of the statute and specific transactions and less concern w i t h h o w to save the w o r l d f r o m manipulative taxpayers w o u l d have led to sounder holdings in all these cases" (p. 879). F r o m there Isenbergh m o v e s to larger g r o u n d and argues for w h a t he describes as a formalistic approach to interpretation in general. " H a r d grappling w i t h the facts of a case and the inner w o r k i n g s of a statute, although b o t h difficult and intellectually admirable, is frequently passed off as a trivial or excessively ' f o r m a l ' exercise. For o n e w h o has gotten that far, the slogan o f ' s u b stance over f o r m ' is as g o o d a means as any to clear the intellectual landscape for an inquiry about the 'larger' nature of the statute itself. T h e latter exercise is in fact quite easy, requiring only the assertion of a statutory p u r p o s e that encapsulates one's o w n tastes, either generally or regarding the transaction under scrutiny" (id.). W h y do j u d g e s adopt such an approach? Isenbergh has already given an answer that m i g h t have seemed sufficient: j u d g e s are s o m e t i m e s lazy, and they are often at sea w h e n dealing w i t h c o m p l e x commercial transactions. N o w he adds: "Judges have aspirations. Little attention is d r a w n to those w h o h e w n a r r o w l y to technical rules. T h e painstaking process of examining transactions and statutes to determine w h e t h e r they concord promises little glory. In a society that has always looked to courts for strokes of statesmanship, it is easy e n o u g h to understand a judge's temptation to cut t h r o u g h , rather than unravel, the Gordian k n o t " (p. 882). 25 So at the root of the "substance over f o r m " doctrine lies—judicial ambition (in tax law?). A n d "on the whole, l a w professors do n o t help m u c h either. It is their w o n t to decry ' f o r m a l i s m ' and glorify the ends of 'policy' in the resolution of disputes" (id.). Isenbergh is not a consistent formalist. H e derides "the enshrinement of substance over f o r m " as a " ' m a x i m ' of statutory construction. As w i t h its older relatives . . . it is harder and harder to k n o w which way it will cut w h e n it surfaces" (p. 879). T h e canons of statutory construction are the main props of formalistic reasoning in statutory cases. N o r does the approach that Isenbergh favors—careful scrutiny of the details 24. " M u s i n g s o n F o r m and S u b s t a n c e in T a x a t i o n , " 49 University of Chicago Law Review 859, 8 6 3 - 8 8 4 (1982). S u b s e q u e n t page references to this article appear in the text. For a w a r m e r view of t h e d o c t r i n e , see M a r v i n A. Chirelstein, " L e a r n e d H a n d ' s C o n t r i b u t i o n to the Law of Tax A v o i d a n c e , " 77 Yale Law Journal 440 (1968). 25. T h e m e t a p h o r is n o t well chosen. T h e G o r d i a n k n o t couldn't be u n t i e d — i t had to be cut, as only A l e x a n d e r the Great u n d e r s t o o d .

of the transaction, mastery of the inner w o r k i n g s of the statute, rejection of broad generalizations offered as easy f o r m u l a s for decision—seem equivalent to f o r m a l i s m in any clear sense of that m u c h - a b u s e d term. The approach does not entail denial that there are unintended tax l o o p holes and that courts should try to close t h e m . Yet Isenbergh does n o t merely criticize the intellectually lazy style in w h i c h m a n y j u d g e s perform that office. H e also criticizes the result, as in the f a m o u s case of Gregory v. Helvering,26 Mrs. Gregory was the sole o w n e r of a corporation that o w n e d certain assets which she wanted to sell. If they w e r e either sold directly or distributed to and then sold by her, there w o u l d be a large tax; for example, the distribution of assets to her directly w o u l d be treated as a dividend equal in a m o u n t to the value of the assets and w o u l d be taxable in that amount as ordinary i n c o m e to her. So she f o r m e d a n e w corporation, also wholly o w n e d by her, to w h i c h she had the old corporation transfer the assets in question. She then dissolved the n e w corporation, three days after having created it; received the assets as a liquidating dividend; and then sold t h e m . Because the distribution of the assets to her had literally been pursuant to a "plan of reorganization," and because a liquidating dividend issued pursuant to such a plan is not a taxable dividend, she claimed to be liable only for a tax o n the capital gain she reaped w h e n she sold the assets. T h e Supreme C o u r t held that the " r e o r ganization" had been a sham and therefore Mrs. G r e g o r y was not entitled to treat the distribution of assets pursuant to it as a liquidating dividend exempt f r o m ordinary i n c o m e tax. Isenbcrgh argues that the decision is w r o n g . T h e statute defined a reorganization as a change in corporate f o r m (for example, dissolution) that complied w i t h certain technicalities, all of w h i c h M r s . G r e g o r y had observed. There was thus n o fact to which a court could refer in order to decide w h e t h e r a reorganization c o n f o r m i n g to the statutory definition was invalid; there was only the concept of reorganization that had been created by the statute. This reasoning, quintessentially formalist, treats the tax statute as having created a conceptual w o r l d , like the set of real n u m b e r s or the theorems of Euclid, h a v i n g n o necessary correspondence to the w o r l d of fact. To Isenbergh, reorganization is not a transaction in the w o r l d , b y w h i c h a f i r m advances the business interests of its owners; it is a paper shuffle danced to the tune played by the statute. Ignored is the p u r p o s e of e x e m p t i n g the gains incident to changes in corporate structure f r o m taxation: to facilitate reorganiza26. 293 U . S . 465 (1935), discussed at pp. 8 6 6 - 8 7 0 of Isenbcrgh's article.

tions viewed as purposive business transactions, rather than to encourage the empty transaction in which Mrs. Gregory engaged—a transaction the sole purpose and (had it succeeded) effect of which were to beat taxes. 27 We n o w understand "hard grappling . . . with the inner w o r k ings of a statute" to preclude judicial consideration of the consequences of alternative statutory interpretations. 2 8 M y disparagement of legal formalism, in its specific guise as rule by rules, may seem music to the ears of left-wing critics of liberalism, w h o see any compromise with the strictest formalist conception of law as signifying a crisis of liberalism. 2 9 1 disagree with them, 3 0 while granting that they can cite Hayek and other conservatives in support of their position. It is desirable to minimize the discretion of officials, including judges, but undesirable as well as impossible to eliminate official discretion altogether. It seems obvious that legislators have, and must have, discretion. 3 ' And if, rather than attempting the impossible by specifying in advance every imaginable exception to every rule of conduct laid d o w n by them, they leave it to the courts to create (perhaps within specified bounds) exceptions on an ad hoc basis, this is a detail about the division of labor a m o n g the branches of government rather than an abandonment of the concept of the liberal state. Even if responsibility for formulating basic rules of conduct is left mainly to the courts, as was

27. As explained b y the c o m m i t t e e r e p o r t s in the H o u s e and Senate, C o n g r e s s had " a d o p t e d the policy of e x e m p t i n g f r o m tax the gain f r o m e x c h a n g e s m a d e in c o n n e c t i o n w i t h a r e o r g a n i z a t i o n , in o r d e r that o r d i n a r y business transactions will n o t be p r e v e n t e d . " S. R e p . N o . 398, 68th C o n g . , 1st Sess. 1 4 - 1 5 (1924); H . R . R e p . N o . 179, 68th C o n g . , 1st Sess. 13 (1924). 28. T h e Irish S u p r e m e C o u r t , in the recent case of McDermott v. McGrath, rejected the " s u b s t a n c e o v e r f o r m " d o c t r i n e for Irish tax law; and the Irish p a r l i a m e n t seems unable to figure o u t a legislative m e t h o d of p l u g g i n g the l o o p h o l e created b y the decision. See G e o r g e G u t t m a n , "Irish S u p r e m e C o u r t U p h o l d s Tax A v o i d a n c e S c h e m e , " 40 Tax Notes, J u l y 25, 1988, at 349, 350. 29. See, for e x a m p l e , D u n c a n K e n n e d y , "Legal F o r m a l i t y , " 2 Journal of Legal Studies 351 (1973); R o b e r t o M a n g a b e i r a U n g e r , Knowledge and Politics, ch. 2 (1975); M a r k K e l m a n , A Guide to Critical Legal Studies, ch. 1 (1987); Terry E a g l e t o n , William Shakespeare 35—58 (1986). 30. See m y b o o k Law and Literature: A Misunderstood Relation 105-109 (1988); also N a n cy L. R o s e n b l u m , Another Liberalism: Romanticism and the Reconstruction of Liberal Thought 3 8 - 4 2 (1987). 31. O b v i o u s to me, a n y w a y . B u t H a y e k b e l o n g s to t h e school of political t h o u g h t , G e r m a n in origin (see L e o n a r d Krieger, The German Idea of Freedom: History of a Political Tradition [1957|, esp. pp. 46—125) that stresses limited g o v e r n m e n t rather than p o p u l a r g o v e r n m e n t ; and the m o r e g o v e r n m e n t is limited, the less discretion legislators have. See m y article " T h e C o n s t i t u t i o n as an E c o n o m i c D o c u m e n t , " 56 George Washington Law Review 4, 2 1 - 2 2 (1987).

done in England and America until this century, this does not make liberalism a lie. Were England and America less liberal, in the classical sense of the word, during the heyday of the c o m m o n law than they are in the era of the welfare state? A system of untrammeled official discretion would be inconsistent with the premises of the liberal state, prominent among which is the rule of law—the concept of a "government of laws, not men." But pockets of discretion are not inconsistent with liberalism. We do not have an "illiberal" system merely because judges use discretion to decide difficult cases, juries sometimes nullify criminal statutes, and the executive branch can pardon criminals without giving reasons. The liberal state has always been a mixture of rule and discretion, and the issue should be whether we get a better mixture under liberalism than we would get in the sorts of regime preferred by its critics. But even to state the issue this way is to take labels too seriously. It should not matter whether our legal system is called "liberal." The practical question is whether it is better than a system with even more rules and less discretion or a system with even more discretion, more standards, and fewer rules. 32 Having criticized legal formalism at such length I should make clear that I am not opposed to all formalism. Mathematics is a formalist discipline; so is logic; so are much painting, sculpture, architecture, and literature. Economics has its formalist side; my work in economic analysis of law has been described both as formalist and as antiformalist. 33 Abstracting from particulars is an essential part of science; so in a sense all science, not just economic science, is formalist. All this merely illustrates the poverty of categorization (a typically formalist defect!). The prestige of true formalisms is part of the reason so many judges and legal scholars aspire to make law a formalist discipline.

Scientific

Observation

After logic the most rigorous and objective methods of extending our knowledge involve systematic empirical inquiry, using either controlled experiments or "natural" experiments that can be statistically verified. 32. For g o o d illustrations of l a w y e r s ' "label a n x i e t y " — s p e c i f i c a l l y the a n x i e t y of m o d ern welfare-state liberals to preserve the " l i b e r a l " label for t h e m s e l v e s in the face of the radical Left's e f f o r t to attach the " l i b e r a l " label to the advocates o f laissez-faire—see R o n a l d D w o r k i n , Law's Empire 4 4 0 - 4 4 4 (1986); Richard H . Fallon, Jr., " W h a t Is R e p u b l i c a n i s m , and Is It W o r t h R e v i v i n g ? " 102 Harvard Law Review 1695 (1989). 33. C o m p a r e Wilson, n o t e 4 a b o v e , at 481, w i t h Louis M e n a n d , " O b j e c t i o n O v e r ruled," New Republic, M a r c h 13, 1989, at 31.

A l t h o u g h these m e t h o d s of inquiry provide only tentative support for propositions, and the " t r u t h s " they demonstrate (or at least lead people to accept) often are transitory, the scientific m e t h o d is for most people in m o d e r n society the m o d e l of objective inquiry—such has been the success of science in altering b o t h our w o r l d v i e w and our w o r l d . Scientific m e t h o d plays little role in legal reasoning. Parties to lawsuits lack the time and m o n e y necessary to conduct scientific experiments or other scientific studies. T h e relevant data are often difficult or impossible to obtain. Legal decisions may depend on predictions of extremely l o n g - t e r m consequences, such as the effect of abortion o r capital p u n i s h m e n t on attitudes t o w a r d the sanctity of life, and large-scale social experiments designed to test such predictions are rarely feasible, w o u l d take too long, or w o u l d raise insuperable ethical questions. L a w yers and j u d g e s (not to m e n t i o n jurors) are not trained in the scientific m e t h o d . T h e trial is n o t modeled on scientific inquiry. As w e shall see in the next chapter, authority and hierarchy play a role in law that w o u l d be inimical to scientific inquiry. A n d m a n y legal doctrines are too entrenched to adjust rapidly to changes in scientific understanding. (There is no principle of stare decisis in scientific inquiry.) Society is unable or at least unwilling to allow legal j u d g m e n t s either to be deferred until the results of patient scientific research are available or to be changed the m i n u t e a legal doctrine becomes scientifically obsolete. O n e could perhaps imagine law o n the m o d e l of a science—say, physics. Academic lawyers, corresponding to physicists, w o u l d f r o m time to time discover scientific laws that govern a legal system, such as the less certain law is, the l o w e r the settlement rate will be. Practicing lawyers and j u d g e s , corresponding to electrical engineers, w o u l d e m b o d y these discoveries in the " m a c h i n e r y " of justice, w h i c h already contains devices, such as the pretrial conference, for fostering settlement by reducing uncertainty a b o u t the o u t c o m e o f trial. Innovations would be evaluated by c o m p a r i n g their actual as distinct f r o m their anticipated effects. 14 An objection to the "social engineering" m e t a p h o r is that it seems to invite lawyers to r e m a k e our w o r l d , but w e can avoid this implication by focusing on the role of the engineer as one w h o implements scientific discoveries. O n the theoretical or "scientific" plane, society decides that burglary should be punished m o r e heavily w h e n it

34. T h e n o t i o n of using the scientific m e t h o d to g u i d e social r e f o r m is quintessentially p r a g m a t i c . See, for e x a m p l e , G e o r g e H. M e a d , " T h e W o r k i n g H y p o t h e s i s in Social R e f o r m , " 5 American Journal of Sociology 367, 3 6 9 - 3 7 1 (1899).

takes place at night than w h e n it takes place d u r i n g the daytime. T h e social engineer's j o b is to find a feasible, t h o u g h necessarily an arbitrary, way of distinguishing between day and night. T h e need to m a k e dichotoraous cuts in continuous p h e n o m e n a s h o w s w h y law can never be— and should never aspire to be—an elegant field; and in this too it resembles engineering. Another w a y to m a k e the point is to note that b o t h professions care critically about the application of theories, w h i c h requires w o r k i n g w i t h people and materials as well as with designs and implies a concern w i t h stability, often at the expense of elegance. T h e redundancies that engineers build into structures and machines to reduce the probability of failure correspond (though only very roughly) to the reluctance of legal professionals to countenance abrupt changes and daring departures in law. T h e engineer's preference for a tried-and-truc process or product over one e m b o d y i n g a m o r e exciting technology has its counterpart in rules limiting the reopening of legal j u d g m e n t s on the basis of n e w l y discovered evidence or n e w arguments. A related respect in which lawyers are like engineers is that they try very hard to avoid m a k i n g mistakes. Mistakes in law as in engineering are m o r e likely to have dramatic and immediate real-world consequences (the loss of a case or collapse of a building, for example) than w o u l d a scientific or a mathematical error. Some progress t o w a r d recasting the law in a genuinely rather than merely analogically scientific-technological m o l d is visible. This p r o g ress is due largely to the efforts of economists and economically m i n d e d lawyers. Economics, including the branch k n o w n as e c o n o m i c analysis of law, or "law and economics," really is a science, t h o u g h an i m m a t u r e one (see Chapter 12). T h e practitioners of law and economics are trying with some success to use the m e t h o d s and results of economics to improve our understanding of law and assist in its r e f o r m . Further p r o g ress on this front can be expected. But in part because the scientific fields, such as economics and psychology, u p o n w h i c h a science o f law would have to build are i m m a t u r e , and in part because of the institutional factors noted above, the day is far distant w h e n law can take its place a m o n g the sciences. As w i t h science, so w i t h engineering: engineering marvels are m o r e easily determined to be such than are legal ones, progress in engineering is m o r e dramatic than in law, and the methods of engineers are less problematic than those o f lawyers. A m I pressing the distinction between the technical and legal realms too hard? Is exact inquiry really all that exact? T h e reader m a y bridle at m y serene equation o f exact inquiry w i t h science. N o t only is it u n c o n -

ventional (outside of the philosophy of science) to treat logic, mathematics, and statistics as branches of science; 35 but even natural science (leave aside economics, psychology, and other social sciences) is not always, perhaps not typically, as rigorous, exact, and certain as the term "exact inquiry" connotes. 3 6 Many scientific theories, including natural selection and the "big bang," cannot be verified by experimentation or by any other method of exact observation. 37 Many have been proved false after having been universally accepted; examples are Euclidean geometry as a theory of spatial relations, the geocentric universe, N e w ton's laws of motion, and the luminiferous ether. Many scientific theories—some philosophers of science think all—are temporary or ad hoc constructs to explain phenomena that might be explained in other ways. Many yield results that are only weakly supported. 3 8

35. B u t n o t incorrect. See, for e x a m p l e , W. V. Q u i n e , "Success and Limits of M a t h e m a t i z a t i o n , " in Q u i n e , Theories and Things 148 (1981). It is t r u e that logic and m a t h e m a t i c s deal w i t h entities (such as n u m b e r ) w h o s e o n t o l o g y is m o r e q u e s t i o n a b l e than that of such scientific entities as a t o m s and e l e c t r o m a g n e t i c r a d i a t i o n — i n w h a t sense does a n u m b e r "exist"? B u t the t e n d e n c y of b o t h logical and m a t h e m a t i c a l inquiries to converge, and the role of logic and science in successful scientific theories, are g r o u n d s for believing that logical and m a t h e m a t i c a l inquiries yield " t r u t h . " In c o n t r a s t , ethical, literary, and legal inquiries n e i t h e r exhibit a s t r o n g t e n d e n c y t o w a r d c o n v e r g e n c e n o r play a role in successful scientific t h e o r i z i n g . 36. For an excellent i n t r o d u c t i o n t o the p h i l o s o p h y and m e t h o d o l o g y of science, see D a v i d O l d r o y d , The Arch of Knowledge: An Introduction to the History of the Philosophy and Methodology of Science (1986). 37. Yet it w o u l d be a m i s t a k e to infer f r o m this p o i n t , as the Louisiana legislature did in insisting u p o n the teaching in its p u b l i c schools of "creation science," that " e v o l u t i o n is not a scientific 'fact,' since it c a n n o t actually be o b s e r v e d in a laboratory. Rather, evolution is m e r e l y a scientific t h e o r y or ' g u e s s . ' " E d w a r d s v. Aguillard, 107 S. C t . 2573, 2598 (1987) (dissenting o p i n i o n ) . W h a t a c u r i o u s collocation of religious faith w i t h e x t r e m e scientific positivism! I m a g i n e the fix that religion w o u l d be in if its d e f e n d e r s w e r e required to p r o vide l a b o r a t o r y c o n f i r m a t i o n of miracles. O n the status of e v o l u t i o n as scientific theory, see Michael Ruse, Philosophy of Biology Today, ch. 1 (1988); Florian v o n Schilcher and Neil T e n n a n t , Philosophy, Evolution, and Human Nature 91—105 (1984). O n the debate over " c r e ation science," sec Science and Creationism (Ashley M o n t a g u ed. 1984). 38. T h e f r e q u e n t mistakes, and f r e q u e n t l y s h o d d y m e t h o d s , even of m o d e r n science are painstakingly r e c o u n t e d in D a v i d Faust, The Limits of Scientific Reasoning (1984). F r o m the Left c o m e charges that i d e o l o g y shapes the direction, and s o m e t i m e s even the o u t c o m e , of scientific research. See, for e x a m p l e , R o u s e , n o t e 1 a b o v e ; B r u n o L a t o u r and Steve W o o l ger. Laboratory Life: The Social Construction of Scientific Facts (1979); Feminist Approaches to Science ( R u t h Bleier ed. 1986). T h e s e criticisms m u s t be kept in perspective. N o n e of these w r i t e r s c o n t e n d s that such p r o p o s i t i o n s as that the b l o o d circulates or that tuberculosis is caused by bacteria or that the sun is millions of miles f r o m the earth are j u s t t w e n t i e t h c e n t u r y m y t h s , n o t r u e r t h a n the v i e w s the ancient Greeks held o n these m a t t e r s ; or that the " A r y a n p h y s i c s " of N a z i G e r m a n y had as m u c h t r u t h value as the physics practiced by Einstein. See R o m H a r r e , Varieties of Realism: A Rationale for the Natural Sciences (1986);

The basic problems concerning the objectivity of science are three. The first is the sheer diversity, indeed indefinability, of science, which makes it impossible to survey the entire field, let alone to reduce it to a single methodology or tight set of methodologies. T h e second, little discussed nowadays in academic circles, is that millions of highly intelligent people selectively disbelieve in science, and w h o is to gainsay them? 39 The third problem is that it is unclear h o w a scientific theory can ever be validated. 40 Suppose one deduces f r o m theory A a h y p o t h esis that event X will be observed under certain conditions; and X is observed, which supports A. But the observation will equally support A + B (A might be the theory that the blood circulates, B that human beings have souls), or C (any other theory f r o m which a hypothesis predicting X can be derived logically). Realistically, in order for X to support A + B or C rather than just A, there would have to be an argument that B or C was related to A. But such arguments are not difficult to make. An indefinite n u m b e r of theories, many of them plausible, will be consistent with a given set of observations, and the choice among the plausible theories is likely to be made on utilitarian or practical grounds, such as simplicity and fruitfulness, rather than on epistemic ones. Moreover, the observations relied on to " c o n f i r m " the theory may be ambiguous. "Controlled" experiments suppress features of the natural environment that are deemed irrelevant, in order to isolate the effect of the variable under investigation. But the experimenter may err in the design of the experiment. O n e of the excluded features may be the real cause of the phenomenon observed, and the independent variable that the experimenter wanted to test and that he found to have causal significance may just be a correlate of the omitted variable.

Alan D. Beyerchen, Scientists under Hitler: Politics and the Physics Community in the Third Reich, ch. 7 (1977). M o r e o v e r , the sense in w h i c h such r e f u t e d scientific theories as the Euclidean t h e o r y of space o r N e w t o n ' s laws of m o t i o n are " f a l s e " is a r a t h e r special one; for m o s t p u r p o s e s , i n c l u d i n g m o s t t e c h n o l o g i c a l p u r p o s e s , t h e y are true ( " t r u e e n o u g h " ) . ( M o r e on " t r u t h " in C h a p t e r 3.) For a succinct s u m m a r y of the case for scientific realism, see Richard N . B o y d , " H o w t o Be a M o r a l Realist," in Essays on Moral Realism 181, 1 8 8 189 (Geoffrey S a y r e - M c C o r d ed. 1988). T h e antirealist case is well a r g u e d in A n d r e w Pickering, Constructing Quarks: A Sociological History of Particle Physics (1984). 39. "I believe that every h u m a n b e i n g has t w o h u m a n parents; b u t C a t h o l i c s believe that Jesus only had a h u m a n m o t h e r . " L u d w i g W i t t g e n s t e i n , On Certainty 32e (G. E. M . A n s c o m b e and G. H . v o n W r i g h t eds. 1969) (11239). A m o n g these believers are a n u m b e r of distinguished scientists. See, for e x a m p l e , R. J. Berry, " W h a t t o Believe a b o u t Miracles," 322 Nature 321 (1986). 40. See, for e x a m p l e , C l a r k G l y m o u r , Theory and Evidence 2 9 - 4 8 , 110-290 (1980).

T h e p r o b l e m is n o t solved by trying to g o f r o m data to theory, that is, b y induction. As H u m e pointed out centuries ago, to discover that w h e n e v e r w e observe d a t u m A w e also observe d a t u m B will habituate us to expect B w h e n w e see A but will give us n o rational basis for expecting the pattern to continue. 4 1 Even seeing the sun rise every day does n o t give one a warranted confidence that it will continue to do so— any m o r e than w a k i n g u p every m o r n i n g gives one a warranted confidence that one will live forever—unless one has a theory of w h a t is causing the sun to rise. B u t as j u s t noted, it is unclear that theories can be p r o v e d by confirmation. A n indefinite n u m b e r of theories, m a n y contradictory, will yield the prediction that the sun will continue to rise every day, and each of these theories will thus be confirmed repeatedly. Theories can be falsified, but if w e agree therefore w i t h P o p p e r that the m a r k of a scientific t h e o r y is that it is falsifiable but n o t verifiable (see C h a p t e r 3), w e are driven to the odd conclusion that all scientific theory is conjecture and n o inductions are reliable. All that this discussion demonstrates, however, is the existence of unsolved p r o b l e m s in the philosophy of science, n o t the existence of deep p r o b l e m s w i t h science. 42 A l t h o u g h every bit of what w e n o w believe about the nature of the universe m a y eventually be o v e r t h r o w n , in the m e a n t i m e "science reveals hidden mysteries, predicts successfully, and w o r k s technological wonders." 4 3 Airplanes designed n o t by trial and error b u t b y applying scientific theories "verified" in wind-tunnel experiments actually fly. As this example 4 4 shows, progress in science is c o m patible w i t h the conjectural character of h u m a n k n o w l e d g e . Even if the theories u p o n w h i c h o u r notions of flight are based are someday falsified or superseded, planes designed in accordance with those theories will fly; w e are in n o danger of being t h r o w n back to the Wright brothers' time. As this e x a m p l e shows, w e have only to describe k n o w l e d g e in " k n o w h o w " rather than " k n o w that" terms to see that scientific k n o w l edge has indeed g r o w n steadily. O u r atomic theory m a y ultimately be

41. For a g o o d m o d e r n discussion, see H a i m G a i f m a n , " O n I n d u c t i v e S u p p o r t and S o m e Recent Tricks," 22 F.rkenntnis 5 (1985). 42. For a g o o d discussion, see A r t h u r Fine, " U n n a t u r a l A t t i t u d e s : Realist and I n s t r u mentalist A t t a c h m e n t s to Science," 95 Mind 149 (1986). 43. W. V. O . Q u i n e , " N a t u r a l K i n d s , " in Naturalizing Epistemology 31, 43 (Hilary K o r n blith ed. 1985). 44. W h i c h is f r o m Paisley L i v i n g s t o n , Literary Philosophy of Science 7 4 - 7 5 (1988).

Knowledge:

Humanistic

Inquiry and the

shown to be as off base as Democritus's; but that w e k n o w h o w to do more things than D e m o c r i t u s and other ancient Greek scientists k n e w h o w to do—for example, w e k n o w h o w to m a k e an atomic b o m b — i s beyond question. Natural laws that engineers m a k e use of in building things are m o r e reliable than the scientific theories that are formulated to explain those laws. O n e might be t e m p t e d to argue in like vein that even t h o u g h every legal rule w e live by m a y s o m e d a y be o v e r t h r o w n , m e a n w h i l e the system of rules is as certain, as useful, as well established as the existing body of scientific k n o w l e d g e . S o m e parts of l a w — f o r example, the rule that m u r d e r is criminal—are more solidly established than s o m e parts of science. But it w o u l d be a mistake to conclude that because the epistemological foundations of science are less secure than they once seemed (even as science has b e c o m e an ever m o r e reliable f o u n t of knowledge), fields with even weaker epistemological foundations, such as law, yield, on average, k n o w l e d g e comparable in reliability to scientific k n o w l edge. T h e m e t h o d s by w h i c h scientific k n o w l e d g e is created and, if not verified, at least temporarily supported and t r a n s f o r m e d into useful gadgetry are by and large n o t available to law, not yet anyway. Even if in g o o d pragmatist fashion w e forget about foundations, accept the broadest possible definition of science, and agree w i t h Richard Rorty to w i t h h o l d the n a m e of science f r o m a m e t h o d of belief formation only " i f f o r c e is used to change belief. . . [or] w e can discern . . . [no] connection with o u r ability to predict and control," 4 5 w e shall not confuse law with science. Prediction and control are precisely w h a t science gives us and philosophy, politics, and law d o n o t — e x c e p t (in the case of politics and law) w i t h the aid of force. To summarize, if science does establish certainty, still its m e t h o d s and domain are so different f r o m those of law that the exactitude of science cannot be translated into exactitude in law, while if science docs not establish certainty, then it cannot be used as a f o u n d a t i o n for or m o d e l of legal certainty. In either event w e m u s t look elsewhere for the grounds of that certainty. But is there an elsewhere? Logical positivists

45. Rorty, "Is N a t u r a l Science a N a t u r a l K i n d ? " in Construction and Constraint: I'he Shaping of Scientific Rationality 49, 72 ( E r n a n M c M u l l i n ed. 1988). It m a y be hard to see h o w w e can "predict and c o n t r o l " physical p h e n o m e n a if w e do n o t have a grip o n reality, but this a r g u m e n t for scientific realism is w e a k e n e d by the fact that m a n y r e f u t e d theories have had excellent predictive properties, i n c l u d i n g P t o l e m a i c c o s m o l o g y , w h i c h is still usable for navigating by the stars.

w o u l d say no. T h e y believe 46 that propositions, to be meaningful, must be either analytic, in the sense in w h i c h propositions in logic or mathematics are analytic (true b y definition or something akin to definition), or verifiable by observation; propositions neither analytic n o r verifiable are merely emotive. 4 7 O n this view m o s t legal propositions are either emotive or deduced f r o m e m o t i v e propositions. This conclusion if correct w o u l d w a r r a n t a radical skepticism about the objectivity of law, but it w o u l d be correct only if exact inquiry were the only path to truth. Suppose w e g o to the opposite extreme and accept the increasingly widespread view that the objectivity and certainty of the scientific m e t h od were greatly exaggerated b y previous generations o f philosophers of science—including the logical positivists themselves, w h o s e skepticism about nonverifiable nonanalytic propositions was the obverse of their passionate faith in science. 48 Would this argue a fortiori for the indeterminacy of law? N o t necessarily. O b s e r v i n g that science is successful despite the lack of epistemological foundations as solid as once believed, w e m i g h t conclude that law could be successful even if its foundations are likewise less solid than once believed. (Indeed, in either field, the quest f o r foundations m a y be misconceived.) This possibility can be

46. P e r h a p s I s h o u l d use the past tense; p h i l o s o p h e r s c o n s i d e r logical p o s i t i v i s m t h o r o u g h l y discredited. B u t discredited p h i l o s o p h i e s are r e v e n a n t s (recent e x a m p l e s are social c o n t r a c t a r i a n i s m , A r i s t o t e l i a n i s m , and p r a g m a t i s m — a l l o f w h i c h w e r e w i d e l y considered discredited t h i r t y years ago); and t h e attitudes that m o t i v a t e d logical positivism c o n t i n u e to be influential. T h e m e t h o d o l o g y o f e c o n o m i c s , w e shall see in C h a p t e r 12, has a logicalpositivist flavor. 47. T h e best i n t r o d u c t i o n to logical p o s i t i v i s m r e m a i n s A . J . Ayer, Language, Truth, and Logic (rev. ed. 1946), b u t the best brief s u m m a r y is m o r e than t w o h u n d r e d years old: "If w e take in o u r h a n d a n y v o l u m e ; of d i v i n i t y or school m e t a p h y s i c s , for instance; let us ask, Does it contain any abstract reasoning concerning quantity or number? N o . Does it contain any experimental reasoning concerning matter of fact and existence? N o . C o m m i t it t h e n to the flames: for it can contain n o t h i n g b u t s o p h i s t r y and illusion." D a v i d H u m e , An Enquiry Concerning Human Understanding 165 (3d ed., P. H . N i d d i t c h ed., 1975) (§12, pt. 3). R e m e m b e r that logical p o s i t i v i s m and legal p o s i t i v i s m are n o t t h e s a m e t h i n g , a l t h o u g h H o l m e s for one subscribed to b o t h . 48. C l i f f o r d Geertz, Local Knowledge: Further Essays in Interpretive Anthropology 162—163 (1983), gives a d r a m a t i c e x a m p l e of h o w , v i e w e d f r o m the inside, physics lacks the purity and certainty that l a y p e o p l e associate w i t h it. H e is q u o t i n g a physicist w h o says that " p h y s ics is like life; there's n o p e r f e c t i o n . It's n e v e r all s e w e d up. It's all a question of better, better yet, and h o w m u c h t i m e and interest d o y o u really have in i t . . . A t h e o r y isn't right and w r o n g . A t h e o r y has a sort of sociological p o s i t i o n that changes as n e w i n f o r m a t i o n comes in. 'Is Einstein's t h e o r y correct?' You can take a poll and h a v e a l o o k . Einstein is rather 'in' right n o w . B u t w h o k n o w s if it is ' t r u e ' ? . . . N o b o d y said a n y t h i n g a b o u t T r u t h . Perhaps T r u t h is ' o u t . ' O n e t h i n k s , 'Well, this idea l o o k s bad for or l o o k s g o o d for general relativity."'

glimpsed in the following comparison between induction in science and in law: Sometimes an observation or measurement in the experiment turns out to have been faulty, and so the report of what occurred was incorrect. Sometimes a hidden variable was interfering with the action of the controlled variables in the experiment, so that factors other than those reported were influencing the result. In either of these t w o types of case the replicability of the experiment would be unreliable. A n d in a third type of case the experiment is satisfactorily replicable, but the hypothesis is treated as an idealization to which the real w o r l d only c o n f o r m s under an unrealizable ceteris paribus condition. Similarly a supposed legal precedent m a y be discounted in one or other of three analogous ways. Sometimes the previous case may be s h o w n to have been w r o n g l y decided, or to have been heard by a j u d g e whose decisions are not binding in the instant case, so that the alleged precedent lacks authority. Sometimes the previous case had features that were significantly different f r o m those of the instant case, so that the alleged precedent is irrelevant. And sometimes the conclusion is that the principles underlying the law w o u l d be best codified by our establishing a new nexus of legal roles and relations which w o u l d not quite accord at every point with relevant precedents. 4 9

The analogy is fine; and maybe it is more than analogy—maybe induction is induction, a universal method of reasoning, as inescapable in law as in science. But neither the fact that lawyers, like scientists, use induction, nor the intriguing suggestion that scientists, like lawyers, are j u d g mental rather than mechanical in their use of induction—that is, in the weight they attach to "precedents," viewed as confirming instances— makes law scientific in an interesting sense. What is missing f r o m law are penetrating and rigorous theories, counterintuitive hypotheses that are falsifiable but not falsified (and so are at least tentatively supported), precise instrumentation, an exact vocabulary, a clear separation of positive and normative inquiry, quantification of data, credible controlled experiments, rigorous statistical inference, useful technological byproducts, dramatic interventions with measurable consequences, and above all and subsuming most of the previous points, objectively testable—and continually retested—hypotheses. In law there is the blueprint or shadow of scientific reasoning, but no edifice. This conclusion may seem to overemphasize the observational aspect of science. Science is not only exact observation but also the search for

49. L. Jonathan C o h e n , The Dialogue of Reason: An Analysis (1986).

of Analytical

Philosophy

98

unity in multiplicity (and finding it, most dramatically in the physics of N e w t o n and of Einstein). Science thus has its Platonic as well as its Aristotelian aspect. And law certainly has the former. But this does not make law scientific, or even half-scientific, any more than the shared characteristic of abstraction makes a legal rule a scientific law. The quest for unity in multiplicity is not unique to science. It is also the project of theology and of metaphysics, modes of inquiry that differ f r o m science in declining to subject their theories to the fires of empirical testing— and that is all the difference in the world. In this critical respect law is closer to theology and to metaphysics than to science. Lawyers are not only quick but unashamed to make emphatic assertions on matters of fact (for example, the role of the religion clauses of the First A m e n d m e n t in preserving religious pluralism in the United States) without attempting, desiring, or even being willing to subject those assertions to an empirical test. If wc redescribe "science" as merely (but not trivially) the idea, practice, and ethics of systematic, disinterested inquiry 511 —in essence, an attitude of respect for fact—we may seem to be setting before the lawyer, judge, and law professor an eminently attainable as well as highly worthwhile ideal. We can then begin to see more clearly h o w cases might be the counterpart to scientific data and a potential basis therefore for inferring lawlike behavior in the scientific, not legal, sense of "law." Wc can note the parallel between, on the one hand, the scientist's deduction of consequences f r o m a scientific theory and attempt to discover those consequences in nature and, on the other hand, the judge's comparison of the implications of a legal doctrine with social reality. O f course there is a vast gulf between scientific aspiration and scientific achievement, but the more interesting point, to be explored below, is that practices deeply woven in the fabric of law may place even the aspiration beyond law's reach. 50. A typically p r a g m a t i s t c o n c e p t i o n . See, for e x a m p l e , Uorty, n o t e 45 above.

2 Legal Reasoning as Practical Reasoning

What Is Practical Reason? Set against the methods of exact inquiry are those of "practical reason." Unfortunately the term lacks a standard meaning. It is most often used to denote the methods ("deliberation" and "practical syllogism" are the key expression here) that people use to make a practical or ethical choice, such as whether to go to the theater, or whether to lie to an acquaintance. 1 Practical reason in this sense is action-oriented, in contrast to the methods of "pure reason" by which we determine whether a proposition is true or false, an argument valid or invalid. Practical reason involves setting a goal—pleasure, the good life, whatever—and choosing the means best suited to reaching it. The term is also used, notably by the academic lawyers w h o m I call "neotraditionalists" and discuss in Chapter 14, to denote a methodology for reaching conclusions that relies heavily on the traditions in the particular field of inquiry or endeavor and is actually suspicious of what we usually think of as reason, including, in some versions of traditionalism, practical reason in the first sense. Both of these usages have good Aristotelian credentials. I am going to use the term in a third sense, which overlaps the others and also is Aristotelian. Found mainly in Aristotle's discussions of induction, dialectic, and rhetoric, it denotes the methods

1. For illustrative discussions, see Practical Reasoning (Joseph Raz ed. 1978); A n t h o n y Kenny, Will, Freedom, and Power, ch. 5 (1975); D a v i d P. G a u t h i e r , Practical Reasoning: The Structure and Foundations of Prudential and Moral Arguments and Their Exemplification in Discourse (1963); Michael E. B r a t m a n , Intention, Plans, and Practical Reason (1987); M . T. T h o r n t o n , " A r i s t o t e l i a n Practical R e a s o n , " 91 Mind 57 (1982). T h i s is the sense in w h i c h Vincent Wellman uses "practical r e a s o n i n g " in his p r o j e c t of c o n s t r u c t i n g a j u r i s p r u d e n c e of practical reason. See "Practical R e a s o n i n g and Judicial J u s t i f i c a t i o n : T o w a r d an A d e q u a t e

by which people w h o are not credulous form beliefs about matters that cannot be verified by logic or exact observation. 2 The first usage, which conceives of practical reason as the methodology for deciding what to do, might seem more appropriate than the third usage to a worldly activity like law. But m y focus is on the judicial decision, especially the appellate decision, which is action of a sort but contemplative action. T h e j u d g e is not deciding what to do in his life; he is deciding what the litigants should have done in their lives, and the litigants and society demand a statement of reasons. But this is not correct either; actually the j u d g e is in the uncomfortable position of having both to act and to offer convincing reasons for acting. He does not have the luxury of the pure thinker, w h o can defer coming to a conclusion until the evidence gels. I have been taken to task for ignoring the substantial literature that treats law as a f o r m of practical reason in the first sense, the doer sense. 3 I admit that it is a sense relevant to law, but I am unable to find the content in this literature. Granted that the j u d g e needs analytical methods that will issue in action—what then? The "then" is exploring the resources for decision making that people use when they are in a hurry, or lack the skills or resources for patient, disinterested mathematical or scientific research, or are dealing with a question not amenable to logical or scientific inquiry. And that brings me immediately to practical reason in m y sense.

T h e o r y , " 57 University of Colorado Law Review 45, 8 7 - 1 1 5 (1985). See also J o h n Ladd, " T h e Place of Practical R e a s o n in Judicial D e c i s i o n , " in Nomos VII: Rational Decision (Carl J. Friedrich ed. 1964); and n o t e 3 below. 2. It is h a r d to find a c o m p e n d i o u s m o d e r n discussion of these m e t h o d s , but s o m e useful references are S t e p h e n E d e l s t o n T o u l m i n , The Uses of Argument (1958); Practical Reasoning in Human Affairs: Studies in Honor of Chaim Perelman (James L. G o l d e n and J o s e p h J. Pilotta eds. 1986); C h . P e r e l m a n , The Idea of Justice and the Problem of Argument, chs. 1 0 - 1 3 (1963); H . H . Price, Belief (1969); D . S. C l a r k e , J r . , Practical Inferences (1985); J a m e s L. Kinneavy, A Theory of Discourse 2 3 6 - 2 5 5 (1971); R o n a l d Beiner, Political Judgment 7 2 - 9 7 (1983); Larry A r n h a r t , Aristotle on Political Reasoning: A Commentary on the "Rhetoric" 141-162 (1981); Stephen N . T h o m a s , Practical Reasoning in Natural Language (2d ed. 1981); Larry Wright, Better Reasoning: Techniques for Handling Argument, Evidence, and Abstraction (1982). T h e principal sources in A r i s t o t l e are the Prior Analytics, the Posterior Analytics, the Topics, and the Rhetoric. A neglected m a s t e r p i e c e of the g e n r e is C a r d i n a l N e w m a n ' s Grammar of Assent. Finally, o n the m a n y pitfalls of practical r e a s o n i n g , see Richard N i s b e t t and Lee Ross, Human Inference: Strategies and Shortcomings of Social Judgment (1980). 3. See S t e v e n J . B u r t o n , " J u d g e P o s n e r ' s J u r i s p r u d e n c e of S k e p t i c i s m , " 87 Michigan Law Review 710, 7 2 0 - 7 2 3 (1988). T h e literature to w h i c h B u r t o n refers is illustrated b y R o b e r t Alexy, A Theory of Legal Argumentation: The Theory of Rational Discourse as Theory of Legal Justification, pt. c (1989). A l e x y also discusses i n t e r p r e t a t i o n and o t h e r m e t h o d s of practical reason in m y sense, b u t t h e discussion is e x c e e d i n g l y abstract.

Practical reason in this sense is not a single analytical m e t h o d or even a family of related m e t h o d s . It is a grab bag that includes anecdote, introspection, imagination, c o m m o n sense, empathy, i m p u t a t i o n of motives, speaker's authority, metaphor, analogy, precedent, custom, memory, "experience," intuition, and induction (the expectation of regularities, a disposition related b o t h to intuition and to analogy). T h e r e is duplication in this list, so in a sense the list is t o o long. A n d s o m e of the entries are questionable—either because they straddle the line between exact inquiry and practical reason (induction, for example), or because they belong to the logic of discovery rather than o f j u s t i f i c a t i o n , which limits their utility to judges. S o m e of the entries can be viewed not as belonging to a d o m a i n distinct f r o m exact inquiry but instead as approximations to exact inquiry (but of course that means they are not exact). M a n y of the inferences w e d r a w w i t h the aid of practical reason, whether in everyday life, literary criticism, or legal analysis, are based on crude f o r m s of hypothesis testing, parallel to scientific inquiry though not close to it. In another sense, however, m y list is too short, because s o m e of the entries are multiple rather than single. Consider "intuition." T h e brain imposes structure on o u r perceptions, so that, for example, w e ascribe causal significance to acts w i t h o u t being able to o b s e r v e — w e never d o observe—causality. This "faculty of induction," to w h i c h I alluded in Chapter 1, is one sense of intuition. In a different and m o r e colloquial sense intuition is the sort of inarticulate awareness that I discuss in the next chapter under the heading of "tacit k n o w i n g . " Finally, and m o s t interesting, intuition is the b o d y of o u r bedrock beliefs: the beliefs that lie so deep that w e do not k n o w h o w to question them; the propositions that w e cannot help believing and that therefore supply the premises for reasoning. 4 Miscellaneous and u n r i g o r o u s it m a y be, but practical reason is o u r principal set of tools for answering questions large and small. M a y b e our only set. Logic almost always, and scientific experimentation often, are m e t h o d s o f j u s t i f i c a t i o n rather than of discovery. (Mathematics is both.) But as this b o o k is greatly concerned w i t h m e t h o d s o f j u s t i f i c a tion available to j u d g e s , I take n o c o m f o r t in the fact that logic and experiments, so rarely usable to decide difficult legal cases, are for the most part merely m e t h o d s o f j u s t i f i c a t i o n rather than of inquiry. We shall see that m a n y m e t h o d s of practical reason are also better at generating conclusions or guiding behavior than at providing justifications. T h e y 4. Sce L. J o n a t h a n C o h e n , The Dialogue of Reason: An Analysis 73-117(1986).

of Analytical

Philosophy

too, therefore, let the j u d g e d o w n when it comes time to write the opinion explaining his decision. Sometimes, it is true, practical reason yields as high a degree of certainty as do logical demonstrations. An example of a proposition that is not analytic or verifiable and also not as a practical matter falsifiable, and yet is certain, is that no h u m a n being has ever eaten an adult elephant in one sitting. 5 This is the sort of example that philosophers like to give in order to show that logical positivism is false, but in so doing they also show w h y it refuses to stay dead. Logical positivism may be w r o n g but it is inescapable, because it points, however exaggeratedly, to something real and important—that in areas not susceptible of logical demonstration or empirical verification our knowledge is likely to be meager and insecure. N o t many of the propositions we are interested in establishing are as certain as that no one has ever eaten an adult elephant in one sitting or that cats don't grow on trees—and both these propositions, though not themselves verifiable, are k n o w n to be true only by virtue of scientific investigations, f r o m which it is possible to infer with some confidence that the earth's biological repertoire has never permitted either phenomenon. Without science, these would be purely inductive propositions, and we would be uneasy about them, just as we would be uneasy about the proposition that "none of the people living today existed at the time of N e w t o n " 6 if we had no scientific theory explaining w h y h u m a n cells cannot renew themselves indefinitely. Here is a more difficult question for practical reason: H o w do we k n o w that objects continue to exist when no one is looking at them? O n e answer is pragmatic: W h o cares? It makes no difference whether objects continue to exist when no one is looking at them. Another, suggested by Wittgenstein, is that our certainty that objects continue to exist when no one is looking at them is greater than any ground that could be given for this belief. 7 Put differently, if one doubts the existence 5. T h e e x a m p l e is f r o m Peter D. Klein, Certainly: A Refutation of Skepticism 122 (1981). See also H i l a r y P u t n a m , " P h i l o s o p h e r s and H u m a n U n d e r s t a n d i n g , " in P u t n a m , Realism ami Reason 184, 185-186 (vol. 3 of his Philosophical Papers). P u t n a m ' s e x a m p l e is that cats d o n ' t g r o w o n trees; the s a m e e x a m p l e appears in L u d w i g Wittgenstein, On Certainty 36e ( G . E . M . A n s c o m b e and G. H . v o n W r i g h t eds. 1969) (1282). 6. H a i m G a i f m a n , " O n I n d u c t i v e S u p p o r t and S o m e Recent Tricks," 22 Erkenntnis 5, 13 (1985). 7. See Wittgenstein, n o t e 5 a b o v e , at 17e ( H i l l ) ; Alan R. White, " C o m m o n Sense: M o o r e and W i t t g e n s t e i n , " 40 Revue Internationale de philosophic 313 (1986); R.W. N e w e l l , Objectivity, Empiricism, and Truth, ch. 4 (1986). If (this is the central e x a m p l e in On Certainty) y o u asked m e w h e t h e r I had a r i g h t h a n d , I w o u l d n o t , except in e x t r a o r d i n a r y c i r c u m s t a n c e s — p e r h a p s after an accident—look at m y r i g h t hand to verify that I still had it.

of the external w o r l d one j u s t is n o t playing o u r game. Less politely, one is nuts; the premise of an external w o r l d is constitutive of rationality in much the same w a y that scientific i n d u c t i o n — w h o s e validity also is not demonstrable—is. Suppose a person "based all his m a j o r decisions on visions of the future he has w h e n asleep. F u r t h e r m o r e , he has always been w r o n g . W h e n w e point out this fact to h i m , he replies that he does not care because he has j u s t had a vision that assured h i m that all his future visions will be accurate. Would w e not, on this basis, j u d g e him to be irrational?" 8 Bertrand Russell, however, tried to reason (albeit practically) to the conclusion that unobserved objects exist: The way in which simplicity comes in from supposing that there really are physical objects is easily seen. If the cat appears at one m o m e n t in one part of the room, and at another in another part, it is natural to suppose that it has moved from the one to the other, passing over a series of intermediate positions. But if it is merely a set of sense-data, it cannot have ever been in any place where I did not see it; thus we shall have to suppose that it did not exist at all while I was not looking, but suddenly sprang into being in a new place. If the cat exists whether I see it or not, we can understand from our own experience h o w it gets hungry between one meal and the next; but if it does not exist when I am not seeing it, it seems odd that appetite should grow during non-existence as fast as during existence. And if the cat consists only of sense-data, it cannot be hungry, since no hunger but my own can be a sense-datum to me. Thus the behavior of the sensedata which represent the cat to me, though it seems quite natural when regarded as an expression of hunger, becomes utterly inexplicable when regarded as mere movements and changes of patches of colour, which are as incapable of hunger as a triangle is of playing football.''

Russell's a r g u m e n t was m o r e persuasive before the invention of the motion picture and the d e v e l o p m e n t of q u a n t u m theory, and it relies heavily on a criterion w h o s e relation to t r u t h is obscure: simplicity. Here is another difficult question, one w e shall c o m e back to in Chapter 5, that only practical reason can answer, t h o u g h again n o t well. H o w do w e k n o w that there are other minds, since w e can never observe another person's mind but only w o r d s , actions, brain waves, and other physical p h e n o m e n a ? We use a combination of introspection, observation, and induction. We k n o w (never m i n d h o w ) that w e o u r selves have minds, w h i c h enable us to do things such as plan and con-

8. Brian S k y r m s , Choice and Chance: An Introduction to Inductive Logic 44 (1966). 9. The Problems of Philosophy 23 (1912).

ceal, and w e observe that other people appear to do the same things, which—since they seem at least grossly similar to us in other ways— argues that they possess similar mental equipment. B u t this m e t h o d of inquiry will n o t tell us w h e t h e r animals, some of w h i c h also plan and conceal, b u t in m o r e r u d i m e n t a r y and predictable ways than we, and w h i c h d o not appear to speak, also have minds, or w h e t h e r a n e w b o r n h u m a n infant has a mind. A n d there is an ambiguity in the concept of mind. W h e n w e say of people that they have minds, it is unclear to what extent w e are a t t e m p t i n g to denote a thing and to w h a t extent w e are merely using a shorthand expression for our ignorance of motives, our o w n and others'. O f particular i m p o r t a n c e for law, practical reason can answer some ethical questions with a high degree of certainty. It is almost as certain that killing people for pure sport is evil as it is that cats don't g r o w on trees. O f course, a h u n d r e d years f r o m n o w this particular ethical certitude (which m a y well depend o n man's being a "social animal"—a condition that conceivably could change) m a y be o v e r t h r o w n ; but then a h u n d r e d years f r o m n o w it m a y be possible to g r o w cats on trees. 1,1 N o t only can practical reason yield the occasional ethical certainty; logic and science cannot. This point is overlooked in A r t h u r Leff's influential skeptical writings on law. To s h o w the futility of n o r m a t i v e discourse in law, Leff asserts the impossibility of establishing in a secular age even the most elementary moral propositions: " T h e r e is today n o way of ' p r o v i n g ' that n a p a l m i n g babies is bad except b y asserting it (in a louder and louder voice), or b y defining it as s o . " " T h e example is m a w k i s h and incomplete; the morality of "napalming babies" m a y depend on w h e t h e r it is an inevitable accident in a j u s t war, a readily avoidable accident in such a war, or a deliberate act of terrorization in an injust war. 1 2 But forget all that; the pertinent point is Leff's misunderstanding of the nature of k n o w l e d g e . T h e q u o t e d passage assumes that the only things w e really k n o w are the things that have been proved. Yet if a p r o o f is deductive, the conclusion of the p r o o f will be true (other than by accident) only if the premises are true and the deduction is valid; and

10. Will w e still call t h e m cats w h e n that h a p p e n s ? I t h i n k so; t e s t - t u b e babies are h u m a n , after all. B u t the q u e s t i o n of h o w m u c h and w h a t k i n d s of c h a n g e are consistent w i t h c o n t i n u e d i d e n t i t y is a p r o f o u n d one. 11. A r t h u r Allen Leff, " E c o n o m i c Analysis o f L a w : S o m e Realism a b o u t N o m i n a l i s m , " 60 Virginia Law Review 451, 454 (1974). See also Leff, " U n s p e a k a b l e Ethics, U n n a t u r a l Law," 1979 Duke Law Journal 1229. 12. Leff n o d o u b t had in m i n d t h e V i e t n a m War, and n o d o u b t t h o u g h t it an injust war.

the premises are n o t the result, b u t the beginning, of the proof. If the premises are the result of another proof, this simply pushes the quest for certainty back a step. T h e ultimate premise has to be an intuition— something w e cannot help believing—rather than the conclusion of an earlier proof. If a proof is inductive rather than deductive—and if you believe in inductive " p r o o f s " — i t s validity depends on the accuracy and n u m b e r of observations and on the principles of scientific induction, and hence ultimately on our intuitions about perception, causality, and regularity. O u r m o s t confident k n o w l e d g e , therefore, is intuitive, because intuitions lie at the base of all o u r proofs and reasoning and because it is always possible to m a k e a mistake in the process of p r o o f itself (as by omitting premises). So the fact that w e cannot prove that napalming babies is bad does n o t i m p l y that w e cannot know that it is bad. In fact our intuition that w a n t o n killing is bad is as strong as m a n y of the intuitions on which o u r k n o w l e d g e of the empirical w o r l d is f o u n d e d , and stronger than m a n y conclusions of proofs. Difficulties arise only w h e n different people have different and inconsistent moral intuitions regarding the issue at hand. This p r o b l e m , w h i c h is obscured by the tautological character of so m a n y moral propositions (we k n o w m u r d e r is w r o n g f u l because the w o r d " m u r d e r " means deliberate and unjustified killing), is a serious one, but it has n o t h i n g to d o w i t h the idea, w h i c h is false, that k n o w l e d g e is limited to w h a t can be proved. Let m e change Leff's example. You and I are driving; you are the driver and I the passenger. A child appears in the middle of the road, and you turn to m e and say, "Should I try to avoid killing the child?" This question w o u l d m a r k you as crazy, j u s t as if you told m e that you had discussed this b o o k with Plato last night. So at s o m e level there is epistemological parity b e t w e e n science and morals. O f course, it m a y not be the interesting level. If practical reason can yield k n o w l e d g e about metaphysical, scientific, and ethical issues, not always but sometimes, w h y n o t sometimes about law as well? T h e n j u s t because the m e t h o d s of exact inquiry are rarely usable by j u d g e s it w o u l d not follow that m o s t judicial decisions were willful or arbitrary, in the sense of determined b y the j u d g e ' s personality or temperament, by class bias ("politics"), or by the flip of a coin. 13 T h e

13. N o t i c e that a n y o n e w h o did believe in the radical i n d e t e r m i n a c y o f judicial decision m a k i n g w o u l d think it exceedingly p e r v e r s e that j u d g e s s h o u l d ever i m p o s e sanctions (as they routinely do) for the filing o f f r i v o l o u s cases. See S a n f o r d L e v i n s o n , " F r i v o l o u s Cases: D o Lawyers Really K n o w A n y t h i n g at All?" 24 Osgoode Hall Law Journal 353 (1986).

a r g u m e n t that most judicial decisions are of that character is not well supported. 1 4 N o t only is its theoretical base (Leff-tist skepticism) weak; its empirical support is shallow. It depends heavily on an impression of r a m p a n t indeterminacy that is a result of sampling bias. A sample of cases litigated to j u d g m e n t will be biased in favor of uncertainty because w h e n the o u t c o m e is clear the parties will usually settle the case before trial. Even within this biased sample, not all cases will be uncertain. M a n y cases are litigated a outrance not because the case is difficult but because the parties or their lawyers are obtuse or s t u b b o r n or because of acrimony arising f r o m the underlying dispute or f r o m the litigation itself. But w h e n a litigated case is easy, the j u d g e s often decide it w i t h o u t a published opinion, so the universe of reported appellate cases is heavily weighted in favor of difficult cases. A n d this is above all true of the Supreme C o u r t ' s decisions, which c r o w d the horizon of m a n y academic lawyers to the virtual exclusion of decisions by other courts. Yet, while not all cases are difficult, m a n y are; and while s o m e of these can be resolved by logic, science, or practical reason, a considerable residue of cases is left—and those the very ones the profession cares the most about—against w h i c h logic and science will be unavailing and practical reason will break its often n o n e - t o o - s t u r d y lance. Moreover, the m e t h o d s of practical reason necessary to resolve the difficult case may not be m e t h o d s of legal reasoning in a distinctive sense. Law uses m a n y m e t h o d s o f practical reason. This chapter discusses the t w o that have received the m o s t attention: reliance on authority and reasoning b y analogy. In b o t h the concept of precedent plays a key role. I m a k e t w o main arguments. T h e first is that authority plays a different role in law than in science, being essentially political in the f o r m e r and epistemic in the latter, and indeed that the law's heavy t h o u g h necessary reliance on authority retards the emergence of a scientific ethos in law. T h e second is that reasoning b y analogy is not a distinctively legal m e t h o d of reasoning, and m a y reduce to basing decision on all available i n f o r m a t i o n including i n f o r m a t i o n contained in previous decisions.

14. As p o i n t e d out in J o h n Stick, " C a n N i h i l i s m Be P r a g m a t i c ? " 100 Harvard Law Review 332 (1986); L a w r e n c e B. S o l u m , " O n the I n d e t e r m i n a c y Crisis: C r i t i q u i n g Critical D o g m a , " 54 University of Chicago Law Review 462 (1987); J o a n C. Williams, "Critical Legal Studies: T h e D e a t h of T r a n s c e n d e n c e and t h e Rise of the N e w Langdells," 62 New York University Law Review 429, 4 7 1 - 4 9 5 (1987); Brian Langille, " R e v o l u t i o n w i t h o u t F o u n d a tion: T h e G r a m m a r of Scepticism and Law," 33 McGiU Law Journal 451 (1988); Ken Kress, "Legal I n d e t e r m i n a c y , " 77 California Law Review 283 (1989).

Authority In an age not only of science but of hostility to almost all forms of authority, it is easy to forget h o w many of our beliefs, including scientific ones, are based on authority rather than on investigation. 15 An example is the proposition that (roughly speaking) the earth revolves around the sun. Neither an observable nor a readily inferable fact (unlike the roundness of the earth), it is simply the theory that best organizes the data. Few of us have a first-hand acquaintance with the data or can replicate the reasoning that connects them with the theory. C o m m o n sense and intuition support the discredited geocentric theory. We believe in the heliocentric theory only because scientists are unanimous in believing it and because we are taught to defer to scientific consensus on matters classified as scientific (unless it collides with our religious beliefs!), of which the earth's revolution is today one. O f course such deference is, by and large, a very sensible policy; it has survival value. It is not merely blind obedience, especially since what is to be classified as a scientific question is jointly determined by the scientific and lay communities. "Authority" means something else in law. Legal decisions are authoritative not when they command a consensus among lawyers, corresponding to a consensus among scientists, but when they emanate f r o m the top of the judicial hierarchy. The only parallel between this sort of political authority and the intellectual authority to which laypeople defer in forming scientific beliefs is that judicial decisions made at the top of the hierarchy are somewhat more likely to be correct than those of judges lower down. The higher judges are more carefully selected (on average, not in every instance, of course) and have a broader view as well as the benefit of the lower judges' thoughts on the case and additional briefing and argument by the lawyers. But this presumption of superior correctness is weak. And even if all the judges up and d o w n the line agree, their decisions have much less intrinsic persuasiveness than unanimous scientific judgments have, because judges' methods of inquiry are so much feebler than scientists' methods. (Does anyone doubt, as Justice Robert Jackson once remarked, that if there were a

15. For g o o d discussions, see Richard T. D e G e o r g e , The Nature and Limits of Authority, ch. 3 (1985); W. V. Q u i n e and J. S. Ullian, The Web of Belief 5 4 - 6 3 (2d ed. 1978); Price, note 2 above, lecture 5; P u t n a m , n o t e 5 a b o v e , at 186; C . A. J. C o a d y , " T e s t i m o n y and O b s e r v a t i o n , " 10 American Philosophical Quarterly 149 (1973). For discussions of a u t h o r i t y in law, see note 35 below.

court above the Supreme C o u r t a large fraction of the Supreme Court's decisions would be reversed?) 16 This is w h y the legal profession would look askance at a j u d g e w h o cultivated a close personal relationship with the members of the court above him in the hope that a better understanding of their values and beliefs would enable him to predict their decisions more accurately. Society does not have such confidence in the superior wisdom of the higher judges that it wants their judicial inferiors to abdicate all independent j u d g m e n t . Another reason not to place too much weight on the fact that many judicial decisions are unanimous (even in the Supreme Court) 1 7 is that few judges will write or even note a dissent in every case in which they disagree with the majority. And sometimes when a case is indeterminate but not highly charged ideologically, some, maybe most, members of the court will lack a powerful conviction about h o w it should be decided and will defer to a colleague w h o does have such a conviction—without necessarily agreeing in any strong sense with him. Finally, while there is very little explicit vote trading in appellate courts, judges do make efforts to minimize disagreement with each other and as a result will on occasion go along with the strongly expressed conviction of a colleague, even if their impulse is to disagree. 18 This is particularly likely within factions of a factionalized court. Admittedly it is a legal convention—though one not fully shared by the rest of the c o m m u n i t y or even by the entire legal profession—that a decision foursquare in accord with a recent decision by the highest court of the jurisdiction is "correct" by virtue of its conformity to authority. But it is a weak convention. A lawyer w h o loses a case in the Supreme Court, a j u d g e w h o is reversed by the Court, a law professor commenting on the Court's latest (and let us say unanimous) decision—none of these is speaking nonsense, or even violating professional etiquette, if he says the decision is wrong. O u r legal discourse is not so positivistic that one is forbidden to appeal to a "higher law" even after the oracles of the law have spoken; nor need the appeal be couched in such terms. Even when the latest decision is admitted to be "on all fours" with a string of earlier ones, it is possible to argue that it is incorrect because 16. B r o w n v. Allen, 344 U . S . 443, 540 (1953) ( c o n c u r r i n g o p i n i o n ) . 17. For the statistics o n h o w many, see F r a n k H . E a s t e r b r o o k , " A g r e e m e n t a m o n g the Justices: A n E m p i r i c a l N o t e , " 1984 Supreme Court Review 389. 18. For an empirical s t u d y o f the e x t e n t t o w h i c h recorded dissents u n d e r s t a t e disagreem e n t a m o n g j u d g e s , see J u s t i n J. G r e e n , " P a r a m e t e r s of Dissensus in Shifting Small G r o u p s , " in Judicial Conflict and Consensus: Behavioral Studies of American Appellate Courts 139 (Sheldon G o l d m a n and C h a r l e s M . L a m b eds. 1986).

those decisions were w r o n g too. (Yet in another and equally valid sense, judges usually are correct to follow precedent.) Vast areas of established jurisprudence are weakly grounded, in the sense of resting o n highly contestable precedents; and indeed a c o m m o n s y m p t o m of formalist discourse is to treat a decision as a reason for, rather than as the source of, the holding for which the decision is cited, as if the decision were a theorem in Euclid or the o u t c o m e of a scientific e x p e r i m e n t rather than a political act. T h e feet-of-clay p r o b l e m is conspicuous in the c o m m o n law. A decision m a y be authoritative by virtue of a previous decision that is a u t h o r itative by virtue of a decision previous to it and so on until one reaches the first decision in the chain of precedent—but w h a t then? T h a t decision cannot be validated by reference to binding precedent. It was a policy decision, which is to say a political decision although usually n o t a partisan political decision. T h e political foundations of precedent are i m p o r t a n t to emphasize lest w e be tempted to d r a w an analogy between decision according to precedent and the conservative epistemic procedure s u m m e d u p in Quine's (and before h i m William James's) notion of a w e b of belief. W h e n a discovery in logic or mathematics or science is m a d e that w e accept as true and that is at variance w i t h o u r existing beliefs, w e d o n o t chuck out all the old beliefs; instead w e try to fit the n e w discovery into o u r belief system in the w a y that will cause the least perturbation in the system. We try—it is the efficient p r o c e d u r e — t o a c c o m m o d a t e the n e w discovery by adjusting the periphery of o u r epistemic system rather than by changing the core. (Nonliteral interpretation of Genesis is an e x a m ple.) In a superficially similar sense, the law tries to a c c o m m o d a t e d o c trine to altered conditions w i t h as little change as possible; this is one way of describing stare decisis, the policy of generally abiding b y precedent. All that the comparison shows, however, is that at a sufficiently high level of abstraction there are m a n y analogies b e t w e e n science and law (and abstraction is one of t h e m , as I said in the last chapter). At the operational level there are few. T h e law does n o t cling to the old verities because it is an efficient epistemic strategy to give g r o u n d slowly in areas of apparently warranted certitude, but because stability of legal obligation is an i m p o r t a n t social policy. T h e legal core is n o t a b o d y of principles validated by scientific m e t h o d or by robust c o m m o n sense; it is a set of policy decisions, s o m e m a d e long ago u n d e r different social conditions. To give up, in the face of anomalies, Marbury v. Madison, or Erie R. R. v. Tompkins, or the "right of privacy" decisions on w h i c h the dissenting Justices in Bowers v. Hardwick relied w o u l d n o t cause the same

epistemic wrench as giving up the proposition that 2 + 2 = 4 or that all men are mortal. Authority in intellectual matters is best understood as a transmission belt that carries news of scientific or other intellectual discovery to persons lacking the time or background to verify the discovery themselves and that also authenticates the discovery for them. Authority in law is different. Judicial decisions are authoritative because they emanate f r o m a politically accredited source rather than because they are agreed to be correct by individuals in w h o m the c o m m u n i t y reposes an absolute epistemic trust. The trappings of judicial authority—the robe, the elaborate deference, the solemn rhetoric, and so forth—are clues to the political nature of that authority. Another clue is the doctrine of precedent itself, which in one sense is a refusal to correct errors—a posture that would be thought bizarre in scientific inquiry. T h e hierarchical structure of a legal system and the desire for stability that is encapsulated in the doctrine of stare decisis may advance "justice" in a variety of senses, by making judicial decisions more acceptable to the lay public and by reducing uncertainty, but they impede the search for truth. Justification in a scientific sense involves mounting, meeting, and overcoming challenges. 19 Beliefs that are not challenged tend to be weakly grounded—especially when challenge is actively discouraged. Systems of thought that emphasize hierarchy, tradition, authority, and precedent disvalue the kind of critical inquiry that tests belief and advances knowledge, and as a result the truths that such systems accept are not robust. This is notoriously true of religion, an activity in which the perceived costs of free inquiry are often very high. 20 It is also true of law, and is one reason that the scientific attitude is not at h o m e in the legal enterprise.

19. See, f o r e x a m p l e , C a r l W e l l m a n , Challenge and Response: Justification in Ethics 128, 167 (1971). In J o h n D e w e y ' s w o r d s , " I n e v e r y instance, f r o m passing q u e r y to elaborate scientific u n d e r t a k i n g , the art o f k n o w i n g criticizes a belief w h i c h has passed current as g e n u i n e coin, w i t h a v i e w t o its revision. It t e r m i n a t e s w h e n freer, richer and m o r e secure objects of belief are i n s t i t u t e d as g o o d s of i m m e d i a t e acceptance . . . Starting f r o m one g o o d , treated as a p p a r e n t and q u e s t i o n a b l e , and e n d i n g in a n o t h e r w h i c h is tested and s u b stantiated, t h e final act o f k n o w i n g is acceptance and intellectual appreciation of w h a t is significantly c o n c l u s i v e . " Experience and Nature 4 2 8 - 4 2 9 (1929). O f c o u r s e this is an idealized p i c t u r e of science; real science c o n t a i n s e x a m p l e s of appeals to a u t h o r i t y in a p p r o x i m a t e l y t h e legal sense. See D a v i d L. H u l l , Science as a Process: An Evolutionary Account of the Social and Conceptual Development of Science 374 (1988). 20. See, f o r e x a m p l e , D o u g l a s L a y c o c k and Susan E. Waelbroeck, " A c a d e m i c F r e e d o m and the Free Exercise of R e l i g i o n , " 66 Texas Law Review 1455, 1456-1458 (1988), discussing h o w m u c h a c a d e m i c f r e e d o m is o p t i m a l f o r t h e C a t h o l i c C h u r c h .

To be blunt, the ultima ratio of law is indeed force—precisely w h a t is excluded by even the m o s t latitudinarian definitions of rationality. Rationality "is simply a m e t h o d of being open and curious, and of relying on persuasion rather than force." 2 1 L a w is not characteristically "open and curious," and it relies on force as well as o n persuasion. If you ask h o w w e k n o w that Venus exerts a gravitational pull on Mars, the answer is that the people w h o study these things agree it does. If you ask h o w w e k n o w that the Fourteenth A m e n d m e n t forbids the states to prohibit certain abortions, the answer is that the people w h o have the political p o w e r to decide the issue—namely, the Justices of the Supreme C o u r t — h a v e so determined by m a j o r i t y vote. T h e example of abortion brings to light another i m p o r t a n t difference between authority in science and authority in law. In general, the m o r e scientists, textbooks, and so on that endorse a scientific proposition the more warranted w e are in accepting it. B u t the m o r e cases that endorse a proposition the shakier it m a y be. R e m e m b e r that people are unlikely to litigate a dispute if the judicial o u t c o m e is certain. If an issue keeps being relitigated this m a y be because the litigants are uncertain w h e t h e r the court will stick to its position, or because they cannot believe the court could have meant w h a t it said, or because they think the court is susceptible to pressure. T h e court's reaffirmation of its position in the face of repeated challenges m a y s h o w that the position is robust, but alternatively m a y be the prelude to the a b a n d o n m e n t of the position. All this said, there are m a n y judicial decisions—and m a n y legal p r o p ositions not requiring judicial e n d o r s e m e n t — t h a t n o t only all j u d g e s but all lawyers w o u l d agree w e r e correct. Laypersons are entitled to accept those decisions and propositions as "authoritative" in the a p p r o x i m a t e sense that scientific consensus is authoritative for laypersons. A n d there are m a n y questions about w h i c h scientists disagree, a l t h o u g h n o t too much should be m a d e of this. Scientists have procedures that enable t h e m to answer questions w i t h a high degree of confidence and then m o v e on to other and m o r e difficult questions, so that while at any m o m e n t the scientific c o m m u n i t y is full of controversy there is a sense that scientific k n o w l e d g e is g r o w i n g continuously. T h a t sense is missing in law. We do not think, for example, that although the question whether and to w h a t extent the C o n s t i t u t i o n should be interpreted to protect sexual f r e e d o m is u n d o u b t e d l y a difficult one, the legal c o m munity will eventually answer it and m o v e o n to other questions. T h e

21. Richard Rorty, "Is N a t u r a l Science a N a t u r a l K i n d ? " in Construction The Shaping of Scientific Rationality 49, 71 ( E r n a n M c M u l l i n ed. 1988).

and

Constraint:

question m a y eventually cease to be troubling, but not because it will have been answered to the satisfaction of all reasonable persons. It is not a question but a p r o f o u n d social problem. We cope with, w e manage, w e o u t g r o w — b u t w e rarely solve—such problems. This lack of closure, of convergence, 2 2 of progressivity—the sheer interminability o f so m u c h legal debate—makes the p r o b l e m of legal indeterminacy fundamentally different f r o m that of scientific or mathematical indeterminacy. A n o t h e r difference is that the scientific c o m m u n i t y itself largely determines the field of its inquiries; it is n o t forced to butt its head against a stone wall. 23 J u d g e s decide virtually all issues society flings at them, h o w e v e r intractable the issues m a y be. Occasionally the j u d g e s will balk, notably in cases raising "political questions." But this is a t e r m of art (and one decreasingly invoked); m a n y conventionally justiciable legal disputes are political to their core. Formalists seek to constrict the range ofjusticiable issues, while prudentialists in the style of the late Alexander Bickel want the courts to p o s t p o n e as long as possible intervening in highly charged political issues; I noted in Chapter 1 that Bowers v. Hardwick w o u l d have provided an apt occasion for ducking a hot issue in the n a m e of p r u d ence. But j u d g e s rarely heed such advice n o w a d a y s , and a n y w a y they do n o t escape politics w h e n they d r a w their h o r n s in for political reasons,2*1 j u s t as they do not escape politics w h e n they use political reasoning to create a rule that thereafter can be applied deductively, mechanically. I therefore question David Brink's suggestion that w e can appeal to the example of science to d e m o n s t r a t e that changes in legal rules are consistent w i t h law's being determinate. 2 5 T h e suggestion is, however, based on a plausible criticism of w h a t used to be the standard philosophical account of meaning. In that account the meaning of a w o r d is given b y its definition (the identifying properties conventionally associated w i t h it) and in turn determines the extension, or reference, of the 22. T h e p r o p o s i t i o n that science is " c o n v e r g i n g " on the " t r u t h " is, like o t h e r versions of scientific realism, c o n t r o v e r s i a l . See, f o r e x a m p l e , Ian H a c k i n g , Representing and Intervening: Introductory Topics in the Philosophy of Natural Science 55—57 (1983). It is not essential to m y a r g u m e n t . 23. See Richard W. Miller, Fact and Method: Explanation, Confirmation, and Reality in the Natural and the Social Sciences 133-134 (1987). O f c o u r s e social and political factors play a role in g u i d i n g scientific research; but only in totalitarian societies are scientists forced to do research that they consider scientifically b a r r e n . 24. See Jan G. D e u t s c h , " N e u t r a l i t y , Legitimacy, and the S u p r e m e C o u r t : S o m e Intersections b e t w e e n L a w and Political Science," 20 Stanford Law Review 169 (1968). 25. See D a v i d O. B r i n k , "Legal T h e o r y , Legal I n t e r p r e t a t i o n , and Judicial Review," 17 Philosophy and Public Affairs 105 (1988).

w o r d (the things it names). T h e standard account implies that m o d e r n scientists k n o w n o m o r e about the " a t o m " than D e m o c r i t u s did, because they define the w o r d differently and therefore it has a different extension. This is implausible; w h a t is " o u t there" also affects the m e a n ing of the word. 2 6 Therefore, B r i n k argues, we should not be troubled by the fact that the w o r d s "cruel and unusual p u n i s h m e n t s " meant something different to the authors of the Eighth A m e n d m e n t f r o m w h a t they mean to us; w e simply k n o w m o r e about cruel and unusual p u n ishments than they did. But t h o u g h it seems reasonably clear, at least to one of scientific-realist bent, that the universe contains objects of the sort that D e m o c r i t u s was g r o p i n g for and that m o d e r n scientists have a m u c h better grip on, it is unclear that o u r social universe contains objects or entities—cruel and unusual p u n i s h m e n t s — t o w a r d w h i c h both the authors of the Eighth A m e n d m e n t and m o d e r n j u d g e s and scholars were (are) groping. Punishments there were and are, but the characterizing of t h e m as cruel and unusual is an act that has n o close counterpart in physics. T h e physical e n v i r o n m e n t contributes to scientific meaning; the counterpart contribution of reality to legal m e a n i n g is unclear. N o t always: for example, there are monopolies " o u t there," and thanks to the progress of economic science w e k n o w m o r e about t h e m than the framers of the Sherman Act did. W h e t h e r there are cruel and unusual punishments out t h e r e — w h e t h e r they are m o r e than unstable local cultural artifacts—may be doubted. I d o not mean to depreciate the legal prohibition, or to contend that there are n o applications of it that w o u l d c o m m a n d a consensus a m o n g lawyers and j u d g e s , or to deny that there are scientific questions on which n o consensus has f o r m e d . But the quest for the meaning of the E i g h t h A m e n d m e n t is n o t in m y v i e w fruitfully described as a quest for greater k n o w l e d g e of w h a t has always been sitting out there in s o m e political or social counterpart of physical space. I add that even m y Sherman Act example is a m b i g u o u s , since w e are not sure that the Act's framers w e r e trying to talk about w h a t w e call a monopoly. T h e " i n c o m m e n s u r a b i l i t y " thesis inspired by the w o r k of T h o m a s K u h n is pertinent. In its m o s t radical form 2 7 the thesis denies that there is scientific progress, by asserting that every scientific revolution involves a change in the f r a m e of reference rather than an advance along a c o m m o n path. T h e implication is that f o r Aristotle, Ptolemy, the

26. See Hilary P u t n a m , Representation and Reality 12-14, 32, 36 (1988). 27. To w h i c h K u h n h i m s e l f does n o t subscribe. See T h o m a s S. K u h n , The Structure of Scientific Revolutions 169-170, 198-207 (2d ed. 1970).

medieval Christian church, and even modern sailors, the sun and the other heavenly bodies do revolve around the earth, because that is the simplest and most fruitful cosmology for their purposes. I am dubious. The ambitions of the geocentric school went beyond helping navigators; the geocentric theory was a refutable theory of the structure of the universe, and has been refuted. But the same cannot be said of the "theory" of cruel and unusual punishments held by nineteenth-century judges. It is not a theory that has been falsified as we have learned more about punishment. The idea that what passes as linear progress actually involves a succession of incommensurable frames of reference has greater application to legal and moral issues than to scientific ones. In a traditional Indian frame of reference suttee makes perfectly good sense; in a Western frame of reference it is barbaric. Neither translation nor mediation between these frames of references seems possible. 28 The issue of capital punishment can be debated fruitfully among persons w h o believe that deterrence is the dominant consideration in designing a system of punishments, because such persons occupy the same frame of reference. Whether it can be debated fruitfully among persons who disagree about the values at stake in a criminal justice system or w h o attach radically different weights to those values is unclear.

Reasoning by

Analogy

The heart of legal reasoning as conceived by most modern lawyers is reasoning by analogy. This method of practical reason has an impeccable Aristotelian pedigree, 29 but no definite content or integrity; it denotes an unstable class of disparate reasoning methods. This is an important point, not a quibble. With formal logic playing no role in legal reasoning, reasoning by analogy is the principal candidate for a method that 28. T h i s is the a n s w e r given b y m o r a l relativism, the p r o s and cons of w h i c h are the subject of a vast literature well represented b y Relativism: Interpretation and Confrontation (Michael Krausz ed. 1989). T h e issue b e t w e e n the relativists and the absolutists is well stated in Gilbert H a r m a n , "Is T h e r e a Single T r u e M o r a l i t y ? " in id. at 363. 29. As n o t e d in w h a t is still the a u t h o r i t a t i v e t r e a t m e n t of reasoning b y a n a l o g y in law: E d w a r d H . Levi, An Introduction to Legal Reasoning 1 n. 2 (1949). See also M a r t i n P. G o l d ing, Legal Reasoning 4 4 - 4 9 , 102-111 (1983). For c o m p e n d i o u s philosophical t r e a t m e n t s of analogy, see H i l a r y P u t n a m , The Many Faces of Realism: The Paul Carus Lectures 7 3 - 7 5 (1987); Q u i n e and Ullian, n o t e 15 a b o v e , at 90—95. T h e fallacious quality o f m u c h reasoning by analogy is well discussed in M o n r o e C . Beardsley, Thinking Straight: Principles of Reasoning for Readers and Writers 125 (3d ed. 1966) (§11), and in Brian Barry, " O n A n a l o g y , " 23 Political Studies 86 (1975).

will set lawyers apart f r o m everyday reasoners. Informal logic is an important method of reasoning and lawyers can take pride in being good at it. But in part because logic is more often a critical than a constructive tool in law, lawyers w h o recognize the limitations of logic (not all do) aspire to being good at more than logic, and the " m o r e " is often said to be reasoning by analogy. Aristotle gives the following example: "If then we wish to prove that [for the Athenians] to fight with the Thebans is an evil, we must assume that to fight against neighbours is an evil. Conviction of this is obtained from similar cases, e.g., that the war against the Phocians was an evil to the Thebans. Since then to fight against neighbours is an evil, and to fight against the Thebans is to fight against neighbours, it is clear that to fight against the Thebans is an evil." 30 Hardly. 31 If the war with the Phocians was an evil to the Thebans, the Thebans must have lost. W h y shouldn't the Athenians be optimistic about beating proven losers w h o probably are in a weakened state? Aristotle's example seems, in fact, either nonsense or induction. It is nonsense if taken to mean that if t w o things (a war of Athenians against Thebans and a war of Thebans against Phocians) have one property in c o m m o n (warring against neighbors), they probably have every other property in c o m m o n (such as experiencing an evil) as well. That is like saying that if A and B have dark hair, and A is a w o m a n , B is probably a w o m a n too. Aristotle's example is induction if taken to mean that experience shows that people w h o start wars against their neighbors lose—so Athens had better watch out. Induction will often suggest some regularity, some "law," such as that water boils at 212° Fahrenheit or that it is easier to start a war than to stop one, and then we cast about for a theory that might explain w h y and by doing so might help us avoid false predictions, such as that water will boil at 212° on a mountaintop. Even without a theory we often reason inductively, with more or less reliability depending on the circumstances. I have o w n e d Volvo automobiles (a total of four) since 1963, and I have been generally satisfied with them. I infer f r o m this experience that if I replace m y present Volvo with a new one I probably will be satisfied with the new one too. T h e prior purchases are "precedents" or "analogies" that create a certain like30. Prior Analytics, B k . II, §24, in The Complete Works of Aristotle: The Revised Oxford Translation, vol. 1, p. 110 (Jonathan B a r n e s ed. 1984) (p. 68, col. b, 1. 37, to p. 69, col. a, 1. 19, of the original Greek text). 31. See Ricardo M i g u e l Barrera, "Legal R e a s o n i n g a c c o r d i n g t o E d w a r d H . Levi: A n Epistemological C r i t i q u e " ( u n p u b l i s h e d s e m i n a r paper, U n i v e r s i t y of C h i c a g o L a w School, J a n . 12, 1988).

lihood that I will be satisfied if I b u y another Volvo the next time I am in the market for a car. This m a y be a sensible m e t h o d of reasoning, and it is hardly touched b y the lively philosophical debate, b e g u n by H u m e and n o t yet resolved, over the validity of induction (for o n e m u s t not confuse philosophical and practical doubts about induction, the latter being the sort of d o u b t that can land you in the madhouse). But c o m mon-sense, practical, everyday induction—well illustrated by the inferential procedure of a baby w h o burns himself on a stove—is a bit l o w keyed to be the core of legal reasoning in a sense flattering to the legal procession. C a n w e fancy u p o u r account of induction by speaking of scientific induction instead? 32 If reasoning by analogy is actually induction, m a y b e there is s o m e t h i n g to the nineteenth-century formalists' idea that law is an inductive science (see Introduction). If careful study of court cases revealed that promises were never enforced unless there were both an offer and an acceptance, w e w o u l d infer a rule or principle, corresponding to a law o f nature. H o w e v e r , the f o r m s of science m a y be present w h e r e the spirit is absent. Judges and law professors have a f r e e d o m in their inductive procedures that w o u l d be highly destructive in science. This is the f r e e d o m to reject an observation as normatively unsound. 3 3 It is as if a scientist could say, "I k n o w that the orbit of M e r c u r y is a n o m a l o u s in N e w t o n i a n cosmology, but I shall ignore that ugly fact because it w o u l d be nobler if that orbit w e r e different." O r , "I shall deem it to be different." O r , "I shall change it." Judges often and law professors always are free to reject a n o m a l o u s decisions; they are n o t imperatively required to reconsider their theories in order to a c c o m m o d a t e anomalies. T h e pressure b o t h to theorize and to adjust theory to observations is correspondingly relaxed. A related but deeper p r o b l e m w i t h the "inductive science" concept of law is that the existence of a pattern, although it m a y create an expectation that it will be followed in the future, cannot s h o w that it should be followed. T h e fact that n o promise had ever been enforced w i t h o u t b o t h an offer and an acceptance w o u l d not be a reason for refusing to enforce the next p r o m i s e that lacked one or b o t h elements. T h e inductive m e t h o d can isolate the c o m m o n element in the previous cases but cannot establish its indispensability, j u s t as a rule does not create an obli32. Recall C o h e n ' s c o m p a r i s o n o f scientific and legal i n d u c t i o n , in C h a p t e r 1; see also C o h e n , n o t e 4 a b o v e , at 7 1 - 7 2 . 33. A p o i n t n o t e d by Francis Bacon, the great scientific-legal inductivist. See Paul H . Kocher, "Francis Bacon on the Science of J u r i s p r u d e n c e , " 18 Journal of the History of Ideas 3, 20 (1957).

gation to follow it. Finally, the deepest problem with scientific induction is also a problem with legal induction. Generalizing f r o m observations is perilous. Water will boil at a lower temperature than 212° Fahrenheit at the top of a mountain, some wars against neighbors are w o n by the aggressor, and some Volvos are lemons. Likewise, the fact that ten cases have been decided one way does not prove that the next case, which is bound not to be identical in every respect to any of the previous ones, should be decided the same way too. The selection bias in litigation figures importantly here. Cases identical in all conceivably relevant respects to a string of previous cases are likely to be settled out of court (or never to arise in the first place) rather than to be litigated to the appellate level. Legal induction is likely to fail i n j u s t the cases where it is needed. I have suggested that reasoning by analogy is actually induction, yet the way in which lawyers purport to reason by analogy is often, and misleadingly, syllogistic (technically, enthymematic) rather than inductive. The property lawyer w h o says that oil and gas are analogous to rabbits, deer, and other wild animals is really proposing that the rule governing property rights in such animals—the "rule of capture"—is an instance of a more general rule that subsumes oil and gas: the rule that there are no nonpossessory property rights in fugitive resources. The problem is then to justify the general rule, which cannot be done either syllogistically or analogically. There is still another sense in which lawyers and other practical reasoners can be said to reason by analogy. Analogies, viewed simply as instances similar to the problem at hand (examples, anecdotes) rather than as steps in a logical demonstration or even as the pieces in a regular pattern (the Volvo and contract examples) on which an inductive inference might be based, provide a fund of ideas and information on which to draw in deciding what to do. It is c o m m o n sense that before a staff officer formulates a plan for a military campaign he consider "precedent" in the form of similar campaigns in the past; maybe that is all Aristotle meant by his military example. Previously decided cases supply lawyers and judges with a wealth of facts, reasons, and techniques pertinent to h o w a new case should be decided. (Admiralty's use of the deodand analogy is a dramatic example.) Cases are vicarious experiences. The use of cases as informative analogies must be distinguished f r o m their use as authorities, that is, f r o m the policy of decision according to precedent. All analogies are, f r o m the user's standpoint, precedents— that is, things that go before—whether or not they are authoritative.

Paying attention to precedents thus does not commit one to stare decisis; the issues of authority and analogy are distinct. T h e use of analogy, example, anecdote—and hence of "precedent" in the nonauthoritative sense—is inevitable in fields where theory is weak, as it is in military science, in advertising, in law, and in many other fields of human endeavor. I merely question whether reasoning by analogy, when distinguished f r o m logical deduction and scientific induction on the one hand and stare decisis on the other, deserves the hoopla and reverence that members of the legal profession have bestowed on it. Obviously if one has a case that raises for the first time the question whether automobile manufacturers should be liable for negligence to their ultimate consumers, 3 4 one will want to find out what previous cases said about similar questions of liability, such as the liability of drug manufacturers to the ultimate consumers of drugs. That is not a reasoning procedure special to law; it is like asking other owners of Volvos what their experience has been or seeing h o w other city-states have fared in wars against their neighbors. And one can conduct such canvasses with only the vaguest sense (it need not even be conscious) of what counts as "similar" or "analogous" instances, although a j u d g m e n t of similarity or analogy presupposes some, and possibly extensive, categorization. But unless a precedent is authoritative in the sense of announcing a major premise that cannot be questioned, it can be a source only of data that are anecdotal in character or of reasons, considerations, values, policies. The qualification in the last sentence needs to be emphasized: a previous case may be authoritative by virtue of stare decisis. 35 And the more precedents there are on a particular point the stronger the tug of stare decisis may be (recall the qualification noted in the preceding section). But counting cases to determine the force of stare decisis is not reasoning by analogy; it is weighing authority. These are different—even opposed—methods of appealing to a previous case, and it is remarkable that lawyers and judges run them together so often. When cases are

34. M a c P h e r s o n v. B u i c k M o t o r C o . , 217 N . Y . 382, 111 N . E . 1050 (1916)—Cardozo's classic m a n i p u l a t i o n of p r e c e d e n t — d i s c u s s e d in Levi, n o t e 29 above, at 9 - 2 4 . 35. For i l l u m i n a t i n g discussions of p r e c e d e n t s as authorities, see Frederick Schauer, " P r e c e d e n t , " 39 Stanford Law Review 571 (1987); C h a r l e s W. Collier, " P r e c e d e n t and Legal A u t h o r i t y : A Critical H i s t o r y , " 1988 Wisconsin Law Review 771; M a x Radin, " C a s e L a w and Stare Decisis: C o n c e r n i n g Prdjudizienrecht in Amerika," 33 Columbia Law Review 199 (1933). Schauer emphasizes, as d o I, t h e close parallels b e t w e e n the use o f precedent in law and its use in e v e r y d a y life. See especially 39 Stanford Law Review at 602-604. O n logical c o n u n d r u m s created b y decision a c c o r d i n g t o p r e c e d e n t , see Laurence Goldstein, " S o m e P r o b l e m s a b o u t P r e c e d e n t , " 43 Cambridge Law Journal 88 (1984), and references cited there.

viewed as experiences rather than as authorities, reasoning by analogy is a method of undermining legal certitude (at least initially) rather than of establishing it—a method by which established principles are subjected to continual retesting in the crucible of n e w cases. If the principles survive the tests, they are strengthened; the analogy to fallibilist theories of science should be apparent. But the n e w cases are tests, not mere instantiations, of the principles. When as in this example we think of law on the model of science, we challenge rather than celebrate law's hierarchical and authoritarian character. We come to see law as an uneasy compromise between science, where inquiry is sovereign, and theology, where authority is sovereign. In defense of reasoning by analogy it is tempting to point out that human beings have an innate capacity for recognizing patterns, an innate standard of similarity. 36 This is what enables us to recognize faces after an interval and objects seen f r o m a new angle. A set of cases can c o m pose a pattern. But when lawyers or judges differ on what pattern it composes, their disagreement cannot be resolved either by an appeal to an intuitive sense of pattern or by the methods of scientific induction. N o r will it do to emphasize the importance of analogy in the w o r k ings of the scientific—or the legal—imagination. 3 7 To see one problem as being like another that has already been solved is indeed to place the new problem on the road to solution. But here the difference between a logic of discovery and one of justification kicks in; analogy belongs to the former rather than to the latter. It is one thing to see oil and gas as analogous to rabbits and foxes, and another to justify the same (or different) legal treatment. To take a scientific example, suppose there is a form of radiation that is powerful enough to destroy a malignant t u m o r but that will also destroy the tissues through which it passes en route to the tumor. By analogy to the case of an attacking force that splits into small groups, which enables the fortress to be attacked f r o m all sides at once, the t u m o r can be treated safely and effectively by directing lowintensity beams on it f r o m different directions, so that it receives a

36. See H o w a r d M a r g o l i s , Patterns, Thinking, and Cognition: A Theory of Judgment (1987) (esp. pp. 113-114); W. V. O. Q u i n e , " N a t u r a l K i n d s , " in Naturalizing Epistemology 31 (Hilary K o r n b l i t h ed. 1985). 37. For e x a m p l e s f r o m science, see A n d r e w P i c k e r i n g , Constructing Quarks: A Sociological History of Particle Physics 12, 407 (1984); M a r y Hesse, Revolutions and Reconstructions in the Philosophy of Science, ch. 4 (1980). T h e use of n u m b e r s f o r c o u n t i n g and o t h e r c o m p u tation is in a sense reasoning b y analogy. A n d there are i m p o r t a n t e x a m p l e s in p h i l o s o p h y as well, Plato's cave b e i n g o n l y t h e best k n o w n ; w e shall e n c o u n t e r a n o t h e r e x a m p l e in C h a p t e r 11, in c o n n e c t i o n w i t h J u d i t h J a r v i s T h o m s o n ' s v i e w s o n a b o r t i o n .

higher dose of radiation than d o any of the tissues t h r o u g h which the beams pass. 38 T h e analogy is fine; b u t it is not evidence for the efficacy of the medical treatment and w o u l d n o t be cited in a paper writing up the discovery. It does n o t belong to the logic o f j u s t i f i c a t i o n . T h e mere assertion of an analogy may, it is true, have persuasive force in a psychological sense. M e t a p h o r s arc often persuasive in that sense, and they are a f o r m of analogy. M e t a p h o r is also a f o r m of redescription—an effort to change the w a y things look—and such efforts are i m p o r t a n t to intellectual change, including, as I shall note in C h a p t e r 4, doctrinal change in law. But j u d g e s aspire to m o r e than rhetorically effective, emotionally compelling, or even perspective-altering expression—and will n o t be c o m f o r t e d to be told that m u c h of their reasoning is m e t a p h o r i c — j u s t as they aspire to m o r e than the ability to decide cases on sound but inarticulable grounds. T h e limitations of reasoning b y analogy are b u t highlighted w h e n it is praised as a m e t h o d of m a k i n g sure that j u d g e s take only one step at a time, that is, create law incrementally. 3 9 First, the admonition does not help the j u d g e decide w h e t h e r to take each step; it merely licenses h i m to take small steps. Second, a series of small steps can add up to a giant stride, and although on the one hand m o v i n g incrementally gives j u d g e s a chance to stop as soon as experience demonstrates the error of their ways, on the other hand it m a y conceal f r o m t h e m the m a g n i t u d e of the change they are cumulatively effecting. T h e latter tendency m a y well dominate, given the limited feedback that judicial decision makers receive (see C h a p t e r 3). Third, there is n o metric f o r determining the social, political, or e c o n o m i c "distance" b e t w e e n a prior, "analogous" case and the present case. T h e j u d g e w h o analogizes oil and gas to rabbits and foxes m a y think he is taking a small step; actually he is i m p e d i n g the efficient exploitation of valuable resources. 4 " Fourth and related, a r g u m e n t by analogy and the closely related technique of the

38. See J o h n H . H o l l a n d et al., Induction: Processes of In ference, Learning, and Discovery 2 8 9 - 2 9 5 (1986). 39. See J o s e p h Raz, The Authority of Law: Essays on Law and Morality 180-210 (1979). 40. T h e " r u l e of c a p t u r e , " o n w h i c h see H o w a r d R. Williams and Charles J. M e y e r s , Oil and Gas Law, vol. 1, §§203.1, 204.4 (1988), m a k e s sense w h e n applied to things that are not scarce, such as rabbits and f o x e s in the p e r i o d w h e n the rule was d e v e l o p e d . It makes no sense w h e n applied t o scarce resources, such as oil and gas. B y n o t a l l o w i n g rights to f u t u r e use to be o b t a i n e d , the rule creates an i n c e n t i v e t o exploit the resource as quickly as p o s s i b l e — w h i c h is t o o quickly. See, for e x a m p l e , Richard J. Pierce, Jr., "State Regulation of N a t u r a l Gas in a Federally D e r e g u l a t e d M a r k e t : T h e T r a g e d y of the C o m m o n s Revisited," 73 Cornell Law Review 15, 2 0 - 2 3 (1987).

legal fiction are often used to disguise change as continuity, m a k i n g it difficult to evaluate or even to understand legal development. 4 1 T h e last point illustrates the rhetorical f u n c t i o n of citing previous cases in a judicial decision. O r , rather, rhetorical functions; citations of previous cases are also used to disguise fiat as reason, to establish p r o p ositions not in dispute and therefore not in need of support, and as sources f r o m which to quote general language that either is truistic or is contradicted by general language in other cases, w h i c h the opinion does not cite. Case citations often are used, in other w o r d s , to m a k e an o p i n ion look m o r e solid than it really is. 42 But m y m a i n point, w h i c h I will illustrate with an example of Neil M a c C o r m i c k ' s , 4 3 is that reasoning by analogy, even w h e n it is n o t rhetorical—or e n t h y m e m a t i c , or fallacious, or purely ornamental—is n o t actually a m e t h o d of reasoning, that is, of connecting premises to conclusions. M a c C o r m i c k discusses a case in which the plaintiff, despite the absence of contractual or other u n d e r standings, was suing to recoup the expenses he had incurred in saving the defendant's property. T h e court had previously allowed such recoveries w h e n life rather than p r o p e r t y was in j e o p a r d y ; the question n o w was whether to decide the p r o p e r t y case for the plaintiff too, by analogy to the life-saving case. To answer such a question intelligently requires not skill at spotting analogies but the ability to identify the considerations behind allowing such recoveries w h e n life is at stake, to appraise the force of these considerations w h e n only p r o p e r t y is at stake, and to bring to bear any other relevant considerations. Suppose the earlier cases had rested on the idea that the defendant, if asked before the e m e r g e n c y that endangered his life w h e t h e r he w o u l d be willing to reimburse the expenses of s o m e o n e w h o saved h i m , surely w o u l d have replied, " O f course!" T h e question w o u l d then be h o w plausible such an i m p u t a t i o n of promissory intent was w h e n s o m e t h i n g of lesser value to the o w n e r than life was at stake. T h e earlier cases w o u l d be sources of i n f o r m a t i o n useful in deciding the present case, b u t the decision w o u l d require a fresh ethical or policy j u d g m e n t . T h e distinction between legal precedent as i n f o r m a t i o n and as a u t h o r ity may seem to overlook the fact that the values, considerations, policies, and ethical insights f o u n d in previous decisions of the same or a

41. Sec m y b o o k Law and Literature: A Misunderstood Relation 2 - 4 (1988). 42. O n the m a n i p u l a t i v e character of judicial rhetoric, see Peter G o o d r i c h , Legal Discourse: Studies in Linguistics, Rhetoric, and Legal Analysis, pt. 2 (1987); also Law and Literature: A Misunderstood Relation, n o t e 41 above, ch. 6. 43. F r o m his b o o k Legal Reasoning and Legal Theory 1 6 1 - 1 6 3 (1978).

coequal or a higher court are entitled to greater weight—are more authoritative—in the decision of the present case than are the values, considerations, and so forth that might be gleaned f r o m other sources. But is it more than habit and indolence that causes lawyers and judges to look for the ethical and political materials of j u d g m e n t in previous decisions rather than in scholarly literature, statistical compendia, and everyday experience? Well, it is a little more; the blinkered vision that results may conduce to greater stability in law, which is a good, although not an unqualified one. And the millions of pages of reported cases do contain m u c h information and ethical insight. But they also contain vast amounts of misinformation and obsolete ethics, and there are no reliable techniques for w i n n o w i n g the wheat f r o m the chaff. A careful study would, I predict, show that judges w h o k n o w more about a particular field of law are less deferential toward precedent than equally able (and no more "restrained") judges w h o k n o w less about the same field. This is so even when the greater knowledge of the specialist j u d g e is due simply to a deeper immersion in the case law rather than to a study of nonlegal sources of wisdom as well. That immersion will bring to light conflicts a m o n g the precedents that will undermine the claim of any single one to be authoritative. A field of law looks less tidy to a specialist than to a generalist. A related but weaker hypothesis is that a specialized court will be less deferential to precedent and therefore less (rather than, as generally believed, more) predictable than a generalist court. It is a weaker hypothesis because the precedents to which the specialized court will be asked to defer will for the most part be its o w n precedents, the product of specialists rather than of generalists. A specialist court should not be censured for paying less attention to precedent than the generalist court does. Precedent is only one source of information on which to base j u d g m e n t , and the specialist court has additional sources that the generalist court lacks. Against this it can be argued that in seeking sources of guidance the j u d g e is looking not only for relevant experience but also for relevant commitments, and these are more likely to be found in judicial opinions than in statistical compilations or academic commentary. But the question is the role of precedent in cases where there is no previous decision squarely on point. Why must dicta (that is, the nonauthoritative parts of the judicial opinion) be given more weight than the considered views of scholars w h o may have spent years studying the particular legal problem involved, or of social scientists w h o may have spent a professional lifetime in systematic study of the pertinent social realities, merely because the j u d g e is an official? Even if judges should not feel limited to previous cases in seeking

guidance in deciding novel ones, it can be argued that they should not stray outside the b o u n d s of conventional moral and political opinion in their society—and therefore that an A m e r i c a n j u d g e is not to decide cases on the basis of the ethics of M a r x or Nietzsche, b u t is to stay within the circle m a r k e d out b y the values that have already gained a footing in our legal traditions. I w o n d e r . H o w does a n e w value get into the legal tradition? B y legislation or constitutional enactment only? C a n a j u d g e never be the first person to bring a n e w value, a n e w political or ethical insight, into the law? I am not arguing that academic or other extrajudicial texts should have the same authority that judicial decisions have. A judicial holding normally will t r u m p even a better-reasoned academic analysis because of the value that the law places o n stability, a value often t h o u g h not always p r o m o t e d b y sticking by w h a t has been decided. I a m arguing only that w h e n there is n o holding, w h e n there are only dicta, their weight o u g h t to be determined b y their intrinsic merit rather than b y their official source. A possible c o u n t e r a r g u m e n t is that the core of distinctively legal reasoning lies precisely in distinguishing b e t w e e n precedent as authority (holding) and precedent as i n f o r m a t i o n (dictum). Precedents can be read broadly or narrowly, and deciding w h i c h course to follow in a particular case is often said to be central to the art of lawyering and j u d g i n g . M a y b e , but this has n o t h i n g to d o w i t h logical or quasi-logical deduction, or even w i t h the handling of analogies. If, read as n a r r o w l y as possible, a precedent dictates the decision of a later case, that decision will be a decision based on precedent. If, read that narrowly, the precedent does n o t control the later case, but the court in the later case chooses to read the precedent m o r e broadly so that it will control, the key to the decision is precisely that choice, a choice n o t dictated b y precedent—a choice as to w h a t the precedent shall be. O n c e the choice has been made, the precedent, viewed as authority rather than example, drops o u t of the picture. T h e r e is n o practical difference between on the one hand treating a case as one of "first impression," and on the other hand s u b s u m i n g it u n d e r a previous case after first deciding as a matter of discretion to read the previous case broadly enough to enable the subsumption. 4 4

44. H u m e ' s discussion of r e a s o n i n g b y a n a l o g y in l a w is t h u s u n e x p e c t e d l y apt to c o n t e m p o r a r y A m e r i c a n l a w — a s well as p u n g e n t : "If direct l a w s and p r e c e d e n t s be w a n t i n g , i m p e r f e c t and indirect ones are b r o u g h t in aid; and the c o n t r o v e r t e d case is r a n g e d u n d e r t h e m b y analogical reasoning and c o m p a r i s o n s , and similitudes, and c o r r e s p o n d e n c i e s , w h i c h are o f t e n m o r e fanciful t h a n real. In general, it m a y safely be a f f i r m e d that j u r i s p r u d e n c e is, in this respect, different f r o m all t h e sciences; and that in m a n y of its nicer

T h e more narrowly a precedent is interpreted, the less force it has as rule, so that the choice between broad and narrow interpretations of precedent is a choice between more and less dependence on rules relative to standards and implicates the considerations examined in Chapter 1. Those considerations have nothing to do with logic or with reasoning analogically. This is not to deny that they are weighty considerations. Since rules are a necessary component of a rational legal system and precedents are an important source of rules in the Anglo-American legal system, it would be senseless always to read precedents in the narrowest possible manner; there would be no judge-made rules. Suppose that a previous case—involving, let us say, an interpretation of the estoppel defense to the statute of limitations—is "on all fours" with the present one except that in the earlier one the plaintiff had been left-handed and in the present one the plaintiff is right-handed. Could the previous case be "confined" to left-handed plaintiffs? It could not—but only because there is no consideration of policy or ethics that would justify so narrow an interpretation. So, not only does the decision whether to read a precedent narrowly or broadly have nothing to do with logical or analogical reasoning; even ascertaining the lower bound of reading narrowly is independent of these methods of reasoning. That the distinction between holding and dictum is not hard and fast is shown by the inability of courts to agree on an operational definition of these terms. 4 5 T h e reasons for reading a previous decision narrowly are multitudinous, and their weight will determine h o w much of a previous decision is deemed holding, h o w much dictum. The reasons include the later court's possession of additional information; the earlier court's lack of considered attention to the issue in question, maybe because it was peripheral to the case before that court; the later court's disagreement with the earlier court's analysis (as opposed to outcome—

questions, there c a n n o t p r o p e r l y be said to be t r u t h or f a l s e h o o d o n either side. If one pleader b r i n g the case u n d e r a n y f o r m e r l a w or p r e c e d e n t , b y a refined a n a l o g y or c o m p a r ison; the o p p o s i t e pleader is n o t at a loss to find an o p p o s i t e a n a l o g y or c o m p a r i s o n : and the p r e f e r e n c e given by the j u d g e is o f t e n f o u n d e d m o r e o n taste and i m a g i n a t i o n than o n any solid a r g u m e n t . Public utility is the general object of all c o u r t s of j u d i c a t u r e ; and this utility t o o requires a stable rule in all c o n t r o v e r s i e s ; b u t w h e r e several rules, nearly equal and indifferent, present t h e m s e l v e s , it is a very slight t u r n of t h o u g h t w h i c h fixes the decision in f a v o u r of either p a r t y . " An Enquiry Concerning Human Understanding 308—309 (3d cd., P. H . N i d d i t c h cd., 1975) ( A p p . 3). 45. See discussion in U n i t e d States v. C r a w l e y , 837 F.2d 291 (7th Cir. 1988); also Peter G o o d r i c h , Reading the Law: A Critical Introduction to Legal Method and Techniques 7 2 - 7 3 (1986).

for then the later court would incontestably be rejecting the holding of the previous case rather than anything that could be described as merely dictum); the earlier court's use of overboard language inadvertently encompassing the issue in the present case; the relative unimportance of having fixed rules in the particular field of law involved in the case; and, in short, virtually anything that might make the previous decision an unsuitable or dispensable guide to decision of the present one. T h e absence of all or most of these considerations would argue for a broad reading. All are considerations of policy, and they happen not to lend themselves to a neat formulaic or algorithmic definition. This discussion bears on the issue of the proper weight to be given dicta in comparison with unofficial sources of wisdom. The decision to classify a part of an earlier opinion as dictum is a decision to treat it as relatively nonweighty in the present case. This opens the door to consideration of other sources of guidance. And were it more generally recognized that at least in difficult cases—but they are disproportionately represented in litigation—the main significance of precedents is as information rather than as authority, as challenging instances rather than as confirming ones, judges and lawyers might make greater use than they do of nonlegal and comparative materials, which are important sources of information but not of authority. M o d e r n judicial opinions do cite nonlegal materials, sometimes promiscuously, but the vast majority of these citations reflect the reading of the law clerks rather than of the judges. The preoccupation with precedent as authority may be one of the causes of American judges' insensitivity to the ways in which foreign legal systems deal with problems similar to ours, since foreign decisions have no authority in an American court except in the rare case where a question of foreign law is presented. 46 Too many of our judicial opinions contain unexamined assumptions, conventional and perhaps shallow pieties, and confident assertions bottomed on prejudice and folklore. 47 46. M a y b e P r o f e s s o r G l e n d o n ' s brilliant s t u d y of f o r e i g n a p p r o a c h e s to the a b o r t i o n question will help us o v e r c o m e o u r p r o v i n c i a l i s m . See M a r y A n n G l e n d o n , Abortion and Divorce in Western Law 10—62 (1987). A n interesting p r o p o s a l f o r u p g r a d i n g the a u t h o r i t y of social science findings in law is m a d e in t w o articles b y Laurens Walker and J o h n M o n ahan: "Social Science Research in L a w : A N e w P a r a d i g m , " 43 American Psychologist 465 (1988), and "Social Facts: Scientific M e t h o d o l o g y as Legal P r e c e d e n t , " 76 California Law Review 877 (1988). 47. Here is a typical e x a m p l e of judicial certitude: " N o better i n s t r u m e n t has been devised for arriving at t r u t h than to give a p e r s o n in j e o p a r d y of serious loss notice of the case against h i m and o p p o r t u n i t y to m e e t it. N o r has a better w a y been f o u n d for g e n e r a t i n g the feeling, so i m p o r t a n t to a p o p u l a r g o v e r n m e n t , that j u s t i c e has been d o n e . " J o i n t A n t i fascist R e f u g e e C o m m . v. M c G r a t h , 341 U . S . 123, 1 7 1 - 1 7 2 (1951) ( F r a n k f u r t e r , J . , c o n -

The point to be particularly emphasized is that in a system of precedent it is the later court that has the whip hand, not the earlier court, the court that created the precedent. The later court decides whether to read the earlier decision broadly or narrowly and, if it cannot be narrowed sufficiently to distinguish the present case, whether to overrule it. That court has the power, and it also has more information, just by virtue of coming later. The decision of h o w much weight to give the earlier precedent—whether to apply it at all, and if so how broadly—is a pragmatic decision in which the uncertainty that will be created by a too casual attitude toward past decisions—and the additional work that such an attitude will create for the courts both by requiring more time on each case and, as a result of the greater uncertainty, engendering more cases—are compared with the increased risk of error that an uncritical view of past decisions will create. (This discussion contains an echo of Chapter 1; the balance is much like that which must be struck in deciding whether to follow a rule or create an exception.) Reasoning by analogy is not a technique for striking the balance.

A Note on Legal

Education

Searching the past for relevant experience and considerations, which I contend is the essence of legal reasoning by analogy, is an important and worthy dimension of practical reason. But it is so ordinary—a word I use not as a pejorative but merely to indicate the remoteness of reasoning by analogy f r o m an inferential technique that might have to be taught, as statistical inference or formal logic has to be taught—that one is led to wonder in just what sense the highly inductive, caseoriented, analogy-saturated "Socratic" method of legal instruction actually teaches reasoning at all, beyond honing the student's skills in identifying contradictions. 48 That is an important qualification—but is

curring) ( f o o t n o t e o m i t t e d ) . T h e o n l y s u p p o r t s o f f e r e d f o r these e m p h a t i c and n o n o b v i o u s p r o p o s i t i o n s are a q u o t a t i o n f r o m a speech of D a n i e l Webster and a q u o t a t i o n f r o m an English judicial o p i n i o n . T h e style suggests a d o g m a t i c r a t h e r t h a n an i n q u i r i n g m i n d . It is a characteristic judicial style. 48. T h e " S o c r a t i c " m e t h o d is well discussed in William C . H e f f e r n a n , " N o t Socrates, B u t P r o t a g o r a s : T h e Sophistic Basis of Legal E d u c a t i o n , " 29 Buffalo Law Review 399 (1980); William Epstein, " T h e Classical T r a d i t i o n of Dialectics and A m e r i c a n Legal E d u c a t i o n , " 31 Journal of Legal Education 399 (1981); J o h n O . C o l e , " T h e Socratic M e t h o d in Legal E d u cation: M o r a l D i s c o u r s e and A c c o m m o d a t i o n , " 35 Mercer Law Review 867 (1984); E d w i n W. P a t t e r s o n , " T h e C a s e M e t h o d in A m e r i c a n Legal E d u c a t i o n : Its O r i g i n s and O b j e c tives," 4 Journal of Legal Education 1 (1951). A reader of Plato's dialogues will notice genuine

identifying contradictions all that "thinking like a lawyer" comes to? The "Socratic" method familiarizes the student with legal materials, most of them written in the profession's standard style, which exaggerates the uniqueness and the power of the analytical methods that lawyers and judges use. And it imbues the student with that style, at the same time training him to exploit, by means of logic wielded as a critical tool, the indeterminacies in legal materials. To recur to an earlier "analogy," what law school teaches on the methodological side is a language rather than a method of reasoning: a culture, a vocabulary, a set of representative texts and problems. Courses in foreign languages do not claim to teach methods of reasoning. I have no desire to belittle conventional legal training (which in any event is not all that modern law schools offer). Without it a tedious apprenticeship would be required to equip a person to practice law and would not do so as well. O n e should recognize, however, the similarity of conventional legal training to the apprenticeship system that preceded and for a long time coexisted with it. T h e immersion injudicial opinions and other legal materials that is the hallmark of a legal education provides the student f r o m the outset with a simulacrum of practice; he is like an airline pilot training on a simulator. T h e teacher throws questions at him as would a senior partner and the student answers them by "researching" the legal materials placed before him by the casebook editors. 49

affinities b e t w e e n the m e t h o d of Socrates depicted in the dialogues and the " S o c r a t i c " m e t h o d of law school i n s t r u c t i o n . B o t h Socrates and the Socratic l a w school p r o f e s s o r identify s p o k e s m e n of r e c u r r e n t fallacies and d e m o n s t r a t e b y s h a r p q u e s t i o n i n g that t h e s p o k e s m e n are deeply c o n f u s e d ; b y a process of r e f u t i n g fallacious a p p r o a c h e s t h e true approach emerges. T h e m e t h o d s are particularly close w h e n , as is c o m m o n , the l a w p r o fessor finds a T h r a s y m a c h u s a m o n g his s t u d e n t s w h o t h e r e a f t e r serves as the s p o k e s m a n for the " l a w as politics" school. O f c o u r s e there are also m a n y differences b e t w e e n the t w o m e t h o d s , h a v i n g to do w i t h the difference in subject m a t t e r s , the difference in the social relationship b e t w e e n the Socratic figure and the o t h e r participants in the d i a l o g u e , and the ultimate objectives of the exercise. 49. Veblen e x a g g e r a t e d , yet had the k e r n a l of a g o o d p o i n t , w h e n he said: "In p o i n t of substantial merit the law school b e l o n g s in t h e m o d e r n u n i v e r s i t y n o m o r e than a school of fencing or dancing. T h i s is particularly t r u e of t h e A m e r i c a n l a w schools . . . and it is m o r e particularly true the m o r e consistently the 'case m e t h o d ' is adhered to. T h e s e schools d e v o t e themselves w i t h great singleness t o the t r a i n i n g of p r a c t i t i o n e r s , as distinct f r o m jurists; and their teachers stand in a relation to their s t u d e n t s a n a l o g o u s to that in w h i c h the 'coaches' stand to the athletes. W h a t is had in v i e w is the exigencies, e x p e d i e n t s and s t r a t e g y of successful practice; and n o t so m u c h a grasp of e v e n t h o s e quasi-scientific articles of metaphysics that lie at the r o o t of the legal s y s t e m . W h a t is required and inculcated in the w a y of a k n o w l e d g e of these e l e m e n t s of law is a familiarity w i t h their strategic use." T h o r s t e i n Veblen, The Higher Learning in America 211 (1918).

The cases studied in law school provide not only glimpses of the rules and rhetoric of law but also a kind of surrogate experience of life, and of the particular slices of life (such as crimes, breaches of contract, and racial discrimination) that the lawyer is most likely to encounter in his career at the bar. But I repeat that these voluminous case materials are not a reliable guide to the social phenomena depicted in them. 5 0 The immersion in cases that is so characteristic of both legal education and legal practice may be less a source of strength than a brute necessity indicative of weakness—a stopgap pending better (more scientific, empirical, interdisciplinary) legal theory rather than a superior alternative to such theory. This question could itself be examined empirically. If conventional lawyerly experience increases the power of legal analysis, we might expect a judge's performance (something more readily, though not easily, measurable than the performance of a practicing lawyer, by counting citations to the judge's opinions and the number and percentage of his cases that are reversed) 51 to vary systematically with the nature of his experience before appointment to the bench. We might also expect judges having greater experience as practicing lawyers to outperform judges with less such experience, other things (such as education) being equal. T h e most important thing that law school imparts to its students is a feel for the outer bounds of permissible legal argumentation at the time when the education is being imparted. (Later those bounds will change, of course.) What "thinking like a lawyer" means is not the use of special analytic powers but an awareness of approximately h o w plastic law is at the frontiers—neither infinitely plastic, as a Thrasymachus might think, nor rigid and predetermined, as many laypersons think—and of the permissible " m o v e s " in arguing for, or against, a change in the law. It is neither method nor doctrine, but a repertoire of acceptable arguments and a feel for the degree and character of doctrinal stability, or, more generally, for the contours of a professional culture—a professional culture lovable to some, hateful to others. 52 50. For a c o u r t to use B l a c k s t o n e as a u t h o r i t y f o r the p r o p o s i t i o n " t h a t natural b o n d s of affection lead parents to act in the best interests of their children," as in P a r h a m v. J. R., 442 U . S . 584, 602 (1979), is legal r e a s o n i n g at its m o s t provincial. 51. It should be a p p a r e n t f r o m earlier discussion w h y reversal rates cannot be the only m e a s u r e of judicial p e r f o r m a n c e (just as the best baseball hitter is n o t the one w h o strikes o u t the least). In m a k i n g this p o i n t I a d m i t to vocational bias. 52. Science is a c u l t u r e t o o , b u t a culture of u n f o r c e d i n q u i r y in a w a y that law is not.

3

G^c)

Other Illustrations of Practical Reasoning in Law

Interpretation If a neighbor, seeing your house on fire, phones you and says, "Your house is on fire," you had better be able to decode the message. To do so you will need a certain linguistic competence, of course—a competence in the comprehension and manipulation of symbols. But you will also need to k n o w something about the speaker's seriousness and reliability, that is, about character, capacity, and intentions. The process of understanding is thus not a logical process, although lawyers and judges often pretend it is, but a matter of understanding people, practices (such as living in houses), and the physical environment (the consequences of fire)—forms of understanding that depend on sharing the same basic life experiences. Alternatively, it is an imaginative process enabled by this sharing. We understand a message by putting ourselves in the speaker's shoes. 1 We imagine that we are seeing a house burn and telling the o w n er about it, and the congruence between the speaker's intentions and our imaginative reconstruction—the success of the latter—is what enables the communication to succeed. The role of imagination in understanding is one reason you can decode the sentence, "I'll eat my hat," as ironic (another is that it is a well-known idiom). You k n o w you'd never try

1. This is not to say that the listener m u s t be able to re-create in his m i n d an i m a g e in the speaker's m i n d . Suppose, to b o r r o w an e x a m p l e f r o m Gerald M a c C a l l u m (based o n Wittgenstein—see The Blue and Brown Books 3 |1958]), that y o u ask y o u r assistant to fetch you all the ashtrays he can find for a m e e t i n g at w h i c h there will be m a n y s m o k e r s but y o u d o n ' t k n o w h o w m a n y (an e x a m p l e that r a t h e r dates M a c C a l l u m ' s fine article). It w o u l d be idiotic of him to return e m p t y - h a n d e d , w i t h the excuse that he w a s n o t sure w h i c h of the ashtrays he had seen w e r e the ones y o u had in y o u r m i n d w h e n y o u dispatched h i m o n

to eat a hat (even though it's smaller than an elephant), and you assume that I am enough like you in this regard not to try either. 2 The difference between a logical proposition and a communication can be clarified with the aid of the syllogism "All men are mortal" f r o m Chapter 1. Its conclusion—that Socrates is mortal—is a logical proposition. But if you say to me, "Socrates is mortal," your statement is a communication, and to grasp its meaning I will need to k n o w a lot more than the rules of logic and the contents of the dictionary. The statement may have no intended meaning; you may be a parrot repeating what you heard, without comprehension. I may misunderstand the reference: you may be referring to your pet gerbil, "Socrates," rather than to its Athenian namesake. There may be confusion about the meaning of particular words: by "mortal" you may mean "fallible" rather than bound to die. There may be uncertainty about the m o o d of the utterance: by emphasizing the first w o r d and pitching the third word higher you may be expressing skepticism about the proposition rather than warranting its truth. As the last example suggests, the danger of misunderstanding a spoken communication is reduced by the fact that the speaker's inflection and facial expression help dispel ambiguities in his words; it is almost as if inflection and facial expression were additional words. Although inflection does not always dispel ambiguity—it can inject ambiguity, just as additional words can—a listener can seek clarification f r o m the speaker. T h e potential for misunderstanding a written communication is usually greater. N o t always: documents often are drafted at leisure, with special care taken to achieve clarity. But the potential for misunderstanding is great if the document, however carefully written, was written by someone w h o has since died or by a committee (all of whose members may be dead, too), or is in a foreign language, or was written hundreds of years ago. Interpreting the Constitution involves all these problems except translation f r o m a foreign language—and in some respects eighteenth-century English is a foreign language. Even interpreting recently enacted statutes can be a daunting task, because the authors may not have foreseen and addressed—or may have foreseen and decided not to address—the question of meaning or application that

the e r r a n d . See M a c C a l l u m , "Legislative I n t e n t , " in Essays in Legal Philosophy 237, 256-257 ( R o b e r t S. S u m m e r s ed. 1968). T h e visual i m a g e r y in m y fire e x a m p l e is thus inessential. 2. M a c C a l l u m ' s e x a m p l e of the fallacy o f l i t e r a l - m i n d e d n e s s has the assistant c o m p l y i n g w i t h the request to fetch all t h e ashtrays he can find b y r i p p i n g s o m e off the walls and stealing others. See id. at 256-257.

has arisen. This last point helps s h o w w h y the interpretation of judicial opinions is a less serious p r o b l e m f o r j u d g e s than the interpretation of statutes is. If the previous opinion did not address a question, that is a good reason for not treating the opinion as authoritative o n the question and for seeking the answer elsewhere. Whether one thinks of success in c o m m u n i c a t i o n as being dependent on the sender's and the recipient's experience of the same relevant practices (such as " s m o k i n g " and "getting things" in M a c C a l l u m ' s examples in notes 1 and 2) or on the recipient's exercise of imagination—and these may c o m e to the same t h i n g — t h e social distance b e t w e e n sender and recipient is critical. N o r m a l l y a person can readily understand w h a t he himself is saying (even if it is complex), because he already is, as it were, in his o w n shoes. Yet w h e n he reads s o m e t h i n g he w r o t e m a n y years ago, at a time w h e n he m a y have been in a sense a different person, his attempt at imaginative reconstruction m a y fail. O l d married couples understand each other's f r a g m e n t a r y utterances better than a stranger would, because married people g r o w to be alike b y sharing the same experiences (and also b y tailoring language to their precise c o m m u n i cation needs—the analogy is to customizing a w o r d processor's f o r mats). It is the same w i t h identical twins, w i t h people w h o have been brought up together, and w i t h people educated in the same way. T h e closer knit the c o m m u n i t y that includes the speaker and his audience, the easier the interpretive task. Equivalently, the m o r e h o m o g e n e o u s the interpretive c o m m u n i t y , the l o w e r the costs of o v e r c o m i n g the inevitable "noise" in the channels of c o m m u n i c a t i o n : another illustration of the importance of cultural h o m o g e n e i t y in fixing m e a n i n g and thereby securing w h a t passes for objectivity. O n e can see w h y the idea that understanding is imaginative reconstruction rather than simple decoding was a R o m a n t i c idea: 3 the R o m a n 3. N o t a b l y Schleiermacher's. See Richard E . Palmer, Hermeneutics: Interpretation Theory in Schleiermacher, Dilthey, Heidegger, and Gadamer 84-97 (1969). B u t it is older; it was Vico's idea before it was Schleiermacher's (see, f o r e x a m p l e , R. G. C o l l i n g w o o d , The Idea of History 65 [1946]), and Aristotle's b e f o r e it was Vico's. A n d in b e t w e e n A r i s t o t l e and Vico was the English j u d g e E d m u n d P l o w d e n — s e e E y s t o n v. S t u d d , 2 P l o w . 459, 467 (1574), discussed in Warren L e h m a n , " H o w to I n t e r p r e t a D i f f i c u l t S t a t u t e , " 1979 Wisconsin Law Review 4 8 9 — n o t to m e n t i o n the earlier E n g l i s h j u d g e s discussed in T h e o d o r e F. T. P l u c k nett, Statutes and Their Interpretation in the First Half of the Fourteenth Century (1922). See H a r o l d D e x t e r Hazeltine, " G e n e r a l Preface," in id. at v, xxii. A r i s t o t l e h i m s e l f had applied the idea to law: "All law is universal b u t a b o u t s o m e t h i n g s it is n o t possible to m a k e a universal s t a t e m e n t w h i c h will be correct. In t h o s e cases, t h e n , in w h i c h it is necessary to speak universally, b u t n o t possible to d o so correctly, t h e l a w takes the usual case, t h o u g h it is n o t i g n o r a n t of the possibility of error. A n d it is n o n e the less correct; f o r t h e e r r o r is

tics were trying to break d o w n the barriers between individuals. But the idea has only limited utility for the modern judge, particularly in constitutional interpretation. The modern j u d g e has little in c o m m o n with the draftsmen of the Constitution. It is futile for him to try to put himself in their place in order to figure out whether they would have wanted to strike d o w n laws forbidding abortions, sodomy, antitakeover statutes, affirmative action, the special-prosecutor law, or laws authorizing the censorship of student newspapers or pornographic videocassettes. T h e relevant practices are not ones we share with the framers. They did not have the experiences we have, they did not k n o w what we k n o w (and we have forgotten much of what they knew), and we haven't a clue to h o w they would have fitted our experiences to their values. Even the interpretive c o m m u n i t y that consists of the members of a recent legislature and the judges asked to interpret the legislature's enactments is not at all like one person, a married couple, or a group of friends chatting over lunch. 4 T h e Romantic ideal of an imaginative coalescence between writer and reader, when transposed to the legal setting, may be not only impossibly strenuous but unnecessarily so. Even though the hypostatization of a "legislative intent"—a group mind, when even the concept of a single mind is problematic (see Chapter 5)—is an insult to philosophy, statutes and constitutional provisions are purposive utterances. And often their purposes can be discerned f r o m text and context (including what the draftsmen may have said about the text in a committee report or hearings, or in floor debate) and used to answer an interpretive question in

n o t in the l a w n o r in the legislator b u t in the n a t u r e of the t h i n g . . . [ T h e r e f o r e ] it is right, w h e n the legislator fails us and has erred b y over-simplicity, to correct the o m i s s i o n — t o say w h a t the legislator h i m s e l f w o u l d have said had be been present, and w o u l d have put into his l a w if h e had k n o w n . " Nicomachean Ethics, Bk. V, §10, in The Complete Works oj Aristotle: The Revised Oxford Translation, vol. 2, pp. 1795-1796 (Jonathan Barnes ed. 1984) (p. 1137, col. b, 11. 12-24 of the G r e e k text); to similar effect see Rhetoric, Bk. I, §13, in id., vol. 2, p. 2188 (p. 1374 of the Greek text, col. a, 11. 18-36; col. b). See also R a y m o n d B. M a r c i n , "Epieikeia: E q u i t a b l e L a w m a k i n g in the C o n s t r u c t i o n of Statutes," 10 Connecticut Law Review 377 (1978). J e r e m y B e n t h a m , Of Laws in General 164-165 (H. L. A. H a r t ed. 1970), contains an interesting r e f o r m u l a t i o n of Aristotle's p o i n t . 4. A p r o b l e m that H a r t and Sacks elided in their influential version of interpretation as i m a g i n a t i v e r e c o n s t r u c t i o n b y a s s u m i n g that the legislators are j u s t like the j u d g e s . "A c o u r t s h o u l d try to p u t itself in i m a g i n a t i o n in t h e position of the legislature w h i c h enacted the m e a s u r e . . . It s h o u l d a s s u m e , unless the c o n t r a r y u n m i s t a k a b l y appears, that the legislature w a s m a d e u p of reasonable p e r s o n s p u r s u i n g reasonable p u r p o s e s reasonably." H e n r y M . H a r t , Jr., and A l b e r t M . Sacks, The Legal Process: Basic Problems in the Making and Application of Law 1414-1415 (tentative ed. 1958).

a way that advances the cooperative purpose set on foot by the enactment. It may not be necessary to "enter" a legislator's "mind" in order to understand and follow the command in a legislative enactment. It may, however, be necessary to project in the imagination the consequences of alternative interpretations; and the interpretation that has, all things considered, the better consequences may by virtue of that fact be the "correct" interpretation. The fact that interpretation is a mysterious process, distinct from logic and scientific observation, is not in itself a challenge to law's objectivity. The process by which scientists choose which hypotheses to test, the process Charles Sanders Peirce called "abduction," is mysterious too, but we know it works, because we can verify the results. We can often verify the results of communication as well. If I send out invitations to a party at a specific time and address, I am hypothesizing that the recipients (or some of them) will appear at that time and that address. If they do, the hypothesis is supported, and with it the theory of workable communication from which it is derived. But the type of communication that we call a statute or a constitution cannot be verified, a problem that produces the multiplicity of insecure approaches to statutory interpretation explored in Chapters 9 and 10.

Means-End

Rationality

The weatherman has forecast rain, and I must decide whether to take an umbrella when I leave the house. I will consider (very rapidly, perhaps unconsciously) the probability that the forecast is correct, the discomfort of being rained on, the bother of carrying the umbrella, and the probability of losing it. This type of analysis, called cost-benefit analysis by economists and means-end rationality (sometimes "deliberation") by philosophers of practical reason, is important in every department of thought and certainly in legal reasoning. The choice between alternative legal rules often depends on deciding which makes a closer fit with some underlying goal; we saw this in discussing interpretation. But the decisional process is policy or ethical analysis rather than anything unique to law. When the goal is agreed on and it is clear which of t w o alternative rules, interpretations, or applications is better suited to achieving it, then what I am calling policy or ethical analysis will conduce to a determinate outcome—which is fine but does not identify a distinctive method of legal reasoning. Consider the old jurisprudential chestnut about whether the benefi-

ciary in a will w h o kills his testator should be permitted to inherit. 5 T h e case is s o m e t i m e s analyzed as a conflict between the principle that donative intentions should be h o n o r e d and the principle that no one should be allowed to profit f r o m his o w n w r o n g d o i n g . 6 A n alternative approach dissolves the conflict and demonstrates b o t h the utility of means-end rationality in legal reasoning and the close relationship b e t w e e n it and imagination, the essential faculty for exploring alternative possible states of the w o r l d . 7 T h e approach is to ask w h e t h e r if the testator had t h o u g h t about the possibility of being m u r d e r e d by his heir he w o u l d have added to his will a clause disinheriting the murderer. (Notice the analogy to the m e t h o d used in the preceding chapter to determine w h e t h e r s o m e o n e w h o saves another's life or property should be entitled to r e i m b u r s e m e n t of his expenses by the rescued person.) A l m o s t certainly the testator w o u l d have included such a clause, so his donative intentions are h o n o r e d by forbidding the heir to inherit, and there is n o conflict b e t w e e n legal principles. (This is "imaginative reconstruction" writ small.) This conclusion, although highly satisfactory, o w e s n o t h i n g to legal reasoning in any distinct sense. A n o t h e r alternative approach, also illustrative of means-end rationality, asks whether disinheriting the m u r d e r e r is a suitable adjunct to the other sanctions for murder. Either approach provides a f r a m e w o r k for answering such f u r ther questions as w h e t h e r the heir should be disinherited if he killed the testator accidentally rather than deliberately. I have treated the m u r d e r i n g - h e i r p r o b l e m as one of adjusting t w o c o m p e t i n g principles of c o m m o n law, but the analysis of Riggs itself (note 5) is complicated b y the fact that the m u r d e r e r was the beneficiary u n d e r a will rather than an heir in the technical sense, that is, one w h o inherits if there is n o will. T h e victim had complied w i t h all the statutory formalities f o r a bequest, and the statute contained n o provision invalidating bequests to a testator's murderer. In order to rule against the m u r d e r e r the court had to interpret the statute as containing an implied provision to that effect. M o r e realistically, w h a t the court did was to graft an exception o n t o the statute, the better to carry out in proper Aristotelian fashion—the m a j o r i t y opinion in Riggs cites Aristo5. T h e best k n o w n of the m u r d e r i n g - h e i r cases is R i g g s v. Palmer, 115 N . Y . 506, 22 N . E . 188 (1889). 6. See, f o r e x a m p l e , R o n a l d D w o r k i n , Law's Empire 15-20 (1986); and for criticism, C h a r l e s Silver, " E l m e r ' s Case: A Legal Positivist Replies to D w o r k i n , " 6 Law and Philosophy 381 (1987). 7. For a g o o d discussion o f the i m a g i n a t i v e d i m e n s i o n o f deliberation, see J o h n Dewey, Human Nature and Conduct: An Introduction to Social Psychology 189-209 (1922).

tie's discussion of epieikeia (equity)—the desires the legislators w o u l d have had regarding the question if they had foreseen it. Vincent Wellman has argued that the fitting of means to ends is, and rightly so, the m e t h o d of justification used in law. Fidelity to precedent is just another consideration—another policy or principle—to be placed in the balance in deciding w h e t h e r a particular o u t c o m e w o u l d be a suitable means to the judicial end. 8 So far, so g o o d . But Wellman believes he has identified w h a t is at once a distinctive and an adequate f o r m of legal reasoning, w h e n actually w h a t he has d o n e is describe the m a k i n g of policy j u d g m e n t s under conditions of often radical uncertainty. Because such j u d g m e n t s , m a d e u n d e r such conditions, are unreliable, one is not surprised to find h i m a c k n o w l e d g i n g that his t h e o r y of law as practical reason "denies that legal statements are true or false." 9 His article illustrates b o t h halves of m y thesis: that legal reasoning is n o t special and that it often does n o t yield determinate outcomes. Given the identity of means-end rationality to cost-benefit analysis and the close relationship noted in C h a p t e r 1 b e t w e e n logical reasoning and economic models, such as Learned H a n d ' s f o r m u l a f o r negligence, one can understand w h y economics has m a d e such inroads into law in recent years: the implicit structure of m u c h legal reasoning is economic. The resistance to this d e v e l o p m e n t is understandable too. Lawyers and judges are reluctant to admit the degree to w h i c h legal reasoning is pervaded by policy considerations, e c o n o m i c or otherwise. R u p e r t Cross said that " w h e n a j u d g e does consider the pros and cons of each party's case [in a case not ruled by precedent], there is n o recognized n a m e for the reasoning." 1 0 T h e disconcerting characteristic of this analytic m e t h o d that dare not speak its n a m e is that different j u d g e s are apt to w e i g h the pros and cons differently. Even logic can be s u b s u m e d under means-end rationality. T h e l a w yer's critical use of logic is a m e t h o d of p r o m o t i n g consistency. C o n s i s tency is valued—logic is valued—because it helps with fitting means to ends. Such fitting is the core of rationality, and logic and rationality are closely related. Unfamiliar t h o u g h the t e r m is to m o s t lawyers and judges, means-end rationality is closer to the center of the legal enterprise than logic, a t e r m m u c h bandied about by the profession. O r than 8. See Vincent A. Wellman, "Practical R e a s o n i n g and Judicial J u s t i f i c a t i o n : T o w a r d an A d e q u a t e T h e o r y , " 57 University of Colorado Law Review 45, 87-115 (1985). 9. Id. at 108. 10. Precedent in English Law 194 (3d ed. 1977). T h e full title of C r o s s ' s b o o k is p e r t i n e n t ; as w e shall see in the next chapter, E n g l i s h j u d g e s d o n o t need t o m a k e policy j u d g m e n t s as o f t e n as A m e r i c a n j u d g e s .

reasoning by analogy. T h e key step in deciding w h e t h e r the rule of capture should be extended f r o m rabbits and foxes to oil and gas, or w h e t h e r the salvor's right to a reward f r o m life salvage should be extended to p r o p e r t y salvage, is to extract a goal f r o m the previous cases or f r o m other sources and then determine which decision in the n e w case will p r o m o t e that goal m o s t effectively and at least cost. I have classified means-end rationality under practical reason, but an alternative classification, w h i c h w o u l d g r o u p interpretation, analogy, and authority together w i t h logic, and exact observation together with means-end rationality, will help us distinguish between formalistic and pragmatic reasoning. T h e lawyer's h o p e for the m e t h o d s in the first g r o u p is that they will enable using existing k n o w l e d g e to answer a n e w question with minimal investigation of the new question. If all A's are B's, and C is an A, then w e can conclude that C is a B w i t h o u t examining C to see w h e t h e r it is a B. O r , if w e interpret a statute to forbid X conduct, then to decide w h e t h e r the statute applies to Y all w e need to determine is w h e t h e r Y is an instance of X , and w e m a y be able to determine this w i t h only a superficial examination of Y. If w e had great confidence in reasoning b y analogy, w e m i g h t decide that oil and gas are subject to the rule of capture w i t h o u t k n o w i n g anything more about t h e m than that they f l o w rather than sit quietly in one place. In contrast, means-end rationality and exact observation entail direct scrutiny of the n e w question, although existing k n o w l e d g e provides the indispensable j u m p i n g - o f f point for that scrutiny. T h e t w o groups of reasoning m e t h ods thus differ w i t h respect to willingness to confront the n e w question on its o w n terms rather than to insist on answering it exclusively on the basis of old experience. T h e first g r o u p is backward-looking, in the m a n n e r of Blackstonc, the second f o r w a r d - l o o k i n g , in the manner of B e n t h a m . Lawyers tend to be b a c k w a r d - l o o k i n g . T h a t is w h y they tend to value precedent as authority, rather than as testing instance: as a reason to act, rather than as i n f o r m a t i o n potentially useful in deciding h o w to act.

Tacit

Knowing

Philosophers as different f r o m each other as Michael Polanyi and Gilbert Ryle have emphasized that s o m e of o u r m o s t complex thinking is tacit, u n c o n s c i o u s . " T h e mathematical f o r m u l a for adjusting one's weight on 11. Sec, for e x a m p l e , Michael P o l a n y i , " T h e Logic of Tacit Inference," in Polanyi, Knowing and Being: Essays 138 ( M a r j o r i e G r e n e ed. 1969); Michael Polanyi and H a r r y

a bicyclc to keep f r o m falling is highly complex, yet without knowing the formula people learn to ride bicycles. People follow the incredibly intricate rules of language use (going far beyond what is taught in the name of grammar or syntax or vocabulary) without having any conscious knowledge of those rules, leading N o a m Chomsky, Jerry Fodor, and others to conclude that people must have a substantial innate facility for language. Many distinguished writers "write with their pen," and the examples are not limited to literature. 12 So much "thinking" is unconscious that the very concept of " m i n d " becomes problematic. Tacit knowledge is important in legal reasoning. Lawyers develop a feel, not fully articulable, for what types of argument are in the legal ballpark, and what are not ("thinking like a lawyer," again). A comparison can be drawn to the native English speaker w h o corrects the learner's description of a "red large barn." T h e native speaker is unlikely to be aware of any rule governing the order of adjectives; "red large barn" just doesn't sound right to him. There are plenty of logical " m o v e s " in law that just don't sound right to the experienced lawyer. O n e also speaks of lawyers w h o have "good j u d g m e n t " — s o m e ineffable compound of caution, detachment, imagination, and c o m m o n sense. But that is a different phenomenon. Sagacity and j u d g m e n t o w e little to legal training and experience (and much to age), being qualities that are brought to bear on legal methods, materials, and experiences rather than created by them. Young lawyers do not have better j u d g ment than persons of similar age and intelligence in other walks of life. The fundamental difficulty with using the concept of tacit knowledge to defend a view of law as determinate or legal reasoning as distinctive is that unless the possession of such knowledge is stipulated (as in m y language example, in which a native English speaker is defined vis-a-vis a novice as one w h o k n o w s the tacit rules of English), it can be gauged only by observing the uses to which it is put. We measure the bicycle rider's tacit knowledge by watching him ride; if he keeps falling d o w n ,

Prosch, Meaning 46-65 (1975); Gilbert Ryle, The Concept of Mind, ch. 2 (1949) ( e m p h a s i z i n g the difference b e t w e e n " k n o w i n g h o w " and " k n o w i n g t h a t " ) . B u t cf. J o h n Searlc, Minds, Brains, and Science 51-52 (1984). 12. I give s o m e e x a m p l e s of u n c o n s c i o u s literary creativity in m y b o o k Law and Literature: A Misunderstood Relation 231-232 (1988). W i t t g e n s t e i n said that w h e n he sat d o w n to w r i t e p h i l o s o p h y he had n o idea w h a t he was g o i n g t o say. See Culture and Value 17e (G. H. v o n Wright ed. 1980). H e r e is the t e s t i m o n y of a n o t h e r d i s t i n g u i s h e d p h i l o s o p h e r : "it is only as I w r i t e that 1 discover w h a t I t h i n k , and this is a l m o s t n e v e r w h a t I t h o u g h t w h e n I b e g a n . " D o n a l d D a v i d s o n , " P o s t s c r i p t t o Replies," in Essays on Davidson: Actions and Events 253 (Bruce V e r m a z e n and Merrill B. H i n t i k k a eds. 1985).

we conclude that he lacks it. What are the counterparts in law to the rider w h o keeps his balance and the rider w h o keeps falling down? The market for legal services provides some criteria for evaluating the performance of practicing lawyers, but the criteria are imperfect because information about causality is sparse; surprisingly little is k n o w n about successful technique in law. E n o u g h is known, however, to make clear that skill in legal reasoning, as measured for example by law-school exams, is only a part of that technique, so that observing successful lawyers in action provides incomplete and often misleading evidence of skill at legal reasoning. T h e evaluation of judicial performance is also difficult, which in turn makes it difficult to determine which judges are well endowed with the requisite tacit knowledge and which poorly endowed. T h e underlying problem is that so little is k n o w n about the consequences of legal decisions. N o t only are the usual methods of scientific verification unavailable but so are the commonplace observations and experiences that enable us to correct our everyday behavior, whether in riding a bicycle, changing a fuse, or assembling a piece of equipment. C o m m o n sense cannot answer the question whether the preservation of political or religious freedom requires a broad reading of the First A m e n d m e n t , or whether the exclusionary rule is needed as an adjunct to the tort remedies against unreasonable police searches. 13 Often when people have difficulty evaluating the output of a process they evaluate its inputs instead. This may be one reason for the emphasis placed by employers (particularly academic employers) on the lawyer's formal credentials. It is easier to determine whether a lawyer did well in law school than whether he is a good lawyer. But it is difficult to monitor even the inputs into judging, let alone its outputs. In this country, judges usually are middle-aged when first appointed to the bench and their academic performance many years earlier is an uncertain predictor of their judicial performance, though not a worthless one. A disproportionate though very small n u m b e r of distinguished judges have been stellar law students (examples that come immediately to mind are Brandeis, Frankfurter, and Friendly), but some stellar law students have turned out to be undistinguished judges, and in any event it would be a mistake to stock the bench exclusively f r o m the ranks of the academically gifted, 13. O n the d i f f i c u l t y of e v a l u a t i n g the c o n s e q u e n c e s of law, see, for e x a m p l e , Gerald N . R o s e n b e r g , " P r o t e c t i n g F u n d a m e n t a l Liberties: T h e C o n s t i t u t i o n in C o n t e x t " ( u n p u b lished paper, U n i v e r s i t y o f C h i c a g o , D e p a r t m e n t of Political Science, 1988); Richard A. Posner, " T h e C o n s t i t u t i o n as an E c o n o m i c D o c u m e n t , " 56 George Washington Law Review 4, 27-31 (1987).

even if this were politically feasible. It would deprive the bench of necessary diversity. Academic prowess to one side, we do not k n o w what readily observable traits predict excellent performance in a judge, and only in part is this due to divergent opinions as to what an excellent judge is. This discussion bears on the question whether law is an autonomous discipline. Perhaps, even if the lawyer's analytical tools are no different f r o m those of everyday life, they can be so sharpened by training and experience as to yield a qualitatively distinct product. T h e analogy would be to the difference between an amateur musician and a professional one. But it is not a close analogy. We can distinguish between the two musicians by listening to them play; h o w do we distinguish between lawyers and ordinary reasoners as far as analytical power is concerned? Obviously lawyers are m u c h more comfortable with legal questions. But many questions can be cast as either legal or political (or ethical, or policy) questions—abortion, for example. H o w do we decide whether it is best to formulate them as legal questions and thus give them to lawyers to decide? Although of only limited applicability to questions of legal justification, the concept of tacit knowledge does offer a partial antidote to skepticism about the judicial process. When discussing the process, judges and lawyers, including law professors, often seem to ascribe to judges unrealistic capacities for reflection, detachment, and analysis—capacities inconsistent with the conditions under which judges w o r k and with the criteria by which they are selected. As a description of the conscious thoughts of judges this talk is indeed unrealistic and inflated—but so would be a "model" of the bike rider that showed him making continuous mathematical calculations to avoid losing his balance. By virtue of background, training, and experience both judicial and pre-judicial—in part simply by virtue of being a m e m b e r of society—a j u d g e may be capable of making j u d g m e n t s that can be modeled in highly complex terms even though they are not the terms in which the j u d g e thinks. 14

14. C o m p a r e J o h n D e w e y : " L o n g b r o o d i n g o v e r c o n d i t i o n s , i n t i m a t e c o n t a c t associated w i t h keen interest, t h o r o u g h a b s o r p t i o n in a m u l t i p l i c i t y o f allied experiences, tend to b r i n g a b o u t those j u d g m e n t s w h i c h w e t h e n call ' i n t u i t i v e ' ; b u t t h e y are t r u e j u d g m e n t s , because they are based o n intelligent selection and e s t i m a t i o n , w i t h s o l u t i o n of a p r o b l e m as the controlling s t a n d a r d . " How We Think: A Restatement of the Relation of Reflective Thinking to the Educative Process 124 (1933). A n d c o m p a r e J. H . H e x t e r ' s c o n t r a s t ( p a r a p h r a s i n g C o k e and Hale) b e t w e e n the c o m m o n l a w y e r ' s i m m e r s i o n in and l o n g s t u d y o f particulars, l e a d ing to " e r u d i t e experienced alertness," w i t h the logician's "analytical d e d u c t i v e efficiency." " T h o m a s H o b b e s and the Law," 65 Cornell Law Review 471, 485 (1980).

So there may be great judicial decisions—yet we may be unable to determine which ones they are! This last observation recalls the distinction between the formation and the justification of beliefs, decisions, and actions. Tacit knowing is a way of deciding to (or h o w to) do something, but it is not a way ofjustifying the decision or action. The only justification is in the doing or, as in my language example, in the saying. In correcting the word order in "red large barn" you don't defend the rule that an adjective of size precedes one of color, and probably you are not even aware that there is such a rule. You just say it correctly. When we speak of legal reasoning we normally have in mind the articulation of the principles and other considerations that show that a decision is correct. Such articulation cannot exhibit or give proof of tacit knowing.

Submitting

to the Test of Time

An underestimated device of practical reason is to subject a proposition to the test of time and to accept it if it passes. Though particularly important in aesthetic evaluation," the test also figures in the evaluation of factual (including scientific) and even legal propositions. It is a refinement of the idea that whatever most people think is probably true. The idea itself is unhelpful even if we think "truth" a meaningful and useful concept. Most people are ignorant about most matters, and history is littered with examples of consensus on matters of fact that we now know—or think we know—are false. But the longer a widespread belief persists, surviving changes in outlook and culture and advances in knowledge, the likelier it is to be correct; the intergenerational consensus is more reliable (you can't fool all of the people all of the time). Much of our stock of commonsense knowledge and elementary moral beliefs is validated in this and no other way. The test of time is a commonsense idea; certainly this is how it was presented by its first influential expositor, Samuel Johnson, and by such important successors as George Orwell. 1 6 But it also has an interesting 15. See A n t h o n y Savile, The Test of Time: An Essay in Philosophical Aesthetics (1982); Law and Literature, n o t e 12 a b o v e , ch. 2. A n d f o r its use in law, see E d w i n W. Patterson, " H i s t o r i c a l and E v o l u t i o n a r y T h e o r i e s of L a w , " 51 Columbia Law Review 681, 685 (1951). P a t t e r s o n cites D e w e y and J a m e s , w h i c h s u g g e s t s the congeniality of the test of t i m e to pragmatist philosophy. 16. See J o h n s o n , " P r e f a c e t o the Plays of Shakespeare," in Samuel Johnson: Selected Poetry and Prose 299, 300 (Frank B r a d y and W. K. W i m s a t t eds. 1977); D a v i d H u m e , " O f the Standard of Taste," in H u m e , Essays: Moral, Political, and Literary 226, 231-233 (Eugene F.

philosophical, scientific, and e c o n o m i c provenance. It can be seen to rest on the pragmatic idea that, w i t h the debatable exception of elementary logical and mathematical truths, consensus is the only operational concept of truth that w e have. This is n o t to equate t r u t h w i t h consensus. That w o u l d be a big mistake. It is n o t misusing the w o r d " t r u e " to say that "everyone except m e believes the proposition p, but p is not true." Indeed, it w o u l d stifle inquiry to suppose that w h e n consensus was achieved, truth had been f o u n d . T h e challenging of settled beliefs is an essential spur to intellectual progress. A n d this implies that settled beliefs are often false, as of course they are. To equate t r u t h to consensus w o u l d imply that the earth once was flat, and n o w is r o u n d ; that the sun used to revolve around the earth but n o w the earth revolves around the sun. Taken literally, it w o u l d imply that if h u m a n i t y w e r e destroyed, the heliocentric theory w o u l d cease to be true.' 7 T h e r e is equal paradox,

Miller ed. 1985); O r w e l l , "Lear, Tolstoy, and the F o o l , " in The Collected Essays, Journalism and Letters of George Orwell, vol. 4, pp. 287, 290 (Sonia O r w e l l and Ian A n g u s eds. 1968); and o t h e r references in Law and Literature, n o t e 12 above, at 71-74. For a rcccnt criticism of the test of t i m e in literature, see B a r b a r a H e r r n s t e i n S m i t h , Contingencies of Value: Alternative Perspectives for Critical Theory 47-53 (1988). She argues, for e x a m p l e , that since H o m e r means n o t h i n g to m o s t people in Africa, its survival in Western c u l t u r e m u s t be due to its fulfilling culturally specific f u n c t i o n s . G r a n t e d — b u t the q u e s t i o n is w h y H o m e r , rather than The Egyptian Book of the Dead or The Epic of Gilgamesh? Is it just racism, as a reader of M a r t i n Bernal, Black Athena: The Afroasiatic Roots of Classical Civilization, vol. 1, The Fabrication of Ancient Greece, 1785-1985 (1987), m i g h t infer? O r is there n o t s o m e t h i n g genuinely " i n " the H o m e r i c epics that m a k e s t h e m so r e s o n a n t in Western c u l t u r e even after m o r e than t w o and a half millennia? S m i t h ' s w o r k is part of a g r o w i n g school of " c a n o n b a s h i n g " that raises issues w h i c h r a n g e t o o far afield to be discussed here. For another e x a m p l e of the school, see J a n e T o m p k i n s , " M a s t e r p i e c e T h e a t e r : T h e Politics of H a w t h o r n e ' s Literary R e p u t a t i o n , " 36 American Quarterly 617 (1984). 17. A p a r a d o x i g n o r e d in Peirce's f o r m u l a t i o n : " T h e o p i n i o n w h i c h is fated to be ultimately agreed to b y all w h o investigate, is w h a t w e m e a n b y the t r u t h . " " H o w to M a k e O u r Ideas Clear," in Collected Papers of Charles Sanders Peirce, vol. 5, pp. 248, 268 (Charles H a r t s h o r n c and Paul Weiss eds. 1934). William J a m e s ' s f o r m u l a t i o n — t r u t h is w h a t is useful to believe—also leads to serious p a r a d o x e s . See B e r t r a n d Russell, " W i l l i a m J a m e s ' s C o n ception of T r u t h , " in Russell, Philosophical Essays 127 (1910). B u t Peirce's fuller s t a t e m e n t — " T h e real, then, is that w h i c h , s o o n e r or later, i n f o r m a t i o n and r e a s o n i n g w o u l d finally result in, and which is therefore independent of the vagaries of me and you" ( " S o m e C o n s e q u e n c e s of Four Incapacities," in Collected Papers of Charles Sanders Peirce, a b o v e , vol. 5, pp. 156, 186; emphasis added)—is c o n v e r g e n t w i t h m o r e c o m m o n n o t i o n s of t r u t h , w h i c h likewise stress o b s e r v e r i n d e p e n d e n c e , that is, o b j e c t i v i t y in o n e of its s t r o n g e r senses. See, for example, Hilary P u t n a m , Representation and Reality 109 (1988) ("it is a p r o p e r t y of t r u t h that w h e t h e r a sentence is t r u e is logically i n d e p e n d e n t of w h e t h e r a m a j o r i t y of the m e m bers of the culture believe it to be t r u e " ) . Peirce's stress o n the l o n g r u n w a s related to his view that scientific i n d u c t i o n is a f o r m of statistical s a m p l i n g . T h e larger the sample, the m o r e reliable the findings based o n it.

however, in using current opinion as the criterion of truth, and thus in saying that the earth is r o u n d because our present m e t h o d s of validating propositions s h o w it to be r o u n d although future m e t h o d s m a y s h o w otherwise. This is equivalent to saying that the earth really is r o u n d today but a century f r o m n o w w e m a y k n o w better—quite a confused statement. Pragmatists believe that t r u t h is w h a t free inquiry—unforced, undistorted, and u n i n t e r r u p t e d — w o u l d eventually discover about the objects of inquiry. 1 8 Since the process of inquiry never ends, this implies that truth always lies b e y o n d o u r horizon: it is there, but w e aren't. T h e p r a g m a t i c concept of t r u t h is related to the test of time, but complexly. T h e p r a g m a t i c concept is f o r w a r d - l o o k i n g ; truth is the destination w e have n o t yet arrived at, but u n d e r the right conditions w e can hope to arrive there eventually. T h e test of time is b a c k w a r d - l o o k i n g ; some of our beliefs are true, and they are probably the ones that, having survived the longest, c o m m a n d the m o s t robust consensus. But the t w o approaches are at o n e in referring truth to a process o f belief formation that unfolds over time. In so doing they depart f r o m n o r m a l usage, in w h i c h " t r u t h " and "belief" are sharply distinguished (see note 17). So what? It m i g h t be well to forget about " t r u t h " and speak only of justified or, better (because weaker), w a r r a n t e d belief (and this is the tendency in pragmatist thought), recognizing that there is n o way to distinguish in practice b e t w e e n t r u t h and w h a t one cannot help believing, and n o t i n g that consensus is a legitimate t h o u g h highly fallible m e t h o d ofjustification. This is a skeptical posture b u t n o t of an anguished or radical cast. If absolute t r u t h is n o t in the cards w e should not be upset w h e n w e fail to find it. It is n o t as if w e shall be left w i t h o u t any m o o r i n g s in what w e call reality. T h e pragmatist's skepticism is not the skepticism that doubts the existence of a w o r l d external to h u m a n sensation or even the existence of (as distinct f r o m the knowability of) truth. We m a y agree that the earth is "real," that it is either r o u n d or not round, and that, if it is r o u n d today, almost certainly it was r o u n d 2,500 years ago, w h e n the consensus view was that it was flat. Today's consensus is strongly 18. See p r e c e d i n g f o o t n o t e , and also Richard R o r t y , "Is N a t u r a l Science a N a t u r a l K i n d ? " in Construction and Constraint: The Shaping of Scientific Rationality 49, 71 (Ernan M c M u l l i n ed. 1988); C a r l W e l l m a n , Challenge and Response: Justification in Ethics (1971); William M . Sullivan, " A f t e r F o u n d a t i o n a l i s m : T h e R e t u r n to Practical P h i l o s o p h y , " in Anti-Foundationalism and Practical Reasoning: Conversations between Hermeneutics and Analysis 21 (Evan S i m p s o n ed. 1981).

supported by the kinds of evidence that we have good reason to find persuasive (note the element of circularity in this method of validation, though). We rate the probability that we are right and the ancients wrong on this question very high. Closer to truth w e cannot come, so why worry? There is reason to worry. Views that many people regard as preposterous display remarkable tenacity, yet this is not taken as evidence of their truth. That a majority of the American people believe (though often without full conviction) in magic, miracles, and astrology, and that these beliefs are thousands of years old, has no persuasive force with disbelievers, w h o regard the believers as ignorant and credulous. This is an example of refusal to abide by the test of time, and more broadly of persistent deviation between belief and reality, consensus and truth. Yet although it is a weak method of determining truth, the test of time is not negligible. It helps repair the deficiencies of consensus as a criterion of truth. Precisely because consensus is a provisional, uncertain, mutable criterion, the broader the consensus on a particular matter the greater its reliability. Time enables the consensus to be broadened; it enlarges the franchise. It also enlarges or multiplies our perspectives, which is vital if, like Nietzsche and Wittgenstein, we hold a perspectival theory of truth, which by emphasizing the extent to which a person's outlook is shaped by his culture and historical situation denies our ability to achieve a God's-eye view. An economic and biological metaphor is an alternative to the electoral metaphor for the test of time: in this construction the test of time subjects ideas to a competitive or Darwinian test, which makes any consensus that emerges more convincing (comparison shopping across the centuries). We think a competitive economy more likely to meet consumer needs than a centrally managed one, a decentralized scientific c o m m u nity in a society committed to freedom of speech and expression more conducive to accurate scientific j u d g m e n t s than a scientific c o m m u n i t y that is subject to censorship or that operates under tight governmental control—although evolution in business and in science is not powered by the same blind and random process that is at w o r k in biological evolution. 1 '' When expressed in the biological metaphor the test of time can be seen to be implicit in Popper's philosophy of science—the highly

19. T h i s is a historically i m p o r t a n t d i s t i n c t i o n . T h e mindlessness of biological e v o l u t i o n is the m o s t arresting aspect of D a r w i n i s m . T h e w o n d e r is n o t that m a n e v o l v e d f r o m the l o w e r animals but that he did so w i t h o u t its b e i n g p l a n n e d .

influential, though flawed, 2 " culmination of the fallibilist tradition whose earlier exponents include Bacon and Mill. Competitive and evolutionary in character, Popper's philosophy teaches that the best methodological rule is to "try out, and aim at, bold theories, with great informative content; and then let those bold theories compete, by discussing them critically and by testing them severely." 21 The more tests they survive, the more confidence we can have that they are a close approximation to truth. And the process of severe and thorough testing takes time. N e w t o n ' s laws of motion, until Einstein, had survived repeated tests triumphantly. T h e test of time becomes especially important for objectivity in law if, giving up on the attempt to discover a distinctive methodology of legal reasoning, we describe the lawyer's and the judge's reasoning as the "art" of social governance by rules 22 (which may be just a fancy term for tacit knowledge). When we reflect on the fact that law schools do 20. For illustrative criticism, see H i l a r y P u t n a m , " T h e ' C o r r o b o r a t i o n ' of T h e o r i e s , " in P u t n a m , Mathematics, Matter, and Method 250 (2d ed. 1979) (vol. 1 of his Philosophical Papers). P o p p e r ' s n o t i o n of testing h y p o t h e s e s that h a v e the least antecedent probability seems particularly questionable; and since there is an infinite n u m b e r of such h y p o t h e s e s , his i m p l i e d research p r o g r a m is h i g h l y inefficient. (Granted, it w o u l d not be a s o u n d research s t r a t e g y to test o n l y the h y p o t h e s e s that are consistent w i t h existing k n o w l e d g e . C o n f i r m a t i o n of t h o s e h y p o t h e s e s — t h e likeliest o u t c o m e of such a s t r a t e g y — w o u l d add n o t h i n g to o u r k n o w l e d g e and w o u l d narcotize rather t h a n s t i m u l a t e o u r i n q u i r i n g faculties.) Peirce m a y have been o n the r i g h t track in s u g g e s t i n g that w e are able to choose p r o m i s i n g h y p o t h e s e s t o test because o u r m i n d s , b e i n g themselves the p r o d u c t s of nature, have an i n t u i t i v e grasp of the principles of n a t u r e . See N i c h o l a s Rescher, Peirce's Philosophy of Science: Critical Studies in His Theory of Induction and Scientific Method (1978). This is the project of w h a t is n o w called e v o l u t i o n a r y or D a r w i n i a n e p i s t e m o l o g y . See, for example, Florian v o n Schilcher and N e i l T e n n a n t , Philosophy, Evolution, and Human Nature, ch. 3 (1984); M i c h a e l Ruse, Taking Darwin Seriously: A Naturalistic Approach to Philosophy, ch. 5 (1986). For a nice s u m m a r y of P o p p e r ' s p h i l o s o p h y of science, see D a v i d O l d r o y d , The Arch of Knowledge: An Introduction to the History of the Philosophy and Methodology of Science, ch. 8 (1986); also D a v i d Miller, "Falsification versus I n d u c t i v i s m , " in Applications of Inductive Logic 109 (L. J o n a t h a n C o h e n and M a r y Hesse eds. 1980). T h e D a r w i n i a n interpretation of P o p p e r ' s p h i l o s o p h y of science is stressed in Peter M u n z , Our Knowledge of the Growth o f Knowledge: Popper or Wittgenstein? (1985), and criticized in Ruse, above, at 6 1 - 6 5 . For a t h o r o u g h e x p l o r a t i o n of Peirce's p h i l o s o p h y of science, see C h r i s t o p h e r H o o k w a y , Peirce, chs. 2, 7 - 9 (1985); also W. B. Gallic, Peirce and Pragmatism (1966). 21. " T h e P r o b l e m of I n d u c t i o n , " in Popper Selections 101, 112 (David Miller ed. 1985). T h i s "challenge t h e o r y " of t r u t h parallels J o h n Stuart Mill's influential theory of free speech, w h i c h e m p h a s i z e s the t e n u o u s m o o r i n g s of beliefs that are n o t subjected to the test of controversy. See On Liberty, ch. 2 (1859). T h e parallel is n o t surprising; Mill like P o p p e r was a fallibilist. 22. A n a p p r o a c h s u g g e s t e d b y the late Paul B a t o r in c o n v e r s a t i o n and in unpublished talks.

not teach a distinctive analytic method, on the prominent role of rhetoric injudicial opinions, and on the low voltage of the methods of legal reasoning, we may be led to agree that law is indeed better regarded as an art (or more humbly as a craft, or as a skill such as riding a bicycle or speaking a foreign language) 23 than as a system of disciplined inquiry whose results can be justified in quasi-scientific terms. But just as art is ineffable, so is its critique. Art cannot be reduced to a set of analytic procedures; no more can art criticism. If law is an art, what can be the criteria for a "correct" legal decision? T h e very word becomes a misnomer. We ask not whether a w o r k of art is correct but whether it is beautiful, meaningful, stirring, enriching. Since these words are inapposite to legal rulings, an aesthetic view of law is hard to imagine. But it would in any event require a wrenching alteration in our vocabulary of legal evaluation. It is dizzying to see the autonomy and objectivity of law defended first by treating law as a branch of logic and then by treating it as a f o r m of art. Yet the incongruity is only apparent; the fact that logic always and art often are formalist disciplines illustrates the tenacity of formalism in legal thought. So inconclusive is evaluative criticism of art, and so striking are the vicissitudes of artistic taste, 24 that the test of time is the only criterion of aesthetic excellence that can be depended on to silence most doubters. The more we model law on art, therefore, the more we shall be drawn to the test of time as the criterion of legal soundness. It is in fact a c o m ponent of "traditionary" theories of precedent (Hale and Blackstone), 25 and it also has a role to play in the concept of precedent presented in Chapter 2. O n e thing that can solidify a precedent—that can make it authoritative (or more authoritative) rather than just a source of information—is its endorsement by many judges over a substantial period of time. Other things being equal, a conclusion to which a n u m b e r of different individuals have come—a conclusion (better, a hypothesis) that has survived continual retesting—is entitled to more deference than the 23. Either the craft or the skill m e t a p h o r is p r e f e r a b l e to the art m e t a p h o r , the last s u g gesting as it does a p r e o c c u p a t i o n w i t h elegance and o t h e r f o r m a l p r o p e r t i e s and an i n d i f ference to consequences in the w o r l d . 24. For an i l l u m i n a t i n g case study, see B r u n o S. Frey and Werner W. P o m m e r e h n e , "Is Art Such a G o o d I n v e s t m e n t ? " 91 Public Interest, S p r i n g 1988, at 79. 25. Sec C h a r l e s M . Gray, " E d i t o r ' s I n t r o d u c t i o n , " in M a t t h e w Hale, The History oj the Common Law of England xi, x x x i v (1971); Gerald J. P o s t e m a , " S o m e R o o t s of O u r N o t i o n of Precedent," in Precedent in Law 9, 18—20 (Laurence G o l d s t e i n ed. 1987). C o m p a r e the discussion of tradition sanctified b y t i m e in Paul Veyne, Did the Greeks Believe in Their Myths? An Essay in the Constitutive Imagination, ch. 1 (1988).

conclusion of a single individual. So time can help stabilize legal doctrine. N o t i c e that, f r o m this perspective, the m o r e diverse the judiciary, the m o r e its rulings invite unforced agreement, u n g r u d g i n g deference. But the test of time does not really respond to the needs of the legal system in a d y n a m i c society. First, the test is after the fact: it does not tell the j u d g e h o w to write an opinion that will survive. Second, all but one decision in a chain of precedent m a y have relied uncritically on the first decision, since acquiescence in precedent m a y reflect a desire for stability rather than agreement w i t h the o u t c o m e of the earlier case. Third, the very thing that gives a chain of precedent its strength—its length over t i m e — m a y in a changing society signal obsolescence. Fourth, m o s t of the rulings that interest the profession are too recent to be evaluated by their survival properties. We can use the test to criticize the D r e d Scott decision, Lochner, and the early free-speech cases, but possibly even Brown v. Board of Education, and certainly Roe v. Wade, are too recent to be a d j u d g e d g o o d or bad by reference to their survival. Swift v. Tyson (see C h a p t e r 14) had a " r u n " of a century. A final point is that a case can accrete authority over time regardless of its merit or even the j u d g e s ' desire for stability and e c o n o m y in decision m a k i n g . It can d o this by engendering reliance the protection of w h i c h is m o r e i m p o r t a n t than getting the law j u s t right, 2 6 or by fostering the organization of an interest g r o u p that has sufficient political m u s cle to prevent the case f r o m being overruled. These possibilities, especially the second, bring to the fore a serious objection to the test of time: its failure to distinguish a m o n g m e t h o d s of creating consensus. As Peirce and H o l m e s liked to point out, killing dissenters is a timeh o n o r e d m e t h o d of creating consensus, but not one calculated to get us closer to the truth. W h e n the test of time is conceived in Darwinian terms, this point is obscured because genocide is the m e t h o d of c o m p e tition in nature. Yet, practically speaking, the longer a belief persists, the likelier it is to be true, rather than to have persisted merely because rival beliefs were repressed. Historically, repression has not proved to be effective on other than a t e m p o r a r y basis; the truth will out. Still, the marketplace analogy is closer than the biological one. C o m p e t i t i o n can be depended u p o n to bring about the socially desired price and output 26. An e x a m p l e is the r e a f f i r m a t i o n , in F l o o d v. K u h n , 407 U . S . 258 (1971), of the a n o m a l o u s and o u t d a t e d decision h o l d i n g that m a j o r league baseball did n o t affect interstate c o m m e r c e and t h e r e f o r e w a s n o t subject t o the federal antitrust statutes. See Federal Baseball C l u b v. N a t i o n a l League, 259 U . S . 200 (1922). N o t i n g the social interest in the p r o t e c tion of reliance is a n o t h e r w a y o f m a k i n g the earlier p o i n t that stare decisis reflects a desire for stability as well as a respect for the t h i n k i n g of o t h e r p e o p l e (the earlier j u d g e s ) .

only if competitors are forbidden to employ certain tactics, including violence, fraud, and collusion. Similarly, the test of time is a reliable method of establishing truth only if the process of inquiry is guaranteed to operate without significant political distortions, and this condition is not satisfied in the legal area because judicial decisions become rallying points for political coalitions. The problem goes deeper. The sensible emphasis in law on hierarchy, continuity, and stability sets up a tension with the search for truth. Imagine arguing that scientists should be d r a w n f r o m a c o m m o n social and educational background, because then they will be more likely to agree. Applied to judges this is not a ridiculous argument, although I am not convinced by it. Stability is a desirable quality of law; and because legal reasoning does not have the power to secure agreement that formalists ascribe to it, legal stability is more likely to be achieved if the judges are like-minded, and they are more likely to be like-minded if they are socially and educationally homogeneous. But the more homogeneous they are, the less robust will their conclusions be in an epistemic, though not necessarily in a political or social, sense. In law, truth is traded off against other valued goods, and rightly so. That is not the scientific ideal. Recall the distinction made in the Introduction a m o n g ontological, scientific, and conversational objectivity. The first is in play w h e n there is something "out there" to which our concepts correspond, the second when observations are replicable so that all observers can be brought to unforced agreement on what it is they are observing. These senses of objectivity are related; the fact that scientists of diverse background and politics will agree with each other on most scientific questions, even though their agreement is not coerced, suggests there is something "out there" that is compelling agreement; scientific consensus argues for scientific realism. Objectivity in its third sense, which refers to bringing about agreement on nonverifiable propositions through methods of inquiry that are not exact (replicable), is usually weak. N o t always: the proposition that the unrestricted killing of h u m a n beings is w r o n g c o m mands the sort of broad and unforced consensus that makes it seem "objectively" correct, although it neither has a referent nor is the fruit of exact inquiry. Such instances do not, however, predominate in the areas of primary interest to the legal profession. Yet law needs agreement in order to avoid social chaos, and it does not scruple to coerce the necessary agreement through a variety of political and rhetorical methods ranging f r o m judicial hierarchy to the licensure of lawyers to formalist opinion writing to the enforcement of j u d g m e n t s by sheriffs,

marshals, and if need be soldiers. 27 Force is always in the background. C o m p a r e the Soviet U n i o n ' s use (diminishing, one hopes) of psychiatric hospitals to imprison dissidents w i t h the practice in o u r country, where c o m m i t m e n t s for insanity are hedged about with m a n y procedural restrictions and administered o n a decentralized basis. T h e difference is that b e t w e e n coercive and competitive objectivity. Yet coercive objectivity is and m u s t be an element of law even in a liberal society such as the U n i t e d States. As Justice R o b e r t Jackson put it, " w e act in these matters not by authority o f our competence but by force of o u r c o m missions. Still another factor that limits the usefulness of the test of time for law is the absence of a feedback mechanism. (Astrology m a y be a parallel case. Astrologers are careful to keep their predictions vague, so that falsification is difficult.) T h e test of time is related to the idea of trial and error. Scientists—in Popper's conception, w h i c h is broadly pragmatist, of science—try one theory after another, c o n f r o n t each one with data, discard the ones that the data falsify, and by this process continuously enlarge scientific k n o w l e d g e . Artists, too, try one thing after another; m o s t of their innovations are rejected, but s o m e survive, and those that do seem robust because of their p o w e r to c o m m a n d the interest of diverse audiences. U n l i k e the case of science, however, one is never confident that a rejected artist or art f o r m will stay rejected; there is a continual rediscovery of f o r g o t t e n authors, painters, composers. 2 9 It is m u c h the same in law. Except for a h a n d f u l of dramatic examples such as the D r e d Scott and Plessy decisions, competitive or survivalist notions rarely bring about decisive rejections of legal innovations. This observation is related to the hierarchical character of judicial systems, which limits the scope for competition a m o n g precedents. Long after the demise of Swift v. Tyson and Lochner v. New York, of privity of contract in personal-injury cases, of strict liability in collision cases, of fact pleading, and so on almost w i t h o u t end, discarded doctrines continue to enlist distinguished champions, w h o cannot be silenced effectively by pointing to the rejection of their views in the legal marketplace. And I am

27. C f . Sabina L o v i b o n d , Realism and Imagination in Ethics 172-175 (1983), on "coercive o b j e c t i v i t y " ; and notice the t w i s t that the p r e s e n t discussion gives to A u s t i n ' s t h e o r y of law ( f o l l o w i n g H o b b e s ) as t h e command of t h e s o v e r e i g n , o n w h i c h see Russell H a r d i n , " S a n c tion and O b l i g a t i o n , " 68 Monist 403 (1985). 28. West Virginia State B o a r d of E d u c a t i o n v. B a r n e t t e , 319 U . S . 624, 640 (1943). 29. I a m o v e r d r a w i n g t h e c o n t r a s t b e t w e e n science and art. For C o p e r n i c u s m a y be said to have revived Aristarchus, nuclear physics to have revived p r e - S o c r a t i c a t o m i c theory, and m o d e r n genetics to h a v e rediscovered M e n d e l .

speaking of champions within the legal c o m m u n i t y , that is, within the c o m m u n i t y of recognized experts; it is as if s o m e distinguished scientists believed in astrology. But this is to paint too bleak a picture, and it is time to begin restoring perspective. T h e r e is a limited competitive process in law; it corresponds to the competitive process in science s u m m a r i z e d in the idea of unforced inquiry. T h e w o r l d contains a vast n u m b e r of legal systems, and c o m parisons are possible. O f particular significance is the large n u m b e r of separate legal systems in the U n i t e d States, the consequence of o u r federal system of g o v e r n m e n t . Here the competitive relationship is direct. A legal system is one of the public services that each state offers to its residents and to the people and firms w h o d o business w i t h t h e m . If the system is p o o r — t h e j u d g e s corrupt or incompetent, the laws archaic or cumbrous or unenforced, the courts biased, expensive, or impacted by delay—the state will lose residents and trade. (In principle, " v o t i n g w i t h one's feet" is also possible at the national level, but emigration is m u c h more costly than m o v i n g to another part of the same c o u n t r y and therefore provides a m u c h weaker constraint o n a legal system.) H e r e at last is a feedback mechanism that m i g h t enable legal p e r f o r m a n c e to be monitored and i m p r o v e d . Unfortunately, judicial competition a m o n g states is weak, b o t h because political systems frequently are u n r e s p o n sive to the interests of the citizenry and because law is only one (and, needless to say, an unstudied) factor in decisions regarding migration and trade. Yet if w e go back far e n o u g h — t o trial by battle, for example, or prosecution for w i t c h c r a f t — w e can find plenty of legal practices that n o one defends a n y m o r e . These practices have decisively flunked the c o m p e t i tive process implicit in the test of time, although m o r e because of the rise of science than because of any ability of law to purge itself of prejudice and superstition. A n d a n u m b e r of legal practices that w e r e i n n o vations in their time have passed the test of time. T h e y include the trust, the counterclaim, estates in land, the concept of estoppel, p o w e r s of attorney, the recording of titles, the prohibition of torture, the subjection of senior g o v e r n m e n t officials to the rule of law, habeas corpus (however overextended the current American practice m a y be), the final-judgment rule, s u m m a r y j u d g m e n t , b u r d e n of proof, concepts of standing to sue, of ripeness, and of justiciability generally, the publication of judicial opinions (including dissents), i m p o u n d m e n t , the injunction, the administrative agency, the receivership, rescission, restitution, arbitration, and m a n y more. S o m e of these, t h o u g h , seem a bit c o m monplace; and even such durable innovations as the canons of statutory

construction, tort liability for accidents, the civil jury, cross-examination, pretrial discovery, the exclusionary rule, the privilege against compulsory self-incrimination, limited liability, and punitive damages are vigorously questioned. Some of these innovations, moreover, have changed so m u c h since they were new—the jury, for example—that it is unclear what it means to say they have survived. While it is reasonably certain that law has progressed since the eighteenth century, there is a perfectly good argument to be made—not a conclusive argument, but not an easily refutable one either—that law, taking all its consequences into account (including the opportunity costs of the human resources that have been sucked into the legal profession), has regressed in the United States since 1960. Like legal questions that do not get into court because they can be decided deductively, legal practices and institutions that have been validated by functioning smoothly over a long period of time refute extreme versions of legal skepticism but do not justify complacency about the p o w e r of legal reasoning either to solve the tough problems, whether doctrinal or institutional, in law or to establish the law's progressivity. T h e situation would be improved if law committed itself to a simple functionalism or consequentialism. Suppose the sole goal of every legal doctrine and institution was a practical one. T h e goal of a new bankruptcy statute, for example, might be to reduce the number of bankruptcies and lower interest rates. T h e operation of the statute would be evaluated in terms of these goals, and if the statute failed to fulfill them it would be repealed. Law really would be a method of social engineering, and its structures and designs would be susceptible of objective evaluation, much like the projects of civil engineers. This would be a triumph of pragmatism. And not the first triumph for pragmatism in law. Consider the fate of the rule of capture. M u c h as in scientific investigation and engineering implementation, analogy (from rabbits and other wild animals) was used to generate a hypothesis—that the rule of capture would provide an efficient property-rights regime for oil and gas. The hypothesis was tested by observing the performance of the rule of capture as applied to oil and gas. The hypothesis flunked the test and was rejected: the rule of capture was superseded by statutes requiring unitization and other methods of imparting incentives for efficient development. 3 "

30. See H o w a r d R. Williams and C h a r l e s J. M e y e r s , Oil and Gas Law, vol. 6, §§905, 912 (1988).

But law is not ready to commit itself to concrete, practical goals across the board. Legal innovations are often defended by reference to intangibles such as the promotion of h u m a n dignity, the securing of justice and fairness, and the importance of complying with the ideals or intentions of the framers of the Constitution or of statutes. These goals are too nebulous for progress toward achieving them to be measured. So even when it is apparent that ballyhooed legal innovations have had costly, unintended, and unforeseen consequences, their defenders may be able to fend off proposals for repeal by invoking unquantifiable benefits, as well as by rallying whatever interest groups have coalesced around the innovation. Having done consulting for a manufacturer of buses before I became a judge, I can attest that bus manufacturers are among the most enthusiastic supporters of judicial decrees that require the busing of schoolchildren to achieve public-school integration. As we shall see in Chapter 14, many recent legal innovations in American law appear to have miscarried; that is w h y I said that the cause of legal justice may actually have suffered f r o m the hectic "reforms" of the recent past. But it cannot be proved that these innovations have miscarried, because their goals are too vague to allow a cost-benefit or means-end evaluation; as a result, the record of apparent failure does not emit a clear signal for change.

4 Legitimacy in Adjudication

The Problem of Rational

Prejudgment

We have seen that when the methods available to the legal rcasoncr for making, criticizing, or justifying legal decisions are examined coldly, we see that they are not always distinctively legal and not always very objective, even when a relaxed, pragmatic approach to objectivity is taken. These points may seem to have potentially serious implications for the legitimacy of the judicial enterprise. I think they do, but I must deal with the counterargument that, provided the reasons ranged on either side of a legal dispute are not of exactly equal weight, the judge can always decide in favor of the party who has the stronger case in a technical legal sense. The judge need never resort to personal values, preferences, or politics, even if the stronger case is not overpowering. If this is true, then all the judge needs to avoid the problems discussed at such length in the preceding chapters is a pair of blinders! Metaphor can mislead. It is unclear what exactly it means to "weigh" arguments and, therefore, whether the process of decision in the face of conflicting arguments can be conceived in mechanically computational terms. Even if it can be, it is the wrong approach for a judge to follow. A person should not surrender deeply held beliefs on the basis of a weak argument just because he cannot at the moment find a stronger one in defense of those beliefs. Intuition, itself a method of practical reason, has its claims, and establishes presumptions that the other methods of practical reason may not always be able to overcome. 1 Granted, intu1. See Richard N i s b e t t and Lee Ross, Human Inference: Strategies and Shortcomings of Social Judgment 1 6 7 - 1 6 9 (1980); Gilbert H a r m a n , Change of View: Principles of Reasoning 35—41 (1986). I am u s i n g " i n t u i t i o n " here in the second sense in w h i c h I defined it at the b e g i n n i n g of C h a p t e r 2; so d e f i n e d , it is closely related to tacit k n o w l e d g e , to G a d a m c r ' s

ition should rule only in close cases; that is a requirement of the rule of law. But those cases are the focus of professional attention. It would be cold comfort to a defender of law's objectivity to be told that only close cases are indeterminate. Moreover, whether a case is close may depend in part on the strength of the judge's intuition. He may feel that legal doctrine has gone seriously awry, without being able fully to articulate the sources of his unease. So the category of close cases is ill defined. To make the point differently, the preconceptions that judges bring to cases are not extraneous and impertinent foreign matter. The tabula rasa is not the judicial ideal. Society does not want judges to act as umpires of debating contests or as referees of controversies in law reviews. Many considerations are out of bounds to the judge (various partisan and personal reactions to a lawsuit or the litigants), but they do not include a disinterested conviction that a case really ought to be decided one way even though the lawyer urging the other way has stronger arguments. The upshot, however, is that it will often be extremely difficult to determine whether a legal decision is correct or incorrect. Intuition is inarticulate, disinterest is not objectivity, and the judge's disinterested intuition will often not be verifiable or falsifiable by experience. So what began as an argument for law's objectivity—the possibility of going with the balance of the arguments—turns out to be another argument against it. The irony is deepened when we reflect that intuition in the sense used here is similar to tacit knowing, a method of practical reason that defenders of law's objectivity invoke as a substitute for logic and science without realizing that the more reliance they place on tacit knowing to determine legal outcomes, the more difficult they make the task of justifying those outcomes. For tacit knowing is private and inarticulate, justification public and articulate.

Consensus The power of legal reasoning to generate determinate case outcomes could perhaps be saved by turning law into something else—economics perhaps, or some ethical or political doctrine that might yield definite solutions to ethical or political problems. But no branch of the human-

emphasis on " p r e j u d i c e " as an essential c o m p o n e n t of i n t e r p r e t a t i o n (see H a n s - G e o r g G a d amer, Truth and Method 2 3 8 - 2 6 7 [1975]), and to the Bayesian statistician's c o n c e p t of priors, on w h i c h see C h a p t e r 6 of this b o o k . For e n d o r s e m e n t of that v i e w f r o m an u n e x p e c t e d source, see H e n r y J. Friendly, " R e f l e c t i o n s of a L a w y e r — N e w l y T u r n e d J u d g e , " 71 Yale Law Journal 218, 2 3 3 - 2 3 4 (1961).

ities or the social sciences, applied to such rebarbative materials as those t h r o w n u p in litigated disputes, is likely to achieve determinacy in the close case; in any event the a r g u m e n t s for t r a n s f o r m i n g law into another branch of social t h o u g h t could n o t be based on legal reasoning. T h e broader point is that either there is a strong political consensus that determines legal doctrine or there is n o t — a n d a f r a g m e n t e d political and ethical discourse will n o m o r e yield determinate outcomes than legal reasoning will. To sound a recurrent note in this book, consensus is a necessary condition f o r legal objectivity in all but the weakest sense. N o t m u c h c o m f o r t can be taken f r o m the fact that there are some areas of political consensus; it only seems that they must guarantee a sphere of legal determinacy. T h e r e is, to begin with, s o m e t h i n g odd about using political agreement to g r o u n d epistemological confidence. T h e point is related to an observation about the pragmatists, especially H a b e r m a s . Their view that " t r u t h is w h a t w o u l d e m e r g e as the result of unconstrained inquiry pursued indefinitely . . . partially transforms the epistemological p r o b l e m of distinguishing true f r o m false (warranted f r o m u n w a r r a n t e d , or rational f r o m irrational) beliefs into the political p r o b l e m of distinguishing free inquiry f r o m inquiry constrained and distorted b y the exercise of power." 2 T h e idea that political interference with the process of inquiry must be prevented is less troublesome than the idea that objectivity in law depends on the happenstance that a political consensus has f o r m e d a r o u n d the premises o f j u d i c i a l analysis. What is political consensus but a polite n a m e for the will of the stronger? What, therefore, are j u d g e s w h o render "objective" decisions but people w h o deduce the implications of the p o w e r relations in society? These rhetorical questions are overheated, however, apart f r o m the fact that j u d g e s need n o t prescribe so confining a role for themselves— although if they d o not, their claim to be engaged in an objective reasoning process is weakened. Even if every law m a y be said in s o m e sense to reflect d o m i n a n t public opinion—even if ultimately m i g h t makes legal right—it does n o t follow that all law is amoral. Causes and justifications need not coincide. There w o u l d be n o law against m u r d e r if

2. J o s e p h R o u s e , Knowledge and Power: Toward a Political Philosophy of Science 7, 18-19 (1987). T h e p r o p o s i t i o n that i n q u i r y can be d i s t o r t e d b y " p o w e r " is a m b i g u o u s . If all this means is that political i n t e r f e r e n c e w i t h the m a r k e t p l a c e in ideas can distort inquiry, fine. But if it m e a n s also or instead that t h e a d v a n t a g e s c o n f e r r e d by the possession of wealth can distort inquiry, i m p l y i n g that u n f o r c e d i n q u i r y m a y require either r e d i s t r i b u t i n g wealth or continually i n t e r v e n i n g in t h e m a r k e t p l a c e of ideas, it is t r o u b l e s o m e — w h i c h is not to say it necessarily is w r o n g ; I shall give an e x a m p l e in C h a p t e r 6 of a situation in w h i c h inequality of resources can indeed d i s t o r t legal inquiry.

the dominant groups in society did not desire it, yet that law promotes the general welfare (as do many other civil and criminal laws, a few examples being the laws for the making and enforcing of contracts, for the protection of intellectual property, and for the prevention and redress of fraud and negligence) and no j u d g e need be ashamed of enforcing it. 3 The evidence for the inconclusiveness of legal reasoning is not merely aprioristic. Controversy—which appears to be on the rise—over the politics of judicial appointees is one datum. Another is the trappings of the judicial process—the impressive courtroom, with the j u d g e sitting at an elevated bench and in uniform, and addressed with honorifics unusual in democratic society: all a conscious effort to make law more impressive than it would be if treated purely as a method of rational inquiry. That the point is indeed to impress the litigants and the public, rather than the judges, is suggested by the fact that judges do not wear robes when actually deliberating or making their decisions; and sitting behind a raised bench and being addressed with exaggerated deference are hardly humbling experiences for a judge. Here is another datum on the inconclusiveness of judicial reasoning. 4 American judges today are subject to exquisitely refined and elaborate rules on disqualification for conflict of interest. T h e tiniest potential conflict is disqualifying. This would make no sense if legal reasoning (including the resolution of factual disputes) were as transparent and reproducible as scientific reasoning and experimentation, for then an erroneous decision would be perceived and corrected and the j u d g e ridiculed or removed for having yielded to temptation. T h e legal system must lack confidence in its ability to detect judicial errors. Consistent with this point, the rules on conflict of interest have been growing stricter in lockstep with the decline of consensus in law and the c o n c o m itant growth injudicial discretion. T h e weaker the consensus, the more difficult it is f o r j u d g e s to fix the premises of decision and, by so doing, to make legal reasoning approximate logical deduction. Because legal reasoning is more (only?) cogent w h e n there is a consensus concerning the relevant political and social values, conflict of interest rules are less

3. O n h o w n o r m s e m e r g e , see R o b e r t A x e l r o d , " A n E v o l u t i o n a r y A p p r o a c h t o N o r m s , " 80 American Political Science Review 1095, 1108-1109 (1986); J. L. M a c k i e , " C o O p e r a t i o n , C o m p e t i t i o n , and M o r a l P h i l o s o p h y , " in M a c k i e , Persons and Values: Selected Papers, vol. 2, p. 152 (Joan M a c k i e and P e n e l o p e M a c k i e eds. 1985); F. A. H a y e k , Law, Legislation, and Liberty, vol. 1, p p . 7 2 - 8 1 (1973); E d n a U l l m a n n - M a r g a l i t , The Emergence of Norms (1977). 4. Well discussed in J o h n L e u b s d o r f , " T h e o r i e s o f J u d g i n g and J u d g e D i s q u a l i f i c a t i o n , " 62 New York University Law Review 237 (1987) (esp. p p . 2 4 9 - 2 5 0 , 261).

needed in that setting to prevent bias f r o m operating. The English judiciary is more homogeneous than ours, enabling greater agreement on premises and hence greater reliance on genuinely syllogistic legal reasoning. We would therefore expect the English conflict of interest rules to be less strict than ours, and so they are, although the difference is small. 5 We would also expect English law to be more certain than American law, and this is the general impression, although it would be nice to be able to m o v e beyond impressions. T h e immediate causes of the decline in consensus in our law are several. O n e is the g r o w t h of statutory and constitutional law relative to c o m m o n law. As we shall see in Chapter 8, it is only superficially that the c o m m o n law can be said to have a structure more logical than that of statutory or constitutional law; but what is true is that statutory and constitutional law has expanded into area after area of intense political controversy, an expansion unparalleled in the c o m m o n law sphere. O t h e r causes of the decline in legal consensus are a rise in the diversity of the legal profession since the 1950s, a concomitant decline in political consensus nationwide, and the increase in litigation/' which has multiplied judges and precedents and in particular has made adjudication by the Supreme C o u r t less determinate. T h e Court's capacity for decision making is essentially fixed, but the Court is selecting f r o m an ever-richer population of cases, and as a result an increasing fraction of the cases it decides are difficult rather than easy ones. And as the court of last resort it cannot be, or at least has chosen not to be, ruled by precedent—which is another reason that so large a fraction of the cases it decides are difficult. Moreover, as the ratio of decisions by the Supreme Court to decisions by lower courts shrinks, lower courts are increasingly deprived of authoritative precedent to guide their decision making; so more of their cases fall in the open area, too. T h e decline in legal consensus has deeper causes. Hierarchy and other forms of authority often weaken as peoples become wealthier; at least that seems to be the lesson of American, perhaps of world, history since

5. See S h i m o n Shetrcet, Judges on Trial: A Study of the Appointment and Accountability of the English Judiciary, pt. 5 (1976) (esp. pp. 308-309); de Smith's Judicial Review of Administrative Action, ch. 5 (4th ed., J. M . E v a n s ed., 1980). 6. T h a t increase has been e n o r m o u s , especially in the federal courts, in the last three decades. For statistics to m i d - 1 9 8 3 , see m y b o o k The Federal Courts: Crisis and Reform, ch. 3 (1985). T h e increase has c o n t i n u e d since, as d o c u m e n t e d in the annual reports of the D i r e c t o r of the A d m i n i s t r a t i v e O f f i c e of the U . S . C o u r t s .

World War II. 7 With wealth c o m e f r e e d o m , independence, and mobility, and pari passu a loosening of familial and c o m m u n i t y bonds. Judges, like other authority figures, are trusted less, respected less, unthinkingly deferred to less—another cause of m o r e stringent rules against judicial conflicts of interest. People stand on their rights m o r e and an immediate effect is m o r e litigation, w h i c h can lead to greater indeterminacy by the path j u s t sketched. With the decline of authority and the rise of independent thinking a society also becomes morally heterogeneous, 8 to the point where people within the same political c o m m u n i t y m a y c o m e to inhabit i n c o m m e n s u r a b l e moral universes. T h a t is the situation in this country today w i t h respect to the abortion controversy and the reason the controversy seems to admit of n o rationally d e m o n s t r a b l e resolution—seems literally interminable. 9 This standoff is repeated in a variety of legal contexts involving issues of personal, economic, religious, and sexual liberty. 1 " W h e n the j u d g e reaches an epistemological impasse and of necessity bases decision, w h e t h e r reflectively or unreflectively, o n some ethical or political principle, or public opinion, or whatever, he is unlikely to have escaped indeterminacy. T h e interesting question will then be what accident of psychology or personal history or social circumstances m o v e d h i m to adopt one social or political principle rather than another. T h e nation's legal heterogeneity mirrors its moral heterogeneity. 7. See C l i f f o r d Geertz, Local Knowledge: Further Essays in Interpretive Anthropology 16 (1983). O n the rise in s k e p t i c i s m c o n c e r n i n g p r o f e s s i o n a l i s m , see T h o m a s L. Haskell, " I n t r o d u c t i o n , " in The Authority of Experts: Studies in History and Theory x i i i - x v i i i (Haskell ed. 1984). A d m i t t e d l y , the p a t t e r n is m o r e c o m p l i c a t e d than I h a v e sketched in the text. T h e r e was a b u r s t of c o n f o r m i s m in the 1950s; p e r h a p s w e are a b o u t to sec a n o t h e r . D i s order and dissensus are m o s t visible in u n i v e r s i t y faculties, w h i c h increasingly are u n r e p r e sentative of A m e r i c a n society as a w h o l e . 8. O r at least is revealed to be m o r e h e t e r o g e n e o u s . In a hierarchical society the m o r a l distance b e t w e e n p e r s o n s m a y be great; w e s h o u l d n o t t h i n k of the a n t e b e l l u m S o u t h or Periclean A t h e n s , each w i t h large slave p o p u l a t i o n s , or India w i t h its caste s y s t e m , o r classridden E n g l a n d , as " h o m o g e n e o u s " societies tout court. B u t the official n o r m s , prescribed as they are b y the t o p of the hierarchy, will tend to be m o n o l i t h i c in such societies. A n d it is the official n o r m s that the c o u r t s enforce. 9. See Alasdair M a c l n t y r e , After Virtue: A Study in Moral Theory 6 - 1 0 (2d ed. 1984). 10. J u d g e Learned H a n d once r e m a r k e d : "Values are i n c o m m e n s u r a b l e . You can get a solution only by a c o m p r o m i s e , or call it w h a t y o u will. It m u s t be o n e that p e o p l e w o n ' t complain of too m u c h ; b u t y o u c a n n o t expect any m o r e o b j e c t i v e m e a s u r e . " " A Personal C o n f e s s i o n , " in H a n d , The Spirit of Liberty: Papers and Addresses of Learned Hand 302, 307 (3d ed. 1960). T h i s is t o o s w e e p i n g , if m y analysis in the p r e v i o u s c h a p t e r s is correct; b u t the d o m a i n in w h i c h H a n d ' s o b s e r v a t i o n is accurate is large and, I t h i n k , g r o w i n g . H a n d ' s moral skepticism (on w h i c h see also Michael M o o r e , " M o r e Reality," 1982 Wisconsin Law

Policy versus Pedigree as Warrants for Judicial

Action

Reasonableness as the Judicial Lodestar. Where does all this leave the judge? I can think of no better approach than for judges to conceive of their task, in every case, as that of striving to reach the most reasonable result in the circumstances—which include though are not limited to the facts of the case, legal doctrines, precedents, and such rule-of-law virtues as stare decisis." Bland as this recommendation may seem, 12 it differs from both the orthodox legal view of the judge's task and the various natural law approaches by substituting the humble, fact-bound, policy-soaked, instrumental concept of "reasonableness" for both legal and moral rightness. Yet it also differs f r o m Holmes's conception of the judge as interstitial legislator, 13 or, equivalently, f r o m Article 1(2) of the Swiss Code of 1807, which provides that "if no rule can be derived from the statute, the judge shall decide in accordance with the rule which he would promulgate if he were the legislator." 14 The picture of the judge as an interstitial legislator is both unedifying and, on a realistic view of the legislative process, misleading. It suggests that the only difference

Review 1061, 1066, and references cited there) w a s m o r e than a p a r l o r pose. It played a large role in a series of o p i n i o n s h e w r o t e u p h o l d i n g the refusal to g r a n t U . S . citizenship to resident aliens guilty of m o r a l lapses. See, f o r e x a m p l e , J o h n s o n v. U n i t e d States, 186 F.2d 588, 5 8 9 - 5 9 0 (2d Cir. 1951) ( " p e o p l e differ as m u c h a b o u t m o r a l c o n d u c t as they do a b o u t b e a u t y . . . So it seems t o us that w e are c o n f i n e d t o t h e best guess w e can m a k e of h o w such a poll [a poll o f citizens r e g a r d i n g t h e c o n d u c t in q u e s t i o n — d e s e r t i o n and adultery] w o u l d result"); R e p o u i l l e v. U n i t e d States, 165 F.2d 152, 153 (2d Cir. 1947) ( " o n l y a m i n o r ity of v i r t u o u s p e r s o n s w o u l d d e e m t h e practice [euthanasia] m o r a l l y justifiable, w h i l e it r e m a i n s in p r i v a t e h a n d s , even w h e n the p r o v o c a t i o n is as o v e r w h e l m i n g as it was in this i n s t a n c e " ) . Incidentally, if it is correct that society is b e c o m i n g m o r a l l y m o r e diverse, this is a b l o w to m o r a l realism, j u s t as an increase in d i s a g r e e m e n t a m o n g scientists w o u l d be a b l o w to scientific realism. 11. T h e s u g g e s t e d a p p r o a c h is close t o that p r o p o s e d b y K e n t G r e e n a w a l t , " D i s c r e t i o n and Judicial D e c i s i o n : T h e Elusive Q u e s t f o r t h e Fetters T h a t B i n d J u d g e s , " 75 Columbia Law Review 359, 377 (1975). See also A h a r o n Barak, Judicial Discretion 2 5 - 2 7 and ch. 4 (1989); cf. Wade L. R o b i s o n , " T h e F u n c t i o n s and Limits o f Legal A u t h o r i t y , " in Authority: A Philosophical Analysis 112, 119 (R. Baine H a r r i s ed. 1976); Steven J. B u r t o n , "Judicial D u t y and D i s c r e t i o n : T h e G o o d Faith T h e s i s " ( u n p u b l i s h e d article, U n i v e r s i t y of Iowa Law School, A p r i l 25, 1989). 12. A n d v u l n e r a b l e as it is to m o c k e r y : A m b r o s e Bierce in The Devil's Dictionary defined " r e a s o n a b l e " as "Accessible to the i n f e c t i o n of o u r o w n o p i n i o n s . H o s p i t a b l e to persuasion, dissuasion, and e v a s i o n . " The Collected Writings of Ambrose Bierce 187, 339 (1946). 13. See S o u t h e r n Pacific C o . v. J e n s e n , 244 U . S . 205, 221 (1917) (dissenting opinion). See also B e n j a m i n N . C a r d o z o , The Nature of the Judicial Process 113-115 (1921); J o h n Bell, Policy Arguments in Judicial Decisions 17-20, 2 2 6 - 2 4 6 (1983). 14. I a m i n d e b t e d to G e r h a r d C a s p e r f o r the translation. N o t i c e Article I(2)'s d e p a r t u r e f r o m the f o r m a l i s m characteristic of C o n t i n e n t a l law, m e n t i o n e d in the I n t r o d u c t i o n .

between a j u d g e and a real legislator is that the former fills the gaps left by the latter. If the comparison is taken literally—as no doubt the framers of the Swiss C o d e did not intend—the j u d g e should feel free to engage in poll taking, vote trading, naked interest-group politics, and other c o m m o n practices of legislators. In fact there are important institutional and procedural differences between courts and legislators, and they impose bounds on the domain of the reasonable in j u d g i n g that are not found in legislating. They differentiate the judicial product f r o m that of a legislature but do not dictate the outcome of difficult cases—they may rule out some outcomes, but not all except one. The j u d g e in the difficult case is more a policy maker than a conventional lawyer and within his domain of freedom or discretion may be as free-wheeling as a legislator. But neither is unconstrained, and, more to the point, the constraints are different. N o r is the j u d g e an arbitrator or o m b u d s m a n or village wise m a n — which is to say, a type of dispute settler entitled to ignore "the law." In one sense this is the opposite extreme f r o m thinking of the j u d g e as a legislator and in another sense the same thing—both the arbitrator and the legislator being "above" the law in the sense of not being bound by preexisting legal norms. But at both extremes, if that is what they are, what is distinctive about law and adjudication drops right out of the picture. The circumstances that determine the reasonableness of judicial decisions include statutory language, precedents, and all the other conventional materials of judicial decision making, including such prudential virtues familiar to lawyers as sensitivity to the limits of judicial knowledge and to the desirability of stability in law. These conventional resources and constraints of judicial decision making are foreign to both the legislator and the arbitrator. In many cases the conventional materials will lean so strongly in one direction that it would be unreasonable for the j u d g e to go in any other. But in some they will merely narrow the range of permissible decision, leaving an open area within which the judge must perforce attempt to decide the case in accordance with sound policy—in those grand symbolic cases that well out of the generalities and ambiguities of the Constitution, in accordance with a vision of the good society—while paying due heed to the imprudence of trying to foist an idiosyncratic policy conception or social vision on a recalcitrant citizenry. To repeat a previous point, the open area is not always smaller f o r j u d g e s than it is for legislators. Judges are not as subject to interestgroup pressures and popular sentiment as legislators are, and in particular areas these forces may constrain legislators more tightly than judges are constrained by the constraints on the judicial process. Federal judges,

for example, have as a practical matter more freedom in bringing about changes in the conditions of state prisons than state legislators do. But at the same time the lack of a popular mandate, the lack of fiscal authority, and the requirement of providing a reasoned, written justification for decisions impose constraints on judges that legislators do not face. It is not possible to say, in general, which official—judge or legislator— is less constrained, and therefore more powerful. A j u d g e w h o conceives of his role in difficult cases as that of policy maker rather than that of conduit for policy decisions made elsewhere in the political system need not on that account be a judicial activist. Judicial self-restraint in the sense of hesitation to overturn the decisions of other branches of government may be part of the judge's vision of the good society. But judicial self-restraint is a political theory rather than the outcome of legal reasoning; it cannot be deduced f r o m legal materials or otherwise rigorously (or even very convincingly) derived f r o m them. Those materials may determine h o w broad the area of judicial discretion is, but they will not determine h o w bold or timid the j u d g e should be within that area in making decisions that pinch another branch of government. I shall illustrate m y conception of sound judicial decision making with the example of antitrust law. Step 1 in deciding a tough antitrust case, a case not controlled by precedent or otherwise susceptible of confident j u d g m e n t at a first pass, is to extract (not—it goes without saying—by a deductive process), f r o m the relevant legislative texts and history, f r o m the institutional characteristics of courts and legislature, and, lacking definitive guidance f r o m these sources, f r o m a social vision as well, an overall concept of antitrust law to guide decision. A popular candidate for such a concept today is that of wealth maximization, but it is, needless to say, a contestable choice. Having made this choice (the current Supreme C o u r t has almost but not quite made it for him),' 3 the j u d g e will then want to canvass the relevant precedents and other sources for information that might help in deciding the case at hand. This is step 2. Step 3 is a policy j u d g m e n t — i n some cases, though, it might approximate a logical deduction—resolving the case in accordance with the tenets of wealth maximization. Step 4 returns to the precedents, but they are n o w viewed as authorities rather than merely

15. See, for e x a m p l e , B r o a d c a s t Music, Inc. v. C o l u m b i a B r o a d c a s t i n g System, 441 U . S . 1, 19-20 (1979); Reiter v. S o n o t o n e C o r p . , 442 U . S . 330, 343 (1979); Matsushita Electric Industrial C o . v. Z e n i t h R a d i o C o . , 475 U . S . 574 (1986).

as data; the j u d g e will want to make sure that the policy j u d g m e n t made in step 3 is not ruled out by authoritative precedent. Actually this is the third rather than the second time the j u d g e will have consulted precedents. They must be consulted at the outset to determine whether the case is indeed in the open area; if not, the four-step analysis that I have described is pretermitted. The suggested approach describes the actual, though often implicit, reasoning process that most judges use in tough cases. It also recasts legal analysis in those cases as a f o r m of policy analysis. T h e j u d g e has to make a policy choice, and the choice is dictated by the results of surveying and evaluating the consequences of alternative choices: consequences for the rule of law, for the parties, for the economy, for public order, for civilization, for the future—in short, for society. In n o n c o m mercial settings one might prefer to describe the analysis as ethical analysis rather than as policy analysis. But that would not alter the basic point: strictly legal materials are used only for help in setting an initial orientation and in providing specific data, and later as sources of possible constraints. Antitrust happens to be a field in which there is a large body of nonlegal—specifically, economic—learning upon which judges can draw for policy guidance. In many fields there is not much extralegal learning to draw on. In resolving novel issues of free speech, for example, a j u d g e is likely to be forced back on his personal notions of the proper balance between liberty and order, and on public values crystallized in earlier judicial opinions. But this does not affect m y basic point, which is that the decision of difficult legal cases is very often a f o r m of policy analysis rather than the product of a distinctive methodology of legal reasoning. The policy-soaked reasoning process that I have described with the aid of the antitrust example is reasoned, not arbitrary, but it is more likely to foreclose some outcomes than to generate a unique one, and not too much should be made of being able to rule out some points on a distribution. We would not think much of a weather forecaster w h o assured us that the temperature t o m o r r o w would be between 120° Fahrenheit and — 40° Fahrenheit but was unable to exclude any intermediate possibilities. Within the feasible range of legal outcomes (corresponding to the feasible range of temperatures for the climate and time of year), the observer, depending on his o w n values, policy preferences, temperament, social vision, life experiences, and so forth, will find one outcome more congenial, attractive, or persuasive than another but will not

be able to demonstrate its correctness. In short, decisions in difficult cases often are not "bivalent" (that is, either true or false). 16 Elsewhere I have suggested t w o principles for stabilizing judicial decisions against the problem of indeterminacy. 17 O n e is the avoidance of contradiction by the judge, not only within an individual opinion but across his opinions; the j u d g e must employ his powers of critical logic unflinchingly. This precept may seem so obviously correct as to be trivial; yet if adhered to rigidly it would prevent the j u d g e f r o m changing his mind. Foolish consistency is not an o x y m o r o n . This is another example of h o w limited the domain of logic is; it is not illogical to change one's mind on the basis of new information. The second principle, the "publicity principle" as I called it, is that the judge avow the true grounds of decision. Adherence to this principle will prevent the j u d g e f r o m rejecting consensus views; but since on many important political and social questions there is no consensus, the principle has a weak bite. It is also unenforceable; there is no mechanism for inducing judges to be candid. Other ways of attempting to stabilize legal doctrine include the principle of judicial self-restraint, rules limiting the circumstances in which judges consider themselves free to overrule previous cases (that is, stare decisis), and the conversion of multifactored tests to formulas or algorithms. 1 8 N o n e of these devices will close the open area all the way, and all rest on policy j u d g m e n t s that can be and are contested. A m o n g still other auxiliary principles—as one might call them by analogy to the auxiliary hypotheses of science—that might be used, for good or ill, to stabilize legal doctrine are strict construction, rigid adherence to precedent, favoring the underdog, trying to p r o m o t e private ordering, and insistence on definite, "bright-line" rules. N o n e of these principles can be derived by the methods of legal reasoning. Some are inconsistent with others. All depend on j u d g m e n t s of political theory. They are methods not of closing the open area but of helping us learn to live with it. They impart to judicial decision making not objectivity but pseudo-

16. C f . C a t h e r i n e Z . Elgin, " T h e Relativity of Fact and the O b j e c t i v i t y of Value," in Relativism: Interpretation and Confrontation 86 (Michael K r a u s z ed. 1989); G o r d o n C . F. B e a m , " T h e H o r i z o n of R e a s o n , " in id. at 205; J o s e p h M a r g o l i s , " T h e T r u t h a b o u t Relativism," in id. at 232. For an e x a m p l e f r o m q u a n t u m t h e o r y see Elgin, above, at 92. 17. See The Federal Courts, n o t e 6 a b o v e , at 205—206. 18. O n self-restraint, see id. at 198—222; o n o v e r r u l i n g , see m y article " T h e C o n s t i t u t i o n as an E c o n o m i c D o c u m e n t , " 56 George Washington Law Review 4, 36—37 (1987); and on the c o n v e r s i o n of m u l t i f a c t o r e d tests t o a l g o r i t h m s see A m e r i c a n H o s p i t a l Supply C o r p . v. H o s p i t a l P r o d u c t s L t d . , 780 F.2d 589, 5 9 3 - 5 9 4 (7th Cir. 1986).

objectivity, unless w e are content to equate the objective w i t h the reasonable. Emphasizing as it does the i m p o r t a n c e of policy in judicial decision making, m y discussion in this section m a y seem to i m p l y that it would m a k e sense to d r a w s o m e j u d g e s f r o m the ranks of n o n l a w y e r s — especially appellate j u d g e s , since a substantial fraction of appellate cases will be underdetermined by legal reasoning. 1 9 But there are compelling objections to this idea. First, m a n y of the cases that c o m e u p to an appellate court are determinate (as I noted in C h a p t e r 2), and lawyers can deal with t h e m m o r e proficiently than n o n l a w y e r s can. Second and related, law is a m o n g other things a language. T h e n o n l a w y e r will be at a loss to figure out w h a t the lawyers are talking a b o u t in their briefs and arguments, and is likely to fall u n d e r the sway of his professional colleagues. The third and most interesting point is that even if the nonlawyer has an epistemic contribution to make to appellate judging, a well-functioning legal system involves a trade-off b e t w e e n t r u t h and stability. T h e u n i formity of the legal profession reduces the vigor of legal inquiry b u t helps to stabilize legal obligation. The Pedigree Approach. T h e concept of judicial decision m a k i n g implicit in the preceding chapters and here m a d e explicit raises a question of judicial legitimacy: w h o has licensed j u d g e s to decide cases in accordance w i t h social vision? B u t to state the question this w a y is to appeal covertly to a particular political theory, o n e that regards the j u d g e as an agent of legislators, of constitutional framers, or of earlier j u d g e s and thus insists that every judicial decision be fairly referable to a c o m m a n d by a principal—in other w o r d s , that decisions be pedigreed. 2 0 " T h e Constitution d e m a n d s that all p o w e r be authorized . . . J u d g e s applying the C o n s t i t u t i o n . . . m u s t take their guidance and authority

19. T h e a r g u m e n t is m a d e in Frederick Schauer, " J u d g i n g in a C o r n e r of the L a w , " 61 Southern California Law Review 1717, 1732 (1988), and criticized in L a w r e n c e B. S o l u m , " T h e Virtues and Vices of a J u d g e : A n Aristotelian G u i d e to Judicial Selection," in id. at 1735. C f . D o r i s M a r i e P r o v i n e , Judging Credentials: Nonlawyer Judges and the Politics of Professionalism (1986). 20. T h e a n a l o g y is to the w a y in w h i c h w e learn such t h i n g s as o u r n a m e and birthplace, or that France once had an e m p e r o r n a m e d N a p o l e o n B o n a p a r t e ; w e b u i l d a chain b e t w e e n ourselves and the original e y e w i t n e s s e s . B u t that is a m e t h o d of establishing e p i s t e m i c a u t h o r i t y rather t h a n political authority. T h e m o s t p o w e r f u l c o n t e m p o r a r y a d v o c a c y of w h a t I a m calling the " p e d i g r e e " t h e o r y o f j u d i c i a l l e g i t i m a c y is b y F r a n k E a s t e r b r o o k . See, for example, " M e t h o d , Result, and A u t h o r i t y : A Reply," 98 Harvard Law Review 622 (1985); " T h e Influence o f j u d i c i a l R e v i e w o n C o n s t i t u t i o n a l T h e o r y , " in A Workable Government? The Constitution after 200 Years 170 ( B u r k e M a r s h a l l ed. 1987); " T h e Role o f O r i g inal Intent in S t a t u t o r y C o n s t r u c t i o n , " 11 Harvard Journal of Law and Public Policy 59 (1988).

f r o m decisions made elsewhere. Otherwise they speak with the same authority they . . . and I possess when we fill the law reviews with our speculations and desires: none. And the other branches o w e no obedience to those w h o speak without authority . . . Judges can legitimately demand to be obeyed only w h e n their decisions stem f r o m fair interpretations of commands laid d o w n in the texts." 21 T h e pedigree approach rests on a questionable notion of w h y judicial decisions are and should be obeyed (that is, because they are pedigreed). The main reason they are obeyed is that the consequences of disobedience arc unpleasant; there are heavy sanctions for flouting court orders. Although these consequences depend ultimately on the willingness of the executive branch to enforce judicial decrees and of the legislative branch to pay for these enforcement efforts—indeed, to pay for the courts themselves—it would be naive to suppose that the willingness of the other branches to cooperate with the judicial branch depends on the courts' confining themselves to "fair interpretations of commands laid d o w n in the texts." That willingness may be related to public confidence in the courts. 22 But there is no evidence that such confidence depends on the scrupulousness with which courts confine themselves to fair interpretations of commands laid d o w n in the texts—about which the public k n o w s little—as distinct f r o m notions of justice or fairness that arc independent of fidelity to texts. 23 21. " M e t h o d , Result, and A u t h o r i t y , " n o t e 20 a b o v e , at 628-629. 22. C f . T o m R. Tyler, K e n n e t h A. Rasinski, and K a t h l e e n M . M c G r a w , " T h e Influence of Perceived Injustice o n the E n d o r s e m e n t of Political Leaders," 15 Journal of Applied Social Psychology 700 (1985). 23. O n e s t r a w in the w i n d is t h e public reaction to R o b e r t B o r k ' s n o m i n a t i o n to the S u p r e m e C o u r t : B o r k ' s p o s i t i v i s m seems to h a v e c o u n t e d against h i m in the public eye. A n o t h e r (if slightly inconsistent) s t r a w is the quite a s t o n i s h i n g lack of public awareness of courts, i n c l u d i n g the S u p r e m e C o u r t . See Alan H y d e , " T h e C o n c e p t of L e g i t i m a t i o n in the S o c i o l o g y o f L a w , " 1983 Wisconsin Law Review 379, 408; A u s t i n Sarat, " S t u d y i n g A m e r i c a n Legal C u l t u r e : A n A s s e s s m e n t of S u r v e y E v i d e n c e , " 11 Law and Society Review 427, 4 3 8 - 4 3 9 (1977). For direct evidence that public c o n f i d e n c e in the c o u r t s , such as it is, is i n d e p e n d e n t of j u d i c i a l a d h e r e n c e to craft values—a m a t t e r a b o u t w h i c h the public k n o w s n o t h i n g — s e e R i c h a r d L e h n e and J o h n R e y n o l d s , " T h e I m p a c t of Judicial Activism on Public O p i n i o n , " 22 American Journal of Political Science 896 (1978); G r e g o r y Casey, " T h e S u p r e m e C o u r t and M y t h : An E m p i r i c a l I n v e s t i g a t i o n , " 8 Law and Society Review 385 (1974); R o g e r H a n d b e r g and William S. M a d d o x , " P u b l i c S u p p o r t for the S u p r e m e C o u r t in the 1970s," 10 American Politics Quarterly 333 (1982). " T h e p e o p l e see the S u p r e m e C o u r t as an i n s t i t u t i o n in the n o r m a l c o n t e x t of A m e r i c a n politics. T h e y v i e w the C o u r t as R e p u b licans and D e m o c r a t s , and t h e y j u d g e it in the s a m e o f f h a n d w a y as they do their a c k n o w l edged politicians d u r i n g elections." K e n n e t h M . D o l b e a r e and Phillip E. H a m m o n d , " T h e Political P a r t y Basis of A t t i t u d e s t o w a r d the S u p r e m e C o u r t , " 32 Public Opinion Quarterly 16, 30 (1968). T h e sheer unreality o f t h e legal p r o f e s s i o n ' s professed c o n c e p t i o n of the polit-

If w e switch gears and ask w h y people, official and otherwise, have a duty to obey judicial decisions, the natural answer (which happens also to be the answer given by the natural lawyer) is that they should obey them because they are just. 2 4 But w h a t does " j u s t " mean in this context? Is a decision j u s t by virtue of having a g o o d pedigree? If a statute, or a provision of the Constitution, is u n j u s t , is a decision enforcing that statute or constitutional provision nevertheless j u s t by virtue of its fidelity to its unjust source? A n affirmative answer w o u l d be paradoxical, and w o u l d i m p l y the f u r t h e r paradox that the unjust decision provides the acid test o f j u d i c i a l legitimacy; it makes the very strongest claim to be obeyed by the other branches of g o v e r n m e n t , because it shows that the j u d g e s are indeed faithful agents—only such fidelity could explain the unlovely o u t c o m e . O n e is put in m i n d of A n g e l o in Measure for Measure. "It is the law, n o t I, c o n d e m n y o u r brother. / Were he m y kinsman, brother, or m y son, / It should be thus w i t h h i m " (Act II, sc. 2, 11. 84-86). We are entitled to ask why the "faithful agent" conception of the j u d g e is the right one. To set a g o o d example to other people in agency roles? To restore a needed sense of discipline in public life? To p r o m o t e certainty? To reduce judicial workloads? To spare the courts f r o m political controversy? To allocate g o v e r n m e n t a l functions in accordance w i t h comparative institutional competence? (Maybe legislators are better at making policy than j u d g e s are.) Because the virtues of the rule of l a w — and they are genuine and i m p o r t a n t virtues—cannot, either as a practical or as a theoretical matter, be otherwise attained? S o m e or all of these may be good answers—maintaining public confidence in the courts is not a good a n s w e r — b u t this needs to be demonstrated, not merely asserted. A n o t h e r bad answer, offered by H a m i l t o n in Federalist No. 78, is that j u d g e s w h o h e w close to the original m e a n i n g of the C o n s t i t u t i o n are the authentic oracles of popular sovereignty, since the C o n s t i t u t i o n was adopted by the people of the U n i t e d States. This is artificial even apart f r o m the f r a m e r s ' w e l l - k n o w n distrust of popular g o v e r n m e n t (remember that in the original Constitution, only the H o u s e of Representatives was to be elected directly). E v e r y o n e w h o voted for the C o n -

ical setting in w h i c h c o u r t s o p e r a t e is p o w e r f u l l y argued in Stephen M . G r i f f i n , " W h a t Is C o n s t i t u t i o n a l T h e o r y ? T h e N e w e r T h e o r y and the D e c l i n e of the Learned T r a d i t i o n , " 62 Southern California Law Review 493, 5 0 6 - 5 2 9 (1989). 24. T h e t h e m e of Philip Soper, A Theory of Law (1984). T h i s a s s u m e s that there is a moral d u t y to o b e y law. I am not sure; see C h a p t e r 7.

stitution is long dead, and to be ruled by the dead hand of the past is not self-government in any clear sense. It is true that all judges are required to take an oath "to support the Constitution" and all federal judges another oath to decide cases "agreeably to the Constitution." 2 5 So the lawful j u d g e is constrained by the Constitution. But the question is the nature of the constraint in cases where the Constitution does not provide clear guidance, and it is not a question that the oaths illuminate. It may not be supportive of or agreeable to the Constitution if the judges fold their hands when the text, history, or structure fails to yield an answer to a question of interpretation or application. T h e inadequacy of the "faithful agent" approach is exhibited in Learned Hand's argument against judicial adventurism. "For myself it would be most irksome to be ruled by a bevy of Platonic Guardians . . . I should miss the stimulus of living in a society where I have, at least theoretically, some part in the direction of public affairs." 26 This is an argument not for faithful agency but against constitutional law. The faithful agent will sometimes be enforcing restrictions that irk. If with regard to such restrictions m o d e r n judges allow modern views to influence their decisions, then living people have at least an indirect role in the area of public affairs that is within the scope of the judiciary. But if modern judges are faithful expositors of decisions made by the framers of the Constitution centuries ago and ignore contemporary public opinion, today's citizens are deprived of an opportunity to shape their o w n destiny in areas within the prohibitory scope of the old enactments, except through the cumbersome process of constitutional amendment. T h e issue of the proper freedom for judges is intractable at the theoretical level; it ought to be recast in empirical, pragmatic terms. D o we want judges to play a bigger or a smaller role in the direction and implementation of governmental policy? (And w h o are " w e " in this question?) What are the trade-offs? Which choice would have the better consequences (insofar as we can discern them), all things considered? T h e question whether judges should be passive rather than active, m o d est rather than aggressive, ought to be confronted head-on rather than obscured by endless talk about legitimacy. In fact, consequentialist arguments are often made in support of formalist positions. The formalist will point out for example that judges risk their prestige, their perceived disinterest, perhaps their long-term effectiveness, if they become 25. U . S . C o n s t . , art. IV; 28 U . S . C . §453. 26. The Bill of Rights: The Oliver Wendell Holmes Lectures, 1958 73 (1958).

dragged into politically controversial areas. This is true, but one must ask what the judges are hoarding their prestige, their power to do good, for, and what the costs of this hoarding are. The j u d g e w h o believes that he cannot speak authoritatively except as the agent or mouthpiece of the sovereign is likely to be a "strong" legal positivist, that is, one w h o believes that ethical considerations (broadly construed to include considerations of public policy often thought too mundane to count as ethical) are not proper sources of legal obligation unless they are embodied in a statute or constitution. And inside the strong legal positivist lurks the moral and epistemic skeptic. A recent essay, noting that Holmes was such a one, argues that his "assertion that there is no viewpoint that can claim precedence on the basis of its presumed objectivity" led him to conceive "of his role as judge not as that of an umpire w h o must determine the merits of each case, but as that of a conductor w h o must in each case determine which voice shall be heard . . . The duty of the j u d g e is to consider what person or institution is most entitled to have its viewpoint count in adjudicating the case." 27 The choice of metaphors is a bit jarring, since we are apt to think of a musical conductor as a more "activist" figure than an umpire. However that may be, "conducting" is undoubtedly an important part of the judge's task. And the more skeptical the j u d g e is about his access to the wellsprings of objectivity, the more willing he will be to allow a multitude of voices to be raised. O r will he? If he is denied that access, so are the legislative and executive-branch officials to w h o m he is asked to defer. It seems a standoff. In any event, if the issue is the proper amount of deference by judges to other officials, w h y not say so straight out rather than cast it as one of legitimacy? The "strong" positivist is not a skeptic about one moral value: that of deferring to higher authority. But is it not a warped moral stance to be skeptical about all values except obedience? Is it not rather too Prussian for an American judge? It is one thing to announce that the heavens will fall if judges abandon the m y t h (or perhaps the actuality) of being

27. C a t h a r i n e Wells Hantzis, "Legal I n n o v a t i o n w i t h i n the Wider Intellectual Tradition: T h e P r a g m a t i s m of O l i v e r Wendell H o l m e s , J r . , " 82 Northwestern University Law Review 541, 590-591 (1988). N o t i c e that the skeptical j u d g e m a y w a n t t o delegate as m u c h judicial a u t h o r i t y as possible to n o n j u d g e s ( a d m i n i s t r a t o r s , j u r o r s , etc.), r e c o g n i z i n g that the e x e r cise of such a u t h o r i t y involves an inescapable a m o u n t of sheer will and sensing t h e i n c o n sistency b e t w e e n this characteristic of the judicial process and c o n v e n t i o n a l e x p e c t a t i o n s about the process. H o l m e s , h o w e v e r , p e r h a p s because o f his a d m i r a t i o n f o r the will as a h u m a n faculty (see C h a p t e r 7), was n o t b a s h f u l a b o u t basing j u d i c i a l decisions o n w h a t he liked to call the sovereign p r e r o g a t i v e of choice.

faithful agents of the people and of the people's representatives; this is merely incorrect. It is another thing to say that the heavens should fall. Obedience to rules is j u s t one virtue a m o n g many, 2 8 and it cannot be given its proper w e i g h t w i t h o u t considering the content of the rules and other pertinent social and moral values. It is surprising that a nation which has embraced an ideology of hostility to bureaucrats contains so m a n y j u d g e s w h o apotheosize the bureaucratic virtues. I have said that the pedigree approach needs to be justified, not just asserted. B u t m a y b e n o t too m u c h effort should be devoted to that task. T h e approach m a y have little content or thrust. For it need not preclude a role, even a large one, f o r the judicial exercise of social vision. If framers of statutory and constitutional provisions k n o w , as they must, that there is an o p e n area in j u d g i n g that j u d g e s can close only by bringing in policy preferences, ethical values, and the like, m a y b e they can be taken to have authorized this type of decision in advance. A n o t h e r possibility of course is that they accept it as the unavoidable price of an independent judiciary b u t do n o t desire it. This is conjecture too. They m a y w a n t to give their agents a long tether so that statutes, and especially the Constitution, w h o s e f r a m e r s m a d e it difficult to amend, w o u l d not obsolesce too fast. Legal scholars w h o believe that j u d g e s should be faithful agents tend also to believe that legislators are—and above all, that framers of the C o n s t i t u t i o n w e r e — w i s e and far-sighted people w h o s e c o m m a n d s deserve a conscientious effort at obedience. This belief shades into f r a m e r idolatry—the view that the framers of the Constitution k n e w m o r e about governing t w e n t i e t h - c e n t u r y America than the Justices of the Supreme C o u r t do today. T h e framers may have been an abler g r o u p than any S u p r e m e C o u r t in our history, but they were not clairvoyant. This is n o t to suggest that the general principles embodied in the C o n s t i t u t i o n are obsolete, let alone that a c o n t e m p o r a r y effort to rewrite the C o n s t i t u t i o n w o u l d produce a superior d o c u m e n t (some observers believe that this is w h a t the Supreme C o u r t has attempted to do, and with the predicted result). T h e economic principles expounded by A d a m Smith m a y well be as apt for twentieth-century America as for eighteenth-century England. T h e p r o b l e m b o t h in law and in economics is that the difficult cases are at the level not of principle but of application. A l m o s t everyone can agree with the general principles embodied in the Constitution, precisely insofar as they are general, or with the specifics that d o not matter (for example, h o w m a n y days the 28. See J u d i t h N . Shklar, Legalism:

Law, Morals, and Political Trials 109 (1964).

President has to veto a bill); it is the application of those principles to situations that the framers did n o t foresee and provide for that causes controversy. T h e framers gave us a compass, n o t a blueprint. In any event, the wiser and m o r e far-sighted the framers are assumed to have been, the less rather than m o r e plausible it becomes to i m p u t e to t h e m the view that j u d g e s should n o t exercise independent j u d g m e n t . If the framers were that wise, they k n e w h o w the j u d g e s of the u n p r e c edentedly p o w e r f u l judiciary they w e r e authorizing w o u l d behave. D o not people intend (in a m e a n i n g f u l sense of the w o r d "intend") the n a t ural and probable consequences of their acts? Moreover, the m o s t influential framers were lawyers, and it is unlikely that they greatly feared an "imperial" judiciary. Such fear was n o t entirely absent f r o m their deliberations: the Seventh A m e n d m e n t ' s guaranty of trial by j u r y in federal civil cases attests to anxiety about the p o w e r of federal j u d g e s . But this was a concession to popular feeling—for w h e n have m e m b e r s of the legal establishment complained about courts' being too p o w e r f u l ? The legal realists of the 1920s and 1930s, w h o did complain, and today's conservatives, w h o d o complain, were (are) a distinct m i n o r i t y in the profession. Furthermore, the rights included in the Bill of Rights were for the most part rights that had been created by the English judiciary, on which the provision for a federal judiciary in Article III of the C o n stitution is modeled. Article III envisaged a judiciary even m o r e independent than the English royal courts. T h e f r a m e r s ' distrust was of legislatures. True, Article III was enacted before the Bill of Rights, w h i c h the framers accepted grudgingly; and w i t h o u t the Bill of Rights the role of the federal courts w o u l d have been smaller than it has b e e n — a l t h o u g h it would still have been large, for consider the constitutional decisions of J o h n Marshall's Supreme C o u r t , w h i c h w e r e based not on the Bill of Rights but on the original Constitution. O n e of the difficulties of extracting coherent policy f r o m the C o n s t i t u t i o n is that it is n o t a single document but a palimpsest reflecting confusingly superimposed values of different political factions and of different stages in American political history. A final point is that the framers w e r e revolutionaries, that they exceeded their terms of reference in s u b m i t t i n g the C o n s t i t u t i o n for ratification b y the people, and that the southern states were forced to ratify the Civil War a m e n d m e n t s at g u n p o i n t . T h e "title deeds" of constitutional law are written in blood; the "pedigree" begins in usurpation. Cardozo's suggestion (see Introduction) that the judicial decision w h i c h makes good sense displays on its face the title deed that matters has much to r e c o m m e n d it.

Yet despite its vulnerabilities the pedigree theory can lay claim to being the official theory of statutory and constitutional interpretation, by which I mean the theory that a plurality of judges subscribe to publicly. This may be due in part to the hold of formalist thinking on the legal mind, in part to the desire of public officials to duck responsibility—and shoving off the responsibility on long-dead framers is a convenient dodge—and in part to the vagueness of the pedigree theory. The last t w o points are related; what could be more attractive to judges than a theory o f j u d i c i a l legitimacy that allowed them to do anything they wanted provided they employed a rhetoric determinedly selfabnegating? N o t all official theories are correct. O u r government might not function if the people running it took seriously every bromide about democracy and popular sovereignty. A m o n g alternatives to the pedigree theory, t w o resonate particularly well with the themes of this chapter. One, the conservative (in the attitudinal, not political, sense), points out that American judges seem always and everywhere to have followed something like the four-step approach that I described with the aid of the antitrust example, and infers that it would probably be futile and maybe risky to make them stop. ("Better the devil you k n o w . . .") The other approach, which is pragmatic, argues that if this sort ofjudicial decision making " w o r k s " we should not lose sleep over the fact that it cannot be fitted into a neat table of organization constructed f r o m the Constitution and democratic theory. But h o w do we k n o w it works, and h o w far should we extend it? W h o knows? Maybe the burden of persuasion should rest on whoever is urging a change in settled practices, in either direction—toward a more aggressive judiciary or toward a less aggressive one. O r is this to succumb too readily to the tyranny of the status quo? These t w o theories o f j u d i c i a l legitimacy are not impressive, I admit. Also they are not so remote f r o m the pedigree approach as they may seem, and not only because of the latter's sponginess. The four-step approach that I proposed as an example of pragmatic or "realistic" adjudication does not deny that an independent judiciary creates a potentially serious agency problem or recommend that judges reconceive their role as that of principals rather than agents. It assigns them a creative role but within a f r a m e w o r k in which the j u d g e is in some sense subordinate to the framers of constitutional statutes and, of course, of the Constitution itself. T h e issue is in what sense and h o w best to fulfill this role. The strongest argument for the pedigree approach and against a pragmatic or "realistic" one may itselfbe pragmatic: judges just are not smart

e n o u g h to make wise policy decisions, balancing a m y r i a d of conflicting considerations that include the rule-of-law a r g u m e n t s against balancing. I agree; and rules do reduce the burdens of t h o u g h t . But the choice is not between unwise policy decisions and wise decisions applying rules. Wisdom in applying rules requires a sense of w h e n the rules run out and (what is not necessarily a different question) w h e n it w o u l d be a serious mistake to apply a particular rule "as written." T h e decision to, and h o w to, apply a rule is a policy decision. M o r e fundamentally, h o w do w e k n o w that legislators really are better policy makers than judges? N o doubt they could be—if only they could t h r o w off the y o k e of interestgroup pressures, r e f o r m the procedures of the legislature, and extend their o w n policy horizons b e y o n d the n e x t election. If they cannot d o these things, their comparative institutional advantages m a y be fantasy. To compare real j u d g e s w i t h ideal legislators is to c o m m i t the N i r v a n a fallacy. To s u m up: If the pedigree approach w e r e s o m e h o w compelled— perhaps because it was latent in the definition of adjudication—then w e would have to accept it, as w e accept that the square root of 9 is 3. B u t since it is not compelled, it has to be justified; it has to be s h o w n to be a good thing, or at least a better thing than the alternatives. This the proponents have yet to show. Examples. I begin w i t h an example of "formalist anxiety," the judicial dread of appearing to be m a k i n g policy choices. O f t e n statutes are passed w i t h o u t a provision limiting the time within w h i c h suit u n d e r the statute must be b r o u g h t , that is, w i t h o u t a statute of limitations. When this happens, courts cast about f o r a similar statute that has a statute of limitations and " b o r r o w " that statute of limitations f o r use with the defective enactment (in equity cases, they m a y i n v o k e the j u d g e - m a d e principle of "laches," w h i c h bars a suit unless b r o u g h t within a "reasonable" time). W h y d o j u d g e s d o this, rather than pick a term of years suitable to that enactment? T h e standard answer is that the selection of a statute of limitations is a matter peculiarly within legislative rather than judicial competence. This response cannot be right, however. If it were, it w o u l d be a conclusive a r g u m e n t against b o r rowing. Actually the considerations bearing o n the selection of a statute of limitations are better k n o w n to j u d g e s than to legislators, for those considerations are the effect of passage of time o n the accuracy of a d j u dication, the effect of old claims o n the court queue f o r n e w claims, and the desirability of enabling potential defendants after a k n o w n period of time to go about their business w i t h o u t w o r r y i n g about the possibility of being sued. T h e reason courts will n o t select a period of limitations

is that it is impossible even to appear to reason to a n u m b e r . A j u d g e can give reasons w h y one statute is like another, but he cannot give reasons for selecting 3 years rather than 4 years, or 10 years rather than 8 years, or 180 days rather than 240 days, as the outer limit of bringing a suit. T h e element of free choice, o f discretion, cannot be concealed. Judges w a n t to conceal the fact that m a n y of their choices are free (including the choice of w h i c h statute of limitations to borrow!) in the sense that a different choice could be defended j u s t as strongly. Yet the b o r r o w i n g m e t h o d , viewed as a f o r m of reasoning by analogy, seems thoroughly spurious, quite apart f r o m the general weaknesses of that methodology. Statutes are frequently the product of deals w i t h interest groups (see C h a p t e r 9), so there is n o p r e s u m p t i o n that the statutory limitations period is a bona-fide stab at the optimal period. By b o r r o w i n g it for use in a statute that contains n o limitations period, the court m a y be capriciously projecting an interest-group deal into an area remote f r o m the contemplation of the deal makers and thereby gratuitously disserving the public interest. M y next example of adjudication as policy m a k i n g is Boiling v. Sharpe,29 decided the same day as Brown v. Board of Education, T h e issue was w h e t h e r racial segregation of the District of Columbia's public schools w a s unconstitutional, and the S u p r e m e C o u r t held that it was, finding in the due process clause of the Fifth A m e n d m e n t a guaranty of equal protection. This was a strained interpretation. T h e Fourteenth A m e n d m e n t contains an equal protection clause as well as a due process clause but is inapplicable to the federal g o v e r n m e n t . T h e Fifth A m e n d ment, w h i c h is applicable to (and only to) the federal g o v e r n m e n t , has no equal protection clause; so far as pertains to the Boiling case, it has only a d u e process clause, n o t h i n g in the language of which hints at any prohibition of racial discrimination. A l t h o u g h the term "due process" has vague antecedents that even before the Civil War had led some j u d g e s to suppose it m i g h t contain principles of natural law, 3 " the Supreme C o u r t in m o d e r n times has generally t h o u g h not consistently f o r s w o r n this interpretation in an attempt to set s o m e b o u n d s to con-

29. 347 U . S . 497 (1954). 30. T h e m o s t f a m o u s , a l t h o u g h an a m b i g u o u s , e x a m p l e is the discussion of " l a w of the l a n d " in M u r r a y ' s Lessee v. H o b o k e n Land & I m p r o v e m e n t C o . , 59 U . S . (18 H o w . ) 272, 276 (1855); for criticism, see D a v i d P. C u r r i e , The Constitution in the Supreme Court: The First Hundred Years: 1789-1888 272, 276 n. 304 (1985). O t h e r antecedents are discussed in Daniel A. Farber, "Legal P r a g m a t i s m and the C o n s t i t u t i o n , " 72 Minnesota Law Review 1331, 1350—1353 (1988), and in Suzanna Sherry, " T h e F r a m e r s ' U n w r i t t e n C o n s t i t u t i o n , " 54 University of Chicago Law Review 1127 (1987).

stitutional law. A n d there is the p r o b l e m of identifying the "liberty" or " p r o p e r t y " of which the plaintiffs in Boiling had been deprived. (The due process clause is limited to deprivations of life, liberty, or property.) Furthermore, if the due process clause of the Fifth A m e n d m e n t does forbid racial discrimination, it is hard to see w h y the framers of the Fourteenth A m e n d m e n t , having decided to i m p o s e the due process clause on the states, bothered to add an equal protection clause. If by doing so they were merely trying to m a k e assurance d o u b l y sure, w h y didn't they make the equal protection clause applicable to the federal government, to avoid a negative implication? T h e framers of the Fourteenth A m e n d m e n t m a y well have confined the equal protection clause to the states deliberately. T h e Fourteenth A m e n d m e n t was an assertion of federal p o w e r . States, n o t the federal government, were oppressing black people wholesale. Deliberate or not, it is hard to get around the language of the clause, w h i c h forbids only states to deny persons the equal protection of the laws. A n y o r t h o dox theory of law implies that the equal protection clause was inapplicable to the District of C o l u m b i a schools and the due process clause unavailable to fill the gap. T h e S u p r e m e C o u r t m u s t have felt, h o w e v e r , that there w o u l d be an intolerable anomaly, in a political rather than a conventionally "legal" sense, in allowing the public schools of the nation's capital to remain segegrated w h e n the S u p r e m e C o u r t , sitting in that capital, had j u s t outlawed segregation b y states. It w o u l d have looked as t h o u g h the C o u r t w a n t e d to m a k e sure that the i n c o n v e niences of desegregation w o u l d n o t be visited on its o w n venue. Such a decision not only w o u l d have seemed hypocritical b u t also w o u l d have blurred the message of Brown v. Board of Education and u n d e r m i n e d that decision's moral and political force. Boiling tests the limits of w h a t a court can properly do in the teeth of conventional legal materials. T h e r e is s o m e t h i n g p r o f o u n d l y amiss in a legal system in w h i c h j u d g e s regularly say to themselves or to each o t h er, " T h e law requires o u t c o m e A, but B makes better political sense so we'll go w i t h B . " (They w o u l d never say so publicly.) Yet at s o m e point the o u t c o m e that lacks political sense, that represents bad policy, that has distinctly u n t o w a r d anticipated consequences, m a y by virtue of that fact not be the o u t c o m e required (permitted?) by law. Hans Linde has attacked Boiling as the e p i t o m e of "realistic" judicial decision making, a style he deprecates. 3 1 Yet he devotes half his attack 31. See H a n s A. Linde, "Judges, Critics, and t h e Realist T r a d i t i o n , " 82 Yale Law 227, 2 3 3 - 2 3 4 (1972).

Journal

to conjecturing that the practical consequences of a decision in favor of segregation w o u l d not have been u n t o w a r d — t h a t such a decision might actually have galvanized Congress into dealing responsibly with the segregation p r o b l e m in general and b y doing so m i g h t even have headed off the South's massive resistance to Brown v. Board of Education. T h e reader comes a w a y with an impression, perhaps unintended, that consequences, including w h a t w o u l d normally be considered political consequences, are relevant to the judicial task after all. Linde has no confidence in the ability of j u d g e s to evaluate the consequences correctly, and he has a point there. But w h a t is the alternative? H e proposes "construing the living meaning of past political decisions," 32 but what does that mean? Ray v. Blait*33 provides another illustration o f j u d i c i a l Realpolitik. T h e Alabama legislature had enacted a statute that required presidential electors, w h o u n d e r Alabama law were chosen in a statewide primary, to pledge to vote in the Electoral College for the nominees of their party's national convention. T h a t statute was challenged on the plausible g r o u n d that it t o o k away the electors' independence. T h e framers of the C o n s t i t u t i o n unquestionably intended the Electoral College to be an independent, deliberative b o d y rather than a rubber stamp of the state's voters. 3 4 It was one of several democracy-diluting measures that the framers built into the g o v e r n m e n t a l system they were creating. N e v e r t h e less the S u p r e m e C o u r t upheld the Alabama statute. T h e American people have c o m e to believe that the President is elected by them, not by a cabal of " i n d e p e n d e n t " electors m o s t o f w h o m are u n k n o w n to the people. T h e Alabama statute had been passed in the w a k e of a revolt by "Dixiecrat" electors against T r u m a n ' s election in 1948. T h e prospect of r u n a w a y electors is a deeply unsettling one, containing the seeds of a genuine constitutional crisis; the C o u r t displayed political w i s d o m in declining to provide a n u r t u r i n g soil. T h e fact that Justice Jackson dissented in Ray is a bit of a surprise, since he had been the author a decade earlier of one of the greatest pragmatic opinions in the history of the S u p r e m e C o u r t . I refer to the second flag-salute case, West Virginia State Board of Education v. Barnette.35 Three

32. Id. at 255. B y " p a s t political decisions," he m e a n s those m a d e b y the C o n s t i t u t i o n ' s framers. 33. 343 U . S . 214 (1952). 34. See id. at 2 2 8 - 2 2 9 n n . 15-16, and Federalist No. 68 ( H a m i l t o n ) . 35. 319 U . S . 624 (1943). S u b s e q u e n t page references to this o p i n i o n are in the text.

years earlier, in Minersville School District v. Gobitis,36 the C o u r t had held that the Constitution did not require public school authorities to excuse the children of Jehovah's Witnesses f r o m having to salute and pledge allegiance to the American flag, even t h o u g h these acts violated the Witnesses' conscience. In Barnette, in w h a t m a y be the m o s t eloquent m a j o r ity opinion in the history of the Supreme C o u r t , the C o u r t reached a contrary conclusion and overruled Gobitis.37 Justice Frankfurter, the author of Gobitis, w r o t e a fiery dissent. Jackson acknowledges that his decision is n o t compelled b y the text or history of the Constitution (let alone by precedent): the task of translating the majestic generalities of the Bill of Rights, conceived as part of the pattern of liberal government in the eighteenth century, into concrete restraints on officials dealing with the problems of the twentieth century, is one to disturb self-confidence. These principles grew in soil which also produced a philosophy that the individual was the center of society, that his liberty was attainable through mere absence of governmental restraints, and that government should be entrusted with few controls and only the mildest supervision over men's affairs. We must transplant these rights to a soil in which the laissez-faire concept or principle of non-interference has withered at least as to economic affairs, and social advancements are increasingly sought through closer integration of society and through expanded and strengthened governmental controls. These changed conditions often deprive precedents of reliability and cast us more than we would choose upon our o w n judgment, (pp. 639-640)

T h e offensiveness of the flag salute and pledge to the Jehovah's Witnesses was plain enough; the t o u g h question was w h e t h e r it was o v e r balanced by the needs of national unity. T h e Gobitis opinion had remarked, "National unity is the basis of national security." 38 Describing this remark as "the very heart of the Gobitis o p i n i o n " (p. 640), Jackson makes his o w n opinion pivot on the question—a question factual rather than legal—whether a c o m p u l s o r y flag salute and pledge are effective in

36. 310 U.S. 586 (1940). 37. "We think the action of the local authorities in compelling the flag salute and pledge transcends constitutional limitations on their p o w e r and invades the sphere of intellect and spirit which it is the purpose of the First A m e n d m e n t to our C o n s t i t u t i o n to reserve f r o m all official control." P. 642. Actually Justice Jackson's opinion is a plurality opinion, for only three other Justices j o i n e d it in its entirety; but the verdict of time has awarded it majority status. For a sad c o m m e n t a r y on the decline of judicial eloquence, c o m p a r e the opinions in Barnette with those in Texas v. J o h n s o n , 109 S. C t . 2533 (1989)—the Barnette of the eighties, invalidating a state statute that f o r b a d e desecrating the American flag. 38. 310 U . S . at 595.

p r o m o t i n g national unity. T h e lesson of history, he concludes, including m o d e r n history, is no: Ultimate futility of such attempts to compel coherence is the lesson of every such effort f r o m the roman drive to stamp out Christianity as a disturber of its pagan unity, the Inquisition, as a means to religious and dynastic unity, the Siberian exiles as a means to Russian unity, down to the fast failing efforts of our present totalitarian enemies. Those who begin coercive elimination of dissent soon find themselves exterminating dissenters. Compulsory unification of opinion achieves only the unanimity of the graveyard . . . The First Amendment . . . was designed to avoid these ends by avoiding these beginnings . . . The case is made difficult not because the principles of its decision are obscure but because the flag involved is our own. (p. 641)

T h e reference to " t h e fast failing efforts of o u r present totalitarian enemies" is particularly significant. Gobitis had been decided in J u n e 1940, w h e n the G e r m a n s were riding high—partly, it seemed, because of their impressive national unity ("Ein Volk, Ein Reich, Ein Fiihrer!"). B y J u n e 1943, w h e n Barnette was decided, the G e r m a n s were on the run. Their national unity didn't look so impressive any more, n o r did d e m ocratic disorder seem as reckless as it had three years earlier. Experience seemed to have falsified the premise of Gobitis. Jackson goes on to make a quintessentially p r a g m a t i c observation about the benefits of diversity: "We can have intellectual individualism and the rich cultural diversities that w e o w e to exceptional m i n d s only at the price of occasional eccentricity and a b n o r m a l attitudes" (pp. 641-642). T h r o u g h o u t the opinion the emphasis is n o t on the dogmatics of constitutional law but on the consequence of the c o m p u l s o r y flag salute and pledge, b o t h for the Jehovah's Witnesses and for the rest of the c o m m u n i t y .

How

Are Judges'

Visions

Changed?

I have suggested that political factors, and sometimes social vision, are decisive in the m o s t difficult cases. T h e point can be put m o r e strongly: consequences are never irrelevant in law. If they are sufficiently grave they can sway decision, w h a t e v e r the balance of conventional legal arguments, w h i c h for example strongly favored Justice Jackson's dissent in Ray v. Blair. But j u s t how are consequences or politics decisive? If t w o social visions clash, w h i c h prevails? Equivalently, h o w does a j u d g e choose between c o m p e t i n g social visions? O f t e n the choice will be made on the basis of deeply held personal values, and often these values will

be impervious to a r g u m e n t . Persuasion will figure in s o m e cases but it will be persuasion b y rhetoric rather than b y the coolest f o r m s of reasoned exposition. By definition the latter will n o t arbitrate between competing social visions—a point strikingly illustrated by Hilary Putnam's suggestion that "respectful c o n t e m p t " is the honest and natural attitude t o w a r d a person w i t h w h o m one has p r o f o u n d disagreements over political philosophy. 3 9 T h e j u d g e w h o w a n t s to "sell" his social vision to colleagues or f u t u r e j u d g e s does so by presenting it— often, by presenting himself (the tactic that rhetoricians call the "ethical appeal")—in an appealing, a w i n n i n g , light, in the h o p e of converting the reader to his views. So the liberal m i g h t present himself as h a r d headed, the conservative present himself as compassionate, in order to combat the stereotypes attached to the "liberal" and "conservative" labels and attract followers f r o m the other c a m p or at least cool the ardor of their opposition. O r the contenders m i g h t seek to " e v o k e . . . the core values of [the] audience in a p o w e r f u l and plausible w a y " and argue that those values require a particular resolution to the dispute in question. 4 " Justice Jackson makes m u c h use of this technique in Barnette. I mentioned conversion in the last paragraph advisedly, having in mind the following passage in Wittgenstein: May someone have telling grounds for believing that the earth has only existed for a short time, say since his o w n birth?—Suppose he had always been told that,—would he have any good reason to doubt it? Men have believed that they could make rain; why should not a king be brought up in the belief that the world began with him? And if Moore [G.E. Moore, whose views on skepticism Wittgenstein is criticizing] and the king were to meet and discuss, could Moore really prove his belief to be the right one? I do not say that Moore could not convert the king to his view, but it would be a conversion of a special kind; the king would be brought to look at the world in a different way. Remember that one is sometimes convinced of the correctness of a view by its simplicity or symmetry, that is, these are what induce one to go over 39. Reason, Truth and History 166 (1981). P u t n a m is describing his attitude t o w a r d his colleague Robert Nozick. 40. Michael Walzer, Interpretation and Social Criticism 8 8 - 8 9 (1987). O n the rhetoric of judicial opinions, see m y b o o k Law and Literature: A Misunderstood Relation, ch. 6 (1988); Peter Goodrich, Reading the Law: A Critical Introduction to Legal Method and Techniques, ch. 6 (1986); Goodrich, Legal Discourse: Studies in Linguistics, Rhetoric and Legal Analysis, pt. 2 (1987); cf. Robert A. Ferguson, " H o l m e s and the Judicial Figure," 55 University of Chicago Law Review 506 (1988). T h e increase in recent years in the n u m b e r of difficult cases has been accompanied by a decline in the quality of judicial rhetoric. R e m a r k a b l y little firstrate judicial prose has been written since the 1950s.

to this point of view. O n e then simply says something like: " That's h o w it must be." 41

Transposed to the legal setting, conversion has unsettling overtones. We tend to think of it as a sudden, deeply emotional switch f r o m one nonrational cluster of beliefs to another that is no more (often less) rational, although Wittgenstein's example is of the opposite, and we tend not to think of the fact of conversion as a significant point in favor of the winning faith. Although most lawyers think of themselves as engaged in rational inquiry rather than religious affirmation, the religious impulse is well-nigh universal; it is particularly strong in the United States;42 and in many secular Americans trained in law the impulse gets channeled into veneration of the Constitution as a sacred text and a decision to attend one of the churches at which it is worshiped. It is not only on points of constitutional law that lawyers, judges, and law professors defend positions with a zeal that approaches the religious. Central to the women's movement, which has become influential in several areas of law, is the conversion, or gestalt switch, that consists of taking the traditional w o m a n ' s role (housewife, mother, "better half," etc.) and looking at in a different light—not as division of labor, recognition of essential differences, quietly heroic self-abnegation, or service to the species but as brainwashing, oppression, and slavery. Many w o m e n and some men have been brought in recent years to think of the traditional role of w o m e n in this w a y — n o t by being shown evidence that this is the way things "really" are but by being offered a fresh perspective that, once glimpsed, strikes many with a shock of recognition. Metaphor, narrative, simile, and analogy in its figurative or mctaphoric sense are important terms here. Think of the efforts to defend the abortion decision by comparing the pregnant w o m a n forbidden to abort her fetus to a bystander forced against his will to render nine months of life support to a stranger. (We shall glance at these efforts in Chapter 11.) T h e comparison does not change any of the facts about abortion or even bring new facts to light. Its purpose is to jar people out of their accustomed ways of thinking about pregnancy and abor-

41. L u d w i g Wittgenstein, On Certainly 14e (G. E. M . A n s c o m b e and G. H. von Wright eds. 1969) (H92). See also id. at 81e (11612), and William James, The Varieties of Religions Experience 162 (Frederick H. B u r k h a r d t gen'l cd. 1985). 42. See Gallup Report No. 236: Religion in America 50 Years, 1935-1985 1, 53 (1985).

don. 4 3 Swift put it neatly: "You cannot reason a person out of something he has not been reasoned into." These points are pragmatic. Persuasion and reason tend to merge in a pragmatist view of truth. If what is good or useful to believe or what one just can't help believing is, for all practical purposes, truth, then persuasion as well as proof can establish truth, since persuasion can be a source of tenacious beliefs. But the fusion of reason and persuasion makes the concept of "truth" problematic, as we have already seen, and by doing so it undermines the law's rational pretensions. Yet, like it or not, something like the process described by Wittgenstein does seem to explain—along with changes in judicial personnel (but that is really the same thing)—many of the seismic shifts that have occurred in our law, such as the great expansion of liability on virtually all fronts since the 1950s, the expansion in the rights of criminal defendants and of prisoners, the increased recognition of women's rights, the explosive growth of constitutional law. Consider reapportionment. Until the late 1950s the idea that malapportionment of state legislatures might violate the equal protection clause was virtually unthinkable; twenty years later the idea that it might not had become virtually unthinkable. What changed in the interim? N e w information had not exposed an error in the old ways. There had been no new discoveries about the "real" meaning of the equal protection clause. And there was very little evidence in the 1950s and 1960s, and there is very little today, that legislative malapportionment has significant consequences for public policy. 44 For reasons that are unclear, judges and lawyers started looking at the equal protection clause and the

43. C f . " W h a t M e t a p h o r s M e a n , " in D o n a l d D a v i d s o n , Inquiries into Truth and Interpretation 245, 261—264 (1984). O n the centrality of m e t a p h o r in law, see Steven L. Winter, " T r a n s c e n d e n t a l N o n s e n s e , M e t a p h o r i c R e a s o n i n g , and the C o g n i t i v e Stakes for Law," 137 University of Pennsylvania Law Review 1105 (1989); also references in C h a p t e r 13 of this book. 44. Sec, for e x a m p l e , William H . Riker, " D e m o c r a c y and R e p r e s e n t a t i o n : Reconciliation of Ball V. James and Reynolds V. Sims," 1 Supreme Court Economic Review 39, 4 1 - 5 5 (1982); Larry M . S c h w a b , The Impact of Congressional Reapportionment and Redisricting 196-200 (1988); T i m o t h y G. O ' R o u r k e , The Impact of Reapportionment 159 (1980). A f e w studies have f o u n d a potentially significant effect of r e a p p o r t i o n m e n t o n policy. See in p a r ticular R o g e r A. H a n s o n and R o b e r t E. C r e w , J r . , " T h e Policy I m p a c t of R e a p p o r t i o n m e n t , " 8 Law and Society Review 69 (1973); D o u g l a s G. Feig, " E x p e n d i t u r e s in t h e A m e r i c a n States: T h e I m p a c t of C o u r t - O r d e r e d Legislative R e a p p o r t i o n m e n t , " 6 American Politics Quarterly 309 (1978). N o s t u d y that I a m a w a r e of, h o w e v e r , finds that the strict " o n e m a n one v o t e " approach taken b y the S u p r e m e C o u r t has policy significance.

federal judicial role in a n e w way. 4 5 We exaggerate the efficacy of appeals to objective truths in altering legal doctrine, and in doing so w e exaggerate the i m p o r t a n c e of such truths in law. I d o n o t m e a n to suggest that everything is contingent, "up for grabs," so that the gifted poet or rhetorician can by skill in generating n e w m e t a p h o r s or n e w perspectives alter society. I d o n o t believe in the infinite plasticity of h u m a n nature or social arrangements. Gifted poets and others can a w a k e n people to facts but cannot create facts. In Chapter 13 I offer a "material" explanation for the w o m e n ' s m o v e m e n t . T h e inconsistency of legislative m a l a p p o r t i o n m e n t w i t h democratic theory was n o t s o m e t h i n g that the Supreme C o u r t created in Baker v. Can. It j u s t had not seemed the sort of abuse that federal courts should attempt to correct, in part because the attempt m i g h t (it was feared) precipitate a c o n f r o n t a t i o n w i t h state legislatures. In the event, the fear proved to be u n f o u n d e d ; after several decades of federal judicial activism w e n o w k n o w that states are for the most part delighted and relieved to cede p o w e r and responsibility to the federal g o v e r n m e n t in general and to the federal courts in particular. H a d they not been, Baker m i g h t have b e c o m e as controversial a venture i n j u d i c i a l activism as Roe v. Wade has become. M y point is that the about-face that was Baker v. Can, like the other great turning points in t w e n t i e t h - c e n t u r y American law (and in law, period), was n o t the p r o d u c t of deep reflection on the meaning of the C o n s t i t u t i o n and the c o m m o n law. O f t e n such turning points are n o t even the product of n e w l y obtained i n f o r m a t i o n (and these— meaning and data—are not clearly different things), but instead reflect changing outlooks. We shall consider a dramatic example in Chapter 10: the overruling in Brown v. Board of Education of Plessy v. Ferguson. T h e flag-salute cases are still another example. Frankfurter too had appealed to consequences, in particular the consequences for a proper balance b e t w e e n courts and legislatures if the C o u r t t o o k too aggressive a role in striking d o w n illiberal statutes. S o m e h o w those consequences looked less impressive in 1943 than they had in 1940. Today w e can see that Barnette was the first step t o w a r d a larger and m o r e p o w e r f u l federal judiciary than s o m e believe to be g o o d for the country; and perhaps, too, a greater individualism than is healthy. I suggest not that the step should not have been taken b u t that its best explanation lies in a change of outlook rather than in an acquisition of n e w information, although

45. C f . J a n G. D e u t s c h , " N e u t r a l i t y , Legitimacy, and the S u p r e m e C o u r t : S o m e Intersections b e t w e e n L a w and Political Science," 20 Stanford Law Review 169, 224 (1968).

the course of the war had provided some new, if oblique, information regarding the consequences of the compulsory flag salute. 46 Critical Legal

Studies

I have concentrated in Part I on challenging exaggerated beliefs in the autonomy and cogency of legal reasoning rather than on challenging the opposite exaggeration, the exaggeration of law's indeterminacy and subjectivity. My emphasis reflects the fact that the beliefs on which I have been focusing not only are more ingrained, being c o m m o n to both the liberal and the conservative segments of the legal-political spectrum, but also are more plausible than the opposite view, espoused by some legal realists in previous generations and by many members of the critical legal studies movement in this one, that all law is politics in a narrow and disreputable sense and right-wing politics at that. This view is difficult to evaluate because it is stated in a nonfalsifiable form. Whenever a judge is found acting contrary to his presumed political interests, his behavior is explained as throwing sand in the public's eye—forging even tighter ideological bonds by appearing to be evenhanded. Maybe the judge is doing this. A rule-of-law ideology would not be a persuasive method of mystification and cooptation were it not occasionally employed against the rich and powerful. 4 7 But this makes it exceedingly difficult to distinguish empirically between a genuinely evenhanded administration of the law and an administration permeated by class bias, or even to choose between them normatively. Since judges are human, we cannot expect a perfectly evenhanded administration of justice; and if tilt is therefore inevitable, w h y should an administration of justice subtly tilted in favor of the upper class be thought more disreputable than one subtly tilted in favor of the lower class? This discussion shows, moreover, that the critical legal studies movement is prey to the genetic fallacy. Even if the motives for judges' adherence to rules are sometimes basely political, to the extent the judges do adhere to them law is different from politics. A Peirce-style pragmatist 46. A c o n s e q u e n c e I have n o t m e n t i o n e d is t h e w a v e o f p r i v a t e violence against J e h o vah's Witnesses that f o l l o w e d the Gobitis decision. Yet the w a v e was c o n c e n t r a t e d in the m o n t h after the decision and had abated b y t h e t i m e Barnette w a s decided. See D a v i d R. M a n w a r i n g , Render unto Caesar: The Flag-Salute Controversy 164—165, 169 (fig. 1), 172 (1962). T h e p r o p o s i t i o n that e x p e r i e n c e falsified Gobitis is a r g u e d in R i c h a r d D a n z i g , "Justice F r a n k f u r t e r ' s O p i n i o n s in the Flag Salute Cases: B l e n d i n g L o g i c and P s y c h o l o g i c in C o n s t i t u t i o n a l D e c i s i o n m a k i n g , " 36 Stanford Law Review 675, 722—723 (1984). 47. See E. P. T h o m p s o n , Whigs and Hunters:

The Origin of the Black Act 2 5 8 - 2 6 9 (1975).

(see C h a p t e r 5) w o u l d be inclined to put this m o r e strongly: the p r o p osition that j u d g e s adhere to the rule of law because they are sly politicos and the proposition that they d o so because they are g o o d j u d g e s have the same consequences and therefore the same meaning. T h e "law is politics" school also ignores the existence of easy cases and exaggerates the significance of the indeterminate ones (of which there are indeed plenty) b y insisting that law is not law unless it lives up to its m o s t extravagant formalist billings. Yet between the formalist model, at one extreme, and "conversion," at the other, there are not only the variety of m e t h o d s of practical reason discussed in previous chapters but also discourse in which "agreed-upon criteria for reaching agreement" are not the axis upon which communication turns and the evaluation of disparate views in terms of some accepted framework within which they can be objectively assessed and commensurated with one another is not the organizing aim. Hope for agreement is not abandoned. People occasionally do change their minds or halve their differences as a result of intelligence concerning what individuals or groups of individuals whose minds run on other tracks believe. But "exciting and fruitful disagreement"—how do I k n o w what I think until I see what you say—is recognized as a no less rational process . . . It can also be, less dramatically, a practicable method for living in a situation where dissensus is chronic, probably worsening, and not soon to be removed. 4 8

Disagreement can be rational and creative, or at least cathartic, even if it does n o t lead to consensus or d e m o n s t r a b l e truths; and the positions of the contending parties need n o t be vulgarly political in character or motivation. It can be cathartic simply b y virtue of being a substitute for physical violence. It can elicit i n f o r m a t i o n that m a y reduce the intensity of disagreements that are based in part o n misunderstandings, and it can f u r t h e r reduce that intensity simply by revealing the contestants to each other as serious and rational. B u t it can also intensify disagreements by revealing the true depth of the chasms between persons w h o have different values on f u n d a m e n t a l questions. M a r k Kelman's analysis of Regina v. Cunningham49 illustrates critical legal studies in action. In the course of breaking open a gas meter in the cellar of a h o u s e and stealing the eight shillings that were in the meter, C u n n i n g h a m ripped the meter off the wall, w i t h the consequence— 48. Geertz, n o t e 7 a b o v e , at 2 2 3 - 2 2 4 . 49. 41 C r i m . A p p . 155 (1957); see M a r k G. K e l m a n , " A s s u m e N o t h i n g ! " 22 Stanford Lawyer, S p r i n g 1988, at 18.

unintended by and u n k n o w n to h i m — t h a t the gas main fractured, causing gas to seep t h r o u g h the cellar wall to the house next door; the gas seriously injured a w o m a n w h o lived there. C u n n i n g h a m was convicted of larceny and sentenced to six m o n t h s in prison. H e was also convicted of "unlawfully and maliciously causing a n o x i o u s thing to be taken so as thereby to endanger life," and he was sentenced to five years in prison for that crime. O n appeal he argued that the j u d g e should not have instructed the j u r y that "maliciously" m e a n t simply that the defendant k n e w he had n o business taking the meter. T h e appellate court reversed, holding that the j u r y should have been told to determine w h e t h e r the defendant "foresaw that the removal of the gas m e t e r m i g h t cause injury to someone but nevertheless r e m o v e d it." In Kelman's view, the a p p r o priateness of punishing C u n n i n g h a m f o r " p o i s o n i n g " (as K e l m a n calls malicious endangering) depends on the arbitrary choice b e t w e e n treating h i m "as being in the situation of s o m e o n e dealing w i t h gas meters or someone stealing f r o m gas meters." 5 0 If the former, he is guilty regardless of h o w careful he was; theft is an unreasonable m o d e of dealing with a gas meter and any risk created thereby is undue. If the latter, C u n n i n g h a m should get off if he w a s as careful as the average meter thief. Why the care of the average thief should be t h o u g h t an appropriate benchmark for criminal liability, other than in a society of thieves, escapes me. It implies that if C u n n i n g h a m had been a clumsy repairman rather than a thief, he m i g h t have been punished m o r e heavily than C u n n i n g h a m the thief, because the average m e t e r repairman is m o r e careful than the average meter thief. T h e point of Regina v. Cunningham is that a thief w h o is careless of the consequences of his theft for h u m a n safety is m o r e dangerous to the c o m m u n i t y than a careful thief and should therefore be punished m o r e severely. This m u c h seems unexceptionable; and asking w h e t h e r the thief foresaw those consequences is an indirect w a y of asking h o w dangerous a person he is. T h e difficult question, not so m u c h about the case as about the situation depicted in it— but a question that Kelman's analysis does n o t touch—is w h e t h e r C u n ningham should have been punished because the m e t h o d of stealing f r o m the gas meter endangered h u m a n safety, even if he was too d u m b to realize this. In other w o r d s , should mere negligence be criminally actionable, either generally or in the particular case of a careless thief? This question can be answered by c o m p a r i n g the civil and criminal sanctions for negligence. T h e civil sanction is damages, and since m o s t 50. Id. at 4 6 - 4 7 .

thieves are indigent the threat of a civil sanction f o r negligent injuries inflicted in the course of a theft is unlikely to deter. We need a criminal sanction to d o that, w h i c h means w e need a criminal sanction on top of the sanction for (careful) theft so that the thief will have an incentive to be careful. Readers w h o find this approach to Regina v. Cunningham persuasive m a y nevertheless be inclined to question w h e t h e r the approach is, at b o t t o m , distinguishable f r o m that of critical legal studies. I have offered an e c o n o m i c explanation for the decision, and thereby denied the specialness of law. H a v e I then, all unknowingly, signed on to D u n c a n K e n nedy's a t t e m p t e d expose of the b a n k r u p t c y of conventional legal thought? Teachers teach nonsense when they persuade students that legal reasoning is distinct, as a method for reaching correct results, f r o m ethical and political discourse in general (that is, f r o m policy analysis). It is true that there is a distinctive lawyers' body of knowledge of the rules in force. It is true that there are distinctive lawyers' argumentative techniques for spotting gaps, conflicts, and ambiguities in the rules, for arguing broad and narrow holdings of cases, and for generating pro and con policy arguments. But these are only argumentative techniques. There is never a "correct legal solution" that is other than the correct ethical and political solution to that legal problem. Put another way, everything taught, except the formal rules themselves and the argumentative techniques for manipulating them, is policy and nothing more. It follows that the classroom distinction between the unproblematic, legal case and the policy-oriented case is a mere artifact: each could as well be taught in the opposite way. 51

Until the last sentence, m y only criticism is that K e n n e d y seems to be saying m o r e than he is. Yes, everything in law is ultimately a " w h a t to d o " p r o b l e m , and so in a sense an ethical p r o b l e m . A n d therefore it can indeed be said that there is never a correct legal solution that is other than the correct ethical or, if y o u please, political solution (for ethical questions are themselves political in a society that is ethically diverse). But the impression conveyed b y stating this as a naked proposition is that the correct solution to every legal p r o b l e m is the solution that w o u l d be arrived at b y a careful student of ethics w h o k n e w nothing about law. A n d that is w r o n g . T h e ethicist and the j u d g e are subject to different ethical principles. T h e latter is, and the f o r m e r is not, a decision maker in a system of g o v e r n m e n t , and such a decision maker must be

51. "Legal E d u c a t i o n as T r a i n i n g f o r H i e r a r c h y , " in The Politics of Law: A Critique 40, 47 ( D a v i d K a i r y s ed. 1982).

Progressive

concerned not only with doing substantive justice in the case at hand but also with maintaining a legal fabric that includes considerations of precedent, of legislative authority, of the framing of issues by counsel, of the facts of record, and so forth. All this Kennedy ignores, and the oversight leads him to an erroneous conclusion. For there are unproblematic legal cases; they are so precisely by virtue of the considerations that he ignores. If such cases were taught as policy cases the teacher would not be teaching law, and if the real policy cases—the cases in the open area, where the conventional springs of legal decision making dry up—were taught as unproblematic legal cases the teacher would not be teaching policy where it ought to be taught.

P A R T II

THE O N T O L O G Y OF LAW

5 Ontology, the Mind, and Behaviorism

Ontological

Skepticism

Part I treated the question of law's objectivity primarily as one of epistemology, the branch of philosophy concerned with establishing (or disestablishing) foundations or warrants for knowledge. Epistemology is closely related to ontology, which deals with questions of existence (less provocatively, with what is). The relation lies in the fact that positing entities of debatable ontology is a frequent device for attempting to solve epistemological problems. If we agreed with Plato that there is a Form of Justice accessible to persons having certain aptitudes and experiences, then we might believe that the soundness of a legal decision could be gauged simply by comparing the decision with the Form; and likewise if we thought God had inscribed the principles of justice in the book of nature in a form accessible to human reason. Without going so far, we might think we could guarantee the ability of good judges to decide even difficult cases1 correctly by showing that for every question about law there exists a right answer, even if there is continuing, perhaps interminable, controversy over which answer is the right one. C o n versely, the skeptical position will be strengthened if it can be shown that lawyers and judges encounter serious difficulties in trying to latch on to useful entities. Philosophers have long puzzled over immaterial "things," although their enthusiasm for such debates is beginning to wane. Chicago and 1. I am deliberately a v o i d i n g the m o r e c o m m o n , b u t misleading, expression " h a r d cases." T h e t e r m originally m e a n t cases that t u g at the h e a r t s t r i n g s ; that is its m e a n i n g in the old saw " h a r d cases m a k e bad law." B y s e m a n t i c d r i f t it has c o m e t o m e a n difficult cases, w h i c h makes n o n s e n s e of the old saw. O n l y difficult cases m a k e law, g o o d o r bad. Cases that are easy to decide are so b y v i r t u e of b e i n g c o n t r o l l e d b y existing law.

N e w York, w e m a y agree tentatively, are things, o r assemblages of things. But w h a t about the "fact" that Chicago is west of N e w York— is that a thing, too? W h a t about another relation, the Pythagorean theorem? Is that a thing, despite its apparent lack of a spatio-temporal locus, and, if so, w o u l d it still be one if n o one had ever discovered it? N u m bers and other mathematical entities are objective, in a w a y that pains and colors are not. Yet they are not actual, in a w a y that pains and colors are. So w h i c h are "real"? T h e w o r d s printed on this page are things, but are the propositions they express? A horse is a thing; w h a t about Pegasus, in the sentence (which happens to be true) "Pegasus is a m y t h ical horse"? M i g h t not the assertion that s o m e t h i n g does not exist be t h o u g h t to i m p l y that it is a " t h i n g " of s o m e kind? Heating water to a certain temperature causes it to boil; does this m a k e causation a thing? What about the capacity of water to boil? M o r e generally, are dispositions—color, the speed of an ocelot at rest, the w e i g h t of a massive boulder firmly perched, the t e m p e r of a choleric person—things? Are intentions things? D o w e have a material part (the body) and an i m m a terial part (the m i n d , or if you prefer, the soul)? A rabbit is a thing; is law a thing? A n d w h a t about moral "entities"? Is "goodness" real? "Justice"? It is easy to b e c o m e dizzy thinking about such questions. But they are only the beginning. O n reflection, even the distinction between material and immaterial things blurs. A "solid object"—a table, for instance—is a material thing that consists mostly of e m p t y space (the space within and b e t w e e n its atoms), and a flock of pigeons is a material thing with visible spaces between its constituents. "America's cities" is another discontinuous, more-or-less material thing. But "a pigeon in N e w York and a s p a r r o w in C h i c a g o " is n o t a thing; n o t all sets of n o n c o n t i g u o u s objects count as things. E x p o s u r e to these puzzles makes it difficult to retain a robust faith in a clear-cut ontology, one that contains rabbits, minds, the law, fictional characters, n u m b e r s , and so forth in happy coexistence. O u r use of w o r d s and ideas seems driven by expediency rather than by a striving f o r correspondence with things "as they are." T h e baffling character of questions of existence gives Charles Peirce's pragmatic approach to m e a n i n g — t h e m e a n i n g of a proposition is its consequences, so that propositions that have n o consequences are m e a n ingless—a strong appeal even t h o u g h as a definition of meaning it is forced and polemical. T h e proposition that G o d created the universe but then w i t h d r e w , and has never intervened in its operations and never will, is not meaningless, even t h o u g h it has n o consequences in the sense either that experiments or observation could falsify or confirm the p r o p -

osition or that wc might behave differently if the proposition were true than we would if it were false. It would be m o r e accurate, if less dramatic, to say not that the proposition is meaningless but that it is not worth bothering one's head about. Nevertheless that is an important statement too, and it will give us a handle on dealing with questions of ontology. We can ask, for example, what the consequences are of believing that people do or do not have minds. M a y b e there are no consequences. Maybe the heavy talk in law about such mental entities as intent and premeditation, or such metaphysical entities as causation and justice, is bogus. The reader familiar with philosophical debates may smell a whiff of logical positivism, the view that all propositions can be sorted into one of three bins: the tautological, the empirically verifiable, and the n o n sensical. And a whiff of the antimetaphysicality of the later Wittgenstein and of J. L. Austin. O r the antirealism or conventionalism, already alluded to, that teaches either that our grammar (broadly defined) and perceptual structure divide up the world into convenient slices that need not correspond to the actual structure of things—the view neatly s u m marized by William Blake in " T h e Mental Traveller" as "the Eye altering alters all"—or, alternatively, that things, material and immaterial, are the joint creation of nature and the human frame of reference, which is itself a product of nature. 2 The details, variants, and pedigree of what I am calling with some license the pragmatic approach to meaning, and, by extension, to questions of ontology, need not concern us. All that is important is that the approach be able to improve our understanding of law. Its implications are skeptical, and thus complement the implications of the epistemological approach of Part I. But they are not, or at least need not be, radically skeptical, and again this is in tune with the earlier chapters. Indeed the approach can be used to deflect (although h o w effectively is an open question) the skeptical ploy of asking you to prove that you are not just a brain in a vat, receiving impressions of an external world f r o m a mad scientist w h o controls your access to the sensory world. (This is the

2. T h e literature o n these issues is vast; places to start i n c l u d e H i l a r y P u t n a m , The Many Faces of Realism: The PanI Carus Lectures, lectures 1 and 3 (1987); Willard Van O r m a n Q u i n e , " O n What T h e r e Is," in Q u i n e , From a Logical Point of View: Nine Logico-Philosophical Esasys 1 (2d ed. 1980); Q u i n e , Theories and Things (1981); N e l s o n G o o d m a n , Of Mind and Other Matters, chs. 2 - 3 (1984); B e r t r a n d Russell, The Problems of Philosophy (1912). O f i n c o m p a rable vividness is William J a m e s ' s discussion of o n t o l o g y in The Principles of Psychology, vol. 1, pp. 285-286, and vol. 2, p p . 2 9 1 - 2 9 3 (1890). T h e m i n d ' s c o n s t r u c t i o n of objects is explored f r o m a biological p e r s p e c t i v e in J. Z . Y o u n g , Philosophy and the Brain 8 0 - 8 2 (1987).

modern, technocratic version of the age-old puzzle: H o w do we know there is a real world out there, when all we have is sensations?) Most efforts to defend against the brain-in-a-vat attack founder on the absence of an external reference point with which to compare the experience of being a brain in a vat with the experience of inhabiting the "real" world. The conundrum is cleverly designed to eliminate any such point. If you are a brain in a vat, unlike a person dreaming, who also has (or believes he has!) a waking state, you have by assumption the identical sensations you would have if you were not a brain in a vat—so how do you know whether you are one or not? The pragmatic answer to this is that believing oneself to be a brain in a vat can have no consequences for one's behavior, unless one is crazy. O n e carries on just as before. The belief has no consequences and therefore makes no serious claim to assent. 3 You may be tempted to reply that someone w h o was convinced he was a brain in a vat would realize that his life was meaningless, and that this realization would affect his behavior—he would not exert himself as much as he otherwise would. But h o w could you be convinced you were a brain in a vat? You might entertain this as a possibility but, by hypothesis, there is no way of proving it or even of showing it to be more probably true than false. In fact it is more probably false than true; so it would not be rational to act on it. In further fact we seem incapable of entertaining such a belief except when playing philosophy. T h e qualifications implicit in such language as "rational" and "unless one is crazy" may seem to give the game away, however. If attempting to act on the belief that one is a brain in a vat is taken to demonstrate insanity, and the irrelevance of the belief to action is demonstrated by the fact that only an insane person would act on it, one is arguing in a circle. But the circle can be broken by considering the other beliefs and actions of the insane. A person w h o really believes he is a brain in a vat is bound to have other crazy beliefs. We k n o w they are crazy by watching him act on them. We observe that he is a poor guide to acting rationally in the sense of adapting means to ends. His behavior displays a high

3. An alternative approach, also pragmatic, is to treat the existence o f the external world as simply the most plausible hypothesis about "reality"; Bertrand Russell's discussion o f the o n t o l o g y o f the cat, quoted in Chapter 2, illustrates this approach. Another argument (from the standpoint o f evolutionary epistemology, mentioned in a note in Chapter 3) is that a belief in an external w o r l d is so m u c h more conducive to survival and reproduction than disbelief in it that the belief is "hard-wired" into our brains; w e simply are incapable o f really believing, in the sense o f being prepared to act on, the proposition that there is no external world.

degree of contradiction and futility, and these are reasons for not counting his "vote" on the question whether people are brains in vats. 4 This discussion of rationality casts a sidelight on the test of time. Chapter 3 suggested a voting analogy, noting that time broadens the franchise. Since there are eligibility requirements for voting, the analogy implies that maybe some potential voters in the temporal election should be disfranchised. Yet to use rationality as a criterion f o r j u d g i n g the test of time undermines the claim that the test is the (or an) ultimate foundation of or warrant for our knowledge. A question related to whether there is an external world, but more pertinent to the problems of jurisprudence, is whether other people, or animals, really have minds; for one can never observe another creature's mind. O n one level this is an absurd question—we simply cannot help thinking that there are other minds. 5 But on another level it is a fruitful question, or at least a spur to fruitful speculation. There is after all something deeply puzzling about "the mind." 6 It is not only unobservable but immaterial, yet despite its immateriality it seems to be in control of a material object, the body. Even the concept of mind is unclear. M i n d seems related to consciousness (the set of one's conscious thoughts is a typical dictionary definition of mind), yet it would be odd to say that 4. Cf. D a v i d Pears, The False Prison: A Study of the Development of Wittgenstein's Philosophy, vol. 1, p p 3 2 - 3 3 (1987); and recall the distinction m a d e in C h a p t e r 3, w i t h specific reference to c o m m i t m e n t for insanity, b e t w e e n coercive and c o m p e t i t i v e objectivity. R o r t y takes a m o r e skeptical v i e w : " t h e limits of sanity are set b y w h a t we can take seriously. This, in turn, is d e t e r m i n e d b y o u r u p b r i n g i n g , o u r historical s i t u a t i o n . " " T h e P r i o r i t y of D e m o c r a c y to P h i l o s o p h y , " in The Virginia Statute for Religious Freedom: Its Evolution and Consequences in American History 257, 267 (Merrill D . P e t e r s o n and R o b e r t C . V a u g h a n eds. 1988). 5. In Wittgenstein's characteristic f o r m u l a t i o n , "just try to k e e p h o l d o f this idea [that the people a r o u n d m e are a u t o m a t a , lack consciousness, even t h o u g h t h e y b e h a v e the s a m e w a y as usual] in the m i d s t of y o u r o r d i n a r y i n t e r c o u r s e w i t h o t h e r s , in t h e street, say! Say to yourself, for e x a m p l e : ' T h e children o v e r there are m e r e a u t o m a t a ; all their liveliness is mere a u t o m a t i s m . ' A n d y o u will either find these w o r d s b e c o m i n g quite m e a n i n g l e s s ; o r you will p r o d u c e in yourself s o m e k i n d of u n c a n n y feeling." L u d w i g W i t t g e n s t e i n , Philosophical Investigations 126e (3d ed., G. E. M . A n s c o m b e t r a n s . , 1968) (H420). C o m p a r e H u m e ' s view of skepticism: " t h a t all his [ B i s h o p Berkeley's] a r g u m e n t s , t h o u g h o t h e r w i s e intended, are, in reality, merely sceptical, appears f r o m this, that they admit of no answer and produce no conviction. T h e i r o n l y effect is to cause that m o m e n t a r y a m a z e m e n t and i r r e s o lution and c o n f u s i o n , w h i c h is the result o f s c e p t i c i s m . " An Enquiry Concerning Human Understanding 155 n. 1 (3d ed., P. H . N i d d i t c h ed., 1975), (§12, pt. 1). 6. As d e m o n s t r a t e d in Gilbert Ryle's classic, The Concept of Mind (1949). For a serviceable recent discussion of the m i n d - b o d y p r o b l e m , see M y l e s B r a n d , Intending and Acting: Toward a Naturalized Action Theory (1984). A n o t h e r recent discussion, o n e that is b o t h fascinating and v e r y m u c h in the spirit of Ryle's w o r k in d o u b t i n g the e x p l a n a t o r y value of the idea of the m i n d , is A r t h u r W. Collins, The Nature of Mental Things (1987).

people have no minds while they are sleeping or that all the things people do without conscious thinking—which might include writing an inspired passage of music or poetry 7 —are mindless. Much, perhaps most, thought is unconscious; recall the discussion of tacit knowing in Chapter 3. Maybe, therefore, the idea of the mind has no consequences, at least no interesting or pertinent ones. I am not a philosophical skeptic and will not pretend to be agnostic about whether I am the only person in the world with a mental life. Obviously most adults and older children can and do speak without vocalization (that is, can "conceal their thoughts") and f o r m mental images. But this barebones concept of mind, which essentially equates mind to consciousness, is different f r o m the idea that there is a something, the "mind," which is the locus of intentions, the invisible puppeteer, the inner man or w o m a n . It is that idea which may have no consequences for law and should perhaps be discarded, despite the law's emphatic (but, I shall argue, shallow) c o m m i t m e n t to it. An alternative approach, also broadly pragmatic, is to ask why we cannot help thinking that other people have minds. I suggest that we often use the w o r d " m i n d " (either in the weak sense of consciousness or in the strong sense of intentionality and control) not to name a thing, not to make an ontological assertion, but to cover our ignorance of certain causal relationships. 8 Dispel the ignorance, and the concept of mind ceases to have consequences and can be—and what is more interesting, is—discarded. For example, we are more likely to impute a mind to a cat than to the most powerful computer, even though the computer will beat the world's smartest cat at chess every time/' We think we k n o w more about the causality of a computer's operations than about the causality of a cat's behavior. Wc are not sure w h y a cat j u m p s onto one 7. O n the i m p o r t a n c e of the u n c o n s c i o u s in creativity, see, for e x a m p l e , The Creative Process: A Symposium ( B r e w s t e r Ghiselin ed. 1952). 8. C o m p a r e W i t t g e n s t e i n ' s " b e e t l e " e x a m p l e : " S u p p o s e e v e r y o n e had a b o x w i t h s o m e thing in it: w e call it a 'beetle'. N o o n e can l o o k into a n y o n e else's b o x , and e v e r y o n e says he k n o w s w h a t a beetle is o n l y b y l o o k i n g at his b e e t l e . — H e r e it w o u l d be quite possible for e v e r y o n e to have s o m e t h i n g different in his b o x . . . B u t s u p p o s e the w o r d 'beetle' had a use in these people's l a n g u a g e ? — I f so it w o u l d be used as t h e n a m e of a t h i n g . " W i t t g e n stein, n o t e 5 above, at lOOe (11293). B u t w e d o n o t say that w e k n o w w h a t the m i n d is only b y l o o k i n g at o u r o w n m i n d ; w e d r a w inferences in the m a n n e r suggested in C h a p t e r 2. 9. Actually, a cat is m u c h m o r e intelligent than a n y c o m p u t e r — i n d e e d , the w o r l d ' s m o s t p o w e r f u l c o m p u t e r p r o b a b l y has n o m o r e c o m p u t a t i o n a l p o w e r than a cockroach. (This will change.) C o m p u t e r s are s p e c i a l - p u r p o s e calculators, unlike animals, and t h r o u g h specialization can " b e a t " animals (including people) at certain specific, well-defined, " r o u t i n izable" tasks. See Philip E l m e r - D e W i t t , "Fast and S m a r t : D e s i g n e r s Race to Build the S u p e r c o m p u t e r s of the F u t u r e , " Time Magazine, M a r c h 28, 1988, at 54.

person's lap rather than another's, w h y it meows in a certain way, why sometimes it purrs and at other times it flattens its ears; we are not sure that all this is just instinctual, programmed. (Maybe zoologists are sure; if they are, I would expect them to attribute less mental activity to a cat than does the man in the street.) We impute a mind to a cat in the hope that we can predict and therefore influence the cat's behavior in the same way that we try to anticipate and adjust to people's behavior by assuming they think the way we do. This would not be a plausible strategy for dealing with computers, at least as they are currently designed. This analysis suggests that as human beings learn more about the world, the number of posited mental entities or states diminishes—and so we observe. Ancient and primitive peoples often impute minds to "inanimate" objects, such as the sea. This is notable in Homer; and it is unlikely that he or his audience regarded Poseidon as merely a fictional construct. 1(1 Ignorance about nature made the imputation of mind to "inanimate" objects plausible. T h e sea behaves in a tempestuous and unpredictable fashion, a little like a person; maybe it is a person, and therefore can be placated the way a wrathful, powerful person sometimes can be. Once we learn that the causality of a tempest is different from that of a human temper tantrum, we cease imputing mental activity to the sea. If we understood the stock market better, we would cease personifying it. Aristotle's physics treats objects in nature much as if they were animate beings, with goals; today we are more likely to treat animate beings on the model of objects.

Mental and Other Metaphysical

Entities in Law

I am about to cross the line from general ontology to legal ontology, and it may be useful to indicate at the outset what we shall be encountering on the other side. I shall begin with the question of mental entities in law—intent, premeditation, "free will" (as in the principle that a defendant's confession, to be admissible in a criminal trial, must be the product of the defendant's free will). I shall argue, in the spirit of the preceding section, that these are entities of distinctly dubious ontology (along with such other well-worn courtroom entities as "causation" and "reputation"), and that the law recognizes that at some level and is much less mentalistic than legal semantics implies. I propose that we can do without the concept of mind in the strong sense even when analyzing 10. 1 grant that this is a difficult q u e s t i o n . See Paul Veyne, Did the Greeks Believe in Their Myths? An Essay on the Constitutive Imagination (1988).

judicial behavior. T h e next chapter asks w h e t h e r there exist "right answers" either to controversial legal questions or to the difficult factual questions that arise in litigation; to overstate slightly, m y answer is no. C h a p t e r 7 gives the same answer to the question w h e t h e r law itself is an entity. I propose instead that it is an activity, t h o u g h I recognize the incompleteness of the activity theory. Mens Rea. Oliver Wendell H o l m e s believed, m u c h in the spirit of the first section of this chapter, that the role of mental states in law d i m i n ishes as law becomes m o r e sophisticated, reflecting the progress of scientific k n o w l e d g e ; this is a m a j o r t h e m e of his classic, The Common Law (1881). Recall that in early law a deodand—it m i g h t be the wheel of a cart that had run over and killed s o m e o n e — w a s often punished as a criminal, o n the theory that it had an evil will, w h i c h had caused it to kill. F r o m these and other examples H o l m e s argued that as law matures, liability—even criminal liability—becomes progressively m o r e "external," that is, m o r e a matter of conduct than of intent. There is plainly s o m e merit to the idea. For example, if w e had a complete model of the criminal act, so that w e could predict a crime with one hundred percent accuracy f r o m i n f o r m a t i o n about a person's genes and upbringing, w e p r o b a b l y w o u l d n o t require p r o o f that the act had been "intended," in order to punish the actor. M a y b e w e w o u l d talk about people being p r o g r a m m e d to kill rather than deciding to kill, and m a y b e w e w o u l d n o longer use the w o r d " p u n i s h m e n t . " We w o u l d deal with criminals as w e deal w i t h unreasonably dangerous machines, which implies that, w h e r e feasible, w e w o u l d intervene before the crime was committed. H o l m e s foretold such a d e v e l o p m e n t . " In doing so he was standing the concept of deodand on its head: instead of treating dangerous objects as people, he was p r o p o s i n g to treat dangerous people as objects. O n e could argue that as law becomes m o r e sophisticated, states of m i n d should play an ever larger role in liability. O u r understanding of the m i n d m a y i m p r o v e — m a y b e w e will learn to read minds. But m a y b e there is n o t h i n g to read, or m a y b e w e are not interested in what the m u r d e r e r was thinking w h e n he pulled the trigger. If we take seriously the actor's adage that n o m a n is a villain in his o w n eyes, w e can expect to find, if w e ever succeed in peering into the murderer's mind, an elaborate, perhaps quite plausible, rationalization for his deed. But so what? We w o u l d punish h i m all the same. 1 2 T h e social concern is w i t h the deed

I t . In his essay " T h e P a t h o f the L a w , " 10 Harvard Law Review 457, 470-471 (1897). 12. C o m p a r e A. D . N u t t a l l , " D i d M e u r s a u l t M e a n t o Kill the A r a b ? — T h e Intentional Fallacy Fallacy," 10 Critical Quarterly 95 (1968).

(whether impending or already committed) rather than with the mental state that accompanies it. The one mental element presupposed by this view of law is the capacity of the potential lawbreaker to understand the legal threat and the other relevant circumstances in which he is operating, including any threat posed by the victim. This may seem a significant qualification of my antimentalist stance, because it opens up the possibility of excuses based on mistake or insanity. In fact the only sort of mistake likely to exculpate an offender is the reasonable mistake, which is simply the inference f r o m average behavior; and the only type of insanity that is likely to exculpate is the insanity that manifests itself in behavior—in "acting crazy"—apart f r o m behavior during the crime itself. We can use formal or informal methods of statistical inference to d r a w inferences f r o m observed behavior, but w e cannot discover private thoughts, at least not without torture and probably not with it, for the intentional component of human action may not be conscious. So while it is true that even a behaviorist (or determinist—I use the terms interchangeably) view of law assumes that the persons whose behavior we want to constrain k n o w what the law requires in the situation in which they find themselves, as well as what that situation is factually (what choices the actor has, and so forth), the ability to c o m prehend does not require a mind that harbors intentions and directs muscles. Many animals can understand orders. T h e distinction between comprehension and volition is the distinction suggested earlier between a weak and a strong sense of "mind." A parallel distinction is between rational and mental activity—equivalently, between a rational decision and thinking. The laboratory rat that, faced with a choice between a larger and a smaller quantity of food, chooses the larger is being rational, but this characterization does not c o m m i t us to the idea that the rat has a "mind." T h e person w h o refrains f r o m committing a crime because, when the threat of punishment is added to the other disutilities of such conduct, the net expected utility f r o m the deed is negative is being rational too. And unlike the rat, he must possess language, to understand the threat and choose in light of it (but so, in a sense, must a dog taught to heel—it must understand commands). But whether and h o w he thinks the choice through and "makes up his m i n d " is of no concern in a behaviorist conception of law, the conception Holmes bequeathed us. That conception is prominent in the economic analysis of law. William Landes and I have proposed an analysis of intentional torts that dispenses with the notion of "intent" other than as a proxy for certain

characteristics of the tortious act, notably a large gap between the cost ' of the act to the victim and the small or even negative cost to the injurer of not committing the act. 13 T h e difference between negligence and deliberate wrongdoing, on this view, is that the negligent person is one w h o does not devote enough resources (care, or whatever) to avoiding inflicting injury, whereas the deliberate wrongdoer is one w h o expends resources on inflicting it. It is the difference between the person w h o takes another's umbrella by foolish mistake and the person w h o is "careful" to take an umbrella that is not his. In the w o r d "deliberate," however, is concealed a notion of purposive activity—and can a nonmentalist account of purpose be given? It can. "Purpose" describes the fitting of means to ends. You do A because that is the least costly way of bringing about state B, the state that confers the greatest net benefits on you. What we were doing, then, was substituting what is rational for a person to do for what the person intends to do, 14 and this substitution was a natural one because the process by which hypostasized mental states give way to behavioral hypotheses is central to economic analysis. T h e "utility function" in economics is a concept parallel to "intent" in law. It summarizes the tastes, values, preferences, and objectives of the individual characterized by the function. A goal of economic research is to change as many of the elements of the utility function— mysterious mental entities—as possible into parameters, which can be measured. 1 5 So one might begin by suggesting that some people have a "taste" for obtaining a college education but then show that this taste is instrumental to a more general goal; call it income maximizing. The propensity to attend college will n o w be seen as a function of the cost of college and of the effect of college in raising one's lifetime income. 16 13. William M . Landes and R i c h a r d A. Posner, The Economic Structure of Tort Law 1 4 9 189 (1987). 14. For similar a p p r o a c h e s f r o m a philosophical angle, see N u t t a l l , n o t e 12 above; A n t h o n y Kenny, " I n t e n t i o n and P u r p o s e in L a w , " in Essays in Legal Philosophy 146, 159161 ( R o b e r t S. S u m m e r s ed. 1968); Daniel C . D e n n e t t , The Intentional Stance, ch. 2 (1987). 15. A particularly b o l d e x a m p l e of this a p p r o a c h is G e o r g e J. Stigler and Gary S. Becker, " D e G u s t i b u s N o n Est D i s p u t a n d u m , " 67 American Economic Review 76 (1977). 16. See G a r y S. Becker, Human Capital: A Theoretical and Empirical Analysis, with Special Reference to Education (2d ed. 1975). T h e r e is e x t e n s i v e empirical s u p p o r t for the e c o n o m i c m o d e l of h i g h e r e d u c a t i o n ; for a small s a m p l e of the literature, see W. W. M c M a h o n , " E x p e c t e d Rates of R e t u r n to E d u c a t i o n , " in Economics of Education: Research and Studies 187 ( G e o r g e P s a c h a r o p o u l o s cd. 1987); C h r i s t o p h e r A. Pissarides, " F r o m School to U n i v e r s i t y : T h e D e m a n d f o r P o s t - C o m p u l s o r y E d u c a t i o n in B r i t a i n , " 92 Economic Journal 654 (1982); Alex M a u r i z i , " R a t e s o f R e t u r n t o D e n t i s t r y and the Decision to E n t e r D e n t a l S c h o o l , " 10 Journal of Human Resources 521 (1975); F r a n k A. Sloan, " T h e D e m a n d for H i g h e r Education: T h e C a s e of Medical School A p p l i c a n t s , " 6 Journal of Human Resources 466 (1971); M a r k C.

Ideally, one could predict whether people would go to college even if one knew nothing about their thoughts on the subject; and then one might stop talking, in analytical w o r k at least, about people "wanting" to go to college or "thinking about" going to college. People would still have desires and thoughts, but these would be strictly epiphenomenal. 1 7 I have not mentioned free will, and this may seem a serious omission. That many of the economists w h o p l u m p most strongly for a determinist approach to h u m a n behavior are libertarians may seem to d e m o n strate a deep fissure in the economic model; and h o w can a behaviorist approach handle such classic intrusions of "free will" thinking into law as the principle that a criminal's "involuntary" confessions cannot be used against him at his trial? It may be that in fastening on the measurable dimensions of human behavior, the economists are overlooking its most important dimension. In trying to answer these questions I begin by setting to one side the brand of determinism that asserts or, more commonly, implies that everything in history, including m y writing this sentence, was determined at the m o m e n t of the big bang—that not only did everything that is or will be come f r o m that explosion but that the whole course of history was s o m e h o w p r o g r a m m e d into it, so that a sufficiently p o w erful mind observing the event could have foretold everything to come. But merely to suppose that there is randomness in the universe—even to suppose that in some sense not everything that happens is caused 18 — does not make plausible the idea of free will in its strong sense of selfcaused human action. A random action is not free; a person w h o made every decision in his life by flipping a coin would be a practicing detcr-

Bcrger, "Predicted F u t u r e E a r n i n g s and C h o i c e of C o l l e g e M a j o r , " 41 Industrial and Labor Relations Review 418 (1988); Y o s h i - f u m i N a k a t a and C a r l M o s k , " T h e D e m a n d for C o l l e g e E d u c a t i o n in P o s t w a r J a p a n , " 22 Journal of Human Resources 377 (1987). 17. Y o u n g , n o t e 2, above, at 7 3 - 7 4 , presents fascinating biological e v i d e n c e that c o n scious intentions are indeed e p i p h e n o m e n a l — t h a t they f o l l o w r a t h e r t h a n p r e c e d e the brain's " d e c i s i o n " to initiate " v o l u n t a r y " action. T h e m o r e basic p o i n t , well a r g u e d in C o l lins, note 6 above, ch. 6, is that o u r ability to give a reason for an action in t e r m s of a goal does not m a k e the goal the cause of the action. In like vein, C a t h e r i n e F i t z m a u r i c e and K e n Pease, in The Psychology of Judicial Sentencing 39 (1986), n o t e that p e o p l e m a y infer their reasons for b e h a v i n g as they do f r o m o b s e r v i n g that v e r y b e h a v i o r ; this is deliberation as rationalization. 18. T h e apparent implication, s t r o n g l y resisted b y Einstein, of q u a n t u m theory. T h e story is well told in A r t h u r Fine, The Shaky Game: Einstein, Realism, and the Quantum Theory (1986). For an i n g e n i o u s p r a g m a t i s t c o n s t r u a l of q u a n t u m theory, illustrating the utility of p r a g m a t i s m in dealing w i t h — o r eliding—difficult o n t o l o g i c a l questions, see D u g a l d M u r doch, Niels Bohr's Philosophy of Physics 2 2 4 - 2 3 5 (1987).

minist rather than a free man, because he w o u l d have submitted himself to an external force. This implies, and one observes, that a person w h o acts in accordance w i t h compelling reasons is normally t h o u g h t to be acting freely; 19 s o m e h o w , r e m o v i n g the r a n d o m element—reducing uncertainty about the consequences of action—enhances f r e e d o m . We think of "deliberation" as the characteristic behavior of a free person, but, as H o b b e s argued, the " d e " of "deliberation" is like the " d e " of "deemphasize"; b y deliberating w e n a r r o w our choices, ideally to one. If t h r o u g h deliberation or cognate analytical methods all uncertainty about the consequences of alternative courses of action can be removed, the choice is likely to be " p r e d e t e r m i n e d " ; for uncertainty is w h a t principally makes a choice difficult rather than easy. If you ask a m a t h e m a tician w h a t the second derivative of x2 + 1 is, you know he'll say 2. T h e significance of consciousness m a y be not as a mechanism of choice b u t as a m e c h a n i s m for enabling a person to "see" m o r e clearly the pros and cons of alternative actions. H e makes representations of those actions, and this enables h i m to estimate the benefits and costs of alternative actions w i t h o u t actually p e r f o r m i n g those actions. In m a n y circumstances this is a m o r e efficient procedure than trial and error. But it is possible that once consciousness or imagination has enabled the person to f o r m a clear idea of the advantages and disadvantages of alternative courses of action, his choice a m o n g t h e m is determined; he goes w h e r e the balance of advantages leans. Consciousness, in this account, is a means of acquiring i n f o r m a t i o n rather than the agency of free choice; 20 indecision is the state in w h i c h a person has n o t obtained the optimal a m o u n t of information; and a free person is one w h o does not decide w h a t to do until he has obtained the optimal a m o u n t of information bearing on his choices, and w h o then acts rationally. Stated dif-

19. See T h o m a s N a g e l , The View from Nowhere, ch. 7 (1986); and o n the d i l e m m a s of " f r e e w i l l " generally, see D a n i e l C . D e n n e t t , Elbow Room: The Varieties of Free Will Worth Wanting (1984); A n t h o n y Kenny, Will, Freedom, and Power (1975) (especially ch. 8); P. F. S t r a w s o n , " F r e e d o m and R e s e n t m e n t , " in S t r a w s o n , Freedom and Resentment, and Other Essays 1 (1974); B e n s o n M a t e s , Skeptical Essays, ch. 2 (1981); Peter van I n w a g e n , An Essay on Free Will (1983); G a r y W a t s o n , " F r e e A c t i o n and Free Will," 96 Mind 145 (1987); Ted H o n d e r i c h , " O n e D e t e r m i n i s m , " in Essays on Freedom of Action 187 ( H o n d e r i c h ed. 1973); Ryle, n o t e 6 a b o v e , ch. 3; Responsibility, Character, the Emotions: New Essays in Moral Psychology (Ferdinand S c h o e m a n ed. 1987); E r n e s t N a g e l , The Structure of Science: Problems in the Logic of Scientific Explanation 592—606 (1961). 20. It is n o t i n c o n s i s t e n t w i t h this v i e w that if y o u " k n o w " w h a t a person is a b o u t to d o and tell h i m , he m a y d o s o m e t h i n g else; f o r the i n f o r m a t i o n y o u have i m p a r t e d alters his e n v i r o n m e n t . If y o u k n e w h o w h e w o u l d react to y o u r telling h i m w h a t he was g o i n g to do, y o u could predict his b e h a v i o r , as parents o f small children learn to do.

ferently, w e have desires, and w e have beliefs—formed w i t h the help of consciousness—about h o w to fulfill those desires; and the conjunction of the desires and beliefs determines, w i t h o u t need to posit a faculty of free will, o u r volitional behavior. 2 1 T h e concept of choice that I am e x p o u n d i n g resembles the economist's idea of decision under uncertainty, 2 2 or w h a t I have been calling "rational decision." It also resembles Quine's v i e w that an act is "free insofar as the agent's motives or drives are a link in its causal chain. Those motives or drives m a y themselves be as rigidly determined as y o u please." 23 O n either approach there is n o entity that can properly be called "free will." Adjectival rather than substantive, the t e r m describes certain behavior, being in this respect like the w o r d "luck." " L u c k " is not a thing that people possess; it n o m o r e has a locus in space and t i m e than the speed of a j a g u a r at rest has. It is a w a y of characterizing people's experiences after the fact. Similarly, "free will" is n o t a thing b u t a description of behavior n o t w h o l l y constrained b y forces external to the motives and drives of the actor. If I a m t h r o w n o n t o m y neighbor's land, I do not act of m y o w n free will; if I a m not t h r o w n b u t step o n t o his land, I d o so act. Admittedly this solution to the p r o b l e m of free will leads to p a r a d o x es: a d r u g addict is free and, for that matter, so is a sparrow. Yet this is not so odd as it sounds. We m a y simply understand the motives or drives of creatures w h o s e motives or drives are relatively simple (sparrows) or w h o are in the iron grip of a single o v e r m a s t e r i n g drive (addicts) better than w e understand the motives and drives of the average m a n or w o m a n . A n d it is i m p o r t a n t to distinguish prediction f r o m control. If the causality of the sparrow's behavior is simple e n o u g h , w e may be able to predict that behavior w i t h great accuracy. B u t the sparr o w is f r e e — f r o m us. (And so w i t h the addict.) It is n o t free f r o m its instincts and circumstances, however, and m a y b e w e are n o m o r e free f r o m our instincts and circumstances, vastly complicated t h o u g h they are. Just as one can be free w i t h o u t consciousness of f r e e d o m (I a m assuming the sparrow does n o t have consciousness), one can be unfree yet think one is free. K n o w i n g the addict's weakness, I m a y be able to 21. Actions 22. Events 23. above,

See D o n a l d D a v i d s o n , " A c t i o n s , Reasons, and C a u s e s , " in D a v i d s o n , Essays on and Events 3 (1980). See D o n a l d D a v i d s o n , "Replies to Essays I - I X , " in Essays on Davidson: Actions and 195, 199-201 (Bruce V e r m a z e n and Merrill B. H i n t i k k a eds. 1985). W. V. Q u i n e , " T h i n g s and T h e i r Place in T h e o r i e s , " in Theories and Things, n o t e 2 at 1, 11. See also H u m e , n o t e 5 a b o v e , at 95 (§8, p t . 1, 1173).

control his behavior as effectively as I could if m y brain rather than his gave directions to his muscles. O n e can be free in the H u m e - Q u i n e sense yet determined in an equally good sense. Even if we had "true" free will—a faculty of unfettered choice—it would not guarantee our freedom, for it might lack causal efficacy, as when I "choose" to stay in a room not k n o w i n g that it is locked f r o m the outside. A final paradox, closely related to m y earlier remark on random choice, is that an unpredictable choice may be less free than a predictable one. A person w h o chooses irrationally—who given a choice between what he likes and what he dislikes chooses the latter—is not a free person, but rather a person w h o is incapable of exercising free will. Whether the conception of free will that I am defending—one that essentially equates free will with rational choice in the sense of fitting means to ends, however those ends be chosen and whether or not the "fitting" involves conscious mental activity—is right or wrong, complete or incomplete, edifying or unedifying, it has the merit of summarizing what is unquestionably an influential, perhaps dominant, strain in the law's actual as opposed to ostensible approach to controlling human behavior. What progress the law has made in controlling behavior, particularly in the criminal area, has come f r o m the replacement of mentalist with behavioral explanations—not only in dealing with deodands but also, to take a less exotic example, in determining h o w severely to punish particular offenders and offenses. Criminologists have developed models of recidivism that enable them to predict, on the basis of characteristics such as drug addiction, lack of education, youth, and criminal record that have nothing directly to do with the criminal's thoughts or conscious choices, which criminals are likely to commit crimes after being released f r o m prison. 24 The behavioral model works better than any mental model of criminal propensities.

24. T h e accuracy o f these m o d e l s is the subject o f a vast literature, m o s t l y finding that their accuracy is o n l y fair. See, for a small s a m p l e of the literature, Peter S c h m i d t and A n n D r y d c n Witte, Predicting Recidivism Using Survival Models 157-160 (1988); Stephen P. Klein and Michael N . C a g g i a n o , The Prevalence, Predictability, and Policy Implications of Recidivism 3 7 - 3 8 (1986); S t e p h e n L. Biebcr et al., " P r e d i c t i n g C r i m i n a l R e c i d i v i s m of Insanity A c q u i t tees," 11 International Journal of Law and Psychiatry 105 (1988); Peter B. H o f f m a n and J a m e s L. Beck, " R e c i d i v i s m a m o n g Released Federal Prisoners: Salient Factor Score and FiveYear F o l l o w - U p , " 12 Criminal Justice and Behavior 501 (1985); A n d r e w v o n Hirsch and D o n M . G o t t f r e d s o n , "Selective Incapacitation: S o m e Q u e r i e s a b o u t Research D e s i g n and E q u i ty," 12 Review of Law and Social Change 11, 4 4 - 4 5 (1983-84); K e v i n N . Wright, T o d d R. Clear, and Paul D i c k s o n , " U n i v e r s a l Applicability of P r o b a t i o n R i s k - A s s e s s m e n t Instruments: A C r i t i q u e , " 22 Criminology 113 (1984); F . J . Spellacy and W. G. B r o w n , "Prediction of Recidivism in Y o u n g O f f e n d e r s after Brief I n s t i t u t i o n a l i z a t i o n , " 40 Journal oj Clinical

Even when the criminal law seems most mentalist, as in punishing premeditated crimes more severely than impulsive ones, the underlying policy and even the m o d e of implementation may be nonmentalist. T h e criminal w h o premeditates is more likely to succeed in his criminal aim (and thus do more harm) than is the impulsive criminal, and the premeditator is also harder to apprehend and punish and therefore less likely to be punished. O n both counts effective deterrence requires a heavier punishment if he is caught. 25 A further point is that the would-be criminal w h o plans his crime—that is, w h o projects in his imagination the alternative courses of action of committing and refraining f r o m c o m mitting the crime—may be more responsive than the impulsive criminal to the threat of punishment, because he sees in his mind's eye the full array of costs and benefits, including punishment, before deciding which course to follow. Punishing him may therefore be socially more productive than punishing the impulsive criminal. N o t only do we punish the premeditating criminal more severely for reasons having nothing to do with free will, but in deciding whether a crime is premeditated we e m p l o y — o x y m o r o n though it may seem—a behaviorist account of deliberation. 26 We examine the circumstances of the crime: Was it concealed? Had the criminal made arrangements for a getaway? Had he obtained the means of committing the crime in advance? Were those means suitable to the end (suitably lethal, in the case of a murder)? Did the criminal have m u c h to gain f r o m the crime? From these circumstances a model of a deliberating criminal could be constructed—an "objective" reconstruction of the criminal's m o t i vational experience, created by attributing to him a certain type of rationality. But the step is unnecessary, or perhaps merely heuristic. "Conscious" choice can be redescribed in nondeliberative terms as costbenefit analysis, utility maximization, or means-end rationality; even some philosophical accounts of deliberation do this. 27 It thus is possible, if paradoxical, to understand premeditation in behavioral terms, as a

Psychology 1070 (1984); Terrill R. H o l l a n d et al., C o m p a r i s o n and C o m b i n a t i o n of Clinical and Statistical P r e d i c t i o n s of R e c i d i v i s m a m o n g A d u l t O f f e n d e r s , " 68 Journal of Applied Psychology 203 (1983); H o w a r d Wainer and A n n e M . B. M o r g a n , " R o b u s t E s t i m a t i o n of Parole O u t c o m e , " 19 Journal of Research in Crime and Delinquency 84 (1982). 25. See m y article " A n E c o n o m i c T h e o r y of t h e C r i m i n a l L a w , " 85 Columbia Law Review 1193, 1222-1223 (1985). 26. See the fascinating discussion in N u t t a l l , n o t e 12 a b o v e . 27. See, for e x a m p l e , D . S. C l a r k e , J r . , Practical Inferences 15-69, 1 0 9 - 1 2 6 (1985); D a v i d son, n o t e 21 above.

r o u n d a b o u t w a y of describing the criminals w h o are most likely to succeed. T h e persistence o f mentalist language in law m a y merely bespeak the cultural conservatism o f the legal enterprise. T h e approach that I a m suggesting does not eliminate the mental elem e n t entirely. Planning is mental—an affair of consciousness. I deny not the existence of mental p h e n o m e n a but the utility for law of the concept of m i n d in w h i c h intentions and free will figure. " T h e division of acts into s o m e f o r w h i c h a m a n is regarded as responsible, and others for which he is not, is part of the social apparatus of reward and punishm e n t : responsibility is allocated w h e r e rewards and p u n i s h m e n t s have tended to w o r k as incentives and deterrents." 2 8 A n d being social rather than philosophical in purpose, the allocation of responsibility need not f o l l o w the division b e t w e e n free and coerced acts. A person w h o freely, indeed eagerly, kills in self-defense is excused f r o m criminal liability, but n o t the killer w h o w o u l d n o t have killed had he n o t been raised in a p o o r h o m e b y harsh parents. P u n i s h m e n t w o u l d be counterproductive in the first case, b u t not in the second. Society has n o desire to license people w h o have had a bad u p b r i n g i n g to kill, and it believes that the threat of p u n i s h m e n t (which to be credible m u s t be carried out w h e n it fails to deter) will deter m o s t such people f r o m killing. B o t h cases involve the exercise of free will in Quine's sense. In contrast, one w h o kills another in unreasonable fear f o r his o w n safety, or w h o has sexual intercourse w i t h an underage female w h o m he reasonably believes to be of age, is guilty of a crime, even t h o u g h in neither case does the criminal k n o w he is c o m m i t t i n g a criminal act. These are examples of the strictliability c o m p o n e n t in criminal law. T h e r e are also explicit strict-liability crimes, mainly involving the sale of adulterated o r unsafe foods or drugs. If m y analysis is correct, there is n o a n o m a l y in sometimes i m p o s i n g criminal liability on the p u r e of heart or the e m p t y of mind. T h e criminal law is an i n s t r u m e n t of social control, and—I am contending—treats people as objects, n o t as Kantian subjects. T h e "cold and calculating" killer is not punished m o r e severely than the impulsive killer because his will was freer. H e m a y have been acting u n d e r a heavier c o m p u l s i o n f r o m his past, his genes, his psyche, or his upbringing. To decide w o u l d require a type of investigation that is not 28. Q u i n e , " S m a r t ' s P h i l o s o p h y and Scientific R e a l i s m , " in Theories and Things, n o t e 2 above, at 92, 9 4 - 9 5 . (Ryle, n o t e 6 a b o v e , at 69, m a k e s a similar p o i n t . ) T h e p r o o f of this particular p u d d i n g (that legal r e s p o n s i b i l i t y and free will can be and are divorced) is the fact that w h e n e x p e r t s o n free will discuss practical q u e s t i o n s of responsibility, such as criminal responsibility, t h e y o f f e r analyses n o different f r o m t h o s e of l a w y e r s i g n o r a n t of philosophy. See, f o r e x a m p l e , D e n n e t t , n o t e 19 a b o v e , at 158-165.

conducted in criminal trials, and that a determinist would consider a snipe hunt. Either the courts are deeply deceived, or the point is simply that the criminal w h o plans his crime in advance is on average more dangerous than the impulsive criminal because he is more likely to do harm and more difficult to apprehend, and dangerousness is a reason for punishing one criminal more severely than another. H e is punished, in Nietzsche's term, as "an irresponsible piece of fate." It is true that people w h o commit crimes under an irrational compulsion are not punished. They are not deterrable by the threat of punishment, because they cannot bring the expected cost of punishment to bear in "deciding" whether to commit the crime. They should be and are taken out of circulation, because they are highly dangerous, but there is nothing to be gained by going through the rigmarole of punishment. In contrast, there is a gain f r o m punishing strict-liability crimes such as statutory rape; the threat of such punishment encourages the would-be offender to steer well clear of the protected class, and thereby protects the class more securely. T h e most difficult case to explain is that of p u n ishing the person w h o acts out of an unreasonable fear—how is he different f r o m the person w h o acts f r o m an irrational compulsion? Evidentiary problems aside, it may be that we think the threat of punishment will induce sane people to act with greater care in circumstances of apparent danger, but not insane people. T h e benefits of punishment are therefore greater in the first case. Deterrence presupposes a capacity to make decisions and act on them—presupposes, in a word, rationality, though no stronger sense of free will—and evidently our modern conception of criminal punishment is linked to deterrability and through it to rationality. Purely incapacitative or even retributive theories of punishment might sever that link. Behaviorist approaches to criminal responsibility disturb thoughtful students of law and morals. (In contrast, the "reasonable m a n " standard of tort law and the "objective" theory of contracts raise few hackles.) We are not comfortable thinking of people on the model of animals or automata, or in supposing that the criminal justice system "thinks" of them in those terms, although that is largely what the system does, and it does so for t w o excellent reasons: we cannot peer into people's minds, at least not with the clumsy tools of legal procedure, and if w e could w e are not at all sure that we would find the intentions, malice, premeditation, or other entities that the mentalist language of law invites us to expect. T h e behaviorist approach seems to leave no r o o m for appeals to conscience, for a sense of guilt, of remorse; it seems to strip the moral as well as the distinctively h u m a n content f r o m the criminal law. T h e

pragmatic reply is, So what? There are no realistic grounds for fearing that speculations in the philosophy of mind are likely to affect respect for, let alone observance of, law. A behaviorist metalanguage will not sap people of their sense of moral responsibility—and not only because most people will remain unaware of the metalanguage and its implications. Philosophers w h o believe in determinism behave in their personal lives just like other people. If freedom is an illusion, it is one of those illusions (like the illusion of three-dimensional perception after we have learned that the retina is two-dimensional) that we cannot shake off no matter what our beliefs or opinions are. 29 And there is no mystery about the survival value of the illusion (if that is what it is) of freedom, and thus no mystery about w h y it m a y be hard-wired into our brains. The feeling of freedom forces us to act deliberately, in the sense of gathering as much relevant information as possible concerning the pros and cons of alternative courses of action before we act. Admittedly I am defending behaviorism (or determinism) merely against the charge that it will have bad effects, whereas the main charge is that it reflects an impoverished conception of h u m a n nature. It lies at the farthest possible remove f r o m Kant's influential view that human beings should not treat each other just as objects. 3 " But the question is not what is good for us to believe on the score of freedom and determinism but what premise is good for the law to proceed on. Behaviorism is the only practical working assumption for law, and its dangers have been exaggerated. Moreover, even if behaviorist approaches to criminal law pose serious dangers, so do mentalist ones. The heavy reliance on confessions in medieval and early modern law helps explain the surprising frequency of nonexistent criminal acts, such as acts of witchcraft and sorcery. Perhaps we should neither mentalize nor moralize criminal law. Although behaviorism is a modern idea, it makes a smooth if precarious fit with Enlightenment ideas of personal freedom and autonomy— which seems doubly paradoxical: because freedom is the antithesis of determinism and because in treating people as objects behaviorism is the antithesis of Kant's insistence on the autonomy of the human subject. The paradox diminishes w h e n w e descend f r o m the theoretical to the practical plane. Confining governmental regulation to external behavior 29. For a s t r i k i n g e x a m p l e , see H o w a r d M a r g o l i s , Patterns, Thinking, and Cognition: A Theory of Judgment 38 (1987) (fig. 2.1) (the "Lincoln g r i d " ) . A n d c o m p a r e M e i r D a n - C o h e n , " D e c i s i o n R u l e s and C o n d u c t Rules: O n A c o u s t i c Separation in C r i m i n a l Law," 97 Harvard Law Review 625 (1984). 30. For an e l o q u e n t s t a t e m e n t , see " F r e e d o m and R e s e n t m e n t , " n o t e 19 above, at 9.

leaves the realm of t h o u g h t and feeling as one of personal autonomy. T h e separation is precarious, however, because n o t h i n g in behaviorism bars the w a y to using Pavlovian conditioning to prevent antisocial behavior—and that is "messing w i t h m i n d s " in a dramatic way. Coerced Confessions. M e n t i o n of confessions brings to m i n d a particularly difficult question relating to the question of free will: the distinction between " v o l u n t a r y " and " i n v o l u n t a r y " confessions, the f o r m e r being admissible in the criminal trial of the person w h o confessed, the latter not. 3 1 Even if criminal behavior is n o t "mentalist" in a strong sense, there is n o denying that m o s t criminals are aware of w h a t they have done (even if, to repeat, they w e r e n o t " f r e e " to avoid d o i n g it) and can be induced to confess it. A n d m a n y crimes could n o t be solved except at exorbitant, and sometimes at infinite, cost unless the criminal confessed. If the means used to coerce a confession are so painful or threatening ("confess or I'll kill you on the spot") that the person being interrogated w o u l d find it less costly to confess even to a crime that he had not c o m m i t t e d than to u n d e r g o the interrogation, w e w o u l d almost certainly want to disallow the confession, and this regardless of considerations of free will. T h e costs of obtaining evidence of such dubious reliability w o u l d in all but the rarest cases exceed the value of the evidence, given that there usually are other means of establishing guilt. But the m o d e r n law goes further. It excludes reliable confessions induced by threats, promises, or misrepresentations w h e n the pressure exerted by these methods is deemed unreasonable even t h o u g h n o t necessarily so extreme as to be likely to induce an innocent person to confess. For example, the police m i g h t promise a m u r d e r e r that he w o u l d be let off with a light sentence if he confessed. T h e promise w o u l d if believed sway a murderer, but p r o b a b l y n o t an innocent person (of sound mind—a potentially significant qualification) w h o had been falsely suspected. Nevertheless such a promise, if false, w o u l d m a k e the confession induced by it inadmissible, on the g r o u n d that it was "involuntary." In a society as depressingly rife with crime as that of m o d e r n America, 31. O n the legal position, see, for e x a m p l e , B e c k w i t h v. U n i t e d States, 425 U . S . 341 (1976); S c h n e c k l o t h v. B u s t a m o n t e , 412 U . S . 218, 2 2 3 - 2 2 7 (1973) (not a confession case, but a particularly g o o d s u m m a r y of c o e r c e d - c o n f e s s i o n doctrine); S p a n o v. N e w York, 360 U . S . 315 (1959); Frazier v. C u p p , 394 U . S . 731 (1969); H a y n e s v. W a s h i n g t o n , 373 U . S . 503 (1963); J u r e k v. Estelle, 623 F.2d 929 (5th Cir. 1980); o t h e r cases succinctly s u m m a r i z e d by J u d g e Friendly in U n i t e d States ex rel. H u g h e s v. M c M a n n , 405 F.2d 773, 776 (2d Cir. 1968); Fred E. Inbau, J o h n E. Reid, and J o s e p h R Buckley, Criminal Interrogation and Confessions 3 0 8 - 3 2 6 (3d ed. 1986); J o s e p h D . G r a n o , " V o l u n t a r i n e s s , Free Will, and t h e L a w of C o n f e s s i o n s , " 65 Virginia Law Review 859 (1979); S t e p h e n J. Schulhofer, " C o n f e s s i o n s and the C o u r t , " 79 Michigan Law Review 865 (1981).

the rule against involuntary confessions m a y seem a luxury w e can't afford. I will n o t address that issue. T h e point of philosophical interest lies in the connection b e t w e e n the legal idea of voluntariness and the philosophical idea o f free will. T h a t there is a connection is n o t h i n g new; but I shall argue that the connection is w i t h the H u m e - Q u i n e concept of free will, a concept determinist in spirit. Recall that a r a n d o m choice is not free. By definition it is one made u n d e r radical ignorance of its consequences. If the police use threats, lies, false promises, or the like to alter the e n v i r o n m e n t in which the criminal suspect m u s t decide w h e t h e r or n o t to confess, this is like forcing the suspect to roll the dice; f o r a false sense of the consequences of choice is n o better than n o sense. T h e point is not that the promises, threats, or misrepresentations cause the suspect to confess and by doing so negate his free will. A free choice is n o t an uncaused one. O t h e r w i s e courts w o u l d have to exclude m o s t statements obtained in interrogations, because the interrogation is a cause of the confession elicited by it— clearly so w h e n the suspect is in custody and is being interrogated by a r m e d police officers skilled at o v e r c o m i n g objections, playing on the suspect's anxieties, exploiting slips, and otherwise taking advantage of a favorable setting. 3 2 Yet although interrogation in these circumstances m a y m a k e confession far m o r e likely, it does not confront the suspect w i t h a false e n v i r o n m e n t that prevents a reasoned choice—a weighing of the actual (psychic as well as practical, of course) consequences of alternative courses of action free at least f r o m manufactured uncertainty. Despite the mentalist language in which the law of coerced confessions comes wrapped, 3 3 the theory as I have articulated it is consistent with

32. T h e fascinatig Inbau treatise, n o t e 31 above, pt. 1 and a p p e n d i x , is a " h o w t o " h a n d b o o k o n e x t r a c t i n g a confession b y raising the suspect's anxiety level to the point w h e r e it exceeds his p e r c e p t i o n of the adverse c o n s e q u e n c e s of confessing (and the interr o g a t o r is t a u g h t to m i n i m i z e t h o s e consequences). See also E d w i n D . Driver, " C o n f e s s i o n s and the Social P s y c h o l o g y of C o e r c i o n , " 82 Harvard Law Review 42 (1968). T h e cases reco g n i z e t h e i n h e r e n t l y coercive character o f police i n t e r r o g a t i o n but are correct f r o m a philosophical as well as practical s t a n d p o i n t in r e f u s i n g to j u m p to the conclusion that no confession o b t a i n e d b y such i n t e r r o g a t i o n s h o u l d be admissible. See, f o r example, O r e g o n v. M a t h i a s o n , 429 U . S . 492, 495 (1977) (per c u r i a m ) . 33. " C o n f r o n t e d w i t h the express threat of c o n t i n u e d i n c o m m u n i c a d o d e t e n t i o n and induced b y the p r o m i s e of c o m m u n i c a t i o n w i t h and access to family, H a y n e s u n d e r s t a n d ably chose to m a k e and sign t h e d a m n i n g w r i t t e n s t a t e m e n t ; given the unfair and inherently coercive c o n t e x t in w h i c h m a d e , that choice c a n n o t be said t o be the v o l u n t a r y p r o d u c t of a free and u n c o n s t r a i n e d will." H a y n e s v. W a s h i n g t o n , 373 U . S . 503, 514 (1963). M a y b e the past tense w o u l d be m o r e a p p r o p r i a t e f o r the s t a t e m e n t in text; the S u p r e m e C o u r t ' s affection for the " f r e e and u n c o n s t r a i n e d w i l l " a p p r o a c h to confession law is rapidly w a n ing. See n e x t f o o t n o t e .

the overall behaviorist approach of the criminal law. 34 Remember that rational choice is not the same as free will in the strong sense. A rational choice is objectively right; no assumptions are made about the chooser's freedom or subjectivity. A behaviorist approach to the law of confessions may appear to do better with misrepresentation than with the creation of psychic distress (for example, holding the suspect incommunicado). T h e latter f o r m of pressure or coercion is more accurately described as placing a price on refusing to confess than as preventing a rational, informed decision on whether or not to confess, although if the price is high enough it may elicit an unreliable confession. There is, however, a third alternative. The pressure may interfere with clear thinking (in the nonmentalist sense in which a pigeon might be said to be "thinking" of which bit of food to eat next), may distract and confuse, undermining the power of reasoned choice even if the pressure is not so severe that it would be likely to elicit a false confession. 35 The law's c o m m i t m e n t to preventing the obtaining of confessions by misrepresentation is not absolute; a certain amount of trickery is permitted in interrogations. In particular, exaggeration of the other evidence that the police have gathered against the suspect, a tactic designed to make him feel that he has nothing to lose by confessing, is permitted. The main reason, I conjecture, is that ordinarily the police will exagger34. T h e C o u r t has said that a criminal d e f e n d a n t has n o " r i g h t . . . to confess to his crime only w h e n totally rational and p r o p e r l y m o t i v a t e d . " C o l o r a d o v. C o n n e l l y , 479 U . S . 157, 166 (1986). " W e h a v e n e v e r read the C o n s t i t u t i o n to require that the police s u p p l y a suspect w i t h a flow of i n f o r m a t i o n to help h i m calibrate his self-interest in d e c i d i n g w h e t h e r to speak or stand b y his r i g h t s . " M o r a n v. B u r b i n e , 475 U . S . 412, 422 (1986). Judicial tolerance of psychological pressure t o extract a confession is well illustrated b y Miller v. F e n t o n , 796 F.2d 598 (3d Cir. 1986). As this and m a n y o t h e r cases m a k e clear, the statement in B r a m v. U n i t e d States, 168 U . S . 532, 5 4 2 - 5 4 3 (1897), that to be v o l u n t a r y a confession m u s t n o t have been " e x t r a c t e d b y any sort of threats or violence, n o r o b t a i n e d b y any direct or i m p l i e d p r o m i s e s , h o w e v e r s l i g h t , " is n o l o n g e r a u t h o r i t a t i v e . 35. T h e possibility of eliciting false confessions b y relatively m i l d f o r m s o f p s y c h o l o g ical pressure s h o u l d n o t b e u n d e r r a t e d , h o w e v e r . See William Sargent, Battle for the Mind: A Physiology of Conversion and Brain-Washing, ch. 9 (1957). T h e i n t e r r o g a t o r ' s o b j e c t is to raise the suspect's anxiety level, and an i n n o c e n t p e r s o n m a y suffer f r o m anxiety, and his anxiety is likely to be exacerbated by police q u e s t i o n i n g . T h i s m i g h t seem to i m p l y that innocent people w o u l d be more likely to confess than g u i l t y ones. N o t so, at least in a system such as o u r s w h e r e (see n e x t chapter) f e w i n n o c e n t p e o p l e are c o n v i c t e d . K n o w l edge of this fact will tend to allay the i n n o c e n t p e r s o n ' s anxiety. In a d d i t i o n , he has m o r e to lose f r o m confessing than the guilty p e r s o n , for there is m o r e likely to be o t h e r evidence of guilt, besides a confession, against a guilty p e r s o n than against an i n n o c e n t one. B u t all this assumes that the i n n o c e n t p e r s o n is rational and r e a s o n a b l y well i n f o r m e d , even if anxious, and the a s s u m p t i o n is n o t a l w a y s w a r r a n t e d .

ate the strength of the other evidence only if that evidence is indeed w e a k — w h y exaggerate otherwise? B u t this implies that they really need the confession. A n d the trickery allowed is not the sort that is likely to elicit a false confession. T h e careful reader will discern the rudiments of cost-benefit analysis here. W h e n the desire for effective law enforcement o u t w e i g h s the aversion to the use of a confession obtained by trickery to convict, the j u d g e s call the confession "voluntary." L a w follows philosophy only so far as it is expedient to do so. T h e r e m a y be a f u r t h e r point. In Julius Caesar, Caesar's decision to go to the Capitol on the fatal day is procured by the flattery and misrepresentations of the conspirators, w h o s e skillful playing o n Caesar's vanity overcomes his wife's fears. Yet w e w o u l d not be inclined to deny that the decision was an act of free will by Caesar. T h e decision was in c h a r a c t e r — p r o f o u n d l y so—and that seems e n o u g h . It m a y likewise be the case that w h e n police interrogators procure a confession by playing to the weaknesses of a suspect's character, w e nevertheless are inclined to think of the confession as authentically the suspect's, and w e therefore allow it to be admitted against him at his trial. It m i g h t be argued that the decision to confess can never be rational, because invariably the criminal thinks better of the decision later and tries to w i t h d r a w his confession. B u t the impression of u n i f o r m retraction is j u s t another illustration of h o w sample bias complicates legal analysis. T h e only litigated confession cases are those in which the defendant had second t h o u g h t s about his decision to confess, true; but only a small percentage of criminal cases are litigated, m o s t being disposed of on the basis of a guilty plea. N o r need regret s h o w that the regretted decision was not rational. T h e confession m a y have been impelled by stings of remorse that fade w i t h time. T h e balance of costs and benefits that determines decision in a behaviorist m o d e l of h u m a n action changes over time; so opposite decisions m a d e at different times m a y be consistent. T h e person w h o (whether innocent or guilty) confesses under a credible threat to kill h i m on the spot unless he confesses is striking a rational balance b e t w e e n the costs and benefits of confession; should his confession therefore be admissible? It should not be; but the reason is that the rational choice u n d e r such a threat is to confess regardless of guilt, rather than that the threat prevents a rational choice f r o m being made. T h e choice induced b y the threat is perfectly rational but the confession that results f r o m the choice is unreliable and therefore inadmissible. A similar case w o u l d be o n e in w h i c h the police promised a ne'er-do-well suspected of a m i n o r crime $100,000 to confess and fully intended to fulfill

their promise. T h e r e w o u l d be n o misrepresentation and n o other interference with rational choice, b u t there w o u l d be a serious danger of eliciting a false—although entirely voluntary—confession. Like the threat in the previous case, the p r o m i s e in this one w o u l d m a k e the confession inadmissible. In b o t h cases the confession is voluntary, yet unreliable and therefore properly excluded. What if in either case the confession is corroborated? T w o situations must be distinguished. In the first there is merely other evidence of the confessor's guilt. This evidence m a y m a k e it m o r e likely than n o t that the confession is true, but it does not m a k e the confession itself a reliable piece of evidence, any m o r e than " c o r r o b o r a t i v e " evidence w o u l d enhance the reliability of a confession k n o w n to have been forged. In the second and m o r e interesting situation, the confession produces leads that validate its truthfulness; for example, it m a y disclose facts that only the perpetrator of the crime could have k n o w n . T h e exclusion of such a confession f r o m the defendant's trial—a confession b o t h voluntary in the H u m e - Q u i n e sense and reliable—is, like the exclusion of the leads themselves, justifiable only on a deterrence rationale. We w a n t to discourage a class of confessions that w e believe to be unreliable o n average, so w e forbid the police to derive any benefit f r o m obtaining a confession in that class, even a confession w h o s e reliability is d e m o n strable. This strategy makes sense f r o m a behaviorist standpoint only if the reliable confessions are a small fraction of the total class, but perhaps that is the case. Let us compare m y explanation of the legal principles g o v e r n i n g the admissibility of confessions w i t h the alternative explanations. T h e alternative already examined and f o u n d w a n t i n g is that the legal system has subscribed to a strong sense of free will. A n o t h e r possibility is that involuntary confessions (involuntary in the H u m e - Q u i n e sense, that is) are inadmissible because they are likely to be unreliable. This explanation fails on t w o counts. First, m a n y involuntary confessions are so well corroborated by physical or other evidence that their reliability is not open to serious question, although the qualification well corroborated is important. C o r r o b o r a t i o n m a y s h o w n o m o r e than that the defendant k n e w s o m e t h i n g about the crime, perhaps because he participated peripherally in it. (So corroboration shares the infirmities of i n d u c t i o n — is perhaps the same thing.) Second, if reliability were the underlying concern in these cases, w e would expect the standard for admissibility to be—reliability, period, for that is a simpler and m o r e familiar criterion for admission of evidence than voluntariness is. We could n o t explain a case like Townsend

v. Sain,36 w h e r e the S u p r e m e C o u r t held involuntary a confession procured b y a truth serum. If a t r u t h s e r u m w o r k s , the confession procured by it will be reliable. B u t the t r u t h s e r u m will deprive the person interrogated of his p o w e r of choice b y preventing him f r o m weighing the pros and cons of confessing and going w h e r e the balance inclines. This is a g o o d illustration of the difference between an involuntary and an unreliable confession. Still another explanation for the exclusion of involuntary confessions is that w e are horrified b y brutal police m e t h o d s . But a truth serum is not brutal, and f e w misrepresentations are brutal; yet confessions obtained by a t r u t h s e r u m or misrepresentations (especially promises of lenient treatment) are deemed involuntary and are inadmissible. T h e p r o p o n e n t of the police-methods theory m a y retreat to the a r g u m e n t that it is not the brutality but merely the felt inappropriateness of particular interrogative m e t h o d s that c o n d e m n s t h e m . T h a t begs the question. T h e pattern of the confession cases is m o r e perspicuous once the m e n talist, nondeterministic concept of free will is discarded. This is a pragmatic reason for seeking to g o b e y o n d the mentalist language in which the j u d g e s articulate their reasons. To the complaint that the H u m e Q u i n e v i e w is u n e d i f y i n g — t h a t w h a t is gained in clarity is lost in s y m b o l o g y and r e s o n a n c e — m y reply is as before: s h o w m e w h a t practical h a r m will ensue f r o m greater realism about law in general and confession law in particular. Other Applications. If the above analysis is correct, the defense of duress in contract law also has n o t h i n g to do w i t h free will. A promise to pay, extracted at the point of a gun, is a product of free choice, just as w i t h the parallel case of confessing in the face of a mortal threat. We refuse to enforce the p r o m i s e because w e don't w a n t to encourage such threats, n o t because they deprive people of free will. A n o t h e r entity the law could d o w i t h o u t is "reputation." W h e n it is viewed as a thing, the tendency is to think of it as an asset of the person " w h o s e " reputation it is. F r o m there it is but a small step to thinking that s o m e o n e w h o reveals a true b u t unflattering fact about a person is taking s o m e t h i n g a w a y f r o m h i m and should be liable for the resulting harm. But w h y should people have a legal right to deceive others about their true qualities? W h a t is actually involved is n o t taking something f r o m the person w h o is h a r m e d by disclosures about his character or past conduct; it is facilitating i n f o r m e d transactions with the person by 36. 372 U . S . 293, 3 0 7 - 3 0 8 (1963).

dispelling misconceptions that actual or potential transacting parties may have. And notice h o w recognition of a right to deceive would undermine the policy behind the rule on coerced confessions. Denying the "thingness" of reputation is the key step in demystifying the legal concept of privacy. 37 It enables one to see that, insofar as privacy is more than just the desire not to have one's peace and quiet disturbed by loud music or by telephone solicitors and other pests, it is a desire to be able to deceive the persons with w h o m one transacts in economic, marital, and other markets. Yet the favorable connotations of the term are doubtless what induced the Supreme C o u r t to appropriate it for a line of decisions (involving mainly contraception and abortion) that involve the interest neither in solitude nor in deception, but in sexual liberty. All this is not to deny that privacy and sexual liberty are genuine goods. Most people, myself included, would prefer the burden of unmasking the self-aggrandizing deceptions of the persons with w h o m they deal to being forced to forego such deceptions themselves—to parade as it were naked before the world. I am objecting merely to the reification of the concept of privacy. Continuing the struggle against metaphysical entities in law, Landes and I have argued that causation in tort law is a dispensable category, because, as Quine might say, the question whether to impose liability on conduct belongs to the social apparatus of blame and punishment. 3 8 We related our approach to the strain in the modern philosophical literature on causation [that] regards the cause of an occurrence as whatever antecedent condition, necessary or sufficient, is most significant in relation to the reasons for which causal ascription is being made. A chemist might ascribe fire to the presence of oxygen in the atmosphere; the fire marshal, to a failure to comply with some fire regulation. Thus, one must k n o w the purpose of causal ascription in tort law before one can say what causation in that law means. If the purpose of tort law is to p r o m o t e economic efficiency, a defendant's conduct will be deemed the cause of an injury w h e n making him liable for the consequences of the injury will p r o m o t e an efficient allocation of resources. When it would not p r o m o t e efficiency for the defendant to behave differ-

37. As I a t t e m p t to d o in m y b o o k The Economics of Justice, pt. 3 (1981). 38. See Landes and Posner, n o t e 13 above, ch. 8; also Steven Shavell, Economic Analysis of Accident Law, ch. 5 (1987). T h i s is n o t t o d e n y causality: it is to d e n y that the l a w requires a concept, definition, or d o c t r i n e of causation.

ently, the cause of the accident will be ascribed to an "act of G o d " or some other force on which liability cannot rest. 39

There are striking examples of the dispensability of causation in criminal law as well—in particular, the punishment of attempts and conspiracies that cause no harm at all but demonstrate the dangerousness of the perpetrator with sufficient reliability to warrant taking him out of circulation for a while. The debunking of metaphysical entities in law is not new. It was a pastime of the legal realists. 40 Their favorite target was the corporation. The corporation, being not a thing or person but the name for a pattern of contractual relationships, indeed has a dubious ontology. The realists' lesson has not been learned. Most lawyers and judges still think that one can speak meaningfully of powerful or wealthy corporations, or of placing taxes on corporations rather than on persons. But a corporation is not a rich person, natural or artificial; it is a network of relationships among people (shareholders, workers, customers, suppliers, executives), some wealthy, some not. The difference between the individual income tax and the corporate income tax is not that people pay the former and institutions the latter, but that different people pay, and often at different rates, under the two taxes. After the dust settles, it may turn out that a tax on corporate income operates as a regressive tax on consumers. The realists' emphasis was different: not on the social costs of corporate liability but on the injustice of corporate privileges. This was just the leftist streak in legal realism; it was nothing organic to a skeptical, antimetaphysical, pragmatic conception of law.

Behaviorism

and the Judicial

Perspective

In commending a behaviorist approach to law I may seem to be assuming a stance not only radically inconsistent with my own profession as a judge, which requires me to make decisions, implying mental activity and the exercise of free will, but, more important, inconsistent with any plausible theory of the judicial process. The behaviorist perspective is

39. Landes and Posner, n o t e 13 a b o v e , at 229. textual c a u s a t i o n , " see C h a r l e s M . Yablon, " T h e Studies and the P r o b l e m of Legal E x p l a n a t i o n , " 6 Ernest N a g e l , n o t e 19 a b o v e , at 582-587; cf. J o h n 8 Journal of Legal Studies 419 (1979).

For g o o d discussions of this sort of " c o n I n d e t e r m i n a c y o f the Law: Critical Legal Cardozo Law Review 917, 9 2 5 - 9 2 9 (1985); B o r g o , " C a u s a l P a r a d i g m s in Tort Law,"

40. See, for e x a m p l e , Felix S. C o h e n , " T r a n s c e n d e n t a l N o n s e n s e and the Functional A p p r o a c h , " 35 Columbia Law Review 809 (1935).

external rather than internal. H o w can it be adequate to the self-conscious activity of a judge? There is no inconsistency. Parents, economists, psychologists, marriage counselors, and probation officers all have the experience of being able to predict correctly what another person will do even when the person himself is undecided. This is just the sparrow and addict examples writ large. "We may imagine we feel a liberty within ourselves; but a spectator can commonly infer our actions f r o m our motives and character."41 An ambiguity lurks here. We have genuine "liberty" in the sense that most of our actions are free f r o m gross coercion of the sort found in duress ("your money or your life") cases. That liberty is H u m e - Q u i n e free will. The spurious liberty of which H u m e is speaking here is the sense we sometimes have in making choices of being free f r o m the constraints of instinct, culture, character, and other immutable or relatively immutable features of our person and environment. In fact the interaction of those features may, without our realizing it, determine all our choices. To return to m y examples, the point is not that the expert or the grown-up k n o w s the contents of another's mind better than the other does—an irrelevant, perhaps even meaningless, suggestion—but that the expert or parent has a method of prediction that the person whose actions are being predicted either does not command, for want of the necessary training or experience, or cannot use on himself because of his emotional involvement; it is easier to be analytical about other people than about oneself. And there is something deeper. People systematically misrepresent their motivations to themselves. Almost everyone pretends to himself that he is less concerned with money and status, less selfish, more courageous, more ethical, than he actually is. What Regan said of Lear applies to everyone: " H e hath ever but slenderly k n o w n himself." Judges are no exception. Their self-image of stern self-discipline and self-denial, of rigorous impartiality—a self-image fostered by psychological factors discussed below—may not be consonant with reality. The internal perspective on judging may not provide the best explanation for what judges do. A simple behaviorist model may have greater explanative force. Most judges may be quite predictable even t h o u g h n o judge thinks himself predictable. Yet it must be admitted that only limited progress has been made in predicting judges' votes on the basis of ideology, personality, social 41. David H u m e , A Treatise of Human Nature 408 (L. A. S e l b y - B i g g e ed. 1888) (Bk. II, pt. 3, §2).

class, party affiliation, or other characteristics, and this mainly (as one would expect) in areas where judges have the most discretion, that is, are least cabined by rules—such areas as Supreme Court decision making and criminal sentencing. 42 And courts, as distinct f r o m individual judges, really are unpredictable in an important sense. The difference is again due to sample bias. If parties to a dispute converge in their predictions of h o w the court will decide their case, they will almost certainly settle rather than litigate. T h e sample of litigated cases is drawn f r o m those disputes in which the court's resolution is difficult to predict—as well it may be even if most judges are quite predictable. For example, in a court with seven members, if three always vote a stereotyped "conservative" line and three a stereotyped "liberal" line, the unpredictability of the remaining j u d g e will make the court as a whole unpredictable. O n e needs only a fraction of unpredictable judges in a fraction of cases to make courts unpredictable. This may seem to imply that the problem of unpredictability is less acute at the trial level, where the judges (in all but a trivial fraction of cases) sit alone rather than in panels. But i n j u r y trials there are in effect many judges. And the possibility of appeal may render the ultimate outcome of the litigation uncertain even if there is no j u r y and the trial j u d g e is predictable. Also, 42. T h e recent literature of judicial b e h a v i o r i s m is illustrated b y Judicial Conflict and Consensus: Behavioral Studies of American Appellate Courts (Sheldon G o l d m a n and Charles M . L a m b eds. 1986); R o b e r t A. C a r p and C . K. R o w l a n d , Policymaking and Politics in the Federal District Courts (1983); G l e n d o n S c h u b e r t , Political Culture and Judicial Behavior (1985); Philip E. Tetlock, J a n e B e r n z w e i g , and J a c k L. Gallant, " S u p r e m e C o u r t Decision M a k i n g : C o g n i t i v e Style as a P r e d i c t o r of Ideological C o n s i s t e n c y of V o t i n g , " 48 Journal of Personality and Social Psychology 1227 (1985); M i c h a e l J. Saks and Reid Hastie, "Social P s y c h o l o g y in C o u r t : T h e J u d g e , " in Judgment and Decision Making: An Interdisciplinary Reader 255 (Hal R. A r k e s and K e n n e t h R. H a m m o n d eds. 1986); M a r t h a A. M y e r s , "Social B a c k g r o u n d and the S e n t e n c i n g B e h a v i o r of J u d g e s , " 26 Criminology 649 (1988); A n t h o n y C h a m p a g n e and Stuart N a g e l , " T h e P s y c h o l o g y of J u d g i n g , " in The Psychology of the Courtroom 257 ( N o r b e r t L. K e r r and R o b e r t M . B r a y eds. 1982); J a m e s L. G i b s o n , " J u d g e s ' Role O r i e n tations, A t t i t u d e s , and Decisions: A n Interactive M o d e l , " 72 American Political Science Review 911 (1978); H a r o l d J. Spaeth and Stuart H . Teger, " A c t i v i s m and Restraint: A C l o a k for the Justices' Policy P r e f e r e n c e s , " in Supreme Court Activism and Restraint 277 (Stephen C . H a l p e r n and C h a r l e s M . L a m b eds. 1982); J e f f r e y A. Segal, " P r e d i c t i n g S u p r e m e C o u r t Cases Probabilistically: T h e Search and Seizure Cases, 1962-1981," 78 American Political Science Review 891 (1984); S. Sidney U l m e r , "Social B a c k g r o u n d as an Indicator to the Votes of S u p r e m e C o u r t Justices in C r i m i n a l Cases: 1947—1956 T e r m s , " 17 American Journal of Political Science 622 (1973); V l a d i m i r J. K o n e c n i and E b b e B. E b b e s e n , " T h e M y t h o l o g y of Legal Decision M a k i n g , " 7 International Journal of Law and Psychiatry 5 (1984). O n e i m p l i cation of these studies, I n o t e s e l f - s e r v i n g l y — f o r like m o s t j u d g e s I feel the t u g of the " f a i t h f u l a g e n t " c o n c e p t i o n — i s that i n t e r m e d i a t e appellate j u d g e s are m o r e disciplined and restrained than either trial j u d g e s or s u p r e m e c o u r t j u s t i c e s because they have the least discretionary p o w e r .

trial j u d g e s have a lot of a v o w e d l y discretionary authority—over scheduling, admission of evidence, and other matters. I am not denying that j u d g e s exercise free will in the sense of making choices that are voluntary rather than the p r o d u c t of misrepresentation or disorientation. I a m denying that judicial introspection, and a fortiori j u d g e s ' avowals concerning the nature o f j u d i c i a l decision m a k i n g , are good explanations for judicial action. It is a mistake to take at face value descriptions of j u d g e s as engaged always in a search for " t h e " correct answer, rather than as exercising discretion u n d e r the influence of p e r sonal values and preferences determined b y t e m p e r a m e n t and selective life experiences rather than b y a considered, s o m e h o w self-chosen j u d i cial philosophy. 4 3 A teenager m a y honestly feel that he is deliberating over a choice of colleges that his parents k n o w is foreordained by factors of which the teenager himself is unaware. Adults deceive themselves as well. We should distrust not only the self-serving descriptions of the judicial mind b y the j u d g e s themselves but also the descriptions of that mind by academics w h o , w h e n they are n o t merely advocating a p o w erful judiciary, 4 4 or inferring the judicial mentality f r o m the rhetoric of judicial opinions—an illustration of the difficulty o f obtaining k n o w l edge of other minds—are projecting o n t o persons w h o generally lack the background, t e m p e r a m e n t , tastes, and aptitudes of academics the academic's vision of w h a t it w o u l d feel like to be a j u d g e . Here is a typical account, by a distinguished student of jurisprudence: "What j u d g e s d o say and write in justification o f their decisions rather appears as an attempt to s h o w that these decisions are dictated by authoritative legal standards; since they are j u s t as well able to follow jurisprudential discussions as a n y o n e else, it w o u l d seem that if the prevailing view that j u d g e s are legislators is correct, w e m u s t conclude that they are either stupid, i g n o r a n t o f the jurisprudential issues involved, o r hypocritical. N o n e of these conclusions is very appealing." 4 5 N o t "appealing" maybe, but implausible? Judges are n o t well versed in j u r i s prudential controversies, are n o t always intelligent, and are n o t free f r o m that f o r m of hypocrisy (if that is the right w o r d ) w h i c h consists of 43. For empirical evidence that the reasons j u d g e s give for their decisions are o f t e n r a t i o nalizations of actions caused by p s y c h o l o g i c a l factors o f w h i c h t h e y are u n a w a r e , see Fitzmaurice and Pease, n o t e 17 a b o v e , ch. 3. 44. T h a t is, g o v e r n m e n t by l a w y e r s . T h e n e x t best t h i n g to g o v e r n i n g is t o be g o v e r n e d by people like oneself; so it is natural, quite apart f r o m financial incentives, that m a n y academic and o t h e r l a w y e r s f a v o r the a g g r a n d i z e m e n t of the j u d i c i a r y at t h e e x p e n s e of the o t h e r branches of g o v e r n m e n t . 45. Rolf Sartorius, "Social Policy and Judicial Legislation," 8 American Philosophical Quarterly 151, 158 (1971).

adopting a public face not altogether consistent with one's innermost feelings. Indeed much of what judges say about their jobs in speeches and opinions partakes of the same falsity that characterizes other political discourse. Just as elected officials believe that their effectiveness depends on the public's thinking they serve the public interest rather than their personal interests or those of the most powerful special interests, so most judges believe, without evidence (indeed in the face of the evidence of public ignorance about courts noted in Chapter 4), that the judiciary's effectiveness depends on a belief by the public that judges are finders rather than makers of law. Although judges will occasionally lift the veil a bit, as I am doing in this chapter, they rarely level with the public— and not always with themselves—concerning the seamier side of the judicial process. This is the side that includes the unprincipled compromises and petty jealousies and rivalries that accompany collegial decision making, 4 6 the indolence and apathy that life tenure can induce, the flickers of ambition for different or higher office (judicial or otherwise), the b o r e d o m and burnout that heavy caseloads over a long period of years can induce, the pervasive reliance of modern American judges on ghostwriters (most judicial opinions nowadays are drafted by law clerks), the isolation f r o m normal h u m a n concerns that is experienced by people

46. H o l m e s is said to have described t h e Justices of the S u p r e m e C o u r t as "nine scorp i o n s in a b o t t l e . " A n alternative m e t a p h o r is an a r r a n g e d m a r r i a g e by indifferent parents in a s y s t e m w i t h n o divorce. J u d g e s do n o t pick their colleagues; m o s t of those colleagues serve until r e t i r e m e n t ; and the a p p o i n t i n g authorities give little w e i g h t to the interest in judicial collegiality in d e c i d i n g w h o m to a p p o i n t . R e g a r d i n g p r o b l e m s of collcgiality 1 am s p e a k i n g of appellate rather t h a n trial c o u r t s , since trial j u d g e s v e r y rarely sit in panels. For a vivid case s t u d y of o n e of m a n y f a m o u s f e u d s b e t w e e n colleagues o n an appellate court, sec M e l v i n I. U r o f s k y , " C o n f l i c t a m o n g the B r e t h r e n : Felix F r a n k f u r t e r , William O. D o u g las, and the Clash of Personalities and P h i l o s o p h i e s on the U n i t e d States S u p r e m e C o u r t , " 1988 Duke Law Journal 71. O n the p s y c h o l o g y of j u d g i n g , see A n d r e w S. Watson, " S o m e Psychological Aspects of the Trial J u d g e ' s D e c i s i o n - M a k i n g , " 39 Mercer Law Review 937 (1988). E f f o r t s to relate judicial decisions to the p e r s o n a l p s y c h o l o g i c a l m a k e u p of the j u d g e have generally f o u n dered. For a r e v i e w of such e f f o r t s — w h i c h m a y be said to h a v e b e g u n w i t h J e r o m e Frank, Law and the Modern Mind (1930), a b o o k that attributes legal f o r m a l i s m to authoritarian personality and r e c o m m e n d s that all j u d g e s u n d e r g o psychoanalysis—see C h a m p a g n e and Nagel, n o t e 42 a b o v e , at 262—264; L e o n Shaskolsky Sheleff, " T h e Illusions of L a w — P s y choanalysis and J u r i s p r u d e n c e in H i s t o r i c a l P e r s p e c t i v e , " 9 International Journal o f Law and Psychiatry 143 (1986). B u t t h e s t u d y b y Tetlock, B e r n z w e i g , and Gallant, cited in n o t e 42 above, presents, at last, p e r s u a s i v e evidence that p e r s o n a l i t y systematically affects judicial decision m a k i n g . A n d there is a g o o d deal of s u g g e s t i v e evidence, r e g a r d i n g the i m p o r t a n c e of leadership ability, personality, and socialization, in the studies collected in Judicial Conflict and Consensus: Behavioral Studies of American Appellate Courts, n o t e 42 above; for a s u m m a r y of the s u m m a r i e s , see " E p i l o g u e , " in id. at 275—291.

w h o spend all day, day after day, observing the w o r l d t h r o u g h the prism of law's "artificial reason," and the desire, conscious or not, to shape the law to one's personal values. 47 These are the occupational hazards of j u d g i n g , and they resemble those of bureaucracy. N o t all j u d g e s succ u m b to them; n o n e s u c c u m b s to all of t h e m (at least at the same time), if only because s o m e of the temptations are inconsistent w i t h others. But it is a mistake to credit the O l y m p i a n depictions of j u d g e s b y the judges themselves or by academics and other lawyers w h o believe in a p o w e r f u l judiciary and w a n t the public to be c o m f o r t a b l e w i t h the concept. From the fact that j u d g e s and their defenders in the profession emit a continuous stream of disclaimers that j u d g e s exercise p o w e r or (in other words) behave like other public officials, w e should n o t infer that judges are that special. This conclusion is reinforced by a n u m b e r of additional points. First, a j u d g e may feel constrained—may feel ruled rather than ruling—yet this feeling may be due to principles such as self-restraint and stare decisis that are (1) chosen by j u d g e s on (2) contestable g r o u n d s — t w o facts that j u d g e s often forget. Second, because policy and ethical considerations are permissible elements o f j u d i c i a l decision in o u r legal culture, a j u d g e may lack a sharp sense of w h e r e legal reasoning leaves off and policy j u d g m e n t or social vision (essential tools for resolving the indeterminate case) begins. Indeed, the j u d g e is likely to lack a sharp sense of the line between the pure judicial hunch and a w e l l - f o u n d e d proposition of law. Third, neither the conditions of j u d g i n g n o r the m e t h o d s of selecting judges—including self-selection—would lead one to expect the deep introspection and moral insight that academic literature often attributes to judges. Being a j u d g e , like being a military c o m m a n d e r , is an "ensemble" j o b . To be done well it requires a combination of qualities, of which reflective intellect is only one, and moral insight only another. O n e w o u l d not expect a j u d g e or a c o m m a n d e r to be as intellectually able, on average, as a person pursuing a parallel activity on a purely intellectual plane, such as legal scholarship or military science, even if the selection of j u d g e s were nonpolitical—which in this c o u n t r y it is not. 47. A desire i m p e d e d , h o w e v e r , b y the r a n d o m o r d e r in w h i c h cases c o m e b e f o r e j u d g e s and by the great variety o f cases. T h e s e t h i n g s m a k e it difficult for a j u d g e to set an a g e n d a for altering public policy. H e n c e a d i s c r e t i o n a r y j u r i s d i c t i o n , such as that possessed b y the S u p r e m e C o u r t , m a k e s it easier for a c o u r t t o set an a g e n d a . Specialized c o u r t s , t o o , have an easier t i m e than generalist c o u r t s in t a r g e t i n g particular policies for change. O n the o t h e r hand, o n a generalist c o u r t a j u d g e w i t h specialized k n o w l e d g e in a particular field m a y o v e r a w e his colleagues in a w a y that w o u l d be i m p o s s i b l e o n a specialized c o u r t . T h e t e n dency is held in check b y the practice, c o m m o n in state s u p r e m e c o u r t s , of assigning o p i n ions to individual j u d g e s in rotation, m a k i n g it m o r e difficult for a j u d g e to b e c o m e (or remain) a specialist.

N o r should we overlook the narrowing effect on the lawyer's and judge's mind of a professional lifetime spent reading judicial opinions. T h e judge's essential activity, moreover, is the making of a large n u m ber of decisions in rapid succession, with little feedback concerning their soundness or consequences. People w h o are uncomfortable in such a role—and perhaps they are the most introspective, sensitive, and scrupulous people—do not become judges, do not stay judges, or are unhappy judges. T h e j u d g e w h o is comfortable in his role makes the best decision he can and does not look back. 48 Denied the luxury of being able to withhold decision until persuaded by objectively convincing arguments that the decision will be correct, he no more wants to wallow in uncertainty and regrets than a law student wants to retake an exam in his mind after having taken it in the examination room. T h e starting point of Peirce's pragmatism is the idea that people hate being in a state of doubt and will do whatever is necessary to m o v e f r o m doubt to belief, and this is as true of judges as of anyone else. 49 A closely related point is that since no one likes to think he is making a lot of mistakes, the psychology of j u d g i n g is characterized by a belief that one is almost always (some judges think always) right. This point is an illustration of the theory of cognitive dissonance—the theory that people (and ostriches) will do almost anything to relieve the mental distress caused by an incongruity between their deepest beliefs, assumptions, and aspirations on the one hand and reality on the other. 50 Most means of avoiding or resolving cognitive dissonance (of which the most c o m m o n is a simple denial of the disturbing reality) are "irrational," so persons w h o resort to them normally pay a price—but not judges. There is no penalty for judicial complacency—for ignoring Holmes's dictum that "certitude is not the test of certainty." 51 Like other people, moreover, judges want to diffuse responsibility for their unpopular, controversial, or simply most consequential actions. (This is also related to the desire to wield power effectively.) They do this by persuading them48. For a n o t a b l e e x c e p t i o n , see D a v i d P a n n i c k , Judges 2 (1987), r e p o r t i n g that a j u d g e left m o n e y in his will to a litigant w h o m he decided he had ruled against erroneously. 49. See " T h e Fixation o f Belief," in Collected Papers of Charles Sanders Peirce 223, 2 3 1 233 (Charles H a r t s h o r n e and Paul Weiss eds. 1934). See also William J a m e s ' s essay, " T h e S e n t i m e n t of Rationality," in William James: The Essential Writings 25, 32 (Bruce W. Wilshire ed. 1984). 50. See, for e x a m p l e , Elliot A r o n s o n , " T h e T h e o r y o f C o g n i t i v e Dissonance: A C u r r e n t Perspective," in Cognitive Theories in Social Psychology 181 (Leonard B e r k o w i t z ed. 1978); G e o r g e A. A k e r l o f and William T. D i c k e n s , " T h e E c o n o m i c C o n s e q u e n c e s of C o g n i t i v e D i s s o n a n c e , " 72 American Economic Review 307, 3 0 8 - 3 1 0 (1982). 51. " N a t u r a l L a w , " 32 Harvard Law Review 311, 312 (1918).

selves and others that their decisions are dictated b y law rather than the result of choice. T h e false sense of constraint parallels the false sense of liberty noted by H u m e . If we are realistic about j u d g e s , w e shall refuse to take seriously the suggestion that S u p r e m e C o u r t Justices should think of themselves as the secular counterparts of the O l d Testament prophets, and act accordingly. 52 Yet to the extent that j u d g e s are n o t constrained by positive law, a role of moral leadership will on occasion be thrust o n t h e m , w h e t h e r to be played well (as by Justice Jackson in Barnette) or badly. Which is not to say that even in those cases j u d g e s are " f r e e " in an ultimate sense—a proposition that w o u l d be contrary to the entire thrust of this chapter. T h e y are constrained, all right, b u t b y their genes and u p b r i n g ing, their t e m p e r a m e n t , fears, and ambitions, rather than b y rules. T h e decision to obey a rule is compelled not by the rule itself but by the consequences of disobeying it. N o r is it the case that m o s t j u d g e s actually feel constrained all of the time. Judges feel a m i x t u r e of f r e e d o m and constraint in proportions that vary w i t h the nature and rank of the court, the nature of the particular decision to be made, the nature of the case, and the judge's t e m p e r a m e n t . In summary, j u d g e s decide cases w i t h greater confidence than the realities o f j u d i c i a l decision m a k i n g permit, and speak and write w i t h m o r e confidence than they feel. This situation is unlikely to change w i t h o u t p r o f o u n d , and not necessarily desirable, changes in the political system. I emphasize this qualification in order to m a k e clear that in urging a "naturalistic" view of the j u d g e I a m n o t engaged in criticism. I a m describing the conditions of adjudication that m a k e it unrealistic to believe that o u r j u d g e s can render objective decisions in difficult cases, if "objective" is to mean m o r e than m u d d l i n g t h r o u g h in the conduct of an all-too-human enterprise. As always it is essential n o t to go overboard and ignore the elements that keep the judicial process f r o m r u n n i n g off the rails into utter subjectivity, "pure politics." In emphasizing the gap between w h a t j u d g e s do and w h a t they say they do I have been skating perilously close to the fallacy, itself mentalist in character, of reducing k n o w l e d g e to conscious knowledge. T h e j u d g e w h o s e conscious t h o u g h t s are banal m i g h t still be a great j u d g e , j u s t as a person can be a great billiards player w i t h o u t 52. See Michael J. Perry, The Constitution, the Courts, and Human Rights: An Inquiry into the Legitimacy of Constitutional Policymaking by the Judiciary 97—110 (1982). I n o t e in passing that this suggestion m a y m i s c o n c e i v e the social role of t h e p r o p h e t s . T h e y w e r e social critics—see Michael Walzer, Interpretation and Social Criticism, ch. 3 (1987)—a s t r a n g e role in w h i c h to cast a society's m o s t p o w e r f u l judicial officials.

knowing N e w t o n ' s laws of motion. But the parallel is not exact. We k n o w the billiards player is great by observing him play, and win, in competition with other players; we do not have as objective a method of evaluating judicial performance. A related point is the difference between the internal and the external perspective of a j o b . From the internal perspective most jobs—private and public, glamorous and mundane, dangerous and safe, powerful and powerless—are pretty much the same, the j o b holder being concerned at the conscious level mainly with the mundane incidents (income, taxes, touchy co-workers and subordinates, difficult suppliers or customers, the physical conditions of work, and so on) that are c o m m o n to all jobs. 5 3 The letters and diaries of great writers and artists are illuminating in this regard. The fact that a person's conscious preoccupations are trivial does not make the person or the person's work product trivial. We must also be w a r y of confusing cause and justification. The motives driving a great scientist to succeed might be envy, greed, or thirst for glory. That would not undermine the validity of his scientific discoveries. T h e same is true, in principle, of law. But because we rightly mistrust our ability to verify the soundness of legal decisions, we are concerned that candidates for judicial office have the moral and temperamental qualities that conduce to good judging. Yet we are not sure what those qualities are. Moreover, they are difficult to screen for. And, once appointed, the j u d g e is subject to the psychological tugs—which operate to a greater or lesser extent on all persons thrust into a judicial role—noted earlier. Finally, the effect of professional n o r m s and criticisms must be considered. T h e belief that judges are constrained by law, that there is more to law than the will to power, is a deeply ingrained feature of the legal culture. And this makes the expectation that judges will behave in accordance with that belief to an extent self-validating. A j u d g e w h o flouts this expectation is likely both to feel uncomfortable and to attract professional criticism, which will make him more uncomfortable (although judges' ability to shrug off criticism as being motivated by politics, envy, ignorance, or sour grapes should not be underestimated). Feeble retribution, perhaps? But here the institution of judicial indepen53. N o t i c e that I a m i n v e r t i n g the usual a r g u m e n t — t h a t a p r e d i c t i o n t h e o r y of judicial b e h a v i o r (on w h i c h f u r t h e r see C h a p t e r 7), or o t h e r behaviorist explanation, is u n f a i t h f u l to the internal p e r s p e c t i v e of the j u d g e . F e w p e o p l e w h o w r i t e a b o u t that perspective have been j u d g e s .

dence is i m p o r t a n t and, paradoxically, m a y have the effect of constraining j u d g e s m o r e effectively than if they w e r e dependent on the good will of the legislative or executive branches. B y r e m o v i n g the most p o w e r f u l incentives to action, judicial independence brings to the fore incentives that m i g h t otherwise hold little sway. Certainly one w h o is skeptical about free will will tend to regard institutions such as judicial independence or academic tenure as altering rather than eliminating incentives. T h e designers of those institutions seem, consistent w i t h the analysis in this chapter, to have acted on an implicitly behaviorist conception of the judicial process. T h e desires to be comfortable w i t h oneself and to avoid academic criticism f r o m professors are, I suspect, p o w e r f u l motives only w h e n m o n e y is not a m a j o r factor. B u t that is the situation of j u d g e s . It is not unique. We notice that people seem to behave morally in situations in which the costs of behaving morally are small, but to respond to incentives in situations in which those costs are large. O n e w a y to encourage moral behavior, therefore, is to reduce its costs, w h i c h is d o n e in the judicial arena by trying to m a k e judicial decisions independent of the j u d g e s ' incentives. A n o t h e r paradox is that j u d g e s in a system of political a p p o i n t m e n t may be less political than j u d g e s in a civil service type of judiciary, such as w e find on the C o n t i n e n t . Judges w h o are appointed because of friendship with legislators or other politicians, because of political services that they have rendered, or because of their o w n political influence have m o r e p o w e r than ordinary civil servants w o u l d have to w a r d off political interference w i t h the judiciary. T h e combination in our federal courts of (1) political a p p o i n t m e n t to judgeships carrying (2) life tenure may maximize the independence of these courts and hence their ability to act as bulwarks of liberty, although not their "professional" quality, their craft skills. In this example w e can see the line between law and politics fraying badly. Judicial independence, with or w i t h o u t political criteria of appointment, has of course a negative aspect. T h e careful circumscription of judicial temptations and ambitions reduces the j u d g e ' s drive to get things right; tenure is an anodyne. T h e j o y s of p o w e r and f a m e are a factor in the balance, too; and as the legal culture becomes even m o r e diverse, the possibility g r o w s that virtually any judicial posture—if consistent, but perhaps even m o r e so if inconsistent!—will attract praise. T h e fracturing of the legal culture has r e m o v e d a check on judicial willfulness. Yet this is not necessarily an a r g u m e n t against such fracturing,

against diversity. Although the more uniform the legal culture—the more like-minded the judges—the easier it will be to fix the premises for decision and therefore to identify and criticize incorrect decisions, the agreement of the like-minded is not a powerful warrant for truth. The dilemma is fundamental. Legal objectivity, it seems, can be purchased only with a uniformity that can be deadly for intellectual inquiry. Yet a diverse judiciary makes painfully apparent the weakness of the law's methodology for resolving difficult issues.

6 Are There Right Answers to Legal Questions?

Questions of Law Ronald Dworkin has flung d o w n the gauntlet to legal skepticism by arguing that there are right answers to even the most difficult and controversial legal questions. In a well-known essay he asks whether the persistence of controversy over the correct outcome of a difficult case demonstrates that there are no right answers to the legal questions posed by such cases, and he answers no. 1 Controversy continues over whether Richard III ordered the little princes killed. Nevertheless, he either did or did not, 2 so there is a right answer to the question notwithstanding the controversy over it, and we would arrive at that answer if only we knew more of the facts; so it is with the law (p. 120). 3 1. Ronald D w o r k i n , "Is T h e r e Really N o R i g h t A n s w e r in H a r d C a s e s ? " in D w o r k i n , A Matter of Principle 119 (1985). S u b s e q u e n t page references to this essay are in the text. 2. Well, not really; there are i n t e r m e d i a t e states, such as e n c o u r a g e m e n t , ratification, and c o n d o n a t i o n , and there is also the possibility that the little princes died of natural causes, or in an accident, r a t h e r t h a n b e i n g killed p u r s u a n t to an o r d e r ( w h e t h e r or n o t such an order was given). H e r e is a better e x a m p l e o f a q u e s t i o n that has a r i g h t a n s w e r ( p r o v i d e d "blades of g r a s s " and " b o u n d a r y " have exact d e n o t a t i o n s ) that is unlikely ever to be discovered: "It is either t r u e or false that there w a s an o d d n u m b e r of blades of grass in H a r v a r d Yard at the d a w n of C o m m e n c e m e n t Day, 1903." " W h a t Price Bivalence?" in W. V. Q u i n e , Theories and Things 31, 32 (1981). I have already q u e s t i o n e d (see C h a p t e r 4) w h e t h e r bivalence is a p r o p e r t y of decisions in difficult cases. 3. T h e a r g u m e n t is offered in passing, and I m a y be m a k i n g t o o m u c h of it, especially since D w o r k i n later retracted the m e t a p h y s i c a l i m p l i c a t i o n s of the " R i g h t A n s w e r " paper, a l t h o u g h not the r i g h t - a n s w e r thesis itself. See R o n a l d D w o r k i n , Law's Empire v i i i - i x (1986). For additional criticisms of the thesis, see J a c o b P a u l j a n z e n , " S o m e F o r m a l A s p e c t s of Ronald D w o r k i n ' s R i g h t A n s w e r T h e s i s , " 11 Manitoba Law Journal 191 (1981); Michael S. M o o r e , " M e t a p h y s i c s , E p i s t e m o l o g y , and Legal T h e o r y , " 60 Southern California Law Review 453 (1987); K e n t G r e e n a w a l t , " D i s c r e t i o n and Judicial D e c i s i o n : T h e Elusive Q u e s t for the Fetters T h a t B i n d J u d g e s , " 75 Columbia Law Review 359 (1975).

But "the law" does not have the same ontological status that Richard III and the little princes have. Although what exactly it means to have knowledge of the past is a nice philosophical question, there is no reasonable doubt that Richard III and the little princes were real people and that the princes were killed by someone. That the question "by w h o m ? " is a question about the past no more makes it unanswerable than the fact that the theory of evolution is a theory about the past makes that theory unscientific. But if we ask, for example, whether the Fourteenth A m e n d m e n t forbids racial segregation of public schools, we are not asking a question about, or only about, historical events. We might be asking such a question if we thought the only basis for an affirmative answer would be that the draftsmen and ratifiers of the Fourteenth A m e n d m e n t had meant to order the courts to forbid public school segregation. T h e question whether they issued such an order would be like the question whether Richard III ordered the little princes killed. But it is not Dworkin's theory of constitutional interpretation that we should ask whether a challenged practice was in the conscious thoughts of the framers. He wants courts to consider the level of generality of the provision, the values that inform it, the course of events unforeseen by the framers, the judicial decisions that have interpreted it. Ascertaining "the law," for D w o r k i n as for most other legal thinkers, is not a simple matter of asking whether the framers issued an order long ago to modern judges to do this or that. The law is not the same kind of entity as the events of Richard Ill's reign. (Maybe it is no kind of entity, as we shall see.) Even to ask whether "it is true" that the Fourteenth Amendment means this or that has an odd ring. I am not arguing that controversy is proof per se of indeterminacy. Often—this is a central insight of pragmatism—it is stimulus and precursor to solution. But that observation yields little comfort once we distinguish between determinacy in principle and determinacy in practice. Even if there must be a right answer to the question about Richard III and the little princes, if the answer is inaccessible it is as if there were no right answer, or as if (as a pragmatist might believe) the question were meaningless. T h e situation is even worse in law. We can wait patiently—if need be forever—for the historians to discover the truth about Richard; but a legal case must be decided when it arises, and the decision may engender reliance that precludes reconsideration when the "right answer" is finally discovered. D w o r k i n moves quickly off the Richard III example and tries two other tacks. T h e first is to contend that a question of law is not indeter-

minate so long as no observer would describe the question as a "tie" (p. 144). Here metaphor impedes understanding. T h e arguments in difficult cases cannot be "weighed," so it is indeed a w k w a r d to describe them as being in equipoise. Yet a dispute in which it is impossible to tell which side has the stronger argument is indeterminate. 4 Second, and at much greater length, D w o r k i n argues that correctness in law is similar to correctness in literary interpretation (pp. 138—143). A literary interpretation succeeds to the extent that it accounts for the relevant data— the incidents, language, and other aspects of the w o r k . Likewise, suggests Dworkin, a legal interpretation succeeds to the extent that it accounts for the relevant ethical and political data that i n f o r m legal decision making. In both domains, then, the literary and the legal, there are right and w r o n g answers. But this argument overlooks the fact that, despite the scientific f o r m of the inquiry into "fit"—an interpretation is proposed as a hypothesis, and verified or falsified by its success in explaining the data furnished by the text—a literary interpretation is not verifiable by the methods of exact inquiry, because the data do not lend themselves to experimental, statistical, or otherwise exact observation. N o more does a legal interpretation, which is w h y in Chapter 3 I classified interpretation as a problem in practical reason. The literary example is particularly inapt because of the peculiar ontology of fictional characters. In investigating the question whether Richard III ordered the little princes killed we can bring to bear our normal understanding of h o w people respond to various stimuli as well as our knowledge of English political history. But no law of literary composition requires that fictional characters behave "realistically" or have coherent life stories. 5 If, to b o r r o w an example of Wittgenstein's, there is a picture of a teakettle with steam coming out of the spout, does this mean that the kettle in the picture contains boiling water?

4. See J o h n M a c k i e . " T h e T h i r d T h e o r y of L a w , " in M a c k i e , Persons and Values: Selected Papers, vol. 2, pp. 132, 137-138 (Joan M a c k i e and P e n e l o p e M a c k i e eds. 1985). 5. T h i s is the i r o n y and the p a t h o s of Keats's speculation, in " O d e o n a Grecian U r n , " a b o u t the t o w n f r o m w h i c h the p e o p l e p o r t r a y e d o n the u r n had c o m e to attend the cerem o n y depicted o n it: W h a t little t o w n b y river or sea shore, O r m o u n t a i n - b u i l t w i t h peaceful citadel, Is e m p t i e d of this folk, this p i o u s m o r n ? A n d , little t o w n , t h y streets f o r e v e r m o r e Will silent be; and n o t a soul to tell W h y t h o u art desolate, can e'er r e t u r n .

Consider the f a m o u s question, D o the Macbeths in Shakespeare's play have children? 6 T h e r e are a m b i g u o u s hints going b o t h ways, but no direct evidence. Macbeth's obsessive concern w i t h f o u n d i n g a dynasty is incomprehensible unless he and his wife have or are planning to have children, yet if they d o have children, it is such an i m p o r t a n t fact that it w o u l d be b o u n d to be discussed in the play. A n d they d o not seem y o u n g e n o u g h to be planning a family. Anyway, it is quite impossible to visualize the M a c b e t h s as parents, actual or prospective. So in one sense they m u s t have children and in another and equally p o w e r f u l sense they m u s t not. This w o u l d be unsatisfactory in history but causes no distress to the audience o (Macbeth, for w h o m the rules of logic, of psychological realism, of history, are happily suspended. This carefree o n t o l o g y will n o t do f o r law. If there is n o d e m o n s t r a b l y right answer, does this mean there is also n o d e m o n s t r a b l y w r o n g one? It does if there are only t w o possible answers (yes and no), but n o t otherwise. If there is neither a d e m o n strably right n o r a d e m o n s t r a b l y w r o n g answer, does this mean there is n o answer w o r t h giving? N o t necessarily. T h e r e is a difference between saying "kale tastes b a d " and saying "I d o n ' t think Richard III killed the little princes," or "I don't think the audience for Macbeth should think the Macbeths childless," or "I d o n ' t think the due process clause forbids racial segregation in the District of C o l u m b i a ' s public schools." It is possible, indeed obligatory, to give reasons for the last three statements a l t h o u g h n o t for the first. T h e difference m a y reflect a belief that reasoning about these issues s o m e h o w advances the ball. But this may not be saying a great deal. T h e belief m a y be w r o n g . O r "reasoning" here m a y carry the persuasive or rhetorical sense noted in the discussion in C h a p t e r 4 of competition a m o n g social visions (which is not to say that rhetorical " c o n v e r s a t i o n " b e t w e e n advocates of opposed social visions is worthless). So far as giving reasons for legal j u d g m e n t s is concerned, w e m a y simply feel qualms about d o i n g things that hurt other people unless w e give a reason—some reason, perhaps not very reasoned—for our action. Yet w e allow juries to deliver verdicts w i t h o u t any statement of reasons. This is s o m e evidence that the felt obligation to reason to a j u d g m e n t is m o r e f u n d a m e n t a l than the obligation to give reasons for the j u d g m e n t ; f o r it has been considered unethical since the trial of Justice Bridlegoose in Gargantua and Pantagruel for a j u d g e to flip a coin in order to determine the w i n n e r of a case that the j u d g e considers a toss-

6. See the c o m p a c t discussion in William E m p s o n , Essays on Shakespeare 142-143 (1986).

up. 7 Is this j u s t judicial public relations? O r is the fear that the j u d g e will give up thinking too soon? A f u r t h e r point, w h i c h m a y explain w h y judges (but not jurors) normally are required to state reasons for their rulings, is that even m a d e - u p reasons m i g h t contribute to the policy debate. After all, the reasons lawyers give in advocating a client's cause often are made u p rather than sincere, yet j u d g e s find t h e m helpful anyway. H o w e v e r these matters be sorted out, there is a gulf between saying that discussion of moral, historical, political—and difficult legal—questions is interesting or even socially productive and saying that there are in principle correct answers to all or m o s t such questions. T h e idea that there are correct answers to all or m o s t moral questions is called " m o r a l realism," 8 and in a heterogeneous society this is likely to seem even less plausible than a belief that there are right answers to even the m o s t difficult legal questions. D w o r k i n is a moral realist. 9 O r so at least it seems w h e n his belief that there are right answers to legal questions is j u x t a posed with his belief that j u d g e s should "identify legal rights and duties, so far as possible, on the assumption that they were all created b y a single a u t h o r — t h e c o m m u n i t y personified—expressing a coherent c o n ception of justice and fairness." 1 " For these are moral terms. I want to suggest three m o r e respects in w h i c h D w o r k i n ' s c o m p a r i s o n of statutory (and constitutional) interpretation to literary interpretation actually undermines the quest for right answers in law: First, it is easy to find literary examples of utterly indeterminate questions: not perhaps w h e t h e r the Macbeths had children but w h e t h e r they had fair or swarthy complexions, blue or b r o w n eyes, or w e r e b o r n in

7. M a y b e the l a w has been t o o hasty in rejecting " a l e a t o r y " decision m a k i n g in all cases. For an interesting discussion, see R o b e r t H . M n o o k i n , " C h i l d - C u s t o d y A d j u d i c a t i o n : J u d i cial Functions in the Face of I n d e t e r m i n a c y , " 39 Law and Contemporary Problems 226, 289— 291 ( S u m m e r 1975). 8. A position v i g o r o u s l y d e f e n d e d in Michael M o o r e , " M o r a l Reality," 1982 Wisconsin Law Review 1061. See also N o t e , "Relativistic J u r i s p r u d e n c e : S k e p t i c i s m F o u n d e d o n C o n f u s i o n , " 61 Southern California Law Review 1417 (1988). N o t i c e o n c e again the potential c o n f u s i o n caused b y the fact that p h i l o s o p h y and law use different t e r m i n o l o g i e s . T h e legal c o u n t e r p a r t to m o r a l realism is n o t legal realism b u t legal f o r m a l i s m . For this reason, and also because the o p p o s i t e of m o r a l realism is m o r a l relativism, it m i g h t be b e t t e r to call moral realism " m o r a l a b s o l u t i s m . " See Gilbert H a r m a n , "Is T h e r e a Single T r u e M o r a l i t y ? " in Relativism: Interpretation and Confrontation 363 (Michael K r a u s z ed. 1989). 9. See Larry Alexander, " S t r i k i n g Back at the E m p i r e : A Brief S u r v e y of P r o b l e m s in D w o r k i n ' s T h e o r y of L a w , " 6 Law and Philosophy 419, 425 n. 14 (1987). 10. Law's Empire, n o t e 3 a b o v e , at 225.

leap years. There is an infinite n u m b e r of questions like these." They are ignored because nowadays no serious person is interested in them. (I recognize that this explanation is circular.) But suppose a law were passed entitling anyone with such a question to compel the Modern Language Association to answer it.' 2 Then we would see "litigation" over indeterminate questions with a vengeance. Many legal questions are indeterminate in a similar sense. The authors of the relevant texts have not provided the reader with enough information to enable an answer. But it would be extraordinary if the judges said, "We won't decide your case because we don't have enough information to figure out what the right answer is." They may resolve the indeterminate case by placing the burden of persuasion on one party, but that just regresses the debate to w h o should be assigned the burden—that is, w h o should lose. Imaginative writers might seem to have greater license than legislators to omit essential clues or to decide what may count as a relevant question about the text. Shakespeare was writing for a theater audience, which would not have time to ponder every implication in the text; so he could leave gaps, as he did with the question of the Macbeths' children. Legislators do not have that luxury. Yet often, as the price of obtaining majority agreement, they deliberately leave a question unanswered. Even more often do they do this inadvertently. The result in cither case is a gap that our conventions of adjudication require the judges to fill rather than allow them to ignore, as their literary counterparts can do. Second, people with similar educational backgrounds, politics, religious beliefs, and so forth will tend to interpret a w o r k of literature convergently, while people with different backgrounds will tend to interpret it divergently; and it is m u c h the same with law. A legal profession whose members have the same social and educational background, politics, religion, professional experience, and the like will tend to agree on the premises for decision; and, to repeat a chord struck frequently in this book, the more agreement there is on premises, the more legal reasoning can follow the syllogistic model. Then law will appear—will, in

11. See the interesting discussion in S i m o n B l a c k b u r n , Spreading the Word: Groundings in the Philosophy of Language 2 0 3 - 2 1 0 (1984). 12. C f . Frank H . E a s t e r b r o o k , " T h e Influence o f j u d i c i a l R e v i e w o n C o n s t i t u t i o n a l T h e o r y , " in A Workable Government? The Constitution after 200 Years 170, 174 ( B u r k e M a r shall ed. 1987).

a sense, be—objective, impersonal. 1 3 In just the same way, objective interpretation presupposes an interpretive community that is h o m o g e neous. These conditions for objectivity do not exist in the American legal community today any more than they exist in its literary community. Reflecting larger fissures in society, the legal community is politically and culturally divided. Its heterogeneity, interacting with the predominance of forms of law that give judges great discretion—common law, constitutional law, and statutes enacted by legislatures in which party discipline is so weak that many statutes are the product of complex, hard-fought, often unprincipled compromises—has created a legal terrain in which the consensus that might fix the premises for decision and allow law to proceed on logical tracks is unattainable. Third, the closest analogy of literary to statutory interpretation is the occasional attempt to predict f r o m a writer's life and works h o w he would have responded to some contemporary issue. If George Orwell had lived longer (he died in 1950), would he have sided with the United States in Vietnam? Would he be a neoconservative today? A Green? A fan of Gorbachev?' 4 Such questions, whose fatuity should be selfevident, resemble questions such as whether the framers of the Constitution would be opposed to wiretapping, or "dial-a-porn," or malapportionment, or flag burning, or abortion, if they were living today. Experience with literature suggests that texts and biographical insights cannot be reliably projected into the remote or even, as the case of Orwell shows, the near future, and it therefore provides no support for defenders of law's objectivity.

Questions of Fact The ontology of legal fact-finding, at least, may seem unproblematic. Something happened, and it is the j o b of the court to find out what. Reconstructing the past can be difficult, of course, as the question about Richard III and the little princes shows, but can there be anything spe-

13. T h i s is the abiding lesson of studies of the E n g l i s h c o m m o n law, illustrated b y Brian Simpson, " T h e C o m m o n L a w and Legal T h e o r y , " in Legal Theory and Common Law 8 (William T w i n i n g ed. 1986), and of c o m p a r i s o n s of E n g l i s h and A m e r i c a n law. See, for example, P. S. Atiyah and R o b e r t S. S u m m e r s , Form and Substance in Anglo-American Law: A Comparative Study oj Legal Reasoning, Legal Theory, and Legal Institutions (1987). 14. See J o h n R o d d e n , The Politics of Literary Reputation: George" Orwell (1989).

The Making and Claiming of "St.

daily troublesome about that? I think there can be, though I do not accept the crude "fact skepticism" of some of the legal realists. 15 T h e past cannot be recovered, and this makes it difficult to verify or falsify hypotheses about it. 16 N o t in every case, to be sure; if the murder suspect tells the police that he buried the victim's body at a certain place, and the police go there and, lo and behold, there it is, this is confirmation that he was telling the truth. It is an example of successful prediction—about the past; but much scientific prediction, too, is really postdiction; many of the events "predicted" by astronomers occurred billions of years ago. There are nevertheless a large number of legal cases in which the question of what happened is indeterminate, and must be resolved by a decision on w h o shall bear the burden of producing evidence or of persuading the trier of fact; bluntly, w h o shall lose in cases of doubt. It might appear that the problem of factual indeterminacy would be greatly exacerbated by the importance that so many legal doctrines seem to place on state of mind as an element of liability; but we saw in the last chapter h o w the legal system elides this difficulty by adopting a behavioral approach to mental phenomena. Yet in a sense these are dodges rather than solutions. Even the burial case is less straightforward than it appears to be. We must ask what exactly is being predicted when the murder suspect tells the police where he buried the victim's body: only that the body will be found at the indicated place. T h e fulfillment of the prediction is consistent with the hypothesis that the suspect is indeed the murderer, but like scientific confirmation in general it is also consistent with other hypotheses, such as that the true murderer told the suspect where the body was buried. A distinct problem arises f r o m the heavy but largely unremarked reliance that the law places on being able to determine counterfactuals. Counterfactual conditionals 17 pervade our thinking about causes and 15. See William T w i n i n g , " S o m e Scepticism a b o u t S o m e Scepticisms," 11 Journal of Law and Society 137 (1984). In d i s t i n g u i s h i n g factual f r o m legal q u e s t i o n s 1 do n o t m e a n to s u g gest that there is a b u l k h e a d b e t w e e n t h e m . To a c o n v i n c e d materialist, legal questions are at b o t t o m q u e s t i o n s of fact. For an interesting discussion, see Peter Tillers, " T h e Value of E v i d e n c e in L a w , " 39 Northern Ireland Legal Quarterly 167 (1988). 16. A p o i n t stressed, in the legal setting, in Paul Bator, "Finality in C r i m i n a l Law and Federal H a b e a s C o r p u s f o r State P r i s o n e r s , " 76 Harvard Law Review 441 (1963). O n the philosophical issues i n v o l v e d in k n o w l e d g e of the past, see, for e x a m p l e , A r t h u r C . D a n t o , Narration and Knowledge (1985). A n d o n the reliability of historical testimony, see Stephen M . Stigler, " J o h n C r a i g and t h e P r o b a b i l i t y of H i s t o r y : F r o m the D e a t h of C h r i s t to the Birth of Laplace," 81 Journal of the American Statistical Association 879 (1986). 17. O n w h i c h see, f o r e x a m p l e , Richard O t t e , " I n d e t e r m i n i s m , C o u n t e r f a c t u a l s , and C a u s a t i o n , " 54 Philosophy of Science 45 (1987); Causation and Conditionals (Ernest Sosa ed. 1975); D a v i d Lewis, Philosophical Papers, vol. 2, pt. 6 (1986).

consequences. O f t e n they are unproblematic. When we say that X shot Y, we mean that if X had not pulled the trigger Y would not have been shot. But suppose instead the question is whether X discharged Y because Y is a w o m a n . That is, suppose we have to answer a " w h y " question rather than just a " w h a t " question—the " w h a t " question in this case, Was she discharged? being trivial. Maybe Y admits she did things that would have been grounds for discharge but claims she would not have been discharged were she a man. The task for the trier of fact is to determine what would have happened in the imaginary, the counterfactual world that is just like our world (if one can get over the fact that it does not exist) except that in it Y is a man rather than a w o m a n . Does this mean the trier of fact must ascribe to the counterfactual Y every characteristic of a man except for a man's primary and secondary sexual characteristics? If so, however, would she still be Y in anything but name? Indeed, would she still be a w o m a n ? For w o m e n may differ systematically f r o m men in other ways besides the obvious physical differences (see Chapter 13). Putting these questions aside, one can determine whether Y would have been fired if she were a man only by comparing her treatment with that of men employed by X . But all people are different. Even if men and w o m e n do not differ systematically with respect to attributes relevant to performing the j o b in question, none of the men will be identical to a Y shorn of her female identity, and the job-discrimination laws do not forbid arbitrary decisions—only decisions traceable to the sex (or race, or other protected characteristic) of the employee. If the w o r k were so routinized, so uniform, that f r o m the employer's standpoint the employees really were interchangeable, provided only that they satisfied readily observable m i n i m u m performance criteria, the determination of the elusive counterfactual question about Y would be answerable with some confidence. O r if there were a great n u m b e r of male and female workers, statistical evidence could be used to generate a more or less reliable probabilistic j u d g m e n t . But in cases where neither condition is satisfied, the outcome is likely to depend on the burden of proof, or on the court's biases and prejudices, rather than on "the truth." Thus far I have assumed that the court is seeking truth and only truth, but this is an unrealistic assumption, especially in the American legal system. 18 To pick a humble example, evidence of repairs made by the defendant to his premises after an accident in which the plaintiff was 18. See A t i y a h and S u m m e r s , n o t e 13 a b o v e , ch. 6; M i r j a n R. D a m a s k a , The Faces of Justice and State Authority: A Comparative Approach to the Legal Process 1 1 9 - 1 2 5 (1986).

injured is not admissible to show that the condition of the premises before the accident was hazardous. The evidence is relevant, but if it were admissible injurers would be discouraged f r o m making repairs after an accident. T h e fact finder's concern with other values besides factual accuracy parallels the judicial rule maker's concern, stressed in previous chapters, with other values besides a rule's truth value. I do not mean that the American system is uninterested in factual truth, but only that the goal of truth is in competition with other goals, such as economy, preserving certain confidences, fostering certain activities, protecting constitutional norms. A goal that receives less attention than it should is to provide catharsis for people with grievances, real or imagined. People w h o think their rights have been violated are indignant, and they want a f o r u m to vent their indignation. It has been the observation of many trial judges—and I have observed the same thing in the occasional trials that I have conducted—that even when a plaintiff loses his case he feels better than if he had never been able to air his grievance in a public f o r u m in which the grievance was, if not vindicated, at least taken seriously and the defendant, his oppressor, put to the bother and anxiety of defending himself. This cathartic or therapeutic function of litigation is related to the roots of law in revenge and to the corrective-justice theory of law, both discussed in Chapter 11. Succinctly, the procedural system seeks the optimal trade-off between accuracy and cost, the latter broadly conceived. A cost weighted heavily in the English and American legal traditions is that of a judicial bureaucracy, and as a result great effort has been made to minimize the size of that bureaucracy. In contrast to the Continental legal systems, primary responsibility for developing the facts of the case is hived off on private lawyers; truth seeking is made competitive. This is not necessarily a bad idea, apart f r o m the intensely practical point that Americans have little aptitude for organizing and staffing high-quality civil services. We recall that Popper's theory of scientific progress emphasizes competition; and there is a connection between competition in developing facts and the classical liberal tradition that exalts competitive markets generally. O n e strain in liberal thought is skepticism about claims to possess the ultimate truth about anything. To such skepticism a marketplace notion of truth in the courts—adjudication as a struggle between competing hypotheses presented and supported by private lawyers—is congenial. But the analogy between science and the adversary system has limited force. As David Luban emphasizes, scientists do not compete—not routinely at any rate—by moving to suppress probative evidence or by

trying to rattle fellow inquirers by cross-examination. 1 9 T h e adversary system does not much resemble the concept of unforced inquiry that is the pragmatists' ideal and the scientists' ethic. Furthermore, the c o m petitors in our privatized, competitive system of justice often have markedly and irremediably unequal resources. M o s t criminal defendants lack the resources to hire counsel equal in skill and experience to the public prosecutor, and public subvention of the cost of counsel for indigent criminal defendants has not been sufficiently generous to close the gap. 20 N o t all scientific research centers have the same resources, of course. But they compete for resources, whereas a criminal defendant cannot, just by virtue of having a good case, procure high-priced legal talent to defend him. Rather than equalize the resources of prosecutors and defendants, the legal system has erected formidable procedural obstacles to conviction. These have succeeded in reducing the probability of convicting innocent persons to an extremely low level, but the price is that many guilty persons are acquitted (especially those w h o can afford to hire top-quality lawyers), or are never charged, or are allowed to plead guilty to crimes much less serious than those they actually committed. Even in civil as distinct f r o m criminal law, the privatization of fact gathering and proof is problematic, because of a dearth of reliable information in the market for legal services. Lawyers' performance is difficult to evaluate, making competition among lawyers highly imperfect. But problems of information can also plague a system in which the responsibility for gathering the facts is, as in Continental legal systems, vested in investigative magistrates and other public officials; both their superiors and the public may have difficulty monitoring their performance. 2 1 The only practical way to solve such problems is to make only simple facts legally relevant, but this " r e f o r m " would require a wholesale redefinition of legal rights and wrongs.

19. See " T h e A d v e r s a r y S y s t e m E x c u s e , " in The Good Lawyer: Lawyers' Roles and Lawyers' Ethics 83, 9 3 - 9 4 ( D a v i d L u b a n ed. 1984). T h i s is n o t t o d e n y t h e existence of s o m e pretty r o u g h c o m p e t i t i o n a m o n g scientists, as e m p h a s i z e d in D a v i d L. H u l l , Science as a Process: An Evolutionary Account of the Social and Conceptual Development of Science (1988). 20. See, for e x a m p l e , Stephen J. Schulhofer, " C r i m i n a l Justice D i s c r e t i o n as a R e g u l a tory S y s t e m , " 17 Journal of Legal Studies 43, 5 3 - 5 6 (1988). 21. For a debate o v e r the relative m e r i t s of the A n g l o - A m e r i c a n and C o n t i n e n t a l m e t h ods of d e t e r m i n i n g facts, see R o n a l d J. Allen et al., " T h e G e r m a n A d v a n t a g e in Civil P r o cedure: A Plea for M o r e Details and F e w e r Generalities in C o m p a r a t i v e S c h o l a r s h i p , " 82 Northwestern University Law Review 705 (1988); J o h n H . L a n g b e i n , " T r a s h i n g The German Advantage," in id. at 763; R o n a l d J. Allen, "Idealization and C a r i c a t u r e in C o m p a r a t i v e Scholarship," in id. at 785.

T h e United States is alone a m o n g the civilized nations in continuing to use juries routinely in civil trials, and our rules of evidence are shaped by the perceived (more precisely, the conjectured) limitations of jurors' understanding. In part we preserve the civil j u r y merely because we distrust the judges, sometimes justifiably. In part, perhaps we preserve it in order to make the judicial role seem more objective than it is, by consigning difficult questions to laypersons and thereby reducing the number of cases in which judges have to decide indeterminate questions. And it is precisely the factually difficult questions that the j u r y decides. If the question is easy—if a rational j u r y could not fail to answer it a particular way—the j u d g e will take the case away f r o m the j u r y by granting a motion for a directed verdict or for j u d g m e n t notwithstanding the verdict. (Compare the pardon power, which provides an escape hatch f r o m legal j u d g m e n t s that is located outside the formal legal system.) O n this analysis, the survival of the civil j u r y is due not to the populist streak in American culture, as is generally believed, but to formalist anxieties. T h e combination of juries and poor judges may not be a happy one, epistemologically speaking. The high regard in which the civil j u r y is held in some quarters is suspect. Its most influential supporters, besides lawyers for tort plaintiffs, are trial judges, and their support is largely unrelated to the accuracy of fact-finding by juries. They like the jury system because, to begin with, the burden of decision is shared: the j u r y assumes the primary responsibility for the decision—and judges like to deflect responsibility f r o m themselves; and when the j u d g e is not the decider, he is not required to write out his findings of fact and conclusions of law, which is a bother. Also j u r y trials are more dramatic and exciting than bench trials, and provide the j u d g e with a larger audience as well as with a larger field for the exercise of his people skills and his dramaturgical and managerial aptitudes. And judges have a natural tendency to think that "their" juries—the juries they have instructed, in the trials they have managed—get things right, since if they did not it might reflect poorly on the judge's conduct of the trial. This tendency (a nice example of resolving cognitive dissonance) is reinforced by the fact that many trial judges do not pay close attention to the evidence in a jury trial, because they are not the triers of fact in such trials; the judge's natural tendency therefore is to agree with the jury, which (whatever the other failings of trial by jury) does play close attention. Confidence in the civil j u r y should be undermined by the fact that when people are free to design their o w n methods of determining facts, including the same type of historical facts that are at issue in civil trials,

they never "invent" the jury. For example, there are no juries in arbitration, the private counterpart to public adjudication. And the "judges"— that is, the arbitrators—are often specialists, which is a count against our heavy reliance on generalist judges and randomly chosen jurors. The rules that frame the jury's inquiry, moreover, are inexplicable except on the hypothesis that we are dealing with a fact-finding body that is as gullible as it is inexperienced. O n e may not be willing to go as far as Ambrose Bierce, w h o said that "most m o m e n t o u s actions, military, political, commercial and of every other kind, are daily undertaken on hearsay evidence. There is no religion in the world that has any other basis than hearsay evidence." 22 But one could hardly deny that the hearsay rule—like the best-evidence rule, the rules of direct examination and of cross-examination, the rules on expert witnesses, the rules governing the selection and deliberations of juries, the rule that excludes evidence of a criminal defendant's criminal record unless he takes the stand (in which event his record can be used to attack the credibility of his testimony), and the other characteristic rules of American legal procedure that are intended to protect the simplest jurors f r o m being confused—is a second-best solution to the problem of factual uncertainty. T h e solution is dictated by our reliance on inexpert triers of fact facing competitive producers of evidence. These are not the rules we would expect to find in an inquiry unconstrained by the extrinsic considerations that have shaped our legal procedure. They are remote f r o m scientific procedure. The fact that juries, unlike judges, do not give any justification for their decisions is a dead giveaway; a requirement that jurors explain their votes would be a source of profound embarrassment to the legal system. Another reason for questioning the capacity of our—perhaps of any— legal system to discover " t r u t h " is the tendency of the litigation process

22. The Devil's Dictionary, in The Collected Writings of Ambrose Bierce 270 (1946) (entry for " I n a d m i s s i b l e " ) . H e added: " N o t h i n g in a n y existing c o u r t w a s ever m o r e t h o r o u g h l y p r o v e d t h a n t h e charges of w i t c h c r a f t and sorcery for w h i c h so m a n y suffered d e a t h . " Id. at 271. T h i s is a r e m i n d e r , if any is needed, o f the fallibility o f l e g a l f a c t - f i n d i n g . O n f o r e n sic p r o o f of w i t c h c r a f t , see, for e x a m p l e , B a r b a r a J. Shapiro, Probability and Certainty in Seventeenth-Century England: A Study of the Relationships between Natural Science, Religion, History, Law, and Literature, ch. 5 (1983); S a n f o r d J. Fox, Science and Justice: The Massachusetts Witchcraft Trials (1968). T h e legal s y s t e m c o n t i n u e s to be c r e d u l o u s w i t h regard to certain types of evidence. See, for e x a m p l e , D . Michael Risinger, M a r k P. D e n b e a u x , and Michael J. Saks, " E x o r c i s m of I g n o r a n c e as a P r o x y f o r R a t i o n a l K n o w l e d g e : T h e Lessons of H a n d w r i t i n g Identification ' E x p e r t i s e , ' " 137 University of Pennsylvania Law Review 731 (1989).

to select for trial precisely those cases in which the facts are uncertain. 23 I discussed a parallel tendency with regard to legal issues earlier in this book, noting that it exaggerates the degree to which law is uncertain. But just as that tendency ensures a high degree of legal uncertainty in the appellate process, so the tendency to select for trial those cases in which the facts are in doubt ensures a high degree of factual uncertainty in the trial process. T h e cases that go to trial because the parties are unable to predict h o w the j u d g e or j u r y will resolve the key factual disagreements are also the cases that the j u d g e or j u r y will have the greatest difficulty deciding correctly. This may not matter too much. As long as the courts can decide the easy cases correctly, and litigants and potential litigants k n o w this, at least the system will be resolving, albeit indirectly, a large fraction of all relevant factual questions correctly. Nevertheless, the tendency of litigated cases to turn on difficult factual questions, w h e n combined with the difficulties that fact finders, especially in the American legal system, have in answering such questions, may make judicial opinions a mine of misinformation. If factual uncertainty is disproportionately characteristic of litigated cases (from which the reported cases are drawn), then, given the difficulty of dispelling such uncertainty by the methods of litigation, we can expect the factual recitals in published judicial opinions to be w r o n g m u c h of the time. In addition, the facts cited in an opinion are often merely those contained in the plaintiffs complaint, which, having been dismissed on some legal ground, will be taken for purposes of appeal to be truthful, even if fanciful. Further, the appellate process tends to bleach out the facts of the dispute. By the time the case reaches the appellate court many of the facts of the dispute may have become irrelevant; as a result, the appellate opinion may paint a highly truncated and even misleading picture of what actually happened. And especially in cases where there is no published dissent, judicial opinions exemplify "winners' history." The appellate court will usually state the facts as favorably to its conclusions as the record allows, and

23. T h e r e is an e x t e n s i v e literature o n t h e d e t e r m i n a n t s of the decision to go to trial r a t h e r t h a n settle, e m p h a s i z i n g uncertainty, w h e t h e r legal or factual, as a necessary a l t h o u g h n o t sufficient c o n d i t i o n . See, f o r e x a m p l e , William M . Landes, " A n E c o n o m i c Analysis of the C o u r t s , " 14 Journal of Law and Economics 61 (1971); Patricia M u n c h D a n z o n and Lee A. Lillard, " S e t t l e m e n t O u t of C o u r t : T h e D i s p o s i t i o n o f Medical Malpractice C l a i m s , " 12 Journal of Legal Studies 345 (1983); Richard A. Posner, Economic Analysis of Law §21.5 (3d ed. 1986); G e o r g e L. Priest and B e n j a m i n Klein, " T h e Selection of D i s p u t e s for Litigation," 13 Journal of Legal Studies 1 (1984); Steven Shavell, "Suit, Settlement, and Trial: A Theoretical Analysis u n d e r A l t e r n a t i v e M e t h o d s f o r the Allocation of Legal C o s t s , " 11 Journal of Legal Studies 55 (1982).

often more favorably. This unedifying practice reflects both the psychology of judging (having persuaded himself that a particular outcome is correct, the j u d g e or judge's law clerk will tend unconsciously to screen out facts that support a contrary outcome) and the formalist style. An opinion would not look as powerful, as confident, as certainly right if it acknowledged the strength of opposing views. T h e tendency I have described is abetted by the reluctance of academic commentators to expand their study of cases beyond judicial opinions. Rarely will the commentator get hold of the briefs and record to check the accuracy of the factual recitals in the opinion. All this would be of relatively little importance were it not that lawyers' and particularly judges' knowledge of the world, or at least of the slice of the world relevant to legal decision making, derives to a significant degree f r o m judicial opinions. O n e of the distinctive features of judges as policy makers—and it should be clear by n o w that judges in our system are, to a significant degree, policy-making officials—is that they obtain much of their knowledge of h o w the world works f r o m materials that are systematically unreliable sources of information. 2 4 Since certainty is unattainable in most trials, should the question of legal proof be explicitly characterized as one of probabilistic determination? If so, what theory of probability should be adopted? 25 Probability is a mysterious entity, like number, but we can skip most of the mysteries 26 and merely sketch t w o of the rival concepts. O n e is frequen24. For an illustration of the significance o f this p o i n t in t h e l a w of privacy, see m y b o o k Law and Literature: A Misunderstood Relation 4—5 (1988), discussing M e l v i n v. Reid, 112 Cal. A p p . 285, 297 Pac. 91 (1931), a f a m o u s p r i v a c y case w h o s e " f a c t s , " d r a w n entirely f r o m the plaintiff's implausible c o m p l a i n t , have s h a p e d the t h i n k i n g o f g e n e r a t i o n s of p r i v a c y scholars. I discuss a parallel e x a m p l e , Tuttle v. Buck, in C h a p t e r 8 of t h e present b o o k . 25. T h e issue is t h o r o u g h l y canvased in " B o s t o n U n i v e r s i t y School of L a w S y m p o s i u m on Probability and Inference in the L a w of E v i d e n c e , " 66 Boston University Law Review 377 (1986). ( M o s t of the papers in the s y m p o s i u m appear in Probability and Inference in the Law of Evidence: The Uses and Limits of Bayesianism [Peter Tillers and Eric D . G r e e n eds. 1988].) See also Richard E g g l e s t o n , Evidence, Proof and Probability (2d ed. 1983); J o s e p h L. G a s t w i r t h , Statistical Reasoning in Law and Public Policy (2 vols. 1988); The Evolving Role of Statistical Assessments as Evidence in the Courts (Stephen E. Fienberg ed. 1989). T h e p r o b l e m s of judicial evidence are discussed f r o m a philosophical s t a n d p o i n t in L. J o n a t h a n C o h e n , The Probable and the Provable, pt. 2 and ch. 18 (1977). See also F e r d i n a n d S c h o e m a n , " C o h e n on Inductive P r o b a b i l i t y and the L a w of E v i d e n c e , " 54 Philosophy of Science 76 (1987); L. J o n a t h a n C o h e n , " O n A n a l y z i n g the S t a n d a r d s of Forensic E v i d e n c e : R e p l y to S c h o e m a n , " 54 id. at 92 (1987). 26. For an u n c o m m o n l y lucid discussion, see L. J o n a t h a n C o h e n , An Introduction to the Philosophy of Induction and Probability (1989). O t h e r g o o d discussions include J. L. M a c k i e , Truth Probability and Paradox: Studies in Philosophical Logic, ch. 3 (1973); Terrence L. Fine, Theories of Probability: An Examination of Foundations, ch. 9 (1973); C l a r k G l y m o u r , Theory and Evidence, ch. 3 (1980).

cy. To say that the probability that a randomly selected lightbulb is defective is .001 means, to a "frequentist," that if a sequence of randomly selected lightbulbs, produced under identical conditions, is tested, the relative frequency of defective lightbulbs will, in accordance with the law of large numbers, tend toward .001 as the n u m b e r tested increases without limit. T h e frequentist approach is of limited utility to the law because its conditions (random selection, identical conditions, many tests) are difficult to satisfy in a trial setting; one exception is where guilt or liability turns on fingerprint evidence or some other f o r m of evidence the accuracy of which has been determined by statistical studies of large samples. T h e second approach, the Bayesian, views probability in subjective terms. If a bookmaker lays odds of 9 to 1 against a particular horse's winning, then as far as the bookmaker is concerned the probability of that horse's winning is . 1. (Of course, the decision to place a bet depends on more than the odds—it depends, for example, on the bettor's attitude toward risk. But I ignore this complication and thus forgo discussion of risk preference or risk aversion.) Subjective probability unquestionably plays an important role in litigation. For example, each litigant must decide whether to settle or to go to trial, and in making that decision he will calculate, consciously or unconsciously, the odds of winning if he goes to trial. But does Bayesianism provide a good model of the decisional process of a j u d g e or jury? 27 For example, does (should) a jury convict a criminal defendant provided its members would lay odds of at least 9 to 1 or perhaps 19 to 1 against the innocence of the defendant? This seems unlikely, not only because the jury, unlike the litigant or the gambler, is not wagering anything on its assessment of the chances of guessing right, but also because it is not instructed by judge or counsel about odds. M a n y judges do not try to give the j u r y even a verbal explanation of the standard of proof in criminal cases, "beyond a reasonable doubt"; that is left to the jury's imagination. 2 8 The j u r y is making a decision under uncertainty, of course, but it is not invited to

27. See Lea B r i l m a y e r and Lewis K o r n h a u s e r , " Q u a n t i t a t i v e M e t h o d s and Legal Decis i o n s , " 45 University of Chicago Law Review 116, 135-148 (1978), for a h e l p f u l discussion. 28. O n the o d d i t y of r e f u s i n g t o i n s t r u c t t h e j u r y o n the m e a n i n g o f w h a t m i g h t appear t o be t h e m a s t e r c o n c e p t f o r their deliberations, see the m a j o r i t y and c o n c u r r i n g opinions in U n i t e d States v. Hall, 854 F.2d 1036 (7th Cir. 1988); also E g g l e s t o n , n o t e 25 above, ch. 9. T h e entire issue o f e f f e c t i v e c o m m u n i c a t i o n w i t h j u r i e s is deeply p r o b l e m a t i c . J u r y instructions usually are c o u c h e d in legal j a r g o n ; it is u n c e r t a i n h o w well the average j u r o r u n d e r s t a n d s t h e m . See Walter W. Steele, J r . , and Elizabeth G. T h o r n b u r g , "Jury Instructions: A Persistent Failure to C o m m u n i c a t e , " 67 North Carolina Law Review 77 (1988).

consider the decision in that light; and perhaps that is for the best. In so saying, I a m speaking of h o w these questions should be presented to a jury; I am not opining o n the ultimate merits o f Bayesian versus frequentist (or other) theories of probability. M o s t criticisms of Bayesian probability theory m a y well be either ill i n f o r m e d o r incomplete. 2 9 N o r am I contending that only betting cultures understand decision under uncertainty, and then only in explicit betting situations. People m a k e decisions under uncertainty in their o w n lives every day. B u t j u r o r s are making decisions that affect other people's lives. Essentially the j u r y is told in criminal cases n o t to convict unless it is certain of the defendant's guilt, w i t h the proviso that it is n o t to insist on a degree of certainty unreasonable in the circumstances (that is, it is to put aside unreasonable doubts): the kind of certainty that attends propositions such as that 2 + 2 = 4 or that cats d o n o t g r o w on trees or that n o person b o r n before 1800 is still alive. O b v i o u s l y this approach, which used to be called " m o r a l certainty" and meant "the kind of certainty that is adequate for practical life" 3 0 —and is thus an important example of the use of practical reason in law—is at b o t t o m probabilistic. Nevertheless it may be better than telling j u r o r s about probabilities, betting odds, confidence intervals, and the like, at least until w e either have a populace w i t h s o m e education in statistical inference or alter o u r criteria for selecting j u r o r s . I am not, however, arguing that law's "expressive" f u n c t i o n w o u l d be undermined 3 1 if w e faced u p to the fact that the determination of facts in criminal cases is probabilistic rather than certain and b r o u g h t the best scientific tools to bear o n the question. M y only concern is that juries m i g h t be confused. I a m n o t sure law has an expressive f u n c t i o n — t h a t is, a function of creating or reinforcing a set of social n o r m s o r an ideology. T h a t law affects behavior n o t only directly, by creating rewards and sanctions, b u t indirectly, b y altering attitudes and t h r o u g h t h e m behavior is an article of faith for m o s t legal professionals, especially j u d g e s and professors. But consistent w i t h the lack of scientific curiosity that is so m a r k e d a characteristic of legal t h o u g h t , the legal profession

29. As argued in the c o m m e n t s b y S t e p h e n E. F i e n b e r g and D a v i d H . K a y e in t h e B o s ton U n i v e r s i t y s y m p o s i u m , n o t e 25 a b o v e , at 651, 657, 693, 701. 30. G l e n n Shafer, " M o r a l C e r t a i n t y , " 5 Encyclopedia of Statistical Sciences 623 (1985). 31. As argued recently in C h a r l e s N e s s o n , " T h e E v i d e n c e or the E v e n t s ? O n Judicial P r o o f and the Acceptability o f Verdicts," 98 Harvard Lau/ Review 1357 (1985), and a t t a c k e d in Michael J. Saks and R o b e r t F. K i d d , " H u m a n I n f o r m a t i o n P r o c e s s i n g and A d j u d i c a t i o n : Trial b y H e u r i s t i c s , " in Judgment and Decision Making: An Interdisciplinary Reader 213 (Hal R. Arkes and K e n n e t h R. H a m m o n d eds. 1986).

has for the most part neither participated in conducting nor even paid any attention to (even to the extent of criticizing) studies designed to confirm or refute the existence of such consequences. 32 T h e lack of evidence that these consequences exist—even the evidence that they do not exist 33 —has failed to shake the profession's faith. T h e stubbornness of that faith is surprising because so much legal discourse tacitly assumes the opposite—that law is a social follower rather than a social leader. A police search infringes the Fourth A m e n d ment if it violates the "reasonable expectations of privacy" of the person searched. This criterion would be circular if people were assumed to take f r o m the law itself their cues on h o w much privacy to expect. The criterion for applying the doctrine of promissory estoppel (whereby a promise may be legally enforceable even if not supported by consideration) is whether the promisee reasonably relied on the promise. It would be circular if people relied on precisely those promises they knew the law would enforce. In general, the commercial law of this country, 32. Sociologists and p s y c h o l o g i s t s h a v e c o n d u c t e d a n u m b e r of such studies. See, for e x a m p l e , G a r y B. M e l t o n and Michael J. Saks, " T h e L a w as an I n s t r u m e n t o f Socialization and Social S t r u c t u r e , " in The Law as a Behavioral Institution 235, 2 5 5 - 2 6 3 (Gary B. M e l t o n ed. 1986); Richard D . S c h w a r t z , " L a w and N o r m a t i v e O r d e r , " in Law and the Social Sciences 63 (Leon Lipson and S t a n t o n W h e e l e r eds. 1986); The Impact of Supreme Court Decisions: Empirical Studies (2d ed., T h e o d o r e L. B e c k e r and M a l c o l m M . Feeley eds., 1973); A d a m P o d g o r e c k i et al., Knowledge and Opinion about Law (1973); R o b e r t L. Kidder, Connecting Law and Society: An Introduction to Research and Theory, ch. 6 (1983). T h e expressive or e d u c a t i v e f u n c t i o n o f l a w is discussed f r o m a p h i l o s o p h i c a l s t a n d p o i n t in Joel Feinberg, Harmless Wrongdoing 2 9 4 - 3 0 0 (1988) (vol. 4 of his treatise The Moral Limits of the Criminal Law). 33. For e x a m p l e , Berl K u t c h i n s k y , " ' T h e Legal C o n s c i o u s n e s s ' : A Survey of Research o n K n o w l e d g e and O p i n i o n a b o u t L a w , " in P o d g o r e c k i et al., n o t e 32 above, at 101, 112— 120, p o i n t s o u t that m a n y studies have f o u n d that criminals have a p p r o x i m a t e l y the same k n o w l e d g e of and beliefs a b o u t t h e legal s y s t e m as l a w - a b i d i n g p e r s o n s do. T h i s is s o m e evidence that h a v i n g the " r i g h t " a t t i t u d e t o w a r d l a w does n o t s t r e n g t h e n social n o r m s . A s t u d y b y K i r k R. Williams, J a c k P. Gibbs, and M a y n a r d L. E r i c k s o n , " P u b l i c K n o w l e d g e of S t a t u t o r y Penalties: T h e E x t e n t and Basis of A c c u r a t e P e r c e p t i o n , " 23 Pacific Sociological Review 105 (1980), finds that a l t h o u g h p e o p l e h a v e a p r e t t y g o o d k n o w l e d g e of s t a t u t o r y penalties f o r c r i m e , their p e r c e p t i o n s reflect their preferences as to a p p r o p r i a t e sanctions, and t h o s e preferences in t u r n are reflected in the statutes. So it is p u b l i c o p i n i o n that determ i n e s t h e penalties, r a t h e r t h a n t h e penalties that d e t e r m i n e p u b l i c o p i n i o n . A s t u d y by Sally E n g l e M e r r y , " E v e r y d a y U n d e r s t a n d i n g s of the L a w in W o r k i n g - C l a s s A m e r i c a , " 13 American Ethnologist 253 (1986), finds that even w h e n p e o p l e o b t a i n first-hand k n o w l e d g e of the o p e r a t i o n s of o u r u n d i g n i f i e d local courts, t h e y retain their faith in the rule of law. T h e s e and o t h e r studies ( s o m e cited in C h a p t e r 4) u n d e r m i n e the v i e w that preserving m y t h s a b o u t t h e legal s y s t e m is necessary to secure the m a x i m u m feasible compliance w i t h law. A n o t h e r t y p e of attitudinal i m p a c t is illustrated b y the effect of the Gobitis decision ( C h a p t e r 4 again) in s t i r r i n g u p violence against J e h o v a h ' s Witnesses. But the impact was short-lived.

codified in the U n i f o r m Commercial Code, follows commercial custom, rather than custom following law. All this is not to say that law never alters attitudes, but it does underscore the uncertainty about the question. However this larger question be answered, there seems no reason to believe either that the legal profession's efforts to exaggerate the certainty of legal determinations (both legal and factual) fool the public or that such deception if successful would serve a worthwhile social purpose. It is strange that an important part of American government should seek to rule by mystery, in the manner of Dostoevsky's Grand Inquisitor, rather than by informed consent. 34 Strange or not, it hardly seems necessary in this instance. Even in a civil case, where the j u r y is told to render j u d g m e n t for the plaintiff if the plaintiff has proved his case by a "preponderance of the evidence," it seems that what the j u r y is being asked to do is not to lay a bet but to decide whether it believes the plaintiff's side of the case. It cannot be sure, of course, and it need not be as confident as in a criminal case, but it is not simply deciding w h o m it would bet on—plaintiff or defendant—if it were placing an even bet. That may be w h y the plaintiff w h o presents no or very little evidence will lose even though in the absence of any evidence the j u r y might think it slightly more likely than not that the plaintiff had a good case. 35 Stated differently, the j u r y wants to understand what happened—wants to be able to build a causal chain from the defendant's acts to the plaintiff s harm—rather than to estimate probabilities, because people are more comfortable with causal than with statistical explanations. 36 An alternative explanation, however, is consistent with the probabilistic approach: in the absence of special circumstances, the plaintiffs inability to come up with evidence suggests that he has no case—that the defendant's case is the more persuasive, the case to "bet on." If the plaintiff puts in some evidence, the defendant's failure to put in evidence of his o w n may, in turn, tell against him. O f course, in either case, the inference may be unwarranted; the missing evidence may simply be unavailable.

34. T h e legal p r o f e s s i o n has k n o w n for centuries t h a t legal f a c t - f i n d i n g is probabilistic. See Shapiro, n o t e 22 above, at 178—182, 187. B u t it is reluctant to share t h e k n o w l e d g e w i t h the lay public. 35. For an interesting discussion, see M a r y D a n t , " G a m b l i n g o n t h e T r u t h : T h e U s e of Purely Statistical E v i d e n c e as a Basis for Civil Liability," 22 Columbia Journal of Law and Social Problems 31 (1988). 36. C f . L. J o n a t h a n C o h e n , The Dialogue of Reason: An Analysis of Analytical Philosophy 157-187 (1986).

H o w pervasive is the problem of factual error in law? For starters, there is the well-known fact that many guilty people are acquitted, because of the heavy burden that the prosecution bears. Whether many innocent persons are convicted is likely to depend on the extent and care with which prosecutors screen their cases. 37 If prosecutors never prosecuted people w h o were not in fact guilty, the standard of proof could be as low as one wanted and still no innocent people would be convicted. Conversely, if prosecutors engaged in no screening but instead prosecuted anyone against w h o m a complaint was lodged, the percentage of innocent persons convicted might be significant, since the j u r y need only be reasonably certain of the defendant's guilt to convict. H o w much screening is done by prosecutors probably depends to a large extent on their resources relative to the a m o u n t of crime. The lower this ratio the more carefully prosecutors will screen, and the ratio is extremely low in this country today. The implication, which is counterintuitive, is that fewer innocent people will be convicted in a society with a high (and, especially, a rising) crime rate than in one with a low crime rate. Prosecutorial screening leaves untouched, however—indeed exacerbates—the problem of acquittal of the guilty. T h e more screening there is, the fewer guilty people are convicted; the screening gives them an additional chance to beat the rap. In general, unless the resources devoted to determining guilt and innocence are increased, the only way to reduce the probability of convicting the innocent is to reduce the probability of convicting the guilty as well. In attempting an overall evaluation of the accuracy of our legal system, one must bear in mind the distinction between formal and substantive accuracy. Formal accuracy is easily achieved by delegation—of the fact-finding function, for example—to a body viewed as somewhat apart f r o m the legal system itself; by assignment of burdens of proof; and by choice of the standard of appellate review. 38 Delegation is illustrated by the j u r y — j u r o r s are not legal professionals and judges will not assume responsibility for their mistakes—and by the pardon power. T h e function of burden of proof in achieving formal accuracy is to 37. C f . Glanville Williams, " L e t t i n g O f f t h e G u i l t y and P r o s e c u t i n g the I n n o c e n t , " 1985 Criminal Law Review 115. 38. T h e s e and o t h e r t e c h n i q u e s for r e d u c i n g the a p p e a r a n c e of judicial i n d e t e r m i n a c y are discussed in Rachel F. M o r a n , " R e v i e w Essay: Reflections o n the E n g i m a of I n d e t e r m i n a c y in C h i l d - A d v o c a c y C a s e s , " 74 California Law Review 603, 6 2 9 - 6 3 8 (1986), and Michael A. Rebell and A r t h u r R. Block, Educational Policy-Making and the Courts: An Empirical Study of Judicial Activism 5 0 - 5 4 (1982).

allow a court to reach a definitive result in a case w h e r e it m a y n o t have the faintest idea w h e t h e r the defendant w r o n g e d the plaintiff, and if so h o w seriously. O n e m i g h t think it a reproach to a legal system that it so often has n o clue to the merits of the legal disputes b r o u g h t before it for resolution, but the reproach is deflected by using b u r d e n of p r o o f as a placeholder for the missing k n o w l e d g e . W h e n a reviewing court uses a deferential standard of review (the rationality standard used to review j u r y verdicts, or the abuse of discretion standard used to review m o s t trial rulings and m a n y administrativeagency decisions), again the court is able to render definitive j u d g m e n t s even if it hasn't the foggiest n o t i o n w h e t h e r the rulings that it is reviewing are correct. T h e avoidance of difficult causal issues in criminal cases (not always to the criminal defendant's advantage) is still another of the legal system's techniques for avoiding the implications of its inability to resolve difficult issues of fact; a person can be convicted of a t t e m p t w i t h o u t a determination whether, if the a t t e m p t had n o t been interrupted, the crime that was attempted w o u l d actually h a v e been c o m m i t t e d . A n d so with aborted conspiracies, w h i c h are punishable even t h o u g h they cause no harm. T h e behaviorist character of o u r criminal law (see C h a p t e r 5) may reflect above all the legal system's limited ability to resolve difficult factual questions, as m a y the law's reluctance, despite appearances, to decide questions about intention. T h e celebration by lawyers and j u d g e s of the "fairness" of a system in w h i c h it is t h o u g h t better to acquit ten guilty defendants than to convict one innocent defendant is an a t t e m p t to put a good face on w h a t is actually a confession of systemic ineptitude in deciding questions of guilt and innocence. Finally, in determining whether a trial error was harmless and therefore n o t reversible, w e ask not w h e t h e r the j u r y w o u l d in fact have f o u n d the w a y it did but for the error, but w h e t h e r a rational j u r y w o u l d have so f o u n d . So again w e skirt a difficult counterfactual question, and incidentally reduce the discretionary p o w e r o f j u r i e s . These legal adaptations to factual uncertainty may be intelligent, or at least inescapable, but they m a y also fool the legal profession into thinking it k n o w s m o r e than it does. T h e impossibility of recovering the past w i t h complete confidence by the m e t h o d s of litigation should r e m i n d us of the pragmatists' insistence that scientific inquiry m u s t proceed w i t h o u t assurance that objective truth is attainable. T h e pragmatists' insight, in turn, m a y m a k e us p r o p erly cautious about (for example) adding layers of postconviction review to the criminal justice system in the h o p e of finding o u t for certain whether the rights of the criminal defendant have been violated. T h e

need is to balance what may be slight gains in reducing one type of error (violating the defendant's rights) against the costs in increasing another type of error (mistaken acceptance of the defendant's claim of right), in undermining the finality of the criminal process, and in burdening the courts to the detriment of defendants with valid claims that get lost in the shuffle. T h e problem of the law's weak sense of fact is not confined to facts about particular disputes, but extends to facts on which legal rules are based. For example, under the chief justiceship of Earl Warren the Supreme C o u r t interpreted the Bill of Rights' criminal-procedure provisions, the federal habeas corpus statute, and the due process clause of the Fourteenth A m e n d m e n t expansively; as a result the rights of criminal defendants were multiplied. T h e trend abated in the 1970s, and in recent years some rights of criminal defendants have been curtailed, mainly although not solely by generous interpretation of the concept of harmless error and by narrow interpretation of the Fourth Amendment's protection against unreasonable searches and seizures. (On other fronts the expansion of defendants' rights has continued—but I am painting with a broad brush.) Neither the decisions expanding the rights of criminal defendants nor the decisions cutting back were compelled by the text or history of the relevant provisions. 3 9 N o r had they (outside the area of capital punishment) a powerful moral impetus. They were primarily policy j u d g m e n t s , founded on or at least defended by reference to factual assessments. Liberal judges believe that expanding the rights of criminal defendants neither increases the crime rate significantly nor diverts substantial resources f r o m other social programs to the criminal justice system, and that curtailing those rights exposes the poorest people in our society, m a n y of them innocent, to discrimination and oppression. Conservative judges believe that the expanded rights of criminal defendants have been a factor in rising crime rates and impose heavy costs on the criminal justice system, and doubt that curtailing those rights somewhat will result in the conviction of the innocent or in other social harms. N o one k n o w s w h o is right in this debate. This is a shocking acknowledgment when one reflects that it has been going on for at least h a l f a century. T h e factual issues are difficult, of course, and judges cannot wait for the issues to be resolved before they make their decisions. What is remarkable is h o w little urgency legal professionals feel about resolving these issues. Lawyers and judges are usually conscientious 39. C o m p a r e this discussion to the t r e a t m e n t of r e a p p o r t i o n m e n t in C h a p t e r 4.

about the facts of particular disputes, but they are for the most part uninterested in investigating, or encouraging the academy to investigate, the social facts upon which the soundness of legal doctrine depends. They are content with their intuitions and priors. There is no more striking testament to the absence f r o m law of the scientific attitude.

7 What Is Law, and Why Ask?

Is It a Body of Rules or Principles, an Activity,

or Both?

We are n o w prepared to deal with the largest, although not necessarily the juiciest, ontological question in jurisprudence: the question of what law is. Stated so baldly, the question is virtually meaningless. "Law" is a word like "religion," "time," "politics," "democracy," and "beauty" that can be used without creating serious problems of comprehension but cannot be defined unless the purpose of the definition is understood. If you ask what "law" means in the statement that international law should be taught in law school, you will get a different answer than if you ask whether Judge X knows more "law" than Judge Y, or whether the Supreme Court under Chief Justice Warren was "lawless," or whether Roe v. Wade is "the law." But at least you will get answers. If you ask simply, What is law? you will touch off a jurisprudential debate. The specific questions I have put either address particular issues in pedagogy, judicial behavior, or legal practice, or invite specific comparisons, as in the question, Is primitive law "really" law?—which means, H o w strong is the family resemblance between primitive law and modern law? But the general question What is law? seems to assume—what is most debatable—that "law" is some kind of thing (or collection of things), like " N e w York" or " D o m Perignon" or "salamanders," or perhaps a set of propositions, as in the expression "the law of contracts." It will help to distinguish three senses of the word "law." 1 The first is law as a distinctive social institution; that is the sense invoked when we ask whether primitive law is really law. The second is law as a collection

1. See R o n a l d M . D w o r k i n , " I n t r o d u c t i o n , " in The Philosophy 1977).

of Law 1 ( D w o r k i n ed.

of sets of propositions—the sets w e refer to as antitrust law, the law of torts, the Statute of Frauds, and so on. T h e third is law as a source of rights, duties, and powers, as in the sentence " T h e law forbids the m u r dering heir to inherit." T h e third sense is the focus of this chapter. T h e fact that "the l a w " seems to c o m m a n d and e m p o w e r , to channel and forbid, makes it intuitive to suppose that it is indeed a thing of s o m e sort or perhaps a set of things, specifically rules—for rules c o m m a n d , forbid, e m p o w e r — o r , if n o t rules (or n o t j u s t rules), n o r m s and principles. I shall argue against this conception. H o l m e s was o n the right track w h e n he proposed that law is a prediction of w h a t j u d g e s will do w h e n faced with a specific case. 2 Widely considered self-contradictory, passe, and t h o r o u g h l y discredited, the "prediction t h e o r y " has been written off t o o soon. The Prediction Theory. In denying that law consists of objective, external, impersonal n o r m s , the theory bespeaks a skeptical view of l a w — not in the metaphysical sense of skepticism about the existence of an external w o r l d or other minds, but in the commonsensical sense, w h i c h is the skepticism of the scientist and the pragmatist. H o l m e s ' s j u r i s p r u d ence was pervasively skeptical in this sense. His hostility to the syllogism, that s y m b o l of legal formalism, is one token 3 and his fondness for external standards of liability another, this one related to his skepticism about peering into people's minds. H o l m e s ' s m e t a p h o r for f r e e d o m of speech—the marketplace of ideas 4 —rests on skepticism about the p o s sibility of settling intellectual disputes b y reason, w h i c h could then be embodied in law. This is the same conjectural v i e w of k n o w l e d g e that is found in Popper and the pragmatists and that underlies the test of time 2. See O l i v e r Wendell H o l m e s , " T h e Path of the L a w , " 10 Harvard Law Review 457, 461 (1897) ( " T h e prophecies of w h a t the c o u r t s will d o in fact, and n o t h i n g m o r e p r e t e n t i o u s , are w h a t I m e a n b y the l a w " ) . O n the p r o v e n a n c e of H o l m e s ' s p r e d i c t i o n theory, see M . H . Fisch, "Justice H o l m e s , the Prediction T h e o r y of Law, and P r a g m a t i s m , " 39 Journal of Philosophy 85 (1942). A n d f o r analysis p r o and con ( m o s t l y con), see H e n r y M . H a r t , J r . , " H o l m e s ' P o s i t i v i s m — A n A d d e n d u m , " 64 Harvard Law Review 929, 9 3 2 - 9 3 4 (1951); H . L. A. H a r t , " S c a n d i n a v i a n R e a l i s m , " 1959 Cambridge Law Journal 233; D a v i d H . M o s k o w i t z , " T h e Prediction T h e o r y of L a w , " 39 Temple Law Quarterly 413 (1966); R o b e r t S a m uel S u m m e r s , Instrumentalistu and American Legal Theory, ch. 5 (1982); G e o r g e C . C h r i s t i e , " T h e U n i v e r s a l A u d i e n c e and Predictive T h e o r i e s of L a w , " 5 Law and Philosophy 343 (1986); T h o m a s C . Grey, " H o l m e s and Legal P r a g m a t i s m , " 41 Stanford Law Review 787 (1989). 3. H e liked to say the syllogism c o u l d n ' t w a g its tail (that is, was i n o r g a n i c , dead). See, for example, letter to J o h n C . H . Wu, o f April 1, 1923, in The Mind and Faith of Justice Holmes: His Speeches, Essays, Letters, and Judicial Opinions 419 ( M a x L e r n e r ed. 1943). 4. See A b r a m s v. U n i t e d States, 250 U . S . 616, 630 (1919) (dissenting o p i n i o n ) ( " T h e best test of t r u t h is the p o w e r of the t h o u g h t to get itself accepted in the c o m p e t i t i o n of the market").

as a method of settling disputed questions. 5 T h e characteristically abrupt, incompletely reasoned nature of many of Holmes's most famous opinions 6 is relevant here, as is his view of the j u d g e as an interstitial legislator, since legislators are acknowledged makers of law. Other tokens of Holmes's skepticism about legal reasoning, truth, and goodness are his quintessentially positivistic statement that if the people of the United States wanted to go to hell it was his j o b as a j u d g e to help them get there as quickly as possible; 7 the hands-off stance that he adopt5. C o m p a r e t h e s t a t e m e n t in " N a t u r a l L a w , " 32 Harvard Law Review 40 (1918), that " o u r test o f t r u t h is a reference to either a present o r an i m a g i n e d f u t u r e m a j o r i t y in f a v o r of o u r v i e w , " w i t h Peirce's d e f i n i t i o n o f t r u t h (see C h a p t e r 3): " T h e o p i n i o n w h i c h is fated to be u l t i m a t e l y agreed to b y all w h o investigate, is w h a t w e m e a n b y the t r u t h . " (Notice, t h o u g h , the m o r e o p t i m i s t i c , even idealistic, f l a v o r of Peirce's f o r m u l a t i o n as c o m p a r e d w i t h H o l m e s ' s . ) Parts of t h e Abrams dissent (see n o t e 4 above) seem a l m o s t a paraphrase of Peirce's great essay " T h e Fixation o f Belief," in Collected Papers of Charles Sanders Peirce, vol. 5, p. 223 (Charles H a r t s h o r n e and Paul Weiss eds. 1934). C o r r e s p o n d i n g to H o l m e s ' s v i e w that the o n l y f o u n d a t i o n of belief was w h a t H o l m e s called his "can't helps" ( " W h e n I say that a t h i n g is true, I m e a n that I c a n n o t help believing it," "Ideals and D o u b t s , " in O l i v e r Wendell H o l m e s , Collected Legal Papers 303, 304 [1920]) is Peirce's s t a t e m e n t that " w h a t y o u c a n n o t in the least help believing is n o t , j u s t l y speaking, w r o n g b e l i e f . . . For y o u it is t h e a b s o l u t e t r u t h . " " W h a t P r a g m a t i s m Is," 15 Monist 161, 169 (1905). (For the revival of this v i e w b y a p r a g m a t i s t o f t h e present day, see Richard Rorty, Contingency, Irony, and Solidarity 46—47 [1989]; and recall m y implicit reliance o n it in C h a p t e r 2 to rebut A r t h u r Leff's e x t r e m e m o r a l s k e p t i c i s m . ) T h e possibility that H o l m e s derived his theory of e x t e r n a l s t a n d a r d s of liability f r o m Pcirce is e x p l o r e d in N o t e , " H o l m e s , Peirce, and Legal P r a g m a t i s m , " 84 Yale Law Journal 1123 (1975); see also the extensive discussion of H o l m e s ' s possible intellectual debts to Peirce in R o b e r t a K e v e l s o n , The Law as a System of Signs (1988). H a v i n g n o t e d the affinity b e t w e e n H o l m e s and P o p p e r , let m e c o m p l e t e the circle b y n o t i n g t h e affinity b e t w e e n Peirce and P o p p e r . H e r e is P o p p e r in a nutshell: " T h e scientific m a n is a b o v e all t h i n g s desirous o f learning the t r u t h and, in o r d e r to do so, a r d e n t l y desires to have his present provisional beliefs (and all his beliefs are merely p r o visional) s w e p t away, and will w o r k hard t o a c c o m p l i s h that o b j e c t . " " P r e f a c e to Scientific M e t a p h y s i c s , " in Collected Writings of Charles Sanders Peirce, above, vol. 6, pp. 1, 3. B o t h Peirce and P o p p e r believe that o n l y d o u b t p o w e r s p r o g r e s s — t h a t there is n o inclination to m o v e o f f a p o s i t i o n of n o n d o u b t . 6. See G. E d w a r d W h i t e , " T h e I n t e g r i t y of H o l m e s ' J u r i s p r u d e n c e , " 10 Hofstra Law Review 633, 6 5 0 - 6 5 2 , 664, 6 6 8 - 6 7 1 (1982); D a v i d P. C u r r i e , " T h e C o n s t i t u t i o n in the S u p r e m e C o u r t : Full Faith and the Bill o f R i g h t s , 1 8 8 9 - 1 9 1 0 , " 52 University of Chicago Law Review 867, 900 (1985); C u r r i e , " T h e C o n s t i t u t i o n in the S u p r e m e C o u r t : 1910-1921," 1985 Duke Law Journal 1111, 1145—1155, 1161; and m y b o o k Law and Literature: A Misunderstood Relation 2 8 1 - 2 8 9 (1988). 7. "I h o p e and believe that I a m n o t i n f l u e n c e d b y m y o p i n i o n that it is a foolish law [the S h e r m a n Act], I h a v e little d o u b t that the c o u n t r y likes it and I always say, as you k n o w , that if m y f e l l o w citizens w a n t to go t o Hell I will help t h e m . It's m y j o b . " Letter to H a r o l d Laski of M a r c h 4, 1920, in Holmes-Laski Letters: The Correspondence o f Mr. Justice Holmes and Harold J. Laski, 1916-1935, vol. 1, pp. 2 4 8 - 2 4 9 ( M a r k D e Wolfe H o w e ed. 1953). O f similar character is H o l m e s ' s p r o n o u n c e m e n t , "I hate j u s t i c e , " r e p o r t e d in Learned H a n d , " A Personal C o n f e s s i o n , " in H a n d , The Spirit of Liberty 302, 306 (3d ed. 1960).

ed t o w a r d state and federal legislation (other than that restricting freed o m of speech) that he personally t h o u g h t foolish or worse; the reduction of law to consequences, as in his claim that a contract is merely a promise to pay unless a stated contingency (either p e r f o r m a n c e or s o m e excusing condition) comes to pass, 8 and above all his conception of law as merely a prediction of j u d g e s ' rulings. Although unlikely to posit the existence of so abstract an entity as "the law," H o l m e s realized that since law affected behavior it was "real" in some sense. T h e solution to the d i l e m m a w a s to ask how law affects behavior. T h e state has coercive p o w e r , and people w a n t to k n o w h o w to keep out of the w a y of that p o w e r . So they g o to lawyers for advice. All they w a n t to k n o w is w h e t h e r the p o w e r of the state will c o m e d o w n o n t h e m if they engage in a particular course of action. To advise t h e m the lawyer m u s t predict h o w the j u d g e s , w h o decide w h e n the state's coercive p o w e r m a y be applied to a person, will act if his client engages in the proposed course of action and is sued. L a w is thus simply a prediction of h o w state p o w e r will be deployed in particular circumstances. Law, an abstract entity, is dissolved into physical force, also an abstract entity but o n e that has a m o r e solid ring and, m o r e i m p o r t a n t , can be interpreted in behavioral terms: if I d o X , the sheriff will eventually seize and sell property of m i n e w o r t h $Y. T h e prediction t h e o r y conceives the law as disposition rather than as object. To say of an act that it is unlawful is to predict certain consequences if the act is p e r formed, j u s t as to say that an object is heavy is to predict certain consequences if it is d r o p p e d or t h r o w n . This conception overlooks the people w h o obey the l a w because it is the law, and so it has c o m e to be called the "bad m a n " t h e o r y of law. But the oversight m a y n o t be critical. T h e r e m a y n o t be m a n y " g o o d m e n " in the specific sense of people w h o c o m p l y w i t h laws merely o u t of respect for law, a felt moral obligation to obey it. T h e l o w saliency of courts to Americans (see C h a p t e r 4) and the evidence cited in C h a p t e r 6 against law's having expressive effects are suggestive in this regard. This is not to say that sanctions are the only cause of compliance w i t h law. M a n y people obey the law because they w o u l d n o t profit f r o m breaking it even if there were n o f o r m a l sanctions—they are restrained by habit, conscience (insofar as law tracks morals, as it frequently does), concern w i t h reputation or other considerations of reciprocity, lack of net expected gain w h e n risks and o p p o r t u n i t y costs are taken into 8. See H o l m e s , n o t e 2 above, at 462; see also O l i v e r Wendell H o l m e s , J r . , The Law 300-301 (1881).

Common

account, or sympathy or affection for the potential victims of their wrongdoing. Analytically these people are part of the first group, Holmes's "bad men," w h o obey the law as far as it pays them to do so but no farther (with " p a y " understood broadly, however). Whether there is an independent moral obligation to obey law may thus be an issue of supreme practical unimportance. T h e most frequently encountered objection to the prediction theory is that t h o u g h it may w o r k for the lawyer w h o is advising a client about the legal consequences of a course of action, it will not w o r k for the j u d g e — h e cannot predict his o w n behavior. The objection is formidable only with respect to the judges of the highest court of the jurisdiction. T h e j u d g e of a lower court can predict h o w a higher court would decide his case in exactly the same way that the lawyers in the case can predict h o w that j u d g e (or the higher judges) will decide it. In cases in which federal jurisdiction is based on the diversity of citizenship of the parties, and in other cases in which the court is applying the substantive doctrines of another jurisdiction, the j u d g e has a feasible task in predicting h o w other judges at the same level would decide his case. If a case is novel, the j u d g e can decide it by predicting h o w a majority of other judges probably would decide it. Even a j u d g e of the highest court could try to decide a case predictively, by predicting h o w his predecessors, the authors of the precedents cited to him by counsel in the case, would have decided it. It is easy to understand the lawyer's incentive to conceive of law in predictive terms; that is h o w he will best serve his client's interests. But w h y should judges conceive of their function in similar terms, even if it is feasible for them to do so? N o t all judges do conceive their function thus. Some decide cases in accordance with their o w n view of the law and, by doing so, court reversal. Reversal is after all not so painful. Yet most judges are highly sensitive to being reversed, and for them the prediction theory makes good sense to follow. Weak sanctions can operate powerfully on judges. O n e is criticism, and reversal is a form of criticism (an exception will be noted in a moment). Most judges try to avoid being reversed, and this commits them to the prediction theory. But this analysis has no application to judges of the highest court of the jurisdiction. The theory has greater explanatory force than its critics allow, but it is unusable in areas of profound legal uncertainty. There judges and lawyers can do no better than to predict that the highest judges will decide the case (if it is appealed to them) in accordance with the law or the best

view of the law. We have n o t escaped the ontological question; for what is this entity, the law, that the j u d g e is g o i n g to steer by? We can save Holmes's account b y n o t i n g that if the law is ultimately a prediction of w h a t the highest j u d g e s will do, it is meaningless to ask h o w they can use prediction to discover the law. T h e law is not a thing they discover; it is the n a m e of their activity. 9 T h e y d o not act in accordance with s o m e t h i n g called " l a w " — t h e y j u s t act as best they can. T h e y decide the case, and as a b y - p r o d u c t t h r o w o u t clues to h o w they m i g h t decide the next case. T h e law is the set of hypotheses that lawyers and lower-court j u d g e s propose concerning the regularities in the higher courts' behavior. We can if w e like say that the j u d g e s , in acting—that is, in deciding cases—"make law," and so the law is w h a t j u d g e s d o as well as predictions of w h a t they will do. T h e r e is n o contradiction once the prediction theory is subsumed under a broader activity theory of law. T h e i m p o r tant thing is that law is s o m e t h i n g that licensed persons, mainly j u d g e s , lawyers, and legislators, do, rather than a b o x they pull off the shelf w h e n a legal question appears, in the h o p e of finding the answer in it— thus inviting debate over w h e t h e r the b o x includes only rules, or rules plus principles, or rules, principles, and policies, or perhaps the w h o l e of political morality. T h e c o m m o n law, a l t h o u g h judge-created, is law and so is equity jurisprudence, w i t h its o p e n - e n d e d m a x i m s (the core of what D w o r k i n calls "principles"); so are freewheeling interpretations of statutes and the Constitution; so, too, are exercises of a v o w e d l y discretionary powers by j u d g e s , such as the p o w e r to determine w h a t sentence to impose in a criminal case within the range fixed by the legislature, a discretion curtailed for federal j u d g e s , b u t n o t eliminated, b y the n e w federal sentencing guidelines. E v e n a decision to m a k e law m o r e political, by giving greater p o w e r to the j u r y or selecting j u d g e s in partisan elections or enlarging the p a r d o n p o w e r , is a decision within law—a decision about the kind of law the d o m i n a n t g r o u p s in society w a n t — rather than a decision to shrink law's d o m a i n . If the activity theory is right (I think it is mostly, but n o t entirely, right), the question w h e t h e r the law consists j u s t of rules or also includes the considerations to w h i c h j u d g e s turn w h e n the rules run out or w h e n a b r a n d - n e w rule is created, is a pseudo-question—a gambit, really, in a

9. I a m indebted to R o b i n West for this p o i n t . A d i f f e r e n t " a c t i v i t y " c o n c e p t i o n of law is p r o p o s e d in T h o m a s D . Eisele, " T h e Activity of B e i n g a L a w y e r : T h e I m a g i n a t i v e P u r suit of Implications and Possibilities," 54 Tennessee Law Review 345 (1987).

political and ideological game. Because the w o r d "lawless" is a pejorative, and aggressive j u d g e s w a n t to conceal the exercise of judicial discretion in order to give their decisions a m o r e "objective," less political, and therefore m o r e authoritative ring, c o m m e n t a t o r s w h o approve of an aggressive judiciary define " l a w " as broadly as possible, while those desiring greater judicial restraint define it m o r e narrowly. It is j u s t a semantic g a m e (the g a m e of "persuasive definition"), and it m a y not be an i m p o r t a n t game. Indeed, it m a y j u s t be s h a d o w boxing, because the public to w h o m the contestants are appealing m a y be indifferent to anything b u t the o u t c o m e s of decisions. E v e n the ideological stakes are small. H o w e v e r interesting and i m p o r tant, the question w h e t h e r judicial decision m a k i n g should be m o r e or less freestyle is a question neither about the nature of the law n o r about the f u t u r e of liberalism. T h e law has n o nature, n o essence; and liberalism is b o t h vague e n o u g h and resilient e n o u g h to a c c o m m o d a t e active and passive styles in constitutional interpretation and greater and lesser ratios of rule to discretion. In denying the law's "thingness," I a m n o t saying that it is nothing. It is a practice, easily recognizable in m o s t of its manifestations. But a practice or activity is different f r o m a set of concepts. T h e distinction can help us see that the question w h e t h e r the law consists j u s t of rules laid d o w n by public authorities (traditional legal positivism) or also includes s o m e "higher l a w " that can be used to change the rules, adjust t h e m to n e w circumstances, o r — w h a t a m o u n t s to the same thing—fill gaps b e t w e e n t h e m is misspecified. T h e question assumes that w e are exploring the boundaries of a set of concepts, whereas w e should be asking about a practice. Specifically, w e should be asking whether j u d g e s can decide cases b y m o d i f y i n g old rules or creating n e w ones (and applying their inventions to the case at hand, that is, retroactively) w i t h o u t being assailed b y indignant charges of lawlessness. T h e y can, in o u r system anyway, within broad limits that cannot be determined in advance. Evidently one can still be d o i n g law even if one is not just applying rules. A n d this need n o t mean appealing to a metaphysical or moral entity—"natural l a w " in an ontological sense. I offer several m o r e observations about the prediction theory. First, it cannot be a complete theory, because it implies a slavish subordination of the j u d g m e n t of l o w e r - c o u r t j u d g e s to the will of higher-court judges. T h e first task that the t h e o r y assigns to the lower-court j u d g e s is to predict h o w their superiors w o u l d decide the case, and only if they cannot predict are they allowed to exercise an independent j u d g m e n t . But as noted in C h a p t e r 2, w e w a n t lower-court j u d g e s to exercise some

independent j u d g m e n t . This may be w h y it would be considered inappropriate for a j u d g e of a lower court to predict the decision of his judicial superiors f r o m knowledge of an impending change in the m e m b e r ship of the higher court. T h e lower-court judges are supposed to be subordinates, but not courtiers. Second—and in tension with the first point—the prediction theory implies a more creative scope for the lower-court j u d g e confronted by a seemingly dispositive precedent of a higher court than the theory that law is a set of concepts. T h e latter theory might be taken to imply that the precedent is the law—it is a part of the contents of the box labeled "law." The prediction theory, when viewed normatively, implies that the function of the lower-court j u d g e is to predict h o w the higher court would decide his case. Precedents are essential inputs into the predictive process but they are not "the law" itself, so the lower-court j u d g e w h o has a strong reason to believe that the higher court would not follow its o w n precedent if the case arose today is not being lawless in declining to follow that precedent. Third, and closely related, the prediction theory avoids a paradox created by the notion of law as a set of concepts. Suppose that t w o lower courts, A and B, have before them cases identical both to each other and to a case decided by the higher court m a n y years ago but never overruled or even substantially undermined by later decisions. Nevertheless it is evident to A and B that if the same case arose today the higher court would abandon the precedent because experience has shown that it was a mistake. A, which adheres to the prediction theory, disregards the precedent—and is affirmed. 1 0 B, which adheres to the theory of law as a set of concepts, follows the precedent—and is reversed. O n the prediction theory A is right and B is wrong, and therefore the affirmance of A and reversal of B are just what one would expect in a system in which error is corrected by higher courts. But on the theory of law as a set of concepts A is w r o n g and B is right, producing the paradox that the court that decided the case in accordance with law was reversed as though it had committed an error, while the court that flouted the law was affirmed. T h e fact that a court is permitted to overrule its o w n precedents is thus an argument in favor of the prediction theory, which expresses a forward-looking view of law.

10. T h i s is w h a t the l o w e r c o u r t did in the s e c o n d flag-salute case, discussed in C h a p t e r 4. It disregarded Gobitis, correctly foreseeing that the S u p r e m e C o u r t w o u l d o v e r r u l e that decision, and invalidated the state's c o m p u l s o r y flag-salute law. See Barnette v. West Virginia State Board of Education, 47 F. S u p p . 251 ( S . D . W. Va. 1942).

But, f o u r t h , the idea that every accurate prediction of an overruling identifies a change in " l a w " is difficult to swallow. Suppose that because of the retirement of one j u d g e and his replacement by another it is predictable that a particular precedent will be overruled the next time a case presenting the issue comes before the court. Would it be correct to say that the change in court m e m b e r s h i p had changed the law? T h e prediction t h e o r y implies an affirmative answer, but such an answer is at variance w i t h n o r m a l usage. We think of law not only as w h a t j u d g e s do but also as the criterion f o r evaluating w h a t they do. T h e activity theory is incomplete. Its critics are right that it is an impoverished theory of law. B u t it is only mildly impoverished. W h e n slightly enriched with a w e a k sense of natural law, it becomes the best positive theory of law that w e have. Fifth, the prediction t h e o r y really can (despite m y second point) be n o r m a t i v e for lower-court j u d g e s , especially if the judicial system values stability highly. T h e effect of scrupulous adherence to the theory by such j u d g e s is to concentrate judicial discretion in the j u d g e s of the highest court. T h e r e are fewer of t h e m , they are m o r e carefully selected than the lower-court j u d g e s , they have a broader perspective, and their decisions are subjected to greater scrutiny; for all these reasons they can be expected to exercise discretion m o r e responsibly. Nuremberg and the Limits of Positivism. T h e foregoing analysis bears on the m u c h - d e b a t e d question, w h i c h has long seemed central to the controversy over w h a t law is, w h e t h e r the N u r e m b e r g Trial of the M a j o r War Criminals was legitimate." T h e laws applied b y the Allied tribunal w e r e invented f o r the occasion and thus had n o t existed (other than as ethical principles) w h e n the defendants had violated them; and the defendants had been acting in c o n f o r m i t y w i t h Nazi laws or decrees that w e r e valid according to the rules of recognition of the Third Reich. It is n o t e n o u g h to say that because the N u r e m b e r g Tribunal was a court, its j u d g m e n t s w e r e law. W h e t h e r and in w h a t sense it was a court are difficult, perhaps unanswerable, questions. We cannot, as w e can in dealing w i t h national courts (even N a z i courts), treat its j u d g m e n t s as the activity of licensed individuals and therefore lawful; for w e are not sure w h o had the authority to issue such supranational licenses. If the theory of law as a set of concepts is rejected, as I have argued it should be, and the activity approach is inapplicable by its o w n terms, the question

11. For an excellent recent discussion, see D a v i d L u b a n , " T h e Legacies of N u r e m b e r g , " 54 Social Research 779 (1987). A n o t a b l e earlier t r e a t m e n t is J u d i t h N . Shklar, Legalism: Law, Morals, and Political Trials 1 5 4 - 1 7 8 (1964).

whether the N u r e m b e r g j u d g m e n t s w e r e l a w f u l is meaningless. Rather than beat our heads against the wall w e should consider the pragmatic question w h e t h e r punishing the Nazi leaders using the forms of law was a sensible w a y to proceed. I think it was. It was unthinkable to let those monsters go free, so the question can be recast as w h e t h e r they should have been killed s u m m a r i l y or after a trial. T h e value of the trial, deficient as it was in the elements of due process (adequate w a r n i n g of c r i m inal liability, an unbiased tribunal) was that it enabled a public record to be compiled and gave the defendants a chance to say w h a t they could in their behalf, which for m o s t of t h e m was very little. As a result their moral guilt was established m o r e convincingly in the eyes of the w o r l d than if they had been liquidated hugger-mugger. 1 2 Reading f r o m a distance of thirty years the debate b e t w e e n H . L. A. Hart and Lon Fuller over the legality of Nazi laws, 1 3 1 a m struck b y h o w little was at stake. H a r t is fearful that if law and morals are run together, the moral c o n d e m n a t i o n of the N a z i laws will be diluted. H e does n o t want the moral issue obscured by switching attention f r o m it to the legalistic-sounding question of w h e t h e r the laws w e r e really laws; he does not want people to think they have a moral d u t y to obey any law whose legal validity is unassailable. So he proposes to m a k e the w o r d " l a w " s y n o n y m o u s w i t h positive law, w h i c h the N a z i laws w e r e but the legal principles enforced b y the N u r e m b e r g Tribunal w e r e not. Fuller is concerned that if law is defined so b r o a d l y — b r o a d l y in the sense of including all positive laws, h o w e v e r evil, a l t h o u g h n a r r o w l y in the sense of excluding natural law viewed either as an independent source of legal obligation or as a filter to screen o u t evil positive laws—the legitimacy of the Supreme Court's broad interpretations of the C o n s t i t u t i o n m a y be called into question. H e also argues that if the N a z i laws w e r e law, then in evaluating the wickedness of the Nazis w e shall be forced to give them at least a few points f o r fidelity to l a w — b u t how, he says, can a thoroughly i m m o r a l law give rise to even the m o s t qualified moral claim to obedience? Yet, oddly, Fuller's emphasis is n o t o n the i m m o rality of the Nazi laws but on their procedural irregularity, in particular their secrecy. T h e emphasis is curious if only because m o s t N a z i laws, including the N u r e m b e r g racial laws, w e r e public and regular. It s h o w s

12. Shklar, n o t e 11 above, stresses additional, n o less p r a g m a t i c considerations: the e d u cative effect of the N u r e m b e r g Trial o n the G e r m a n public, and t h e social a d v a n t a g e s of regularized retribution o v e r s p o n t a n e o u s v e n g e a n c e . 13. See H . L. A. H a r t , " P o s i t i v i s m and the Separation of L a w and M o r a l s , " 71 Harvard Law Review 593, 615-621 (1958); Lon L. Fuller, " P o s i t i v i s m and Fidelity to L a w — A R e p l y to Professor H a r t , " 71 Harvard Law Review 630, 6 3 2 - 6 3 3 , 6 4 8 - 6 5 7 (1958).

the lawyer's faith in process 14 (although this is rather a side issue in the debate) and in persuasive definition; Fuller w a n t s to " c o n d e m n [unjust decisions] to grammatical death t h r o u g h definitional execution." 1 5 In Fuller's emphasis on the U . S . Constitution we have a clue that the debate between h i m and H a r t — b e t w e e n an American and an Englishm a n — w a s as m u c h cultural as intellectual. Since the U n i t e d States has a judicially enforceable constitution that contains both open-ended p r o visions such as those guaranteeing equal protection and due process and a n u m b e r of specific prohibitions and guarantees as well, while England has neither a written constitution n o r a practice of judicial review of the validity of legislation, p r o f o u n d l y i m m o r a l laws are apt to be illegal in the U n i t e d States and merely p r o f o u n d l y i m m o r a l in the United Kingd o m . T h e overlap between law and morality is different in the t w o countries. In any event, I disagree w i t h b o t h H a r t and Fuller. T h e w o r d " l a w " is frequently, and harmlessly, used to denote b o t h positive law, however wicked, and natural law 16 (which is to say, approximately, basic political morality). M u r d e r is deliberate unlawful killing, yet w e say without 14. Recall the F r a n k f u r t e r q u o t a t i o n in C h a p t e r 2. T h e peculiar t w i s t that Fuller gives t o this faith is to s u p p o s e that s c r u p u l o u s a d h e r e n c e to the f o r m s of law not only minimizes e r r o r in the application of law but also c o n d u c e s to good law. Yet Shklar, n o t e 11 above, at 17, is correct that " p r o c e d u r a l l y ' c o r r e c t ' repression is perfectly c o m p a t i b l e w i t h legalism." See also id. at 209. T h e e x p e r i e n c e w i t h the e n f o r c e m e n t o f the f u g i t i v e slave acts and o t h e r slave laws in the U n i t e d States b e f o r e t h e Civil War s h o w s this. As to w h e t h e r legal positivism can k e e p j u d g e s in line, Shklar n o t e s that G e r m a n y ' s "official classes, especially its judiciary, w e r e n o t i n d u c e d b y p o s i t i v i s m to accept and a p p l y the law o f the Weimar R e p u b lic. In fact, t h e y s a b o t a g e d it w i t h n o legalistic c o m p u n c t i o n s . Positivism was o n l y i n v o k e d t o rationalize their far greater readiness t o c o m p l y w i t h Hitler's N e w O r d e r . " Id. at 72. H o w e v e r , I shall a r g u e in later c h a p t e r s that o n e aspect o f traditional legalism does seem likely to p r o m o t e liberty and p r o t e c t m i n o r i t i e s : t h e principle of g o v e r n a n c e b y general rules, as distinct f r o m p e r s o n a l , d i s c r e t i o n a r y j u s t i c e . 15. Shklar, n o t e 11 a b o v e , at 107. 16. T h i s needs qualification; natural l a w c a n n o t be s u b s t i t u t e d f o r positive law in all discussions of law. For e x a m p l e , b e t w e e n t w o j u d g e s , o n e of w h o m k n e w slightly m o r e positive law b u t m u c h less a b o u t the sources, t r a d i t i o n , and c o n t e n t of natural law than the other, it w o u l d be e x c e e d i n g l y peculiar to describe the second as k n o w i n g m o r e law than the first, o r to d e n y that the first k n e w m o r e l a w t h a n the second. We m i g h t say that the second j u d g e w a s a b e t t e r j u d g e even t h o u g h he d i d n ' t l^now as m u c h law as the first, but that w o u l d be quite different. Q u e s t i o n s a b o u t legal k n o w l e d g e are questions a b o u t positive law, p e r h a p s because the l a w y e r ' s and j u d g e ' s distinctive k n o w l e d g e is k n o w l e d g e of positive l a w r a t h e r t h a n k n o w l e d g e of political o r o t h e r morality. M o r e o v e r , w e w o u l d be m o r e likely to describe a j u d g e as sensitive to the m o r a l d i m e n s i o n s of his decisions than as knowing a lot a b o u t morality. M o r a l i t y is o n m o s t a c c o u n t s an affair of practical reason in its a c t i o n - o r i e n t e d r a t h e r t h a n c o g n i t i v e sense (see C h a p t e r 2), a l t h o u g h Plato f a m o u s l y disagreed, and I shall n o t e an e x c e p t i o n t o the m o d e r n v i e w in C h a p t e r 12.

semantic strain that the Nazis " m u r d e r e d " the J e w s even t h o u g h the killings w e r e authorized b y the positive l a w of the N a z i state. 17 T h e r e is no contradiction in saying that the Nazis w e r e o b e y i n g positive law yet violating natural l a w — m a n y natural lawyers acknowledge, contrary to Fuller, that even hideously unjust laws are " l a w " in a m e a n i n g f u l sense 18 —or that the C o n s t i t u t i o n has been interpreted as authorizing the courts to use natural law principles to invalidate positive law. O n e does not feel obliged to give the Nazis points for their contribution to upholding the rule of law. If a constitutional a m e n d m e n t w e r e passed re-enslaving black people or ordering t h e m deported to Africa, a federal j u d g e asked to enforce the a m e n d m e n t could gain n o m o r a l absolution by pointing out that he was obeying valid positive law. 19 Inflexible adherence to positive law is less likely to ennoble the adherents than to u n d e r m i n e the rule of law b y m a k i n g the rule-of-law virtues—regularity, predictability, impersonality, self-abnegation—seem grotesque, p a r odic, or i n h u m a n . T h e only dangers in using legal t e r m i n o l o g y to analyze the moral issues raised by N a z i decrees and adjudications lie in supposing that either N a z i laws or the principles administered b y the N u r e m b e r g Tribunal are typical of w h a t w e m e a n w h e n w e speak of "law," or in imagining that the question w h e t h e r either b o d y of " l a w " was really law can be answered. T h e latitudinarian usage of the w o r d s " l a w " and " l a w f u l " — t h e usage that embraces at one end wicked positive laws and at the other end n a t ural law—is not only harmless b u t useful. A b a n d o n i n g it m i g h t be thought to imply, wrongly, that if a j u d g e t h o u g h t the legal a r g u m e n t s on b o t h sides of a case evenly balanced h e should dismiss the case, o n the g r o u n d that it could n o t be decided b y law. O r that at s o m e point in the a r g u m e n t of the really difficult case, counsel should say to the judge, "Judge, w e ' v e finished o u r legal arguments; n o w let's talk a b o u t the nonlegal considerations that o u g h t to influence y o u r decision," or "Judge, w e ' v e d o n e a lot of research, and there's n o law to apply to this

17. T h e u l t i m a t e o r d e r t o d e s t r o y the J e w s w a s a p p a r e n t l y an oral c o m m a n d b y Hitler, but his oral c o m m a n d s w e r e positive law u n d e r the T h i r d Reich's " r u l e of r e c o g n i t i o n . " Cf. Luban, n o t e 11 above, at 7 9 5 - 7 9 6 . 18. See, for e x a m p l e , J o h n Finnis, Natural Law and Natural Rights 3 6 3 - 3 6 6 , 367 n. xii.4 (1980). 19. T h e idea of the j u d g e as an i n d e p e n d e n t m o r a l agent, a c c o u n t a b l e for his decisions and not p e r m i t t e d to hide behind a legislature's—or even b e h i n d " t h e p e o p l e ' s " — s k i r t s , is emphasized in R o b e r t M . C o v e r , "Violence and the W o r d , " 95 Yale Law Journal 1601 (1986), and in R o b i n L. West, " A d j u d i c a t i o n Is N o t I n t e r p r e t a t i o n : S o m e R e s e r v a t i o n s a b o u t the Law-as-Literature M o v e m e n t , " 54 Tennessee Law Review 203 (1987).

case, so let's all go home." 2 0 T h e established approach in this country requires judges (rightly, as we are about to see) to decide even the most difficult cases as best they can, provided the cases are real cases and within the judges'jurisdiction. This implies that any consideration relevant to deciding the case, whether drawn f r o m positive law or natural law sources, is a legitimate input in the manufacture of "law," and that the judge's decision—though of course not i m m u n e to criticism—will have to be pretty crazy before it can fairly be called "lawless." O n e could argue that if the legal issue is so uncertain that no resolution of it would be demonstrably correct, the defendant should not be made to pay damages or incur other costs. But w h y should there be this strong a s y m m e t r y in the rights of plaintiffs and defendants? Even if there should be, it would not follow that the j u d g e w h o went ahead and decided the close case against the defendant was being lawless. A questionable decision, even a w r o n g decision, is not lawless. To call it so is to attempt to resolve by persuasive definition what is properly a question of political theory. Equally, to call the judge's decision lawful is not to commit oneself to belief in strongly objective legal norms outside of those expressed in positive law. There is a deeper objection to refusing to decide borderline cases: it will change the borderline. A case that used to be well within the frontier will become the frontier. T h e law will tend to unravel—for every j u d g e - m a d e doctrine, c o m m o n law or interpretive, was once an extension beyond "existing law." We tend in thinking about law to put it temporally before the resolution of legal disputes, but by doing so we make the mistake (derived f r o m the fallacy of conceiving law as concept rather than activity) of thinking that law exists apart f r o m the process by which legal duties and sanctions are imposed on the persons subject to it. The reverse sequence is more illuminating. A h u m a n society seethes with conflict and dispute, and for certain types of dispute finds it expedient to have a permanent cadre of officials w h o resolve disputes in accordance with official norms. These are the judges, and their business is to resolve disputes in a way that will vindicate those n o r m s and, m o r e fundamentally, satisfy social needs. To lend a necessary regularity and predictability to the process, 20. I am i n d e b t e d for these e x a m p l e s to an u n p u b l i s h e d r e v i e w by A. W. Brian S i m p s o n of R o n a l d D w o r k i n ' s b o o k Law's Empire. T h e A d m i n i s t r a t i v e P r o c e d u r e Act does refer to a c a t e g o r y of cases w h e r e there is no law to apply, see 5 U . S . C . §701 ( a ) ( 2 ) — m e a n i n g n o t that the a d m i n i s t r a t i v e agency is acting lawlessly b u t that the agency's exercise of discretion is n o t judicially reviewable (the matter has been c o n f i d e d to the agency's discretion).

legislatures m a k e rules for the j u d g e s to apply, and the j u d g e s m a k e their o w n rules to fill the gaps (which are s o m e t i m e s e n o r m o u s ) in the legislative product; if there is n o legislature they m a k e all the rules. (Yet in neither case, for reasons stressed in C h a p t e r 4, are the j u d g e s legislators tout court.) B u t usually the m o s t i m p o r t a n t thing is to resolve the dispute. N o t always; sometimes the best resolution, especially w h e n the rules are unclear, is to let the dispute s i m m e r awhile. M a n y disputes, however, have to be resolved at once, even if the rules are unclear or have to be m a d e u p on the spot; and then the j u d g e s d o the best they can, using whatever i n f o r m a t i o n and insight that the lawyers give t h e m or that they can dredge up out of their o w n reading and experiences. This untidy, unrigorous process w e call "law." It m a y influence, t h o u g h perhaps only slightly, the d e v e l o p m e n t of the " m o r a l law," while, in turn, moralistic criticisms o f j u d i c i a l decisions m a y cause the j u d g e s to change legal doctrine; so there is a c o m p l e x interweaving of positive and natural law or, if o n e prefers of law and morality. 2 1 Judges w h o are especially protective of the judiciary's prestige w o u l d prefer not to have to decide difficult, uncertain cases. Such j u d g e s are like generals w h o do n o t w a n t to give battle unless they have a one hundred percent chance of victory. It is precisely to resolve the m o s t difficult, the m o s t uncertain, disputes that w e have judges. C o m p e l l e d to decide such cases, m a n y j u d g e s p r e t e n d — s o m e t i m e s to themselves as well as to the w o r l d — t h a t w h a t they have d o n e is added t w o and t w o and gotten four, so that a n y o n e w h o disagrees w i t h their decision is crazy, or that w h a t they have d o n e is chosen R i g h t over Wrong, so that anyone w h o disagrees w i t h the decision is morally obtuse. In fact they are m o r e likely to have engaged in the same kind of inconclusive practical reasoning, heavily influenced by personal experiences and b y t e m perament, that j u r o r s and politicians and civil servants use to m a k e j u d g ments. T h e distinctive things about the j u d g e s are that their incentives are a bit " p u r e r " than those of m o s t other officials, that their experiences are those of a lawyer, that their reading is d o m i n a t e d by legal materials largely u n k n o w n and incomprehensible to the lay public, and that c o n vention requires of t h e m (as it does n o t of m o s t other g o v e r n m e n t officials) a written justification of their i m p o r t a n t rulings. This last requirement imparts a certain thoughtfulness to w h a t they do, as d o the other procedural constraints o f j u d i c i a l decision m a k i n g as well as the conditions o f j u d i c i a l e m p l o y m e n t .

21. T h e similarity b e t w e e n legal and m o r a l t h i n k i n g is stressed b y Shklar, n o t e 11 above, at 6 0 - 6 2 .

The conception of the judicial function that I am advancing is lowkeyed—so low-keyed perhaps as to make one w o n d e r w h y anyone would feel a moral obligation to obey judicial decrees, 22 or would even respect the courts. O n e response is that as long as the judges are honest and decorous and do not get too far out ahead of dominant public opinion, the public can be expected to respect courts as an integral part of a system of government that seems on the whole to serve us well, or at least better than would the alternatives. We may disagree with particular decisions, but if w e value the concept of an independent judiciary we must take the bitter with the sweet. Another response is that, as far as anyone knows, it is just a lawyers' fancy that public respect for courts is a significant influence on the extent to which a society is law-abiding. Most people are uninformed and incurious about courts, especially those courts lawyers most dither over—appellate courts. Compliance with law is more a matter of incentives than of deference or respect. Still, it is an interesting analytical question whether there can be a moral obligation to obey positive laws that are not somehow underwritten by natural law. T h e argument against such a moral obligation is straightforward: moral obligations are obligations either to obey moral n o r m s or to complete certain voluntary undertakings (for example, promises) to which moral obligation attaches; but positive law, on the positivist view, has no necessary foundation in morals and is not in a realistic sense consented to by those w h o are subject to it. The positivist's argument for a moral duty to obey law is necessarily of the impliedcontractual form: w e would all be better off adhering to all the laws voluntarily (thus economizing on enforcement costs) rather than picking and choosing a m o n g them; so, ex ante—at the constitutional stage or in the original position—we would agree to be law-abiding. But the words "picking and choosing" load the dice. The question rather is whether to suspend moral j u d g m e n t regarding a body of laws that, according to the positivist hypothesis, bears only a contingent relation to morality. If a particular law happens to be immoral, as well it may be, it is hard to see h o w a duty to obey it can be derived f r o m the fact that voluntary compliance with law economizes on transaction costs, unless we place overriding value on public order. And, subject to the same qualification, it is not obvious w h y we should discourage people f r o m making moral j u d g m e n t s about the laws they live under and from acting on those j u d g m e n t s . U n d e r pressure f r o m these points, the moral obligation to obey law, an obligation based on considerations of mutual

22. For a g o o d discussion of this q u e s t i o n , see K e n t G r e e n a w a l t , Conflicts Morality, pt. 2 (1987).

of Law and

advantage and the social contract, dwindles to a weak, easily defeasible obligation to obey laws that are not morally offensive. Implicitly, however, I have been discussing cases in which disobeying the law does not benefit the violator at the expense of others. Even if the law you disobey is silly or vicious, if you are disobeying it in order to take advantage of the law-abiding y o u r conduct invites moral disapproval. You are displaying the disposition that Aristotle called pleonexia—wanting m o r e than y o u r fair share—and that a system of corrective justice seeks to hold in check (see Chapter 11). Since the lawless j u d g e ordinarily is not acting for private gain, on what basis can w e say that j u d g e s have a moral duty to obey the law by enforcing the rules within the j u d g e s ' j u r i s d i c t i o n ? B y raising this question I draw attention to a neglected function of the judicial oath. A solemn promise to administer law w i t h o u t fear or favor, it reinforces w h a t ever general moral duty there is to obey law w i t h the s o m e w h a t clearer moral duty to obey one's promises. B u t w h a t exactly is the promise? Is it a promise to enforce positive law to the hilt, or a promise to do justice? Perhaps a promise to do "justice under law"? B u t this formula m a y allow positive law to yield sometimes to natural law. A further point is that promises are rarely absolute, are usually defeasible. And w e make m a n y promises, and they m a y conflict, forcing us to choose. Whether or not the argument for a moral duty to obey positive law persuades, and whatever the precise scope of such a duty, the enterprise, n o w several thousand years old, of establishing the existence and content of a natural law that underwrites positive law 2 3 is hopeless under the conditions of m o d e r n American society. If the result of this failure is a diminished sense of moral obligation to obey laws, so it will have to be. Even the term "natural l a w " is an anachronism. T h e majority of educated Americans believe that nature is the amoral scene of Darwinian struggle. Occasional attempts are made to derive social n o r m s f r o m nature so conceived, 2 4 but they are not likely to succeed. It is true that a variety of widely accepted norms, including the keeping of certain 23. M o d e r n theories o f natural l a w are t h o r o u g h l y and s y m p a t h e t i c a l l y discussed in D e r y c k Beyleveld and R o g e r B r o w n s w o r d , Law as a Moral Judgment (1986). 24. For recent a t t e m p t s , see Richard A. E p s t e i n , " T h e U t i l i t a r i a n F o u n d a t i o n s of N a t ural L a w " 12 Harvard Journal of Law and Public Policy 713 (1989); P e t e r R a i l t o n , " M o r a l Realism," 95 Philosophical Review 163 (1986). T h e b e s t - k n o w n a t t e m p t r e m a i n s that o f H e r bert Spencer, w h o s e Social D a r w i n i s m i n f l u e n c e d A m e r i c a n l a w in t h e late n i n e t e e n t h and early t w e n t i e t h centuries. See A r t h u r L. H a r d i n g , " T h e G h o s t of H e r b e r t Spencer: A D a r winian C o n c e p t of L a w , " in Origins of the Natural Law Tradition 69 ( H a r d i n g ed. 1954). O n Social D a r w i n i s m generally, see the interesting discussion in M i c h a e l R u s e , Taking Darwin Seriously: A Naturalistic Approach to Philosophy 7 3 - 8 2 (1986).

promises, the abhorrence of unjustified killing of h u m a n beings, and perhaps even the sanctity of property rights, p r o m o t e the adaptation of the h u m a n species to its environment. But so does genocide. Within limits, the destruction of n o n h u m a n species spreads h u m a n genes, and likewise a race that exterminates other races spreads its members' genes. If the Nazis' geopolitical ambitions had been fulfilled, Hitler would have been entitled to kudos f r o m those w h o believe that Darwinism is a source of moral norms. We do not admire human behavior merely because it conforms to genetic programs. There are, of course, alternative conceptions of nature to the Darwinian, a pertinent example being Plato's, in which moral n o r m s are part of the furniture of the universe. But the m o d e r n secular intellect finds Plato's, and the successor Christian, notion of a " g o o d " nature unbelievable. T h e underlying problem of moral objectivity is that there are neither facts to which moral principles correspond (as scientific principles, for example, appear to correspond to things in nature) nor a strong tendency for moral principles to converge. A tiny handful of moral principles—for example, that unrestricted killing of members of one's o w n society is bad—seem essential to social existence, but the rest are conventional and culture-bound to a far greater extent than scientific principles are. 25 If w e put aside nature or other possible sources of ontologically objective moral principles—principles that are "out there" in some sense— the precondition of strong "natural law" (now a misnomer) is that the members of society, whether because they share the same religious beliefs or because they are otherwise culturally homogeneous or in thrall to a powerful ideology, agree on moral principles to a level of detail that enables a legal code to be derived f r o m those principles. Most Americans agree on many things—that unrestricted killing is bad, that Nazis are bad, that incest is bad, that certain forms of discrimination are bad— but not on enough things to prescribe anything like a comprehensive system of legal rights and obligations. In such a society, natural law theorists are exposed to devastating criticisms. 26

25. C f . R. M . H a r e , " O n t o l o g y in E t h i c s , " in Morality and Objectivity: A Tribute to J. L. Mackie 39 (Ted H o n d e r i c h ed. 1985). For a s t r o n g a r g u m e n t (contrary to the position I am taking) for scientific and m o r a l realism, see Richard N . B o y d , " H o w to Be a M o r a l Realist," in Essays on Moral Realism 181 ( G e o f f r e y S a y r e - M c C o r d ed. 1988). O n the general question o f m o r a l realism ( a b s o l u t i s m , o b j e c t i v i s m ) versus m o r a l (or ethical) relativism, see, besides the collection of essays j u s t cited, the f o l l o w i n g collections: Relativism: Interpretation and Confrontation (Michael K r a u s z ed. 1989); Rationality and Relativism ( M a r t i n Hollis and Steven Lukes eds. 1982); Ethical Relativism (John Ladd ed. 1973). 26. See, for e x a m p l e , L l o y d L. Weinreb, Natural

Law

and Justice,

chs. 4 - 8 (1987).

It hardly matters in this regard w h e t h e r one is a moral realist, a conventionalist, an anything. If the c o m m u n i t y is morally heterogeneous, moral principles function as a r g u m e n t s rather than as criteria, and it is then better to speak of natural law considerations than of natural law simpliciter. This is n o t " a n y t h i n g goes" relativism, but merely recognition that m a n y moral disputes cannot be resolved b y peaceful means in a heterogeneous c o m m u n i t y . Legal means are n o t necessarily peaceful means in this sense. If a legal decision has n o roots in moral consensus, it may rise n o higher than the a r m e d force that stands behind it. In a setting of m o r a l "dissensus," the invocation of the emotionally charged term "natural l a w " answers primarily to the rhetorical aim of calling o u r personal opinions law and the laws w e dislike personal o p i n ions. To the extent that our language invites us to f r a m e m a n y of o u r policy considerations in moral language, w e m a y say that natural law, viewed as the s u m of those policy considerations that can be so described, influences positive law. "It is the stuff out of w h i c h h u m a n or positive law is to be w o v e n , w h e n other sources fail." 27 B u t as a repository of frequently w a r r i n g principles, natural l a w in this sense cannot resolve difficult cases. I a m not saying that n o n e of o u r m o r a l beliefs is m o r e than a personal opinion. M a n y of those beliefs are as firmly rooted as o u r epistemic certitudes, as I remarked in discussing A r t h u r Leff's moral skepticism. That is w h y I have not t h o u g h t it necessary to pause to explain w h y I called the Nazis " m o n s t e r s " ; indeed the explanation w o u l d have added nothing interesting to the bare statement. T h e m o r a l beliefs that w e hold with great conviction gain n o analytic strength f r o m having the label "natural l a w " affixed to them; the reasons f o r using the f o r m s of l a w against the Nazis must be sought elsewhere. A n d contested moral beliefs gain n o analytic strength b y being advanced as propositions of natural law. In suggesting that there was little at stake in the Hart-Fuller debate, I mean precisely that the analytic stakes w e r e small. Debates over entities of d o u b t f u l o n t o l o g y or, w h a t is here and usually the same thing, over

Michael M o o r e , in " M o r a l Reality," 1982 Wisconsin Law Review 1061, does a w o n d e r f u l j o b of s u m m a r i z i n g all the o b j e c t i o n s to the existence of o b j e c t i v e m o r a l n o r m s , b u t his e f f o r t to r e f u t e those objections is w e a k , a l m o s t p e r f u n c t o r y . It a m o u n t s to the reiterated o b s e r vation that s o m e m o r a l p r o p o s i t i o n s are as solidly established as s o m e factual p r o p o s i tions—which is t r u e b u t u n h e l p f u l , because t h o s e are n o t t h e m o r a l p r o p o s i t i o n s that m i g h t m s w e r controversial legal questions. Saying that the H o l o c a u s t was a c r i m e against h u m a n ty (which, by the way, is no! a p r o p o s i t i o n that c o m m a n d s a w o r l d w i d e consensus) will rot d e t e r m i n e the a n s w e r t o the q u e s t i o n w h e t h e r a b o r t i o n is m o r a l l y w r o n g . 27. B e n j a m i n N . C a r d o z o , The Nature of the Judicial Process 132 (1921).

the meaning of undefinable words can have profound effects in the "real" world—can cause religious wars, for example. O n e can understand h o w people w h o believe that natural law exists as a set of consistent, definite, and knowable commands are no more willing to allow constitutional text or history or precedent to stand in their way than the positive law of the Nazi regime was allowed to stand in the way of justice at N u r e m b e r g . T h e curiosity about the Hart-Fuller debate is that the debaters seem not to have disagreed on any substantive point. Their disagreement was over the rhetoric in which to express their shared values. T h e debate was at b o t t o m pragmatic, the debaters' ultimate concern being about which vocabulary—that of positive law or that of natural law—is more conducive to the realization of those values. It is an unnecessary choice. O u r dominantly positivistic discourse has enough natural law play in its joints to give us all the rhetorical flexibility we need. Reflection on the Hart-Fuller debate points up t w o flaws in the activity theory of law. First, if as I have argued the use of the term "natural law" is admissible both to describe those (few) ethical or political principles that c o m m a n d a consensus and seem therefore as imperative as positive law or even more so and, more modestly, to summarize the ethical considerations that bear on the decision of close cases, then we have a concept of law that is something other than an activity. For whose activity is it to produce natural law? There are no licensed professionals to point to. T h e relevant concept of law in such a setting is one of law as custom or tradition rather than as a professional activity, or more precisely as a professional activity bounded and shaped by custom, tradition, c o m m u n i t y feeling, and so on. Second and closely related, the debate raises the question h o w one determines when the professional w h o is licensed to make positive law is acting within the scope of his license. If we say that the Nazis " m u r dered" the Jews, is this not tantamount to saying that Nazi judges, prosecutors, executioners, and so forth exceeded the bounds of their licensed activities? If so, this means that the criteria for when a person is licensed to do law are supplied f r o m outside the activity itself. I hinted at this problem earlier w h e n I said that a judge's decision would have to be pretty crazy to be pronounced lawless. If it is so crazy that it is lawless, what does this imply about the criteria for deciding what is law? What does "crazy" mean in this context? Apparently some considerations that a j u d g e might use to decide a difficult case are out of bounds—but w h o fixes the bounds? T h e significance of natural law may lie not in its capacity for resolving the normal issues that arise in a civilized legal system

but in the assistance it provides in d e t e r m i n i n g w h a t a civilized legal system is. If this is correct, have n o t the categories "positive l a w " and "natural law" outlived their usefulness? As a source o f duties, rights, and powers, the law seems best regarded as an activity of licensed professionals (judges and lawyers), cabined by vague b u t p o w e r f u l n o t i o n s o f professional propriety rooted ultimately in social convenience or, equivalently, durable public opinion. Positive law and natural law materials are inputs into the activity w e call law. At this level of generality, Ronald D w o r k i n ' s concept of law converges w i t h mine. While rejecting legal positivism and emphasizing elements in legal reasoning that are emphasized b y natural lawyers as well, such as rights and moral principles, D w o r k i n is careful n o t to describe himself as a natural lawyer. H e evidently w a n t s to avoid w h a t is indeed a j e j u n e dichotomy. His a r g u m e n t at its m o s t general is that the j u d g e is trying to fashion the best o u t c o m e out o f the diverse materials that are at hand, materials that include n o t only precedents and statutory language and other distinctively "legal" materials b u t also the elements of political morality. This is an accurate portrayal of the A m e r i c a n judge, and I agree w i t h D w o r k i n that the j u d g e so p o r t r a y e d is n o t a lawless j u d g e . But I disagree that the j u d g e b e c o m e s lawless w h e n he allows "rights" based o n "principles" to be o v e r c o m e by "policies," where "policy" is defined b y D w o r k i n as a collective goal. M a n y collective goals, ranging f r o m national survival and public order to p r o s perity and social insurance, are n o less deeply w o v e n into the fabric of our political morality than such principles as that equals should be treated equally or that n o person should b e allowed to profit f r o m his w r o n g doing. To distinguish b e t w e e n policies and principles and to link rights with the latter b u t n o t the f o r m e r is arbitrary. T h e r e is n o basis f o r excluding collective goals in determining the scope of legal rights.

Holmes,

Nietzsche,

and

Pragmatism

Holmes w r o t e The Common Law j u s t a f e w years before Nietzsche's great w o r k On the Genealogy of Morals, and b o t h e m p l o y an effective m e t h o d of skeptical analysis: the genealogical. In the Genealogy and o t h er w o r k s Nietzsche tried to u n d e r m i n e the ontological status of C h r i s tian morality by arguing that m o r a l beliefs reflect the needs and circumstances of the d o m i n a n t groups in the c o m m u n i t i e s that h a p p e n to hold them. Morality, in other w o r d s , is relative rather than absolute; in fact, morality is public opinion. The Common Law does the same thing w i t h

law. By tracing legal doctrines to their origins and thus relating each doctrine to a particular constellation of social circumstances, Holmes showed the absurdity of supposing, as did the nineteenth-century formalists against w h o m he was writing, that legal doctrines were unchangeable formal concepts like the Pythagorean theorem. He enforced the lesson of ethical relativism, thereby turning law into d o m inant public opinion in m u c h the same way that Nietzsche turned morality into public opinion. 2 8 T h e genealogical technique continues to be an effective one in law. O f t e n one traces a line of precedents to its source and finds that the first of the line is a mere assertion and that the next merely cited the first, and so on to the latest decision. The genealogical technique can thus show that a rule that seems firmly grounded in precedent actually rests on sand. O r it can show, as Holmes memorably did in discussing the evolution of proceedings to condemn deodands into actions in rem in admiralty, h o w — i n Nietzsche's words—"the cause of the origin of a thing and its eventual utility, its actual employment and place in a system of purposes, lie worlds apart; whatever exists, having somehow come into being, is again and again reinterpreted to new ends." 29 Nietzsche was not the only nineteenth-century genealogist, of course. Higher criticism of the Bible (that is, the critical study of its authorship), the tracing of man's descent of man f r o m the lower animals, and Marx's theory of history are other notable examples of the nineteenth century's penchant for using history to challenge essentialist or transcendental notions. But Nietzsche's expose of the origins of Christianity may be the closest parallel to what Holmes did to law, even though their attitude toward their subject matter was different (and the t w o were even more different of course, in tone, Nietzsche being a most irresponsible writer). M o r e than genealogical method connects these t w o great nineteenthcentury skeptics. They both are sub- or antimentalists. Nietzsche's will to power animates not just man and animals but the rocks and the 28. See, for e x a m p l e , L o c h n e r v. N e w Y o r k , 198 U . S . 45, 75 (1905) (dissenting o p i n ion); H o l m e s , " H e r b e r t Spencer; Legislation and E m p i r i c i s m , " in Justice Oliver Wendell Holmes: His Book Notices and Uncollected Letters and Papers 104 ( H a r r y C . Shriver ed. 1936). R o b e r t W. G o r d o n , " H o l m e s ' Common Law as Legal and Social Science," 10 Hofstra Law Review 719, 746 (1982), n o t e s — u n f o r t u n a t e l y w i t h o u t e l a b o r a t i o n — H o l m e s ' s " N i e t z s c h e a n insight into the p o w e r relations u n d e r l y i n g s y s t e m s of rights and m o r a l s . " 29. On the Genealogy of Morals and Ecce Homo 77 (Walter K a u f m a n n ed. 1967) (Second Essay, §12, of On the Genealogy of Morals).

stars, 30 and Holmes reduces law to force and makes the will the preferred mode of resolving the close case. T h e aphoristic style of both writers may reflect a shared skepticism about the p o w e r of rational thought. The illiberal aspects of Holmes's thought, notably his fondness for war, struggle, and eugenic breeding of h u m a n beings (that is, man as animal), have close parallels in Nietzsche; 31 and it is perhaps natural that people w h o are skeptical about reason should celebrate its antithesis. I do not want to go overboard in comparing H o m e s to Nietzsche. Holmes's free-speech and habeas corpus opinions e m b o d y a humane and democratic perspective that is remote f r o m anything in Nietzsche. And Holmes's militant skepticism and relativism coexist with a tame utilitarianism that would have nauseated Nietzsche. Within the space of a few pages in " T h e Path of the Law," Holmes both suggests that it would be a good thing "if every w o r d of moral significance could be banished from the law altogether, and other words adopted which should convey legal ideas uncolored by anything outside law," and advises lawyers to 30. See the excellent discussion in A l e x a n d e r N e h a m a s , Nietzsche: Life as Literature (1985), and n o t e this typical r e m a r k of Nietzsche's: " H e w h o possesses s t r e n g t h divests himself of m i n d . " Friedrich N i e t z s c h e , Twilight of the Idols and The Anti-Christ 76 (R. J. Hollingdale trans. 1968) ( " E x p e d i t i o n s of an U n t i m e l y M a n , " §14, in Twilight of the Idols). Yet the line b e t w e e n the a n t i m e n t a l and the p a n - p s y c h i c is an u n c e r t a i n one. T h e e x t r e m e s of subjectivity and objectivity, of idealism and realism, tend to m e r g e . T h i s is also s h o w n by the f o l l o w i n g illustration of H o l m e s ' s a n t i m e n t a l i s m , his f a m o u s adage, " W e m u s t t h i n k things not w o r d s . " O l i v e r Wendell H o l m e s , " L a w in Science and Science in L a w , " 12 Harvard Law Review 443, 460 (1899). T h e idea t h a t w o r d s stand in a o n e - t o - o n e c o r r e s p o n d e n c e to things is a f o r m of naive realism that m a y seem r e m o t e f r o m the p i c t u r e I a m d r a w i n g of H o l m e s ; it is certainly r e m o t e f r o m a n y t h i n g in N i e t z s c h e . B u t a h i g h l y referential t h e o r y of language is c o m p a t i b l e w i t h a disdain for a b s t r a c t i o n (as in B e n t h a m ) ; w h a t is n o t a thing, is n o t h i n g . 31. O n Nietzsche, see J. P. Stern, Nietzsche (1979); Law and Literature: A Misunderstood Relation, n o t e 6 above, ch. 3. O n H o l m e s , see, f o r e x a m p l e , E d m u n d Wilson, "Justice O l i v e r Wendell H o l m e s , " in Wilson, Patriotic Gore: Studies in the Literature of the American Civil War 743 (1962); B u c k v. Bell, 274 U . S . 200 (1927). Typical is this s t a t e m e n t in H o l m e s ' s o p i n i o n in M i s s o u r i v. H o l l a n d , 252 U . S . 416, 433 (1920): "It w a s e n o u g h f o r t h e m [the f r a m e r s of the C o n s t i t u t i o n ] to realize or to h o p e that t h e y had created an o r g a n ism; it has taken a c e n t u r y and has cost their successors m u c h s w e a t and b l o o d to p r o v e that they created a n a t i o n . " A n o t h e r link b e t w e e n H o l m e s and N i e t z s c h e is that b o t h admired and w e r e influenced b y R a l p h W a l d o E m e r s o n . O n H o l m e s and E m e r s o n , see Sanford Victor Levinson, " S k e p t i c i s m , D e m o c r a c y , and Judicial Restraint: A n Essay o n the T h o u g h t of O l i v e r Wendell H o l m e s and Felix F r a n k f u r t e r " ( P h . D . diss., D e p t . of G o v t . , H a r v a r d University, April 1969). O n N i e t z s c h e and E m e r s o n , see " T r a n s l a t o r ' s I n t r o d u c tion," in Friedrich Nietzsche, The Gay Science 3, 7—13 (Walter K a u f m a n n trans. 1974); Charles Andler, Nietzsche: Sa vie et sa pensee, ch. 2 (1958).

study economics, where " w e learn that for everything [we want] we have to give up something else, and we are taught to set the advantage we gain against the other advantage we lose, and to k n o w what we are doing w h e n we elect." 32 T h e idea that law ought to be a tool for promoting social welfare understood in economic terms is a moral principle; H o m e s in the t w o passages I have quoted is both rejecting morality and proposing a morality to shape law's future. Holmes is not a nihilist after all, at least not a consistent one; is he a pragmatist? Beware the tyranny of lables—Nietzsche, greatest of nihilists, denounced nihilism. There are significant pragmatist strands in Holmes's thought, and given the salutary vagueness and breadth of the term "pragmatist," no m o r e is needed to label him one. 33 But it would be w r o n g to suppose that every aspect of his thought, even of his philosophical thought, is pragmatist. Holmes's Social Darwinism is not; and a skepticism that teaches that a legal decision has only t w o components, logic and will, or that advocates a simple referential theory of language, or that is hostile to factual inquiry is foreign to most versions of pragmatism—and these are all positions that Holmes (at times, anyway)

32. H o l m e s , n o t e 2 a b o v e , at 464, 474. 33. T h e Fisch and G r e y articles cited in n o t e 2 e m p h a s i z e the parallels b e t w e e n H o l m e s and J o h n D e w e y , t h u s l i n k i n g H o l m e s to the b r o a d e r p r a g m a t i s t school w h i c h Peirce m a y be said to h a v e been a p a r t o f (indeed, t o h a v e f o u n d e d ) , despite the m a n y differences a m o n g h i m , D e w e y , and J a m e s , s o m e n o t e d in Yosal R o g a t , " T h e J u d g e as Spectator," 31 University of Chicago Law Review 213, 251 n. 194 (1964); as n o t e d in M o r t o n White, " T h e R e v o l t against F o r m a l i s m in A m e r i c a n Social T h o u g h t o f the T w e n t i e t h C e n t u r y , " in W h i t e , Pragmatism and the American Mind: Essays and Reviews in Philosophy and Intellectual History 41, 52—53 (1973), D e w e y w a s a great a d m i r e r of H o l m e s — s e e , f o r example, J o h n D e w e y , Experience and Nature 4 1 7 - 4 1 9 (1929); D e w e y even w r o t e a very H o l m e s i a n paper o n legal i n q u i r y : "Logical M e t h o d and L a w , " 10 Cornell Law Quarterly 17 (1924). I n o t e d the affinities b e t w e e n H o l m e s and Peirce in n o t e 5 a b o v e . O n H o l m e s and p r a g m a t i s m generally, Grey's article, n o t e 2 a b o v e , is especially g o o d ; see also C a t h a r i n e Wells Hantzis, "Legal I n n o v a t i o n w i t h i n t h e Wider Intellectual T r a d i t i o n : T h e P r a g m a t i s m of O l i v e r Wendell H o l m e s , J r . , " 82 Northwestern University Law Review 541 (1988); Frederic R o g e r s Kell o g g , The Formative Essays of Justice Holmes: The Making of an American Legal Philosophy, ch. 3 and b i b l i o g r a p h i c a l essay (1984); Marcia J . Speziale, " B y T h e i r Fruits You Shall K n o w T h e m : P r a g m a t i c i s m and t h e P r e d i c t i o n T h e o r y of L a w , " 9 Manitoba Law Journal 29 (1978); E d w a r d J. B l o u s t e i n , " H o l m e s : H i s First A m e n d m e n t T h e o r y and His P r a g m a t i s t B e n t , " 40 Rutgers Law Review 283 (1988). H o l m e s ' s debt t o p r a g m a t i s m is q u e s t i o n e d in H . L. P o h l m a n , Justice Oliver Wendell Holmes and Utilitarian Jurisprudence (1984), but p r i m a r i l y on the basis of dismissive r e m a r k s H o l m e s m a d e a b o u t Peirce and J a m e s . See also Sheldon M . N o v i c k , Honorable Justice: The Life of Oliver Wendell Holmes 4 2 6 - 4 2 7 (1989) (n.4). T h e s e r e m a r k s s h o w m e r e l y that H o l m e s , like m a n y p e o p l e of genius, t e n d e d to exaggerate his o w n originality.

embraced. 34 Pragmatism is a house of m a n y rooms, as we shall see, and a pragmatist jurisprudence is no m o r e committed to the thought of Oliver Wendell Holmes than to that of Richard Rorty (see Chapter 12). O f all Holmes's "heresies" against the pragmatic faith, the most pertinent to this chapter is his "strong" legal positivism, his belief—erratically maintained, to be sure, but emphatically stated in the passage I quoted about the desirability of banishing all moral terms f r o m law— that law and morals should have as little to do with each other as possible (perhaps nothing). T h e belief is false. Law cannot be reduced to rules clearly stated in statutes or constitutions, or to rules plus arbitrary assertions of will. M u c h of the law applied by judges, not to mention the "law" that tells us whether the rules or for that matter the judges themselves are lawful, consists of moral and political considerations. H o w could legal ideas be "uncolored by anything outside the law," when, as Holmes himself emphasized, the law is—and should be— shaped by social needs and interests? T h e same essay states that the premise of a c o m m o n law doctrine "can do no more than e m b o d y the preference of a given body in a given time and place." 35 This is not right either. It denies the possibility that a law might reflect a sufficiently wide and durable interest to be fairly described as moral, as something more than a local and transient preference. Moreover, although the preferences of the dominant groups in society both constrain and give broad direction to law, law is more than simply the e m b o d i m e n t in coercive decrees of the results of the latest public opinion poll or canvass of special interest groups. Holmes seems in these passages in the grip of an overmastering conception of what law must be to count as law. This essentialism is foreign to the pragmatic approach—to which, however, he returned later in the essay, when he suggested that lawyers need a good dose of economics to be able to guide the law in the path of the general social advantage. A similar tension is found in his judicial oeuvre, where candid acknowledgment of the judge's need to make choices based on policy preferences coexists with powerful formalist

34. O n p a r t i t i o n i n g the decisional process i n t o logic and will, see n e x t chapter; o n H o l m e s ' s referential t h e o r y of l a n g u a g e , see n o t e 30 a b o v e ; o n his dislike of factual inquiry, see, for e x a m p l e , letter to H a r o l d Laski of M a r c h 18, 1919, in Holmes-Laski Letters, n o t e 7 above, vol. 1, p. 205 ("1 hate facts"). 35. H o l m e s , n o t e 2 above, at 466. A n d in the s a m e vein, " T h e law can ask n o better justification than the deepest instincts o f m a n . " Id. at 477. W h y n o better? " G o o d " and " i n s t i n c t u a l " are o n different axes.

disclaimers of the relevance of consequences to judicial decision making. 36 Although H o l m e s may be the founder and greatest exemplar of pragmatic jurisprudence, a c o m m i t m e n t to pragmatism neither requires nor even permits a wholesale embrace of his jurisprudential views. 36. For illustrations, c o m p a r e T y s o n & B r o t h e r v. B a n t o n , 273 U . S . 418, 446-448 (1927), and O l m s t e d v. U n i t e d States, 277 U . S . 438, 469 (1928), w i t h A d k i n s v. Children's H o s p i t a l , 261 U . S . 525, 5 7 0 - 5 7 1 (1923); Schlesinger v. Wisconsin, 270 U . S . 230, 241 (1926), and Louis K. L i g g e t t C o . v. B a l d r i d g e , 278 U . S . 105, 115 (1928). T h e s e are all dissenting o p i n i o n s .

P A R T III