Truth, Error, And Criminal Law An Essay In Legal Epistemology

Truth, Error, And Criminal Law an Essay in Legal Epistemology
View more...


Preview only show first 6 pages with water mark for full document please download


P1: JZZ 0521861667pre CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:49 i This page intentionally left blank P1: JZZ 0521861667pre CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:49 Truth, Error, and Criminal Law An Essay in Legal Epistemology This book treats problems in the epistemology of the law. Beginning with the premise that the principal function of a criminal trial is to find out the truth about a crime, Larry Laudan examines the rules of evidence and procedure that would be appropriate if the discovery of the truth were, as higher courts routinely claim, the overriding aim of the criminal justice system. Laudan mounts a systematic critique of existing rules and procedures that are obstacles to that quest. He also examines issues of error distribution by offering the first integrated analysis of the various mechanisms – the standard of proof, the benefit of the doubt, the presumption of innocence, and the burden of proof – for implementing society’s view about the relative importance of the errors that can occur in a trial. Larry Laudan is Principal Investigator at the Instituto de Investigaciones Filos´ oficas, Universidad Nacional Aut ´ onoma de M´ exico. He is the author of many books, including Progress and Its Problems, Science and Values, and Beyond Positivism and Relativism. He is a former divisional president of the American Philosophical Association. i P1: JZZ 0521861667pre CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:49 ii P1: JZZ 0521861667pre CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:49 Cambridge Studies in Philosophy and Law general editor: gerald postema (university of north carolina, chapel hill) advisory board Jules Coleman (Yale Law School) Antony Duff (University of Stirling) David Lyons (Boston University) Neil MacCormick (University of Edinburgh) Stephen R. Munzer (UCLA Law School) Phillip Pettit (Princeton University) Joseph Raz (University of Oxford) Jeremy Waldron (Columbia Law School) Some Other Books in the Series: Larry Alexander (ed.): Constitutionalism Larry Alexander: Is There a Right of Freedom of Expression? Peter Benson (ed.): The Theory of Contract Law: New Essays Steven J. Burton: Judging in Good Faith Steven J. Burton (ed.): “The Path of the Law” and Its Influence: The Legacy of Oliver Wendell Holmes, Jr. Jules Coleman: Risks and Wrongs Jules Coleman and Allan Buchanan (eds.): In Harm’s Way: Essays in Honor of Joel Feinberg R. A. Duff (ed.): Philosophy and the Criminal Law William Edmundson: Three Anarchical Fallacies: An Essay on Political Authority John Fischer and Mark Ravizza: Responsibility and Control R. G. Frey and Christopher W. Morris (eds.): Liability and Responsibility: Essays in Law and Morals Steven A. Hetcher: Norms in a Wired World Heidi M. Hurd: Moral Combat Jody S. Kraus and Steven D. Walt (eds.): The Jurisprudential Foundations of Corporate and Commercial Law Christopher Kutz: Complicity: Ethics and Law for a Collective Age Timothy Macklem: Beyond Comparison: Sex and Discrimination Larry May: Crimes against Humanity: A Normative Account Stephen R. Munzer: A Theory of Property Arthur Ripstein: Equality, Responsibility, and the Law R. Schopp: Justification Defenses and Just Convictions iii P1: JZZ 0521861667pre CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:49 iv P1: JZZ 0521861667pre CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:49 Truth, Error, and Criminal Law An Essay in Legal Epistemology Larry Laudan Universidad Nacional Aut ´ onoma de M´ exico v caxniioci uxiviisir\ iiiss Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo Cambridge University Press The Edinburgh Building, Cambridge cn: :iu, UK First published in print format isnx-:, ,;ï-c-,::-ïo:oo-, isnx-:, ,;ï-c-,::-::,::-, © Larry Laudan 2006 2006 Information on this title: www.cambridg /9780521861663 This publication is in copyright. Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. isnx-:c c-,::-::,::-c isnx-:c c-,::-ïo:oo-; Cambridge University Press has no responsibility for the persistence or accuracy of uiis for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate. Published in the United States of America by Cambridge University Press, New York hardback eBook (EBL) eBook (EBL) hardback P1: JZZ 0521861667pre CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:49 Acquitting the guilty and condemning the innocent – the Lord detests them both. – Proverbs 17:15 As there is the possibility of a mistake, and as it is even probable, nay, morally certain that sooner or later the mistake will be made, and an innocent person made to suffer, and as that mistake may happen at the very next trial, therefore no more trials should be had and courts of justice must be condemned. W. May, Some Rules of Evidence, 10 Amer. L. Rev. 642, at 654–5 (1876) vii P1: JZZ 0521861667pre CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:49 viii P1: JZZ 0521861667pre CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:49 Contents Preface page xi Abbreviations and Acronyms Used xv 1 Thinking about Error in the Law 1 part i the distribution of error 2 The Unraveling of Reasonable Doubt 29 3 Fixing the Standard of Proof 63 4 Innocence, the Burden of Proof, and the Puzzle of Affirmative Defenses 89 part ii flawed rules of evidence and procedure 5 Evaluating Evidence and Procedures 117 6 Silent Defendants, Silent Witnesses, and Lobotomized Jurors 147 7 Confessions, Poison Fruit, and Other Exclusions 171 8 Double Jeopardy and False Acquittals: Letting Felons and Judges off the Hook? 194 9 Dubious Motives for Flawed Rules: The Clash between Values 213 Index 235 ix P1: JZZ 0521861667pre CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:49 x P1: JZZ 0521861667pre CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:49 Preface Every author owes debts more numerous than he can mention. Of some, he is barely aware, though they are no less real for that. More troubling are those that run so deeply that they cannot easily if ever be repaid, and certainly not by the bare acknowledgment of their existence. Still, it remains important to mention them, even if the gesture is brief and fleeting. I first became interested in epistemological issues surrounding the lawabout five years ago, having previously devoted myself to the philosophy of science and applied epistemology. More by accident than by design, my earliest encoun- ters with academic law occurred at the University of Texas, where I often go to consult books unavailable in Mexico, where I work. On one of my annual trips north of the border, I decided to stop into the office of Brian Leiter in the University of Texas Law School. I had, by chance, been reading a classic legal case, In re Winship, a few days before. Leiter and I did not know one another, but something was bothering me and I knew his reputation as one of the few philosophers of law with an interest in questions of proof. After introducing myself, I asked him (more or less): “I can’t make sense of what the court is saying about proof beyond a reasonable doubt. Can you straighten me out?” After puzzling over the relevant passages, he replied candidly: “No.” This book dates fromthat conversation. Probably as much to get me out of his hair as anything else, Brian put me onto LexisNexis, that wonderful repository of all things legal on the Internet. I started reading other Supreme Court cases discussing reasonable doubt, hoping that would set me straight. It did not. This book is the end product of my quest for an answer to that initial and seemingly innocuous question. As these things always do, my puzzle about reasonable doubt mushroomed into worries about a plethora of epistemic notions (the ben- efit of the doubt, the presumption of innocence, the burden of proof, relevance, and reliability) widely used by the judiciary and academic lawyers alike. The nagging worry was that key parts of all these notions (especially proof, rele- vance, and reliability) were being used in ways that were not only nonstandard (at least among philosophers) but also, apparently, deeply confused. The more I xi P1: JZZ 0521861667pre CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:49 xii preface read, the more uneasy I became. Senior jurists, including those on the Supreme Court, often wrote about knowledge and truth seeking in ways that I found foreign and unfamiliar. Sometimes, they seemed plainly wrong. At about this point, I came to know Ron Allen, the Wigmore Professor of Evidence Law at Northwestern, whose work I had read and from which I have learned much. Even when we disagreed, which was not often, I felt that we were in the same conceptual universe, committed to the idea of analyzing a trial as the search for the truth about a crime. Besides, we shared a knee-jerk aversion to the Bayesian project in the law and elsewhere, so I knew he had to be on the side of the angels. A year later, I finally stumbled upon the article that I had been looking for in Leiter’s office that day almost two years earlier: a cogent and sophisticated treatment of the standard of proof beyond a reasonable doubt. It was written by a young legal scholar, Erik Lillquist from Seton Hall Law School, from whom I have also learned much. Fortuitously, some funds from the Institute for Philosophical Investigation at my university made it possible for my colleague Juan Cruz Parcero and me to invite several scholars to the campus for three days of intensive conversa- tions about law and epistemology in December 2003. Apart from Allen and Lillquist, two other scholars attending that meeting made a deep impression on me. They were Michele Taruffo from Pav´ıa and Jordi Ferrer from the Univer- sity of Gerona. Politely overlooking the fact that I was neither a lawyer nor a philosopher of law, both of themheightened my awareness of a number of prob- lems that I had barely stumbled on in my own halting efforts with LexisNexis. Above all, they persuaded me that – where the law of evidence is concerned – the traditional gulf postulated between Roman and Anglo-Saxon law was ill- founded. Both civilian and common law courts face similar problems of proof andevidence, andit hadbeensimplyparochial of me toimagine that anappropri- ate dialogue about evidence could be conducted within the terms of reference of a single legal system. Living and working in Mexico, as I do, reinforced that impression, since I spend much of my time explaining the mysteries and idiosyncrasies of Anglo-Saxon procedure to Mexicans and likewise learning about those of the Mexican system. As I subsequently discovered, Taruffo has written a splendid volume in Italian, The Proof of Judicial Facts, that is, in my judgment, the best current book on the theory of legal proof. (It is a scan- dal, but symptomatic of the problem I just mentioned, that there is no English translation of it.) My examination of the parallels between Mexican and U.S. law has been enormously aided by my friend Enrique C´ aceres of the Institute for Jurisprudence at the National Autonomous University of Mexico (UNAM), whose knowledge of Mexican jurisprudence is more than merely impressive. Two years ago, the Law School at the University of Texas invited me to put together an advanced seminar in legal epistemology. Along with the patient students who suffered through my first shot at writing this book, a very bright P1: JZZ 0521861667pre CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:49 Preface xiii philosopher of law, Les Greene, regularly participated. His sagacious questions saved me from some of the serious errors into which I was falling. Outside the law itself, I must mention my continuing debt to Deborah Mayo’s penetrating analyses of the nature of error and the logic of the design of statistical tests. Closer to home, I am grateful to my colleagues at UNAM, who batted nary an eyelash when I announced to them that I was taking time off for a couple of years from my duties as philosopher of science to learn something about the law. But for their generous provision of time for study-leave, it would have been impossible to write this book. Finally, I want to acknowledge a deep indebtedness to my wife, Rachel, who (among many other things) worked very hard – but with limited success – to make this book intelligible to nonspecialists. Two chapters of this book (2 and 4) are much-altered versions of articles that have appeared or will soon appear in Legal Theory. I remain humbled that the editors of that distinguished journal (Larry Alexander, Jules Coleman, and Brian Leiter) were willing to take a total outsider under their collective wing. Guanajuato, M´ exico 1 August 2005 P1: JZZ 0521861667pre CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:49 xiv P1: JZZ 0521861667pre CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:49 Abbreviations and Acronyms Used BARD: beyond a reasonable doubt BoD: benefit of the doubt BoP: burden of proof CACE: clear and convincing evidence guilt m : material guilt guilt p : probatory guilt innocence m : material innocence innocence p : probatory innocence m: ratio of true acquittals to false convictions n: ratio of false acquittals to false convictions PI: presumption of innocence PoE: preponderance of the evidence SoP: standard of proof xv P1: JZZ 0521861667pre CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:49 xvi P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 1 Thinking about Error in the Law We need hardly say that we have no wish to lessen the fairness of criminal trials. But it must be clear what fairness means in this connection. It means, or ought to mean, that the law should be such as will secure as far as possible that the result of the trial is the right one. – Criminal Law Revision Committee 1 Underlying the question of guilt or innocence is an objective truth: the defendant, in fact, did or did not commit the acts constituting the crime charged. From the time an accused is first suspected to the time the decision on guilt or innocence is made, our criminal justice systemis designed to enable the trier of fact to discover the truth according to law. – Justice Lewis Powell 2 A Road Map If we look closely at the criminal justice system in the United States (or almost anywhere else for that matter), it soon becomes evident that there are three dis- tinct families of basic aims or values driving such systems. One of these core aims is to find out the truth about a crime and thus avoid false verdicts, what I will call the goal of error reduction. A second is premised on the recognition that, however much one tries to avoid them, errors will occur from time to time. This goal addresses the question of which sort of error, a false acquittal or a false conviction, is more serious, and thus more earnestly to be avoided. In short, the worry here is with howthe errors distribute themselves. Since virtually everyone agrees that convicting an innocent person is a more costly mistake than acquit- ting a guilty one, a whole body of doctrine and practices has grown up in the common lawabout howto conduct trials so as to make it more likely that, when 1 Criminal Law Revision Committee, Eleventh Report, Evidence (General) 1972, Cmnd. 4991, at §§62–4. 2 From Powell’s dissent in Bullington v. Missouri, 451 U.S. 430 (1981). 1 P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 2 thinking about error in the law an error does occur, it will be a false acquittal rather than a false conviction. For obvious reasons, I will say that this set of issues directs itself to the question of error distribution. The third set of values driving any legal systemis a more mis- cellaneous grabbagof concerns that donot explicitlyaddress trial error but focus instead on other issues important to the criminal justice system. At stake here are questions about the efficient use of resources, the protection of the rights of those accused of a crime, and various other social goods, such as the sanctity of mar- riage (spouses cannot be made to testify against one another) or preserving good relations with other nations (diplomats cannot generally be convicted of crimes, however inculpatory the evidence). I will call these nonepistemic policy values. Such concerns will figure here because, although not grounded in the truth- seeking project, their implementation frequently conflicts with the search for the truth. Judges and legal scholars have insisted repeatedly and emphatically that the most fundamental of these values is the first: that of finding out whether an alleged crime actually occurred and, if so, who committed it. The U.S. Supreme Court put the point concisely in 1966: “The basic purpose of a trial is the determination of the truth.” 3 Without ascertaining the facts about a crime, it is impossible to achieve justice, since a just resolution crucially depends on correctly figuring out who did what to whom. Truth, while no guarantee of justice, is an essential precondition for it. Public legitimacy, as much as justice, demands accuracy in verdicts. A criminal justice system that was frequently seen to convict the innocent and to acquit the guilty would fail to win the respect of, and obedience from, those it governed. It thus seems fair to say that, whatever else it is, a criminal trial is first and foremost an epistemic engine, a tool for ferreting out the truth from what will often initially be a confusing array of clues and indicators. To say that we are committed to error reduction in trials is just another way of saying that we are earnest about seeking the truth. If that is so, then it is entirely fitting to ask whether the procedures and rules that govern a trial are genuinely truth-conducive. The effort to answer that question constitutes what, in the subtitle of this book, I have called “legal epistemology.” Applied epistemology in general is the study of whether systems of investigation that purport to be seeking the truth are well engineered to lead to true beliefs about the world. Theorists of knowledge, as epistemologists are sometimes known, routinely examine truth- seeking practices like science and mathematics to find out whether they are capable of delivering the goods they seek. Legal epistemology, by contrast, scarcely exists as a recognized area of inquiry. Despite the nearly universal acceptance of the premise that a criminal 3 Tehan v. U.S., 383 U.S. 406, at 416 (1966). P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 Thinking about Error in the Law 3 trial is a search for the truth about a crime, considerable uncertainty and confu- sion reign about whether the multiple rules of proof, evidence, and legal proce- dure that encumber a trial enhance or thwart the discovery of the truth. Worse, there has been precious little systematic study into the question of whether exist- ing rules could be changed to enhance the likelihood that true verdicts would ensue. Legal epistemology, properly conceived, involves both a) the descrip- tive project of determining which existing rules promote and which thwart truth seeking and b) the normative one of proposing changes in existing rules to elim- inate or modify those rules that turn out to be serious obstacles to finding the truth. The realization of a legal epistemology is made vastly more difficult because, as just noted, nonepistemic values are prominently in play as well as epistemic ones. In many but not all cases, these nonepistemic values clash with epistemic ones. Consider a vivid example. If we were serious about error reduction, and if we likewise recognized that juries sometimes reach wrong verdicts, then the obvious remedy would be to put in place a system of judicial review permitting appeals of both acquittals and convictions. We have the latter, of course, but not the former. Every erroneous acquittal eludes detection because it escapes review. The absence of a mechanism for appealing acquittals is patently not driven by a concern to find the truth; on the contrary, such an asymmetry guarantees far more errors than are necessary. The justification for disallowing appeal of acquittals hinges on a policy value. Double jeopardy, as it is known, guarantees that no citizen can be tried twice for the same crime. Permitting the appeal of an acquittal, with the possibility that the appeal would be reversed and a new trial ordered, runs afoul of the right not to be tried more than once. So, we reach a crossroads, seemingly faced with having to choose between reducing errors and respecting traditional rights of defendants. How might we think through the resolution of conflicts between values as basic as these two are? Need we assume that rights always trump the search for the truth, or vice versa? Or, is there some mechanism for accommodating both sorts of concerns? Such questions, too, must form a core part of the agenda of legal epistemology. This book is a first stab at laying out such an agenda. In this chapter, I formulate as clearly as I can what it means to speak of legal errors. Absent a grasp of what those errors are, we obviously cannot begin to think about strategies for their reduction. In Chapters 2 through 4, we examine in detail a host of important questions about error distribution. Chapters 5 through 8 focus on existing rules of evidence and procedure that appear to pose serious obstacles to truth seeking. Those chapters include both critiques of existing rules and numerous suggestions for fixing such flaws as I can identify. The final chapter assays some possible solutions to the vexatious problems generated by the tensions between epistemic values and nonepistemic ones. P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 4 thinking about error in the law A Book as Thought Experiment The two passages in the epigraph to this chapter from Supreme Court Justice Lewis Powell and England’s Criminal LawRevision Committee articulate a fine and noble aspiration: finding out the truth about the guilt or innocence of those suspected of committing crimes. Yet, if read as a description of the current state of American justice, they remain more an aspiration than a reality. In saying this, I do not mean simply that injustices, false verdicts, occur from time to time. Occasional mistakes are inevitable, and thus tolerable, in any form of human inquiry. I mean, rather, that many of the rules and procedures regulating criminal trials in the United States – rules for the most part purportedly designed to aid the truth-finding process – are themselves the cause of many incorrect verdicts. I mean, too, that the standard of proof relevant to criminal cases, beyond reasonable doubt, is abysmally unclear to all those – jurors, judges, and attorneys – whose task is to see that those standards are honored. In the chapters that follow, I will show that the criminal justice system now in place in the United States is not a systemthat anyone concerned principally with finding the truth would have deliberately designed. 4 Anatural way to test that hypothesis would be to examine these rules, one by one, to single out those that thwart truth seeking. And, in the chapters to follow, I will be doing a fair share of precisely that. But, as we will discover, it is often harder than it might seem to figure out whether a given evidential practice or procedure is truth promoting or truth thwarting. In short, we need some guide- lines or rules of thumb for deciding whether any given legal procedure furthers or hinders epistemic ends. Moreover, for purposes of analysis, we need to be able to leave temporarily to one side questions about the role of nonepistemic values in the administration of justice. We will have to act as if truth finding were the predominant concern in any criminal proceeding. In real life, of course, that is doubtful. As I noted at the outset, criminal trials are driven by a host of extra-epistemic values, ranging from concerns about the rights of the defendant to questions of efficiency and timeliness. (Not for nothing do we insist that justice delayed is justice denied.) The prevailing tendency among legal writers is to consider all these values – epistemic and nonepistemic – as bundled together. This, I think, 4 Lest you take my remarks about the lack of a coherent design in the rules of trials as casting aspersions on the founding fathers, I hasten to add that the system now in place is one that they would scarcely recognize, if they recognized it at all. Many of the features of American criminal justice that work against the interests of finding truth and avoiding error– features that we will discuss in detail later on – were additions, supplements, or sometimes patent transformations of American criminal practice as it existed at the beginning of the nineteenth century. Congress or state legislatures imposed some of these changes; judges themselves created the vast majority as remedies for serious problems posed by the common law or abusive police practices. A few date from the late-nineteenth century; most, from the twentieth. P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 Thinking about Error in the Law 5 can produce nothing but confusion. Instead of the familiar form of analysis, which juggles all these values in midair at the same time, I am going to propose a thought experiment. I will suggest that we focus initially entirely on questions of truth seeking and error avoidance. I will try to figure out what sorts of rules of evidence and procedure we might put in place to meet those ends and will identify when existing rules fail to promote epistemic ends. Then, with that analysis in hand, we can turn to compare the current system of evidence rules and procedures with a systemthat is, as it were, epistemically optimal. When we note, as we will repeatedly, discrepancies between the kind of rules we would have if truth seeking were really the basic value and those rules we find actually in place, we will be able then to ask ourselves whether these epistemically shaky rules conduce to values other than truthseeking and, if they do, when and whether those other values should prevail over more epistemically robust ones. Although I ignore such values in the first stage of the analysis, I do not mean for a moment to suggest that they are unimportant or that they can be ignored in the final analysis. But if we are to get a handle on the core epistemic issues that are at stake in a criminal trial, it is best – at the outset – to set them to one side temporarily. If it seems madcap to try to understand the legal system by ignoring what everyone concedes to be some of its key values, I remind you that this method of conceptual abstraction and oversimplification has proved its value in other areas of intellectual activity, despite the fact that every oversimplification is a falsification of the complexities of the real world. Consider what is perhaps the best-known example of the power of this way of proceeding: During the early days of what came to be known as the scientific revolution, Galileo set out to solve a conundrum that had troubled natural philosophers for almost two millennia, to wit, how heavy bodies fall. Everyone vaguely understood that the velocity of fall was the result of several factors. The shape of a body makes a difference: A flat piece of paper falls more slowly than one wadded into a ball. The medium through which a body is falling likewise makes a crucial difference: Heavy bodies fall much faster through air than they do through water or oil. Earlier theories of free fall had identified this resistance of the medium as the key causal factor in determining the velocity of fall. Galileo’s strategy was to turn that natural assumption on its head. Let us, he reasoned, ignore the shapes of bodies and their weights and the properties of the media through which they fall – obvious facts all. Assume, he suggested, that the only relevant thing to knowis howpowerfully bodies are drawn to the earth by virtue of what we would nowcall the gravitational field in which they find themselves. By making this stark simplification of the situation, Galileo was able to develop the first coherent account of fall, still known to high school students as Galileo’s Law. Having formulated a model of how bodies would fall if the resistance of the medium were negligible (which it is not) and the shape of the body were irrelevant (which it likewise is not), and the weight of a body were irrelevant P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 6 thinking about error in the law (which it is), Galileo proceeded to reinsert these factors back into the story in order to explain real-world phenomena – something that would have been impossible had he not initially ignored these real-world constraints. The power of a model of this sort is not that it gets things right the first time around, but that, having established how things would go under limited and well-defined conditions, we can then introduce further complexities as necessary, without abandoning the core insights offered by the initial abstraction. I have a similar thought experiment in mind for the law. Taking the Supreme Court at its word when it says that the principal function of a criminal trial is to find out the truth, I want to figure out how we might conduct criminal trials supposing that their predominant aim were to find out the truth about a crime. Where we find discrepancies between real-world criminal procedures and epistemically ideal ones (and they will be legion), we will need to ask our- selves whether the epistemic costs exacted by current real-world procedures are sufficiently outweighed by benefits of efficiency or the protection of defendant rights to justify the continuation of current practices. Those will not be easy issues to resolve, involving as they do a weighing of values often considered incommensurable. But such questions cannot even be properly posed, let alone resolved, until we have become much clearer than we now are about which features of the current legal regime pose obstacles to truth seeking and which do not. Because current American jurisprudence tends to the viewthat rights almost invariably trump questions of finding out the truth (when those two concerns are in conflict), there has been far less discussion than is healthy about whether certain common legal practices – whether mandated by common lawtraditions or by the U.S. Constitution or devised as court-designed remedies for police abuses – are intrinsically truth thwarting. My object in designing this thought experiment is to open up conceptual space for candidly discussing such questions without immediately butting up against the purported argument stopper: “but X is a right” or “X is required (or prohibited) by the Constitution.” Just as Galileo insisted that he wouldn’t talk about the resistance of the air until he had understood how bodies would fall absent resistance, I will try – until we have on the table a model of what a disinterested pursuit of the truth in criminal affairs would look like – to adhere to the view that the less said about rights, legal traditions, and constitutional law, the better. I said that this thought experiment will involve figuring out how criminal trials could be conducted, supposing that true verdicts were the principal aimof such proceedings. This might suggest to the wary reader that I intend to lay out a full set of rules and procedures for conducting trials, starting from epistemic scratch, as it were. That is not quite the project I have in mind here, since it is clear that there is a multiplicity of different and divergent ways of searching for the truth, which (I hasten to add) is not the same thing as saying that there are multiple, divergent truths to be found. Consider one among many questions P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 Thinking about Error in the Law 7 that might face us: If our aim is to maximize the likelihood of finding the truth, should we have trial by judge or trial by jury? I do not believe that there is a correct answer to that question since it is perfectly conceivable that we could design sets of procedures that would enable either a judge or a jury to reach verdicts that were true most of the time. English speakers have a fondness for trial by jury, whereas Roman law countries prefer trial by judge or by a mixed panel of judges and jurors. For my part, I can see no overwhelming epistemic rationale for a preference for one model over the other. If we Anglo-Saxons have any rational basis, besides familiarity, for preferring trial by jury, it has more to do with the political and social virtues of a trial by one’s peers rather than with any hard evidence that juries’ verdicts are more likely to be correct than judges’ verdicts are. To begin with, I intend to propose a series of guidelines that will tell us what we should look for in deciding whether any particular arrangement of rules of evidence and procedure is epistemically desirable. This way of proceeding does not directly generate a structure of rules and procedures for conducting trials. What it will do is tell us how to evaluate bits and pieces of any pro- posed structure with respect to their epistemic bona fides. It will set hurdles or standards for judging any acceptable rule of evidence or procedure. If you want an analogy, think of how the rules of proof in mathematics work. Those rules do not generally generate proofs by some sort of formal algorithm; bright mathematicians must do that for themselves. What the rules of proof do (except in very special circumstances) is enable mathematicians to figure out whether a purported proof is a real proof. In effect, what I will be suggesting is a set of meta-rules or meta-principles that will function as yardsticks for figuring out whether any given procedure or evidence-admitting or evidence-excluding practice does, in fact, further epistemic ends or whether it thwarts them. What I am proposing, then, is, in part, a meta-epistemology of the criminal law, that is, a body of principles that will enable us to decide whether any given legal procedure or rule is likely to be truth-conducive and error reducing. The thought experiment I have been describing will involve submitting both real and hypothetical procedures to the scrutiny that these meta-principles can provide. When we discover rules currently in place that fail to serve epistemic ends, we will want to ask ourselves whether they cannot be replaced by rules more conducive to finding the truth and minimizing error. If we can find a more truth-conducive counterpart for truth-thwarting rules, we will then need to decide whether the values that the original rules serve (for instance, protecting certain rights of the accused) are sufficiently fundamental that they should be allowed to prevail over truth seeking. If, as Justice Powell says in the epigraph, the system“is designed” to discover the truth, you might reasonably have expected that we already knowa great deal about the relation of each of its component parts to that grand ambition. The harsh reality is that we know much less than we sometimes think we do. Many P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 8 thinking about error in the law legal experts and appellate judges, as we will see on numerous occasions in later chapters, continue to act and write as if certain portions of the justice system that actually thwart truth seeking have an epistemic rationale. Still worse, some jurists and legal scholars attribute error-reducing power to rules and doctrines that, viewed dispassionately, produce abundant false verdicts in their own right. Like Powell, they pay lip service to the mantra that the central goal of the system is to get at the truth, all the while endorsing old rules, or putting in place new ones, that hobble the capacity of that system to generate correct verdicts. So long as jurists believe, as many now do, that certain judicial rules (for instance, the suppression of “coerced” confessions 5 ) promote truth finding – when in fact they do the opposite – there can be nothing but confusion concerning when and if truth seeking is being furthered. One important reason that we know so much less than we should is that the courts in particular, but also the justice system in general, tend to discourage the sort of empirical research that would enable us to settle such questions definitively. In philosophy, my biases lean in the direction of naturalism. That means that I believe that most philosophical issues ultimately hinge on finding out what the facts are. I believe, further, that our methods of inquiry must be constantly reviewed empirically to see whether they are achieving what we expect of them. In writing this book, I have been constantly frustrated by the paucity of empirical information that would allow us to reach clear conclusions about howwell or badly our legal methods are working. Where there are reliable empirical studies with a bearing on the issues addressed here, I will make use of them. Unfortunately, given the dearth of hard evidence, the analysis in this book will fall back on armchair hunches about the likely effects of various rules and procedures far more often than I would have liked. My defense for doing so is simply that one must fight one’s battles with the weapons that one has at hand. I should stress, as well, that I approach these questions as a philosopher, lookingat the lawfromthe outside, rather thanas anattorney, workingwithinthe system. Although I have thought seriously about these issues over several years, I cannot possibly bring to them the competences and sensibilities of a working trial lawyer. 6 What interests me about the law is the way in which it functions, or malfunctions, theoretically, as a system for finding truth and avoiding error. In this role, I amless concerned than a civil libertarian or defense attorney might be with the rights of the accused and more concerned with how effectively the criminal justice systemproduces true verdicts. The analysis offered in this book 5 To see the point of the scare quotes, consult Chapter 7, where we will observe that the majority of “coerced” confessions are not coerced in the lay sense of that term. 6 Accordingly, I ask those readers who know the fine points of the practice of the law far better than I do to overlook the occasional acts of ignorance on my part, of which there are doubtless several, unless they actually impinge upon the cogency of the argument that I am making. P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 Thinking about Error in the Law 9 does not purport to tell juries and judges how to decide a case; such dreadful decisions must depend on the case’s special circumstances and its nuances. Its aim, rather, is the more prophylactic one of pointing out some errors that these fact finders should avoid in the always difficult quest for a true and just verdict. There will be readers who expect any avowedly philosophical treatment of the law to center on issues of morality and rights or on questions about the authority and essence of the law. Such are the themes that have dominated the philosophy of law in the last half-century. The most influential philoso- pher of law in the English-speaking world in the twentieth century, H. L. A. Hart, managed to write a lengthy, splendid book on the philosophy of law (The Concept of Law, 1961) that says virtually nothing about what I am calling legal epistemology. His eminent continental counterpart, Hans Kelsen, did virtually the same thing a generation earlier in his Pure Theory of Law (1934). Readers expecting a similar agenda from me will be sorely disappointed. To them in particular, I say this: If it is legitimate and fruitful for moral philosophers, such as Gerald Dworkin or John Rawls, to focus on the lawprincipally as an exercise in ethics and morality, while largely ignoring the importance of truth seeking in the law (which they famously do), it is surely just as appropriate to look at the law through the lenses of epistemology and the theory of knowledge. Although one is not apt to learn so by looking at the existing philosophical literature on the subject, it is indisputable that the aims of the law, particularly the criminal law, are tied to epistemic concerns at least as profoundly as they are to moral and political ones. This book is a deliberate shot across the bow of the juggernaut that supposes that all or most of the interesting philosophi- cal puzzles about the law concern its moral foundations or the sources of its authority. Principal Types of Error In this initial chapter, I will to begin to lay out some of the analytic tools that we will need in order to grapple with some thorny problems in the theory and practice of the criminal law. As its title already makes clear, this book is largely about legal errors. Since treating the law as an exercise in epistemology inevitably means that we will be involved in diagnosing the causes of error, we need to be clear from the outset about what kinds of errors can occur in a criminal proceeding. Since our concern will be with purely epistemic errors, I should say straight- away that I am not using the term “error” as appellate courts are apt to use it. For them, an “error” occurs in a trial just in case some rule of evidence or procedure has been violated, misinterpreted, or misapplied. Thus, a higher court may determine that an error occurred when a trial judge permitted the introduction of evidence that the prevailing rules should have excluded or when some constitutional right of the defendant was violated. Courts will find that P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 10 thinking about error in the law an error occurred if a judge, in his instructions to the jury about the law, made some serious mistake or other, in the sense of characterizing the relevant law in a way that higher courts find misleading or incorrect. Very occasionally, they will decide that an error occurred if the jury convicted someone when the case against the defendant failed to meet the standard of proof beyond a reasonable doubt. 7 By contrast, I will be using the term “error” in a more strictly logical and epistemic sense. When I say that an error has occurred, I will mean either a) that, in a case that has reached the trial stage and gone to a verdict, the verdict is false, or b) that, in a case that does not progress that far, a guilty party has escaped trial or an innocent person has pleaded guilty and the courts have accepted that plea. In short, for the purposes of our discussion, an error occurs when an innocent person is deemed guilty or when a guilty person fails to be found guilty. For obvious reasons, I will call the first sort of error a false inculpatory finding and the second a false exculpatory finding. There are two important points to note about the way in which I am defining legal errors: First, errors, in my sense, have nothing to do with whether the system fol- lowedthe rules (the sense of “error” relevant for appellate courts) andeverything to do with whether judicial outcomes convict the guilty and free the innocent. Even if no errors of the procedural sort that worries appellate courts have occurred, an outcome may be erroneous if it ends up freeing the guilty or con- victing the innocent. The fact that a trial has scrupulously followed the letter of the current rules governing the admissibility of evidence and procedures – and thus avoids being slapped down by appellate courts for breaking the rules – is no guarantee of a correct outcome. To the contrary, given that many of the current rules (as we will see in detail in later chapters) are actually conducive to mistaken verdicts, it may well happen that trials that follow the rules are more apt to produce erroneous verdicts than trials that break some of them. Accordingly, our judgment that an error has occurred in a criminal case will have nothing to do with whether the judicial system followed its own rules and everything to do with whether the truly guilty and the truly innocent were correctly identified. Second, standard discussions of error in the law – even from those authors who, like me, emphasize truth and falsity rather than rule following or rule breaking – tend to define errors only for those cases that reach trial and issue in a verdict. Such authors, naturally enough, distinguish between true and false verdicts. That is surely a legitimate, and an important, distinction, but it is 7 Courts typically distinguish between errors that, while acknowledged as errors, did not decisively affect the outcome of a trial (called “harmless errors”) and more serious errors, which call for retrial or reversal of a conviction. P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 Thinking about Error in the Law 11 neither the most general nor the most useful way of distinguishing errors. As my definition of “error” has already indicated, I claim that errors occur whenever the innocent are condemned by the system and whenever the guilty fail to be condemned. Obviously, one way in which these mistakes can happen is with a false conviction or a false acquittal. But what are we to say of the guilty person who has been arrested and charged with a crime that he truly committed but against whom charges were subsequently dropped by the prosecutor or dismissed by the judge? These are mistakes just as surely as a false acquittal is. Likewise, if an innocent person – faced with a powerfully inculpatory case – decides to accept a plea bargain and plead guilty, this is an error of the system just as much as a false conviction is, even though the case against the accused is never heard and a jury never renders a verdict. Clearly, this analysis rests on being able to speak about the truly guilty and the truly innocent. Much nonsense has been creeping of late into several discussions, both popular and academic, of the law. For instance, one often hears it said (in a gross misconstrual of the famous principle of the presump- tion of innocence) that the accused “is innocent until proven guilty,” as if the pronouncing of the verdict somehow created the facts of the crime. If it were correct that only a guilty verdict or guilty plea could render someone guilty, then there could be no false acquittals, for it would make no sense to say, as the phrase “false acquittal” implies, that a jury acquitted someone who is actually guilty. Since such locutions make perfect sense, we must reject the notion that a verdict somehow creates guilt and innocence. A second obstacle to talking clearheadedly about guilt and innocence arises from the novel but fashionable tendency to suppose that whether someone is guilty or innocent of a crime simply depends on whether the evidence offered at trial is sufficient to persuade a rational person that the defendant is guilty. The confusion here is more subtle than the former one. It is rooted in the obvious fact that the decision about guilt or innocence made by a reasonable trier of fact will necessarily depend on what he or she comes to learn about the alleged crime. On this view, a verdict is correct so long as it squares with the evidence presented at trial, without making reference to anything that happened in the real world outside the courtroom. One legal scholar, Henry Chambers, has claimed that “what is true is what the [trial] evidence indicates is true.” 8 Contrary to Chambers, I claim that nothing that a judge or jury later determines to be the case changes any facts about the crime. Likewise, I claim that, while what is presented in evidence surely shapes the jury’s verdict, that evidence does not define what is true and false about the crime. Unless this were so, it would again make no sense to talk of a true or a false verdict, so long as that verdict 8 Henry Chambers, Reasonable Certainty and Reasonable Doubt, 81 Marq. L. Rev. 655, at 668 (1998). P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 12 thinking about error in the law represented a reasonable inference from the evidence. Yet, sometimes we come to the conclusion that the evidence presented at trial was deeply unrepresentative of the true facts of the crime. Sometimes, truly innocent people are wrongly convicted and truly guilty people are wrongly acquitted, even though the jury drewthe conclusions that were appropriate fromthe evidence available to them. (Basically, Chambers confuses what I will be calling the validity of a verdict with its truth.) I will be adamant in insisting that the presumption of innocence, properly understood, does not make a guilty person innocent nor an acquittal of such a person into a nonerror. Likewise, I will argue that verdicts don’t make the facts and neither does the evidence presented at trial; they only give official sanction to a particular hypothesis about those facts. Strictly speaking, the only people innocent are those who did not commit the crime, whatever a jury may conclude about their guilt and regardless of what the available evidence seems to show. Likewise, the truly guilty (those who committed the crime) are guilty even if a jury rationally acquits them. “Being found guilty” and “being guilty” are manifestly not the same thing; neither are “being presumed innocent” and “being innocent.” The naive argument to the effect that what we mean when we say that Jones committed the crime is that a jury would find him guilty utterly confuses questions about what is really the case with questions about judgments issued in the idiosyncratic circumstances that we call criminal trials. There are false acquittals and false convictions, and the existence of each entails that verdicts are not analytically true or self-authenticating. Because they are not, we can speak of verdicts as being erroneous, even when they result from trials that were scrupulously fair, in the sense of being in strict compliance with the rules governing such proceedings. By the same token, we can speak of outcomes or verdicts being true, even when they resulted from trials that made a mockery of the existing rules. For future reference, it will prove useful to make explicit the moral of this discussion. In brief, it is legitimate, and in some contexts essential, to distinguish between the assertion that “Jones is guilty,” in the sense that he committed the crime, and the assertion that “Jones is guilty,” in the sense that the legal system has condemned him. I propose to call the first sense material guilt (hereinafter, guilt m ) and the second probatory guilt (guilt p ). Clearly, guilt m does not imply guilt p , nor vice versa. Similarly, we can distinguish between Jones’s material innocence (inno- cence m ), meaning he did not commit the crime, and his probatory innocence (innocence p ), meaning he was acquitted or otherwise released from judicial scrutiny. Again, neither judgment implies the other. With these four simple distinctions in hand, we can combine themin various useful ways. For instance, Jones canbe guilty m but innocent p ; again, he canbe innocent m but guilty p . Either of these situations would represent an error by the system. P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 Thinking about Error in the Law 13 Other Relevant Distinctions among Error Types The most basic distinction we need has already been mentioned: that between false inculpatory and false exculpatory findings. These two types of findings are just what one would expect: A false exculpatory finding occurs when the legal system fails to convict a truly guilty felon. A false inculpatory finding is a conviction of an innocent person. Still, we need to add a couple of other important distinctions to the tool kit of error types. One involves separating valid from invalid verdicts. A verdict of guilty will be valid, as I propose to use that term, provided that the evidence presented at trial establishes, to the relevant standard of proof, that the accused person committed the crime in question. Otherwise, a guilty verdict is invalid. Naturally enough, an acquittal will be valid as long as the conditions for a valid conviction are not satisfied and invalid otherwise. The notion of validity aims to capture something important about the quality of the inferences made by the trier of fact, whether judge or jury. Invalid verdicts can occur in one or both of two ways: a) The trier of fact may give more or less weight to an item of evidence than it genuinely merits, or b) she may misconceive the height of the standard of proof. In either case, the verdict is inferentially flawed. It is crucial to see that the valid/invalid distinction does not map neatly onto the true/false verdict dichotomy. We settle the truth of a verdict (or what I am calling a finding) by comparing it with the facts. That is, Jones’s conviction is true just in case Jones committed the crime. By contrast, we settle the valid- ity of a verdict by comparing it with the evidence presented at trial, asking whether that evidence meets the applicable standard of proof. Just as a deduc- tive inference can be valid even when its conclusion is false (all horses can fly; all stallions are horses; therefore, all stallions can fly), so a verdict can be simultaneously valid and false. Using the terminology of the previous section, it can be a valid verdict that Jones is guilty p , even while it is true that Jones is innocent m . Bythe same token, a verdict of not guiltymaybe validevenif Jones is guilty m . Happily, it sometimes turns out that true verdicts are likewise valid ones and that false verdicts are invalid. But neither of these connections is solid. Sometimes, perhaps often, a jury will produce a valid verdict that is false, that is to say, a verdict that reflects an appropriate inference from the evidence presented at trial but that is factually false. This can occur when the evidence admitted at trial, skewed for whatever reasons, invites a conclusion at odds with what actually happened. But even when the evidence is not skewed or unrepresentative of the crime, there is still plenty of scope for a verdict that is valid but not true. Indeed, the standard of proof guarantees as much. Suppose, for the sake of argument, that the standard of proof is something like 95 percent confidence in guilt. A jury hears a case and concludes that it is 80 percent P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 14 thinking about error in the law likely that the accused committed the crime. Now, the jury, if it acquits, will be producing a valid verdict, for the rules of proof demand acquittal even when the likelihood of guilt is as high as 80 percent. But that valid verdict is likely to be a false acquittal since, by hypothesis, the likelihood that the defendant committed the crime is quite high. Likewise, it is easy to conceive how a jury might produce an invalid verdict that was nonetheless true, although these are apt to be less frequent than cases of valid verdicts that are false. What one hopes to achieve, obviously, is a verdict that is both true and valid. We want jurors to convict and acquit the right people and to do so for the right reasons. Both lack of truth and lack of validity will, as I am using the term “error,” represent serious errors of the system, even though they point to quite different ways in which the systemhas failed. In our efforts to identify the principal sources of error in the legal system, we will be examining rules of evidence and procedure with a view to asking how such rules threaten either the truth or the validity of verdicts. If the outcome of a criminal proceedingis erroneous ineither of these respects –that is tosay, if it is either false or invalid(or both) –the systemhas failed. If one or the other or both types of failure happen frequently, it may be time to change those parts of the system responsible for such errors. In later chapters, we will see that certain practices entrenched in our rules of evidence and procedure tend to produce invalid convictions and acquittals, that is to say, verdicts at odds with what a reasonable person – not bound by those rules – would conclude fromthe evidence available. Other features of the system, by restricting what can count as legal evidence, tend to produce verdicts that, even if valid, are false. The true/false and valid/invalid distinctions reflect the two primary ways in which a trial verdict may go awry: an inadequate (in the sense of unrepresentative) evidence base or faulty inferences from that base. There is a third dichotomy that will prove helpful in thinking about sources of error. It distinguishes those erroneous decisions that are reversible fromthose that are irreversible. For instance, when Schwartz is convicted of a crime, he can appeal the verdict and may persuade a higher court to set that verdict aside. Epistemically, such a reviewmechanismis invaluable as a way of increasing the likelihood that the final result is correct. By contrast, if Schwartz is acquitted, the verdict cannot be appealed, however flawed may have been the reasoning that led the jury to acquit. Other things being equal, irreversible decisions are more troubling sources of error than reversible ones for the obvious reason that there is no machinery for catching and correcting the former while the latter can, in principle, be discovered and rectified. In due course, we will inquire into the rationale for creating a category of decisions, including verdicts themselves, that is wholly immunized from further review and correction. Thus far, our focus on error has been principally with the terminal stage, that is, with erroneous verdicts. But many criminal investigations never get as far as this. Sometimes, police investigations simply run out of steam because P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 Thinking about Error in the Law 15 of lack of clues or bad investigative practices. Although these are errors just as surely as a false acquittal is, they will not be our focus. What will command our attention are those felons who slip through the system, not for lack of incriminatingclues knowntothe police, but whoescape trial because of the ways in which the rules of evidence and procedure impede further pursuit of the case against them. These errors will be as revealing a topic of study as false verdicts are. We need to remind ourselves that a vast number of criminal investigations (probably the overwhelming majority of police inquiries) never reach the trial stage because, although the police have identified a suspect to their own sat- isfaction, someone or other in authority concludes that the case against him is too weak to take to trial. It may be the police themselves who make this deter- mination or it may be the prosecutor. It can be a grand jury that issues a “no bill,” precluding trial. Or it may be an arraigning judge who dismisses the case. At each of these stages, where a decision must be made whether to proceed along the route to trial or not, the participants are bound by an elaborate body of rules of evidence and procedure. Prosecutors who have in hand a confession know that it may be tossed out if there are doubts about its provenance. Similar questions may arise about much of the other evidence seized by police. Even when prosecutors have powerful evidence of a suspect’s guilt, their decision to proceed to trial must be informed by a calculation on their part as to which parts of the evidence they now have in hand will actually be allowed to go before a jury. If there are rules of admissibility that exclude relevant evidence (and much of this book will address itself to rules of precisely this sort), then those rules will exert a weighty influence not only during the trial itself but on all the preliminary decisions about whether to proceed to trial. Even if we leave aside problems generated by the rules of evidence, the standard of proof likewise works to ensure that many parties who are probably guilty never go to trial. Specifically, prosecutors may believe that the evidence against a suspect strongly suggests that he is guilty but that such evidence would probably be insufficient to persuade a jury of his guilt beyond a reasonable doubt. Short on both financial and human resources, prosecutors are unlikely to proceed with such a case. Judge Richard Posner has put the point succinctly: Tight [prosecutorial] screening implies that some, perhaps many, guilty people are not prosecuted and that most people who are prosecuted and acquitted are actually guilty. 9 It puts the importance of this class of problems into vivid perspective if we remind ourselves that there are far more dismissals than acquittals in the criminal justice system. In federal courts in 1999, for instance, there were about 9 Richard Posner, An Economic Approach to the Law of Evidence, 51 Stan. L. Rev. 1477, at 1506 (1999). P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 16 thinking about error in the law eight judge-ordered dismissals for every acquittal. 10 Those writers who focus on the problem of error as if it principally arose in the process of a jury trial itself ignore such numbers at their peril. This is another way of saying that every year hundreds of thousands of suspects are de facto “acquitted” by prosecutors, judges, and grand juries – without ever going to trial. That is as it should be, since many suspects are surely innocent. Dismissal of charges against an innocent person is not a failure of the systembut a success. Afailure occurs, in this context, when a guilty suspect has the case against himdropped prior to trial because relevant evidence of his guilt, although in hand, is thought likely not to be admissible if trial were to ensue. Of the three hundred thousand persons suspected of felonies each year – against whom charges are dropped or dismissed before trial – there is every reason to suspect that a certain proportion of these people are guilty. How large that proportion of failures is cannot be ascertained with confidence since the relevant data are inaccessible; instead, our analysis in this book will attempt to determine weak points in the system that may make such false, pretrial “acquittals” much more common than they need be. A different, more diachronic, way of thinking about the various ways in which failures can occur emerges fromimagining a series of filters that mediate between the crime, at the one extreme, and the jury’s verdict, at the other. There is, to begin with, the crime itself. Jones, let us suppose, mugged Smith and stole his wallet. That event is now past. What survive are traces or remnants of the crime. These include memories of the participants and eyewitnesses and physical evidence of the crime (Jones’s fingerprints on Smith’s wallet, contusions on Jones’s face, and so on). The police will come to find some, but rarely all, of these traces. If they and the prosecutor decide that they have a solid case against Jones, they will next have to persuade a judge or a grand jury that the case is strong enough to go forward. Supposing that all these hurdles have been leapt, the prosecutor will now choose from among the traces known to the police a subset that he intends to enter as evidence at the trial. Jones’s attorney will make a similar decision. At Jones’s pretrial evidentiary hearing, a judge will decide which of these submitted traces can be revealed to the jury. Once the evidence questions are settled, the judge may wrongly decide to dismiss the charges against the accused. Once the trial begins, if it gets that far, the now heavily filtered evidence will be presented and subjected to cross-examination. Once both sides have had their say, the judge will instruct the jury about the relevant law that Jones is alleged to have broken and on the threshold of proof that they should use in deciding whether to convict Jones. 10 See the Department of Justice, Sourcebook of Criminal Justice Statistics for 1999, Table 5.16. For a thorough discussion of this issue, see Samuel R. Gross, The Risks of Death: Why Erroneous Convictions Are Common in Capital Cases, 44 Buffalo L. Rev. 469 (1996). P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 Thinking about Error in the Law 17 There is, obviously, ample scope for error at each of these stages. Important evidence, either inculpatory or exculpatory, may elude the best efforts of pros- ecution and defense to find it. The prosecution may find exculpatory evidence but suppress it, and the defense may be aware of inculpatory evidence that it “forgets” to mention. The judge’s rulings on the admissibility of evidence sub- mitted for trial may end up including evidence that is likely to mislead the jury or excluding evidence that the jury should hear. The grand jury may err in their decision to indict. The defendant may refuse to testify or subpoenaed witnesses with important information may disappear. Witnesses with relevant evidence might not be called because both prosecution and defense fear that their testi- mony would undermine their respective cases. The judge may misinstruct the jury with respect to the relevant lawor botch the instructions about the standard of proof – which occurs more often than you might imagine. (For details, see the next chapter.) Even if all else goes properly, the jury may draw inappro- priate inferences about guilt or innocence from the evidence before them or they may misunderstand the level of proof required for a conviction. Even once the verdict is pronounced, the room for error has not disappeared. If the jury voted to convict, the accused may file an appeal. Appellate courts may refuse to hear it, even when the defendant is innocent. Or, they may take the appeal but reverse it when the verdict is sound or endorse the verdict when it is false. If the defendant is acquitted, double jeopardy precludes further appeal, even if the trial was riddled with acquittal-enhancing errors. Eliminating all these possible sources of error (and I have mentioned only the more obvious) is clearly impossible. The aim of the justice system, realistically construed, should be to attempt to reduce them as far as possible. Current evidential practice in the criminal law, as we will see, often fails to do that. Worse, it frequently increases the likelihood of error deliberately by adopting rules and procedures that prevent the jury fromlearning highly important things about the crime. Relevance versus Admissibility The charge that I have just made can be put in slightly more technical terms, and it will probablybe useful todoso. Inall reasoningabout humanaffairs (andother contingent events), there are two key concepts regarding evidence that must be grasped. One is called credibility or sometimes (as in the law) reliability. As the term suggests, a piece of evidence or testimony is credible when there is reason to believe it to be true or at least plausible. The other pertinent concept is known, in both the law and in common sense, as evidential relevance. The core idea is that a piece of information is relevant to the evaluation of a hypothesis just in case, if credible, it makes that hypothesis more or less probable than it was before. If a certain bit of information, even when credible, would not alter our confidence in a hypothesis one way or the other, we deemit irrelevant to that P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 18 thinking about error in the law hypothesis. In the criminal law, there are always two key hypotheses in play: a) A crime was committed and b) the defendant committed it. Any testimony or physical evidence that would make a reasonable person either more inclined or less inclined to accept either of these hypotheses is relevant. Everything else is irrelevant. Both credibility and relevance are crucial to qualify something as germane evidence. Jurors, above all others, must assess both the credibility and the rele- vance of the evidence they see and hear. For reasons having roots very deep in the common law, however, the judge in a criminal trial is generally not supposed to let judgments of credibility enter into his or her decision about the accept- ability of proffered evidence. This is because the jury, rather than the judge, is by tradition charged with determining the “facts” of the case. Deciding whether eyewitness testimony or physical evidence is credible would, in effect, imply a decision about its facticity. Since that is the province of the jury rather than the judge, the usual pattern is for judges to rule on relevance but not on reliability. This means that when judges make decisions about relevance, they are obliged to think hypothetically; that is, they must ask themselves, “if this evidence were credible, would it have a bearing on the case?” This is why, when a judge admits evidence as relevant, nothing is implied with respect to its credibility. (A sig- nificant exception to this principle occurs in decisions about the admission of expert testimony, where the judge is specifically charged to determine whether the basis for the testimony of the avowed expert is “reliable.” 11 ) American courts at every level of jurisdiction accept this notion of relevance. One of the important and legitimate gate-keeping functions of a judge is to see to it that the jury hears all and only relevant evidence. If American judges stuck resolutely to this principle, they could not be faulted on epistemic grounds since virtually all forms of sophisticated hypothesis evaluation (in science, medicine, and technology, for instance) work with this same notion of relevance. Unfortunately, however, legal texts and the practices of courts routinely flout the relevance-only principle. This is because judges have a second criterion they use, alongside the demand for relevant evidence. It is often known as the admissibility requirement. To be admissible, evidence must not only be relevant; it must also meet a variety of other demands. For instance, the evidence cannot have been acquired by a violation of the rights of the accused. The evidence cannot arise from privileged relations that the accused has had with various professionals or his spouse. The evidence generally cannot have been obtained illegally, even if its being seized violated none of the rights of the accused. The evidence cannot be such that it might inflame the passions of the jurors or unfairly cast the defendant in an unfavorable light. The evidence cannot inform 11 The Supreme Court has held that “the trial judge must ensure that any and all scientific testimony or evidence admitted is not only relevant, but reliable” (Daubert v. Merrell Dow Pharms., 113 S. Ct. 2786, at 2795 [U.S. 1993]). P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 Thinking about Error in the Law 19 the jury that the defendant withdrew a confession of guilt, nor can it refer to admissions of guilt made by the defendant during negotiations about copping a plea. The evidence generally cannot come from a witness whose testimony would be self-incriminating. The jury cannot be informed when key witnesses escaped giving testimony by claiming their Fifth Amendment rights. The jury cannot be told whether the accused cooperated with the police in their inquiries. If the accused does not offer testimony on his own behalf, the judge explicitly instructs the jury to ignore that relevant fact, rather than supposing that the accused may have something to hide. Virtually no one disputes that information of all these sorts is relevant in the technical sense, for it indubitably bears on the probability of the hypothesis that the defendant is guilty. In most jurisdictions, however, these and many other examples of admittedly relevant evidence will not be admitted during the trial. Subsequent chapters will describe many of these exclusionary principles in detail. What we should note here is that every rule that leads to the exclusion of relevant evidence is epistemically suspect. 12 It is universally agreed, outside the law courts, that decision makers can make the best and most informed decisions only if they are made aware of as much relevant evidence as possible. Excluding relevant but nonredundant evidence, for whatever reasons, decreases the likelihood that rational decision makers will reach a correct conclusion. Accordingly, we will want to examine these exclusionary principles carefully to see whether the damage they inflict on our truth-seeking interests are suitably balanced by gains of other sorts. The Case of “Unfairly Prejudicial” Evidence It might be instructive to include here one example of this distinction between relevance and admissibility in order to put some flesh on the skeleton of abstrac- tions with which we have been working. A paradigmatic example of the prob- lems we will be facing throughout the rest of the book is provided by the law’s unimpressive efforts to distinguish between evidence that is “unfairly prejudi- cial” and evidence that is not. At the preliminary hearing preceding a trial, both sides describe the evidence they intend to present at trial and argue about its admissibility. Despite the rule to the effect that the judge should generally admit relevant evidence, the law gives her enormous discretion to exclude evidence, however relevant and however inculpatory, if in her judgment that evidence is of such a sensational or inflammatory nature that ordinary jurors would be unable to assign it its 12 The only time when it is obviously appropriate to exclude relevant evidence is when it is redundant with respect to evidence already admitted. Testimony from two hundred witnesses asserting X is scarcely better than that from two or three credible ones asserting X, unless X happens to be a very bizarre event. P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 20 thinking about error in the law true weight. Specifically, the judge is supposed to conduct a balancing test that ultimately comes down to this question: Is the probative power of this evidence sufficient to offset its prejudicial effects in warping the judgment of jurors? If the answer to that question is affirmative, it should be admitted; otherwise, by law it is to be excluded. To be precise, federal evidence law says: Although relevant, evidence may be excluded if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury, or by considerations of undue delay, waste of time, or needless presentation of cumulative evidence. 13 Key here is the notion of “unfair prejudice.” There are a great many things that courts have held to be apt to prejudice a jury unfairly. They include evidence that the defendant has a bad or violent character, especially vivid and gruesome depictions of the crime, and evidence of the defendant’s association with causes or persons likely to evoke hostility or antipathy from jurors. The same doctrine has been used to justify excluding the confession of a nontestifying codefendant that mentions the defendant’s participation in a crime, 14 graphic photos of the corpse of a homicide victim, 15 and samples of bloodstained clothing of the victim of an assault. 16 The problem, of course, is that information of this kind is often powerful evidence of the defendant’s guilt. Excluding it can weaken the case against the defendant substantially while, if it is really prejudicial, admitting it makes it more likely that jurors may make their decision on purely visceral grounds. Put in slightly more technical language, the judge is required to make a ruling about evidence that, if admitted, may lead to a false conviction while, if sup- pressed, may lead to a false acquittal. As we have seen, the judge is supposed to balance these two concerns against one another and decide about admissibility accordingly. It may help to describe the problem a little more abstractly. In cases of this sort, the judge is called on to decide which of two quantities is greater: the probability of inferential error by the jury if the contested evidence is admitted (which I shall symbolize as “prob [error with e]”) versus the probability of inferential error if the contested evidence is excluded (prob [error excluding e]). The first sort of error represents a potential false conviction; the second, a potential false acquittal. In making her decision about admitting or excluding e, the judge must perform an incredibly difficult task: She must decide on the relative likelihood of the two errors that may arise – that is, she must assign rough-and-ready values to prob (error with e) and to prob (error excluding e). 13 Federal Rules of Evidence, Rule 403. 14 Bruton v. U.S., 391 U.S. 123 (1968). 15 State v. Lafferty, 749 P.2d 1239 (Utah 1988). 16 State v. White, 880 P.2d 18 (Utah 1994). P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 Thinking about Error in the Law 21 It seems doubtful whether this decision can be made objectively. To decide on the values of prob (error with e) and prob (error excluding e), a judge needs much more data than we currently have in hand about the likelihood that particular pieces of evidence (such as vivid, gory photos of the crime scene) will distort a jury’s ability to give such evidence its legitimate weight. Well-designed empirical studies on the prejudicial effects of various sorts of evidence are extremely scarce. Even worse, collecting such information would be inherently difficult since researchers would have to be able to distinguish the emotional impact of a bit of evidence from its rational probative weight. No one has proposed a design for an empirical test subtle enough to make that distinction. I do not mean to convey the impression that this decision about admitting relevant but potentially inflammatory evidence is always insoluble. Sometimes, the problem admits of an easy solution. For instance, the prosecution may have other types of evidence, apparently less unfairly prejudicial, that will permit the state to make its point, in which case the exclusion is no big deal (since the prejudicial evidence here is clearly redundant, and redundancy is always a legitimate ground for exclusion). But what is a judge to do when a principal part of the prosecution’s case involves evidence that, while highly inculpatory, may also appear “unfairly prejudicial” and where no other evidence will do? Consider a hypothetical example: Smith is charged with being a member of a gang that entered a busy restaurant at midday, tossing grenades, firing weapons, and generally creating mayhem. By chance, one patron of the restaurant took photographs during the assault, before he was himself gunned down. One photo in particular is at issue. It shows Smith lobbing a grenade into one corner of the restaurant and also shows, in vivid color, mangled body parts and blood galore and is generally a horribly graphic depiction of the crime scene. The photo obviously passes the relevancy test. It apparently depicts the accused committing the crime with which he is charged. It is not merely relevant but highly relevant. If we suppose that no witnesses survived the mayhem, it is uniquely powerful in placing Smith at the center of things. Unfortunately, however, the judge also considers the photograph to be so vivid and awful that it invites a purely visceral reaction from the jurors. Seeing blood and gore depicted in this manner may, she fears, incline the jurors to rush to judgment rather than considering objectively the other evidence in the case, some of which may be exculpatory. Without the photo, the jury may well acquit Smith since there were no eyewitnesses. With the photo, reckons the judge, they will surely convict. Should the judge admit the photograph into evidence? Currently, that decision is left entirely up to her, with precious little assistance from the law. The guiding legal principle, as we have seen, is that the evidence should be excluded if it is more “unfairly prejudicial” than it is probative. Curiously, the law of evidence includes no canonical definition of when a sample of evidence is “unfairly prejudicial,” apart from this gem of P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 22 thinking about error in the law unclarity in Rule 403: “‘Unfair prejudice’ within its context means an undue tendency to suggest decision on an improper basis, commonly, though not necessarily, an emotional one.” Like pornography, unfair prejudice seems to be the sort of thing that, while it eludes definition, one can recognize when one sees it. But this won’t do. As Victor Gold has noted: “Absent a coherent theory of unfair prejudice, trial courts cannot meaningfully evaluate evidence on or off the record for the presence of unfair prejudice, nor can they conduct the required balancing test.” 17 How, in such circumstances, is a judge supposed to do this “balancing” to decide whether “its probative value is substantially outweighed by the danger of unfair prejudice”? One might argue that this particular rule of evidence is not really so offen- sive epistemologically since in practice it should lead only to the exclusion of inflammatory evidence that is relatively nonprobative. After all, the rule itself seems to concede that, if the evidence is of very high probative value, it could be excluded only in those circumstances where its unfairly prejudicial character was even greater than its probativeness. But there are plenty of actual cases that give one some pause as to how often the weight of highly relevant evidence really is allowed to trump its being even mildly prejudicial. Consider two real examples of the kind of balancing that goes on when trial and appellate judges try to assess unfair prejudice. In a 1994 case in south Texas, Ram´ on Garcia was accused of burgling Charles Webster’s house. Garcia was seen in the house at the time of the burglary by a police officer who had been called to the scene by a neighbor. Taking flight, Garcia was subsequently caught. Police found no contraband on Garcia himself when he was apprehended, but several items stolen from Webster were found on the ground near the site of his arrest. By way of showing intent to commit burglary, the prosecutor introduced evidence that Garcia had arrived at the scene of the crime on a bicycle that he had stolen from a neighboring house two days earlier. The boy whose bicycle was stolen by Garcia testified that he was its owner. Garcia was convicted. His attorney appealed, arguing that the evidence of the stolen bicycle unfairly prejudiced the jury against his client. The appellate court, siding with Garcia, did not deny that the evidence of the stolen bicycle was relevant to the question of whether Garcia intended to rob the Websters but held that its relevance was outweighed by its unfairly prejudicial nature, “particularly so when the State chose to offer the evidence through the child rather than his parents.” 18 The logic of the appellate ruling is a bit tortuous, but here is what seems to be going on: Besides conceding the relevance of the fact that the defendant arrived at the scene of a burglary on a stolen bicycle to the question of Garcia’s inten- tion to rob the Websters, the superior court even seems to grant that evidence 17 Victor Gold, Observations on the Nature of Unfairly Prejudicial Evidence, 58 Wash. L. Rev. 497, at 502 (1983). 18 Garcia v. State, 893 S.W.2d 17, at 22 (Tex. App. 1994). P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 Thinking about Error in the Law 23 concerning the theft of the bicycle might not have been unfairly prejudicial if the testimony about its theft had been offered by an adult. But for a child to testify that his bicycle had been stolen seems to have the court in a tizzy for fear, I suppose, that the jury will conclude that anyone who would steal a bicy- cle from a young boy must be very bad indeed and deserves to be sent to jail, whether he robbed the Websters or not. Are we then to conclude that whenever children have inculpatory evidence to offer, the specter of unfair prejudice is raised and that their evidence should be excluded? I doubt that is the intended moral, but the example certainly suggests how subjective the determination of unfair prejudice can sometimes be. Consider briefly a second case, also from Texas, where the balancing test seems to have gone awry. In 1992, Kenneth Nolen was accused of aggravated possession of methamphetamine. Acting on a warrant, police found Nolen asleep in the bedroom of a friend’s house. Smelling a strong odor, they opened the bathroom door, next to which Nolan was sleeping, and discovered a labora- tory for making amphetamine. Nolen’s prints were found on the lab equipment. To convict someone in Texas of aggravated possession, the state must show that the accused was aware of the fact that the drug in his possession was an illegal substance. Nolen’s attorney suggested that, although his client had indeed been making amphetamines, he did not know that such activity was illegal. To counter that suggestion, the prosecutor sought to introduce evidence that Nolen had been convicted three years earlier of breaking into the evidence room of the local sheriff’s office to steal laboratory equipment suitable for making amphetamines. The prosecutor argued, and the trial judge agreed, that the earlier theft of such equipment was highly relevant to the question whether Nolen knew enough about amphetamines to realize that they were illegal. In the prosecutor’s closing arguments, he insisted that “it’s a reasonable deduction from the evidence that [if] that man is so daring [as] to take lab equipment from the Hood County Sheriff, he certainly knows about am[p]hetamine and the equipment used to produce amphetamine.” 19 The jury convicted Nolen. On appeal, a higher court reversed the verdict, insisting that “it was an abuse of [the judge’s] discretion to determine that the extraneous evidence of burglarizing the sheriff’s evidence shelter was not substantially outweighed by the danger of unfair prejudice to Nolen, confusion of the issues, and misleading the jury.” 20 Howprecisely is it unfairly prejudicial to Nolen – facing a charge of knowingly making illicit drugs and with his prints all over the equipment in question – to show that he had previously stolen such equipment from the sheriff’s office? Since what was in dispute was whether Nolen knewthat making methamphetamine was illegal, we would seemto have here evidence highly germane to the hypothesis that Nolen was knowledgeable 19 Nolen v. State, 872 S.W.2d 807, at 813 (Tex. App. 1994). 20 Ibid., at 814. P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 24 thinking about error in the law about drug making. Not so, says the appellate court, since it is not “deductively certain” that “a man who steals glassware certainly knows the characteristics of a particular chemical compound that may be produced with that type of glassware.” 21 Well, yes. It is (just about) conceivable that Nolen stole equipment for making amphetamines from the evidence room of the sheriff’s department without knowing what such equipment was used for and without knowing that making such stuff was illegal. But we should not be looking for deductive truths in the law. The question is whether Nolen’s previous conviction for theft of equipment for making amphetamines has a powerful evidential bearing on the question of whether he knewthree years later, while making amphetamines, that what he was doing was against the law. For a court to hold that such evidence is more likely to be unfairly prejudicial than relevant strikes me as extraordinarily obtuse. (I am not claiming that cases such as these two are the norm, but, at a minimum, they suggest that the balancing test demanded by the unfair prejudice rule is, at best, problematic.) Surely, a preferable alternative would be to admit all evidence that is gen- uinely relevant, accompanied, where appropriate, by an explicit reminder from judge to jury to bring their critical faculties to bear in evaluating the relation of the evidence to the crime and in keeping their emotional reactions to the evidence firmly in check. Of course, we do not knowhowearnestly juries could or would follow such an instruction. But, for that matter, neither do we really know which kinds of evidence unfairly warp jurors’ judgment and which do not. Since the judge has no robust empirical information about the latter issue, her decision about which potentially prejudicial evidence to include and which to exclude is likely to be every bit as suspect as an emotionally driven verdict from the jury. It is not only the judge who has a role to play here in encouraging the jury to stay on the straight and narrow. One of the functions of the adversarial system is to give each side a shot at undermining or otherwise calling into question the case presented by the other. If there is evidence that the defense regards as misleading or that it suspects may otherwise steer a jury in the wrong direction, it is the job of defense counsel to seek to fit that evidence into a context favorable to the defendant, if that is possible. Failing that, it falls to defense counsel to persuade the jury not to attach more weight to any specimen of inculpatory evidence than it duly deserves. Like the judge, counsel may fail in this task from time to time. Jurors may conceivably rush to judgment for all sorts of inappropriate reasons, despite having been warned of the dangers of doing so. That conceded, if we cannot generally trust jurors to keep their emotions in check, then we should abandon trial by jury altogether. The very idea of trial by jury depends on the presumed fairness and common sense of twelve peers 21 Ibid., at 813. P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 Thinking about Error in the Law 25 of the accused. If jurors cannot generally give vivid but relevant evidence its appropriate weight – having been warned by the judge to do so and having heard counsel for each side make its pitch about the meaning of the evidence – then the system is rotten to the core. Paternalistically coddling jurors by shielding them from evidence that some judge intuits to be beyond their powers to reason about coherently is not a promising recipe for finding out the truth. I use the term “paternalism” deliberately. Recall that, in a bench trial, the same judge who decides on the admissibility of evidence usually acts as the trier of fact. In short, the system trusts judges to be able to see inflammatory and unfairly prejudicial evidence and then to be able to put it into perspective, not allowing it to warp their judgment. By permitting judges in bench trials to see evidence that we would not permit juries to see, we are saying that juries are less reasonable, less objective, or less mature than judges. That may be so; as far as I know, the issue is unsettled. But, settled or not, this is not what judges are for in an adversarial system. Their job, apart from generally maintaining order in the court, is to explain to jurors what the law means. That is what they are trained to do, and there is nothing paternal about that role. It becomes paternal when, out of a systemic distrust of the good sense of jurors, we cast the judge in the role of arbiter on both empirical and policy questions that should not be hers to settle. Put in grander terms, it will be the recurring theme of this book that, leaving redundancy aside, the only factor that should determine the admissibility or inadmissibility of a bit of evidence is its relevance to the hypothesis that a crime occurred and that the defendant committed it. The exclusion of admittedly relevant evidence onthe grounds of its unfairlyprejudicial character is motivated by commendable epistemic instincts. But the rule itself requires judges both to have empirical knowledge that they lack and to make policy determinations (for instance, about the relative seriousness of false acquittals and false convictions) that are beyond their ken. As we have seen, my proposal is squarely at odds with existing practice. In American courts, “mere” relevance, even powerful relevance, does not ensure admissibility. As the U.S. Supreme Court argued in a famous case, Michelson v. U.S.: The State may not show defendant’s prior trouble with the law, specific criminal acts, or ill name among his neighbors, even though such facts might logically be persuasive that he is by propensity a probable perpetrator of the crime. The inquiry is not rejected because character is irrelevant; on the contrary, it is said to weigh too much with the jury and to so overpersuade themas to prejudge one with a bad general record and deny hima fair opportunity to defend against a particular charge. The overriding policy of excluding such evidence, despite its admitted probative value, is the practical experience that its disallowance tends to prevent confusion of issues, unfair surprise and undue prejudice. 22 22 Michelson v. U.S., 335 U.S., at 475–6 (1948). P1: JZZ 0521861667c01 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 14:56 26 thinking about error in the law This token of conventional folk wisdom may be grounded in the “practical experience” of the judiciary. But precious few well-designed empirical studies bear out the claim that properly instructed jurors, exposed to the confrontations typical of the adversary system, are incapable of giving inflammatory or prej- udicial but relevant evidence the weight it rationally deserves. Since that is so, rules of admissibility that trump relevance cannot be shown to further epistemic ends, not even when those rules (such as the one against unfairly prejudicial evidence) are couched in epistemic terms (“preventing a jury’s verdict from being shaped by prejudice rather than the facts”). On the contrary, they almost invariably thwart those ends by keeping obviously relevant evidence out of the courtroom. The proof of that thesis is the thrust of much of the rest of this book. P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 part i THE DI STRI BUTI ON OF ERROR Our dangers do not lie in too little tenderness to the accused. Our procedure has been always haunted by the ghost of the innocent man convicted. It is an unreal dream. What we need to fear is the archaic formalism and the watery sentiment that obstructs, delays, and defeats the prosecution of crime. – Judge Learned Hand 1 1 U.S. v. Garsson, 291 F. 646, at 649 (S.D.N.Y. 1923). 27 P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 28 P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 2 The Unraveling of Reasonable Doubt It is difficult, if not impossible, to so define the term “reasonable doubt” as to satisfy a subtle and metaphysical mind, bent on the detection of some point, however attenuated, upon which to hang a criticism. – Supreme Court of Virginia 1 I find it rather unsettling that we are using a formulation that we believe will become less clear the more we explain it. – Jon Newman, Chief Judge of the U.S. Court of Appeals for the Second Circuit 2 Updating the Road Map As I said in the first chapter, one of the three basic value commitments in the criminal law is to the proposition that, when errors are made, it is far better to acquit a guilty party than to convict an innocent one. In this chapter and the two followingones, we will be examininga cluster of familiar concepts andprecepts, all of which give concrete expression to the importance of distributing errors in line with this sensibility: the standard of proof, the presumption of innocence, the benefit of the doubt, and the prosecutorial burden of proof. As we will see, none of these traditional principles is designed to reduce the likelihood of error in a trial. Their intended result is to distribute errors in a certain way – that is, to ensure that such errors as do occur will be predominantly false acquittals rather than false convictions. What I will call error distribution doctrine in a criminal trial consists of this set of concepts and precepts, all designed to safeguard the fate of the defendant by making it hard to convict any but the most obviously guilty. We all recognize these doctrines, often without thinking very hard about them. Still, recognizing them is one thing; figuring out precisely what they 1 McCue v. Commonwealth, 103 Va. 870, at 1002 (1905). 2 Jon O. Newman, Beyond “Reasonable Doubt,” N.Y.U. L. Rev. 979, at 984 (1993). 29 P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 30 the distribution of error mean and how they relate to one another is quite another. The aim of Part I is to examine these four principles and their effects on error in considerable detail. In one way or another, all confer special advantages on the criminal defen- dant. Thus, the presumption of innocence precludes jurors from attaching any significance to the fact that the defendant has already been found probably guilty in a series of prior hearings and tribunals. The benefit of the doubt insists that, if the verdict is at all a close call, jurors must err on the side of the defendant. The assignment of the burden of proof to the prosecutor further ensures that the defendant will be acquitted unless the prosecutor proves his case. Finally, a demanding standard of proof such as “beyond a reasonable doubt” enjoins jurors to acquit the defendant even if they think he is probably guilty, since only a firm, settled belief in that guilt justifies a conviction. These principles are often seen as motivated by a concern to protect the innocent defendant. Innocence, we might say, is the default assumption. The fact is that such principles function to make it harder to convict any defendant, whether innocent or guilty. Each amounts to putting gentle (and sometimes not-so-gentle) pressure on the scales of justice to skew the verdict in the defendant’s favor. That is not, I hasten to add, meant as a criticism of such doctrines but simply as an observation about their predictable effects. Even though none of the four precepts is driven by the quest for the truth, they nonetheless must occupy a central place in a book about the epistemology of law because, while not grounded in epistemic motives, they have profound epistemic consequences. Those consequences will form the principal focus of the next three chapters. Of these four precepts, the standard of proof is the most important and the most widely discussed. Accordingly, this chapter will focus in depth on analyz- ing the current understanding of proof beyond a reasonable doubt. The upshot of that analysis will be that this notion of proof is grievously inadequate, delib- erately unclear, wholly subjective, and open to about as many interpretations as there are judges, to whom it falls to explain this notion to hapless jurors. Chapter 3 will propose a better way to tackle the problem of articulating a stan- dard of proof, chiefly by exploring the conceptual interdependencies between the standard of proof and the benefit of the doubt. Chapter 4 will conclude our treatment of error distribution questions by looking at both the presumption of innocence and the burden of proof. Introduction At the core of criminal jurisprudence in our time lies a fundamental confusion. Stated succinctly, the problem is that the concept of proving guilt beyond a reasonable doubt (hereinafter, BARD) – the only accepted, explicit yardstick for reaching a just verdict in a criminal trial – is obscure, incoherent, and muddled. One symptom of the problem emerges out of recent research (to be discussed P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 The Unraveling of Reasonable Doubt 31 later in this chapter) on mock juries. This makes it vividly clear that jurors have only the haziest notion of what a “reasonable doubt” is. In theory, that difficulty could be resolved by having judges instruct juries about the meaning of reasonable doubt. Indeed, such instructions are already routine in some legal jurisdictions. Sadly, as we will see in detail shortly, another symptom of the problem is that any serious scrutiny of judges’ oral instructions to jurors shows that trial judges as much as jurors have strikingly discrepant understandings of this key notion. Ordinarily, when confusion of this sort exists at the trial level, we look to higher courts to dispel it. Yet, a careful reading of the opinions of appellate courts that have explicitly sought to clarify this crucial doctrine leads to the inescapable conclusion that not even the keenest minds in the criminal justice system have been able to hammer out a shared understanding about the level of proof appropriate to convict someone of a crime. In sum, the muddle is not only in the minds of jurors but in the written opinions of judges, jurists, and attorneys as well, and at every level from the trial bench to the Supreme Court. Of course, not every conceptual confusion need occasion widespread hand wringing. That we lack fully coherent ideas about free will or the origins of the big bang or how language maps on to the world needn’t worry anyone who does not relish such conundrums. Still, occasionally we discover incoherencies in ideas that profoundly impinge on the affairs of everyday life. Those cannot be so blithely ignored. BARD is surely one of them. The confusion and lack of clarity surrounding BARDimmediately implies that, in any given criminal trial, both the accused and the prosecution, unable to predict what level of proof will be necessary, face a crapshoot. The risk here is not that of the bleeding-heart juror – moved by sympathy to acquit the accused no matter how powerful the evidence – nor that of the law-and-order juror – minded to convict however exculpatory the evidence. It is more systemic than that. The most earnest jury, packed with twelve people desirous of doing the right thing and eager to see that justice is done, are left dangling with respect to how powerful a case is required before they are entitled to affirm that they believe the guilt of the defendant beyond a reasonable doubt. In such circumstances, simply muddling on is not an attractive prospect. Clearly, justice in the sense of fairness and due process cannot be assured, nor is it even likely, in a system where different judges recommend, and different juries use, discrepant standards for guilt and innocence. With the height of the bar for conviction left indeterminate, you have no assurance that any one case at trial will be decided by a jury according to the same standard that a rival jury, handling the identical case, would use. The system, in short, lacks reliability (in the sense of uniformity and predictability). It is thus inherently unjust. More than justice is at stake. If jurors and jurists are unclear about the standard of proof, we lack reason to believe that verdicts have the sort of validity that I described in the first chapter. P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 32 the distribution of error One sign among many of the desperateness of the current situation and of the general judicial uneasiness about the prevailing standard (or should I say mantra?) for conviction is that many appellate courts in the U.S. advise trial judges to tell juries that conviction requires belief in guilt beyond a reasonable doubt full stop, without spelling out what that means. This leaves it to inex- perienced jurors to haggle among themselves about the appropriate height of the bar for conviction. Another more telling sign of the same uneasiness is that England, which has the same common law tradition as our own, has recently abandoned its two-hundred-year-old practice of having judges instruct jurors about the nature of reasonable doubt. Instead, jurors there are simply told that conviction requires that they must be “sure” of the guilt of the accused. England made this change because senior legal theorists concluded that reasonable doubt could be neither defined, nor uniformly understood, nor consistently applied. 3 The targets of this chapter are twofold: Eventually, we will be asking our- selves whether BARD is the right standard to use for furthering the truth- seeking aims of a criminal trial and whether all trials for all crimes should, as the Supreme Court insists, use the same standard of proof. But before we can turn to those questions, a lot of preliminary spadework will be required to get reasonably clear about how BARD has been, and should be, understood. In the next sections, I will describe in detail the nature of the conundrum posed by the current, confusing accounts of reasonable doubt and will suggest some remedies, more or less radical, for this defect. One’s initial unease that the remedies are unconventional should be mitigated by, and balanced against, the desperateness of the status quo, as I will be describing it. Interpretations of Reasonable Doubt Origins BARDhas a long and checkered history. Afewbasic points fromthat history are necessary. Until the end of the eighteenth century and before the introduction of BARD, jurors were sworn – not unlike witnesses – simply to deliver a true verdict about the case before them. Given the presumption of the innocence of the accused, this was generally construed to mean that legitimate guilty verdicts were in order only if the jurors were certain of that guilt. This was, of course, a very exacting standard – too exacting, as it turned out, for philosophers and jurists during the Enlightenment came to realize that in human affairs (as opposed, say, to mathematics or logic), no full certainty was to be had. The next 3 Whether replacing “having no reasonable doubts” with “being sure” is an improvement seems to me to be up for grabs. Both are probably hopeless, as I will argue in the next chapter. The salience of the English example is the evidence it offers that many jurists have come to the conclusion that proof beyond reasonable doubt is incoherent. P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 The Unraveling of Reasonable Doubt 33 best thing, according to such philosophers as John Locke and John Wilkins, was what they called “moral certainty.” They dubbed this sort of certainty “moral” not because it had anything to do with ethics or morality but to contrast it with “mathematical” certainty of the sort traditionally associated with rigorous demonstration. Morally certain beliefs could not be proven beyond all doubt, but they were nonetheless firmand settled truths, supported by multiple lines of evidence and testimony. You could be morally certain, for instance, that Julius Caesar had once been emperor of Rome or that the earth was round, even though there remained room for the thoroughgoing skeptic to point out that we may conceivably be deluded about such matters. What characterized morally certain beliefs was that, although open to the skeptics’ doubt in theory, there was no real or rational grounds for doubting themin practice. Fromthis arose the notion that a guilty verdict required the jury to believe “beyond a reasonable doubt” or “to a moral certainty” in the guilt of the accused. Jurists were quick to notice that, given these new insights in epistemology, the proper criterion for determining guilt should be moral, not full, certainty that the accused committed the crime in question. By the 1850s, the principle that guilt must be established BARDhad become widely accepted wherever Anglo-Saxon, common law traditions of law were in force. In the United States, the BARD rule was the standard for criminal conviction in most state and territorial courts, as well as throughout the federal court system. Accordingly, judges’ instructions to juries, delivered just prior to jury deliberation, uniformly incorporated an admonition to the effect that a guilty verdict could be pronounced only if juries were morally certain or certain BARD that the accused had committed the crime. Chief Justice Lemuel Shaw of the Supreme Judicial Court of Massachusetts nicely summed up the equation between moral certainty and proof beyond a reasonable doubt in a famous case from 1850. Because it was to become the canonical formulation of BARD for almost a century, it is worth quoting in its entirety: Then, what is reasonable doubt? It is a termoften used, probably pretty well understood, but not easily defined. It is not mere possible doubt; because every thing relating to human affairs, and depending on moral evidence, is open to some possible or imaginary doubt. It is that state of the case, which, after the entire comparison and consideration of all the evidence, leaves the minds of jurors in that condition that they cannot say they feel an abiding conviction, to a moral certainty, of the truth of the charge. The burden of proof is upon the prosecutor. All the presumptions of law independent of evidence are in favor of innocence; and every person is presumed to be innocent until he is proved guilty. If upon such proof there is reasonable doubt remaining, the accused is entitled to the benefit of it by an acquittal. For it is not sufficient to establish a probability, though a strong one arising from the doctrine of chances, that the fact charged is more likely to be true than the contrary; but the evidence must establish the truth of the fact to a reasonable and moral certainty, a certainty that convinces and directs the understanding, P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 34 the distribution of error and satisfies the reason and judgment, of those who are bound to act conscientiously upon it. This we take to be proof beyond reasonable doubt. . . . 4 Although language to this effect became standard by the late-nineteenth cen- tury, judges would often expand on and embellish Shaw’s phraseology. Some- times they would say that a belief was true beyond reasonable doubt when it was “highly probable” or when the jurors had an “abiding conviction” about it, or when their “consciences were satisfied” that conviction was the right thing to do. But even in those early days of the BARD rule, many judges inclined to the view that the less said by way of explanation of this relatively new stan- dard, the better. In one classic case dating from 1894 in the Utah Territory, we see precisely such hem-hawing from the bench. The judge in a trial for criminal polygamy, during his instructions to the jury, artfully explained that “a reasonable doubt is not an unreasonable doubt.” Indeed. Armed with this powerful tautology, the jury convicted the defendant. The defendant’s attorney appealed the case, arguing that the judge had failed to explain BARD satisfac- torily. The U.S. Supreme Court upheld the conviction, insisting that the judge’s language “gives all the definition of reasonable doubt which a court can be required to give.” 5 As we will see shortly, this wholly unhelpful response, and its endorsement by the highest court in the land, was a potent augury of things to come. The Current State of Play Although BARD was very widely used in both state and federal courts from the middle of the nineteenth century until the middle of the twentieth, it was there chiefly by custom and common law precedent rather than by law. In 1970, a famous ruling, In re Winship, gave BARD a constitutional imprimatur. Specifically, the Supreme Court found, by a creative act of interpretation – creative in the sense that the Constitution nowhere addresses the question of the standard of proof in criminal trials – that the U.S. Constitution required that all criminal juries were to be instructed that BARD was the applicable bar for conviction. 6 A judge’s failure so to instruct a jury thereby became automatic grounds for reversing a conviction. Further, Winship made clear that every key element of the prosecution’s case must be established BARD before 4 Commonwealth v. Webster, 59 Mass. 295, at 320 (1850). 5 Dunbar v. U.S., 156 U.S. 185 (1894). 6 The Supreme Court itself, in Cage v. Louisiana, argued that In re Winship established that the Due Process Clause of the Fourteenth Amendment “protects the accused against conviction except upon proof beyond a reasonable doubt of every fact necessary to constitute the crime with which he is charged” (498 U.S. 39 [1990]). In dissent in Winship, Justice Black pointed out that “in two places the Constitution provides for trial by jury, but nowhere in that document is there any statement that conviction of crime requires proof of guilt beyond a reasonable doubt” (397 U.S. 358, at 383 [1970]). P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 The Unraveling of Reasonable Doubt 35 conviction was justified. 7 For more than thirty years, then, BARD has enjoyed constitutional status, equivalent to the right to counsel or the right to trial by jury. In the wake of BARD’s new and enhanced status in the criminal justice system, trial judges, appellate courts, and even legislators have been trying to make sense of what the doctrine means and, in more practical terms, how it should be explained (if at all) to puzzled jurors. Until Winship, the most common way of handling it was for judges, following Shaw, to explain to jurors that proof beyond a reasonable doubt meant “belief to a moral certainty.” We have already seen powerful precedent for this gloss since BARD itself had evolved out of the philosophical literature on moral certainty. Since then, however, the Supreme Court has discouraged this usage, holding that the terminology is archaic, unhelpful at best, and misleading, because, as Justice Harry Blackmun put it in a famous dissent, there exists “the real possibility that such language would lead jurors reasonably to believe that they could base their decision to convict upon moral standards or emotion in addition to or instead of evidentiary standards.” 8 The Court opined: “The risk that jurors would understand ‘moral certainty’ to authorize convictions based in part on value judgments regarding the defendant’s behavior is particularly high in cases where the defendant is alleged to have committed a repugnant or brutal crime.” 9 In the same case, Justice Sandra Day O’Connor, writing for the Court, said emphatically that “this Court does not condone the use of the antiquated ‘moral certainty’ phrase.” Justice Ruth Bader Ginsberg ventured that “the phrase ‘moral certainty,’ though not so misleading as to render the instructions [to jurors] unconstitutional, should be avoided as unhelpful in defining reasonable doubt.” 10 These passages already hint at the conclusion that a more detailed scrutiny of Supreme Court rulings bears out, to wit, that the Supreme Court in the last generation has effected a disconnect between the standard of proof (BARD) and the philosophical ideas that originally provided its grounding and coherence. The key question, to which we will return later, is whether BARD, stripped of its embedding in a particular epistemic tradition, is capable of standing independently, that is, whether it is a coherent notion in its own right. With the most common traditional definition – moral certainty – excluded (or at least sternly frowned on), judges and legal scholars have been casting about for other ways of explaining BARD to jurors. I want to review briefly some (though by no means all) of the proposed alternative explications, by way of underscoring just how confused the notion of reasonable doubt has become. 7 While this claim is generally true, there are exceptions to the “every element” rule. One of them involves affirmative defenses and presumptions. For a discussion of the latter, see Chapter 4. 8 From Justice Blackmun’s dissent in Victor v. Nebraska, 511 U.S. 1, at 38 (1994). 9 Ibid., at 24. 10 Ibid. P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 36 the distribution of error All but one of the versions of BARD that I will reprise have been extensively used by trial judges and endorsed (if sometimes later disowned) by appellate courts. The remaining one finds many advocates among academic lawyers but virtually none among judges. 1. BARD as That Security of Belief Appropriate to Important Decisions in One’s Life Ever since the nineteenth century, one popular legal explanation of BARD has involved comparing it with important decisions in the lives of ordinary citizens. Jurors are told that, just as they do not undertake major life decisions except when they are sure of the beliefs undergirding them, so they shouldn’t find the defendant guilty unless their confidence in his or her guilt is as great as the confidence they demand in their own important decision making. The analogy here is that of a prudent person who undertakes an important action, vital to his or her affairs, only when very confident about the beliefs that drive their actions. We find a typical formulation of this idea in the pattern jury instruction recommended by the Fifth U.S. Circuit Court: Proof beyond reasonable doubt, therefore, is proof of such a convincing character that you would be willing to rely and act upon it without hesitation in the most important of your own affairs. 11 Although this instruction to jurors is very common, it is deeply misleading. Ponder howwe typically think about the most important life decisions we make: Do you give up your current job and change employers or careers only when you are virtually certain about the results of the action? Do you decide to marry only when you are certain beyond a reasonable doubt that it will work out? Do you refuse surgery for a life-threatening condition if the doctor tells you that there are risks involved or if there are any rational doubts that it may cure you? Do you hesitate to call the police when you suspect you have heard the sound of someone moving around your living room in the middle of the night, even though you surely have significant and reasonable doubts about whether the noises really came from an intruder? In our ordinary lives, even sizable doubts cause little or no hesitancy to act. Many, perhaps most, of the important life decisions are taken under conditions of significant uncertainty, where doubt is not only rational but rampant. We act even though there is considerable doubt. The analogy between ordinary-life decision making about important matters and the deliberations of a jury in a criminal trial invites jurors to convict on the same sort of precarious beliefs that frequently guide their important practical actions. Ironically, that is precisely what the legal system wants jurors to avoid doing when deciding guilt in a criminal trial. 11 Fifth Cir. Crim. Jury Instr. 1.06 (1990). P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 The Unraveling of Reasonable Doubt 37 For this reason, the Supreme Judicial Court of Massachusetts has forcefully argued against this analogy for understanding BARD, believing that: The degree of certainty required to convict is unique to the criminal law. We do not think that people customarily make private decisions according to this standard nor may it even be possible to do so. Indeed, we suspect that were this standard mandatory in private affairs the result would be massive inertia. Individuals may often have the luxury of undoing private mistakes; a verdict of guilty is frequently irrevocable. 12 2. BARD as the Sort of Doubt That Would Make a Prudent Person Hesitate to Act Eventually mindful of such problems, the Supreme Court has of late been conceding that the analogy just discussed can be misleading and has proposed an alternative. Specifically, the justices have shifted fromtrying to explain what a belief beyond reasonable doubt is to explaining what a reasonable doubt would be. 13 Typically, the point is formulated in this fashion: A rational doubt is the sort of doubt about your beliefs that would cause you to hesitate to act on those beliefs. By contrast, a less than rational doubt (and thus the sort that jurors should ignore) would not cause you to hesitate. The only sort of doubt that should prevent a juror from finding the accused guilty is the sort of doubt that would cause him to hesitate to act in his own personal, but important, affairs. Here is the language proposed in the model jury instruction endorsed by the Second Circuit: A reasonable doubt is a doubt based on reason and common sense – the kind of doubt that would make a reasonable person hesitate to act. Proof beyond a reasonable doubt must, therefore, be proof of such a convincing character that a reasonable person would not hesitate to rely and act upon it in the most important of his own affairs. 14 Sadly, this twist in the analogy does not really get around the previous problem. Are reasonable doubts the only ones that make us hesitant to act? Scarcely. Many people, confronted by major life decisions, fidget and fret even when it is wholly clear and beyond doubt what course of action they should take. A battered wife, knowing perfectly well that her husband will continue beating 12 Commonwealth v. Ferreira, 373 Mass. 116, at 130 (1977). While I agree with the general sentiment expressed here, I do not understand why the Massachusetts high court argued that guilty verdicts are “frequently irrevocable.” Given the universal availability of judicial review of convictions, it would seem fairer to say that they are frequently not revoked. 13 In a landmark 1954 case, Holland v. U.S., the Supreme Court approved the jury instruction described in the preceding section but added, “we think this section of the charge should have been in terms of the kind of doubt that would make a person hesitate to act . . . rather than the kind on which he would be willing to act” (348 U.S. 140 [1954]). 14 Edward J. Devitt et al., Federal Jury Practice and Instructions, §12.10, at 354 (4th ed. 1987). P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 38 the distribution of error her when money is short in the household, will nonetheless often hesitate about extricating herself from the situation. Does that imply, as the Second Circuit would have it, that she has reasonable doubts about her husband’s violent streak? Such deliberative inertia, inherent in many situations and in many decision makers, gives rise to hesitancy and inaction, even when the parties involved have no genuine doubt whatever that their current situation is contrary to their own best interests. It is thus clear that ordinary people often hesitate to act even in the absence of reasonable doubts. A committee of distinguished federal judges, reporting to the Judicial Con- ference of the United States, has criticized all such glosses on BARD that assimilate it to ordinary situations of personal choice and decision making because the analogy it uses seems misplaced. In the decisions people make in the most important of their own affairs, resolution of conflicts about past events does not usually play a major role. Indeed, decisions we make in the most important affairs of our lives – choosing a spouse, a job, a place to live, and the like – generally involve a very heavy element of uncertainty and risk-taking. They are wholly unlike the decisions jurors ought to make in criminal cases. 15 Accordingly, identifying the process of reaching a criminal verdict with the process of practical decision making can only mislead. BARD, after all, is supposed to set a high standard for conviction, higher than we ordinarily insist on in our practical affairs. It is meant to exemplify the old saw that it is better that ten guilty men go free than that one innocent man is condemned. Yet if a judge tells jurors that they can convict if they are as sure of the defendant’s guilt as they are about the beliefs that guide their important practical decisions, this is an open invitation to convict on quite low levels of confidence in guilt. 3. BARD as an Abiding Conviction of Guilt Another common way for judges to describe what it is to believe in guilt beyond a reasonable doubt is to suggest that jurors have such a belief when they have “an abiding conviction that the defendant is guilty.” Here, for instance, is a model jury instruction adopted by California courts: Reasonable doubt is defined as follows: It is not a mere possible doubt; because every- thing relating to human affairs, and depending on moral evidence, is open to some possible doubt or imaginary doubt. It is that state of the case which after the entire comparison and consideration of all the evidence, leaves the minds of the jurors in that condition that they cannot say they feel an abiding conviction, to a moral certainty, of the truth of the charge. 16 15 Federal Judicial Center, Pattern Criminal Jury Instructions 18–19 (1987) (commentary on instruction 21). 16 114 S. Ct, at 1244. The careful reader will have noted that the California language is lifted almost verbatim from Justice Shaw’s instruction of 1850. P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 The Unraveling of Reasonable Doubt 39 So, how does the ordinary person, here and now, determine whether she has a conviction, a moral certainty, which is or is not abiding? Taken literally, an abiding conviction is one that one will have for a long time – as opposed to a transitory belief. But who can say of any inference-based belief that he has just formed that he will not eventually develop second thoughts about it? Ret- rospectively, of course, you can say of someone’s belief, including your own, that it was an abiding conviction. The issue is how, prospectively, a juror can determine that a newly formed belief of hers will be one of those. Since, on pain of incoherence, this cannot be what is meant by an “abiding conviction,” we have to look for its meaning elsewhere than on its surface. The legal liter- ature on this question tends to suggest that, in asking jurors to decide whether their belief in guilt is abiding, the courts really seem to be asking whether the belief is held firmly and unwaveringly. That presumably would be an abiding conviction. On this construal, you can figure out here and nowwhether a fresh belief will be an abiding one by asking whether it is firmly held. Still, does this feature of a belief in guilt show that it is a belief beyond a reasonable doubt? The firmness of a belief – that is, the depth of one’s conviction in it – does nothing to settle whether the belief is rational or founded on the evidence. For instance, I may have an abiding belief that my recently departed Fido is now in dog heaven. I may even go to my grave believing it. Nevertheless, the endurance of this belief or my current tenacity in asserting it has little to do with whether it is rationally well founded. Similarly, a juror may come to an abiding conviction about a defendant’s guilt for all sorts of ill-considered reasons – for example, that the accused is a Scorpio. If it is true, as surely it is, that we don’t want jurors to be deterred from convicting someone because of irrational doubts that those jurors may entertain about the defendant’s guilt, it must be equally true that we don’t want to encourage jurors to rush to convict someone based on irrational suspicions of that person’s culpability. There can be no acceptable exegesis of BARD that fails to make clear that convictions cannot be grounded on ill-considered foundations any more than acquittals can. This gloss fails on those grounds. Saying to a juror that he ought not to convict unless he has strong and abiding convictions about the guilt of the accused is unacceptable because it confuses a (possibly) necessary condition with a sufficient one. It may be that we do want jurors to be persuaded of the guilt of the accused before they convict (although I will raise doubts about this in Chapter 4); but such persuasion is of no probative value unless it derives from a careful reflection on, and wise inferences from, the evidence. True, the model California instruction quoted previously enjoins the juror to “consider” all the evidence; but considering all the evidence and weighing it appropriately are distinct things. Unfortunately, most jury instructions say nothing about the kind of evidence necessary for conviction, focusing instead almost entirely on the subjective state of the juror – a point to which we will return in much detail further along. P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 40 the distribution of error BARD, we must remember, was designed, among other things, to protect a defendant from the rush to judgment that we associate with tribal or vigilante justice. It was meant to codify the meaning of the presumption of the innocence of the accused. Instructing juries that they should not convict unless firmly convinced that the defendant is guilty does nothing to ensure that their beliefs about guilt and innocence will be based on a careful assessment of the evidence of the case. If it is important to stress that doubts, if they lead to acquittal, must be rational doubts, it is likewise crucial to insist that beliefs in guilt, if they lead to conviction, must be rationally well founded. You may think that the problem could be remedied here by combining the language of “abiding convictions” with an insistence that jurors must attend carefully to the evidence in coming to their belief about guilt. That is, the judge might add that the evidence must make the defendant’s guilt overwhelmingly likely or some words to that effect. But then the talk about an abiding conviction is gratuitous. If genuinely reasonable doubt vanishes when the evidence points powerfully to the guilt of the accused, then the juror need only decide whether the evidence presented is of that character. It is irrelevant whether, in addition, he has a firm or abiding conviction about it. It only further muddies the waters that some courts have argued that the language of “firm conviction of guilt” represents a seriously weakening of the BARD standard and that it makes the prosecutorial burden much lighter than it had been when “moral certainty” was required. 17 4. Reasonable Doubt as a Doubt for Which a Reason Could Be Given This construal of BARDappears to be almost tautologous and thus harmless, if bordering on the uninformative. A reasonable doubt is said to be a doubt that is reasonable, that is, for which you could give a reason. On this view, doubts for which no reason can be given fail to be reasonable and thus cannot legitimately block conviction. As obvious as it may seem, this explication of reasonable doubt has been frequently frowned upon, and sometimes explicitly repudiated by, appellate courts. In Burnett v. Nebraska, for example, the bench instructed the jury that a reasonable doubt was a doubt that you could explain to your fellow jurors and 17 Here is the opinion of the Ohio Supreme Court in the face of a legislative redefinition of BARD from moral certainty to a firm conviction: It is apparent that a significant change has been made reducing the test of an abiding conviction to a moral certainty to a present, firm conviction. The changes in the substance of the definition of reasonable doubt reduce the degree of certainty required and eliminate the implication of a steadfast certainty or conviction. The changes are not mere semantics or philosophical differ- ences. They are significant. This significance lends itself to a unique advantage to the state in its argument to the jury and a disadvantage to the accused (State v. Crenshaw, 51 Ohio App. 2d 63 [1977]). P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 The Unraveling of Reasonable Doubt 41 for which some reason could be given in light of the evidence adduced in the case. The judge instructed the jury, in part, “by reasonable doubt is not meant that the accused may possibly be innocent of the crime charged against him, but it means some actual doubt having some reason for its basis.” This seemingly innocuous language was faulted on appeal; the appellate court held that this definition was in error and the guilty verdict was reversed. 18 Again, in a higher court’s review of Carr v. Nebraska, we read “It is error to charge a jury that [reasonable doubt] is a doubt for the having of which the juror can give a reason derived from the testimony.” 19 The Supreme Court holds a similar view, claiming in Young v. Oklahoma, “An instruction containing the phrase ‘a doubt that you can give a reason for’ is wrong.” 20 How, you may ask, can it possibly be reversible error – sufficiently serious to overturn a conviction – to tell jurors that a reasonable doubt is a doubt for which a juror has and can give a reason? Some of the thinking behind the tendency to see this reading of BARD as incorrect comes from an appellate court’s reviewof another case, Morgan v. Ohio, in which the judge had used the “doubt-for-which-you-have-a-reason” language. The appellate court asked: What kind of a reason is meant? Would a poor reason answer, or must the reason be a strong one? Who is to judge? The definition fails to enlighten, and further explanation would seem to be needed to relieve the test of indefiniteness. The expression is, also, calculated to mislead. To whomis the reason to be given? The juror himself? The charge does not say so. And jurors are not required to assign to others reasons in support of their verdict. 21 A Pennsylvania appeals court inclines the same way: In order to be “reasonable,” the doubt must be substantial as opposed to fanciful, but it is not essential that a juror be able to give “some proper reason for entertaining it”; it may exist without his being able to formulate any reason for it. . . . 22 Or we find this similar gem of wisdom from Pennsylvania v. Baker: “A reason- able doubt may exist in the mind of a juror without his being able to formulate a reason.” 23 Some courts fret that it might put undue psychological pressure on a juror to tell him that he needs some reason for his doubts before he can consider themto be reasonable. The U.S. Second Circuit Court voices a concern that this instruction might “intimidate a juror by suggesting that he may be called upon 18 Burnett v. State, 86 Neb. 11 (1910). 19 Carr v. State, 23 Neb. 749, at 750 (1888). 20 Young v. State, 1962 OK CR 70 (1962). 21 Morgan v. State, 48 Ohio St. 371, at 376 (1891). 22 State v. Dauphinee, 121 Pa. Super. 565, at 590 (1936) (emphasis added). 23 State v. Baker, 93 Pa. Super. 360 (1928). P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 42 the distribution of error to explain his doubts, although it surely does not require him to justify them.” 24 Such psychobabble, coming from a politician, might be in keeping with the apparent need for leaders to be seen to “feel the pain” of the electorate. But it is scarcely what we expect from our courts when insisting that jurors make sure that their acquittal is founded on conviction beyond a reasonable doubt. If a juror feels doubtful about guilt but cannot even identify or formulate the reason for that doubt, then how can she possibly decide whether the doubt in question is rational or irrational? A second, to my mind equally specious, line of thinking about why reason- able doubts needn’t be doubts for which you have a reason is linked to the notion that the accused in a criminal trial is not obliged to mount a defense. Requiring a juror to have a reason for acquittal, it is said, is tantamount to requiring the defendant to present a case. Here is the argument of the appellate court in Iowa v. Cohen: Who shall determine whether [a juror is] able to give a reason, and what kind of a reason will suffice? To whom shall it be given? One juror may declare he does not believe the defendant guilty. Under this instruction, another may demand his reason for so thinking. Indeed, each juror may in turn be held by his fellows to give his reasons for acquitting, though the better rule would seem to require these for convicting. The burden of furnishingreasons for not findingguilt established is thus cast onthe defendant, whereas it is on the state to make out a case excluding all reasonable doubt. Besides, jurors are not bound to give reasons to others for the conclusion reached. 25 Precisely the same argument is found in an early, influential Indiana case. There, the appellate court found incorrect an instruction that jurors must have some reason for such doubts as they entertain about the defendant’s guilt. To the contrary, the appellate judges insisted: Such an instruction as the one we are considering can, we think, only lead to confusion, and to the detriment of the defendant. A juror may say he does not believe the defendant is guilty of the crime with which he is charged. Another juror answers that if you have a reasonable doubt of the defendant’s guilt give a reason for your doubt. And, under the instruction given in this cause, the defendant should be found guilty unless every juror is able to give an affirmative reason why he has a reasonable doubt of the defendant’s guilt. It puts upon the defendant the burden of furnishing to every juror a reason why he is not satisfied of his guilt, with the certainty which the law requires, before there can be a conviction. There is no such burden resting on the defendant or a juror in a criminal case. 26 24 U.S. v. Davis, 328 F.2d 864, at 867 (1964). 25 State v. Cohen, 108 Iowa 208, at 214 (1899). 26 Siberry v. State, 133 Ind. 677, at 685 (1893). Since it requires only one juror with reasonable doubts to prevent a conviction, it is quite unintelligible why the justices say that “the defendant should be found guilty unless every juror is able to give an affirmative reason why he has a reasonable doubt of the defendant’s guilt” (emphasis added). P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 The Unraveling of Reasonable Doubt 43 Ultimately, this analysis tears asunder the distinction between doubt and reasonable doubt. We can surely grant that a juror, after hearing all the evi- dence in a case, may still say that she has lingering doubts about the guilt of the accused. What I think that juror cannot possibly do is to certify that those doubts are reasonable doubts unless she can actually formulate them to herself and scrutinize them, preferably with the aid of fellow jurors. If a juror cannot put her finger on any element in the prosecution’s case that seems less than fully convincing, then it is impossible, even for her, to figure out whether her hesitation to accept a verdict of guilty is rational or not. Obviously, the prosecu- tion should not be burdened with having to eliminate all possible doubts from the minds of jurors, only their reasonable doubts. Indeed, that was the motive two hundred years ago for replacing the older criterion (“acquit if you have any doubt about guilt”) with BARD. To the extent that a juror’s doubt cannot be expressed, it cannot be parsed along the axis that separates the rational fromthe irrational. As for the analogy in these passages between reason giving by a juror and defense constructing by the accused, the parallel is a red herring. Requiring a juror to have some reason for doubting the guilt of the defendant in no way requires that reason to have come from testimony offered by the defendant or even from arguments offered by his counsel. Jurors may disbelieve some prosecution witness or doubt some expert offered by the state. Requiring them to be able to express such doubts to fellow jurors, or at least to themselves, is perfectly compatible with the presumption of innocence and the placing of the full burden of proof on the prosecution. Fortunately, not all U.S. courts have been so quick to say that you have a reasonable doubt about guilt even if unable to give any reasons for that doubt. The Supreme Court of Wisconsin has held specifically that “An instruction in a criminal case that a reasonable doubt is a doubt for which a reason can be given based on the evidence in the case, is correct,” adding for good measure, “A doubt cannot be reasonable unless a reason therefor exists, and, if such reason exists, it can be given.” 27 More’s the pity that this exemplary and rudimentary requirement is not the prevalent one, since it emphasizes that jurors must seek reasons for their decisions. Instead, it stands directly opposed to the view of the U.S. Supreme Court that this is an incorrect construal of BARD. 28 I cannot close out this discussion on jury instructions (and misinstructions) on BARD without including my favorite one. Properly deemed unacceptable 27 Butler v. State, 102 Wis. 364 (1899). 28 It is worth remarking, even if one is unsure what to make of it, that the American legal system widely operates on the assumption that the principal players need not give (and perhaps need not even have) reasons for their decisions. Consider, for instance, that counsel can exclude jurors without giving a reason, that trial judges are obliged to give no expla- nation for rulings on objections, that jurors need give no justification for their verdict, and that appellate courts can decline an appeal without offering any rationale. P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 44 the distribution of error by a higher court, it comes from a New York State trial judge. Here is how he instructed a jury about what a reasonable doubt is: It is not a doubt based upon sympathy or a whim or prejudice or bias or a caprice, or a sentimentality, or upon a reluctance of a weak-kneed, timid, jellyfish of a juror who is seeking to avoid the performance of a disagreeable duty, namely, to convict another human being of the commission of a serious crime. 29 5. BARD as High Probability According to this approach, far more popular with legal scholars and the gen- eral public than among judges, the right way to understand belief beyond a reasonable doubt is to say that any such beliefs must be highly probable. One obvious inspiration for this approach is found in the civil law standard of proof. There, courts agree that you should find for one party rather than the other if the “preponderance of the evidence” favors the former. That is normally taken to mean that the trier of fact should favor plaintiff over defendant if the case of the plaintiff is more likely than not, that is, if it has a probability greater than 0.5. The criminal standard, of course, is meant to be much higher than the civil standard, so the relevant probabilities would have to be much higher for conviction (0.9 or 0.95 are commonly cited unofficial estimates). Proponents of this construal of BARD argue that judges should instruct juries that they must acquit unless their confidence or degree of belief in the guilt of the accused is close to certainty (in numerical terms, close to 1.0). Judges, as I have already insinuated, generally reject this formulation of BARD out of hand. It is worth investigating with some care why they do so, especially since, in the next chapter, I will be exploring a version of this approach to the standard of proof. Probabilities, of course, run along a scale from 0.0 to 1.0. It is customary (if not strictly correct) to identify a probability of 1.0 with complete certainty. Since (as BARD recognizes) we cannot be completely certain about anything to do with human affairs, this entails that jurors can never be fully certain of the guilt of an accused party. This naturally invites the suggestion that the standard for conviction could be defined by specifying a threshold probability, a numerical degree of confidence, that belief in guilt must reach before conviction (in both senses of this term) can be justified. Although lower court judges from time to time explain BARD to jurors in these terms, virtually no appellate court in the land will endorse such a gloss of reasonable doubt. On the contrary, they stoutly resist any such quantification. Reversing a conviction in a trial where the judge had defined BARDin probabilistic language, an appeals court in Massachusetts insisted, “The idea of reasonable doubt is not susceptible to quantification; it 29 People v. Feldman, 71 N.E.2d 433, at 439 (1947). P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 The Unraveling of Reasonable Doubt 45 is inherently qualitative.” 30 The Supreme Court of Nevada agrees: “The idea of reasonable doubt is inherently qualitative. Any attempt to quantify it may impermissibly lower the prosecution’s burden of proof, and is likely to confuse rather than clarify.” 31 These, of course, are not arguments but bare assertions. Moreover, this is no argument against using the language of probabilities per se but against a failure to specify that a very high probability is required to convict. The Nevada court’s observation that a judge might define the probability at too low a level is readily remediable by setting the requisite probability quite high. The same short-sightedanalysis afflicts a dissentingopinionof Justices Harry Blackmun and David Souter on the high court when they argued that: The word “probability” brings to mind terms such as “chance,” “possibility,” “like- lihood” and “plausibility” – none of which appear [sic] to suggest the high level of certainty which is required to be convinced of a defendant’s guilt “beyond a reasonable doubt”. 32 Although the bare term “probability” may (but probably does not) have the implications that worry Blackmun and Souter, the phrase “high probability” suggests nothing so weak as mere possibility or plausibility or even bare likeli- hood. Because these arguments are so shabby, I think that the real worry among jurists about specifying probabilities must be sought elsewhere. One source of their worry is this: Any specification of a degree of belief necessary for a finding of guilt (such as 95 percent confidence) involves an explicit admission that wrongful convictions will inevitably occur. For instance, if jurors could somehowdiscover that they had a confidence of 95 percent in the guilt of the accused, this would generally suggest that as many as one in every twenty innocent defendants will be wrongly convicted. While acknowledging in the abstract that no method of proof is infallible and thus admitting in principle that mistakes will occur from time to time, the judiciary has an entrenched resistance to any explicit admission that the system has a built-in tolerance for wrongful convictions. Some elements of this problem come out clearly in Justice William Brennan’s remarks in a case we have already discussed, In re Winship. 33 Brennan worries that the law would lose much of its “moral force” 30 Massachusetts v. Sullivan, 482 N.E.2d 1198 (1985). 31 McCulloch v. State, 99 Nev. 72, at 75 (1983). 32 From the dissent of Blackmun and Souter in Victor v. Nebraska (1994). 33 Justice Brennan wrote: It is critical that the moral force of the criminal law not be diluted by a standard of proof that leaves people in doubt whether innocent men are being condemned. It is also important in our free society that every individual going about his ordinary affairs have confidence that his government cannot adjudge him guilty of a criminal offense without convincing a proper fact finder of his guilt with utmost certainty (In re Winship, 397 U.S. 358, at 364 [1970]). According to a Lexis-Nexis search conducted by the author, more than fifty other appellate rulings since 1970 quote and endorse this incoherent sentiment of Brennan’s. P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 46 the distribution of error if it left room for the suspicion that, even if only rarely, “innocent men are being condemned.” To avoid this problem, he says, the ordinary man in the street must believe that he cannot be sent to jail unless his guilt has been proved “with utmost certainty.” (These are remarkable utterances in an age that has supposedly come to terms with the fallibility of human judgment. But we will leave that curious anachronism to one side.) What should be clear is that identifying BARD with any level of probability less than unity would explicitly acknowledge that the systemofficially condones a certain fraction of wrongful convictions. That, in turn, would supposedly threaten the ordinary person’s faith in the justice system. Better, it seems, to avoid any talk of probability as the standard for conviction than to acknowledge publicly that the system expressly permits incorrect judgments of guilt. Daniel Shaviro has argued that the criminal justice system goes to heroic lengths to reduce the appearance of the possibility of error, even while acknowledging that such a possibility is omnipresent. He suggests that this “may reflect the self-serving interest of lawyers [and, we may add, judges] in promoting the appearance of justice under the legal system,” rather than the true state of affairs. 34 Perhaps the most powerful formulation of the argument against defining reasonable doubt as a specific level of probability comes from the influ- ential English legal writer Thomas Starkie, who put the point this way in 1824: To hold that any finite degree of probability shall constitute proof adequate to the con- viction of an offender would in reality be to assert that, out of some finite number of persons accused, an innocent man should be sacrificed for the sake of punishing the rest – a proposition which is as inconsistent with the humane spirit of our law as it is with the suggestion of reason and justice. 35 A third problem with making BARD explicitly probabilistic is the long-held legal dogma that conviction of the accused requires that a juror’s state of mind be one of “firmbelief in guilt,” that is, the juror must be “fully convinced.” How, you may ask, can you be fully convinced of a proposition that you believe to have a likelihood of (for example) 95 percent? Indeed, the 5 percent uncertainty implies that the juror is less than “fully convinced.” But according to this way of thinking, nothing short of full certainty, a probability of 100 percent, should be enough to convict. No one believes that kind of proof is available. Finally, even if the worries about the broader social message being sent by the use of probabilities could be quelled, it is quite unclear what level of 34 Daniel Shaviro, Statistical-Probability Evidence and the Appearance of Justice, 105 Harv. L. Rev. 531 (1989). 35 Thomas Starkie, Evidence 507 (1824). P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 The Unraveling of Reasonable Doubt 47 probability should be associated with BARD. An English study thirty years ago interviewed judges and jurors about the level of probability that should be required for conviction in a criminal trial. Among judges, fully a third located it between 0.7 and 0.9, almost all the rest pegging it above 0.9. Among jurors, by contrast, 26 percent were willing to convict on probabilities below 0.7 and a bare majority (54 percent) thought that probabilities of 0.9 and above should be required for conviction. 36 The idea that BARD can be associated with a determinate meaning even while judges and jurors locate it at such different points on the scale of probabilities should be a grave worry. In the next chapter, we will examine at length whether the standard of proof can be somehow grafted onto the scale of probabilities. For now, I will leave the last word on this subject to Lawrence Tribe, who insisted, in a classic article on the political importance of not quantifying criteria of proof, that: [BARD] signifies not any mathematical measure of the precise degree of certitude we require of juries in criminal cases, but a subtle compromise between the knowl- edge, on the one hand, that we cannot realistically insist on acquittal wherever guilt is less than absolutely certain, and the realization, on the other hand, that the cost of spelling that out explicitly and with calculated precision in the trial itself would be too high. 37 The Ultimate Act of Desperation: Avoiding Clarification The coexistence of several different definitions of BARD in itself is not neces- sarily a bad thing. After all, most complex notions can be defined in a variety of ways. What is troublesome is that most of these definitions fail to characterize the same underlying idea. Believing firmly in, or having an abiding conviction about, someone’s guilt is not the same as having no doubts for which you can give a reason. And neither of these comes to the same thing as a belief on which you would base important actions in your life. What we face here are not different glosses on the same notion but fundamentally different conceptions of the kind and level of proof necessary to convict someone. To make matters worse, courts have faulted all these definitions as either wrong or misleading or unintelligible. Versions that some courts have found acceptable, even exem- plary, have been dismissed by other courts as violating the constitutional rights of the accused. This situation has prompted many appellate courts to insist that trial judges should not define BARD for jurors in their instructions. At least ten states now hold this view. In some (such as Oklahoma), if a judge offers a jury any 36 John Cohen and Ian Christensen, Information and Choice 62 (1970). 37 Laurence H. Tribe, Trial by Mathematics: Precision and Ritual in the Legal Process, 84 Harv. L. Rev. 1329, at 1375 (1971). P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 48 the distribution of error explanation of what reasonable doubt is, this is automatic grounds for reversing a conviction. 38 By contrast, another fifteen states require judges to define BARD for jurors. We see the same confusion at the federal level. Four of the eleven U.S. circuit courts require a definition, failure to give one being grounds for reversal. Most of the rest hold that judges needn’t define rational doubt. Virtually all appellate courts warn that most efforts to explain BARDeither create confusion in the minds of jurors or else mischaracterize the appropriate burden of proof that is the essence of BARD. As the Seventh Circuit Court of Appeal put it in 1982, “We have repeatedly admonished district courts not to define ‘reasonable doubt’.” 39 Six years later, and faced with lower-court judges who were still explaining reasonable doubt to perplexed jurors, the judges of the Seventh Circuit were even more insistent, asserting that “no attempt should be made to define reasonable doubt.” 40 They argued: “Reasonable doubt” must speak for itself. Jurors know what is “reasonable” and are quite familiar with the meaning of “doubt.” Judges’ and lawyers’ attempts to inject other amorphous catch-phrases into the “reasonable doubt” standard, such as “matter of the highest importance,” only muddy the water. . . . It is, therefore, inappropriate for judges to give an instruction defining “reasonable doubt,” and it is equally inappropriate for trial counsel to provide their own definition. 41 The idea that reasonable doubt must already be clear to all because ordinary people understand each of its two constituent terms is laughable. “Reasonable doubt,” like many other compound terms of art (think of “civil servant” or “black box”), carries a freight not implied by either of its constituents. Recall that many courts have argued that reasonable doubt is not to be understood as a doubt for which you can give a reason. Because that is its most natural rendering in ordinary language, this alone is sufficient to undermine the Seventh Circuit’s claim that the meaning of BARD can be simply “read off ” from its constituent terms. If this argument for avoiding definition is unconvincing, there is a long queue of appellate judges eager to invent other reasons for not defining BARD. Thus, Douglas Woodlock, U.S. district judge for Massachusetts, insisted that it is “a basic philosophical precept that the concept of reasonable doubt has an a priori existence in the minds of all jurors.” 42 If we all understand it a priori, there’s 38 The case establishing that in Oklahoma any explanation of reasonable doubt is a reversible error is Pannell v. Oklahoma, 640 P.2d 568 at 570 (1982). For Wyoming’s similar rule, see Cosco v. Wyoming, 521 P.2d 1345 (1974), at 1346. 39 U.S. v. Martin-Tregora, 684 F.2d 485, at 493 (7th Cir. 1982). 40 U.S. v. Hall, 854 F.2d 1036, at 1039 (7th Cir. 1988). The shifting reactions of appellate courts to the use of BARD instructions is discussed in detail in Jessica Cohen, The Rea- sonable Doubt Jury Instruction: Giving Meaning to a Critical Concept, 22 Am. J. Crim. L. 677 (1995). 41 U.S. v. Glass, 846 F.2d 386 (1988). 42 Smith v. Butler, 696 F. Supp. 748 at 765 (1988). P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 The Unraveling of Reasonable Doubt 49 obviously no need to define it. (Perhaps federal judges should be discouraged from reading Immanuel Kant, or whoever it was who gave Woodlock this silly notion.) In 1994, the Fourth Circuit Court held that, even when the jury asks the judge to define “reasonable doubt,” the judge may properly refuse to do so. 43 The Supreme Court, for its part, has never made up its mind whether reasonable doubt should be defined, holding, as Justice O’Connor wrote in 1994, that the Constitution is noncommittal about whether trial judges may or must define reasonable doubt. 44 It is more than mildly curious that such a large body of judicial opinion would favor a policy according to which juries must be told that conviction for a crime requires belief in guilt beyond a reasonable doubt but must not be told what BARD is, not even when they request a clarification. The explicit ratio- nales for discouraging a definition of BARD avoid admitting that the notion itself is deeply confused. The most common excuse is that BARD is “self- evident” or “self-defining,” and thus not in need of further commentary. Thus, the Seventh Circuit has gone on record that the idea of reasonable doubt is so transparent that definitions of BARD – attempts, as they put it, to make “the clear more clear” – can only confuse. 45 But this is false on its face. Juries frequently request that judges explain to them what reasonable doubt is. That would not occur were it clear and self-evident. More to the point, we have already seen that different judges and different legal jurisdictions have profoundly different understandings of BARD. If judges cannot agree among themselves about this crucial notion, and it is clear that they cannot, it is a dan- gerous act of self-deception (or worse) to suggest that lay jurors, completely unschooled in the law, will have some common, shared understanding of this doctrine. We also have impressive, direct evidence that jurors are deeply confused about BARD. In a study of some six hundred Michigan jurors, Geoffrey Kramer and Doorean Koenig discovered that a quarter believed that “you have a reason- able doubt if you can see any possibility, no matter howslight, that the defendant is innocent.” 46 Not surprisingly, roughly the same proportion of jurors agreed that “to find the defendant guilty, beyond a reasonable doubt, you must be 100 percent certain of the defendant’s guilt.” Appellate courts have rejected both these construals countless times, because not every doubt is a reasonable doubt, nor need jurors be fully certain of guilt. Another study of jurors in Florida discovered that one in four of them believed that when the weight of evidence was equally balanced between prosecution and defense, the defendant should 43 See U.S. v. Reives, 114 S. Ct. at 2679 (1994). 44 Victor v. Nebraska, 114 S. Ct. at 1243 (1980). 45 U.S. v. Lawson, 507 F.2d 433, at 422 (7th Cir. 1974). 46 Geoffrey Kramer and Doorean Koenig, Do Jurors Understand Criminal Jury Instructions? U. Mich. J. L. Ref. 401, at 414 (1990). P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 50 the distribution of error be found guilty! Only half realized that the defendant was not obliged to offer proof of his innocence. 47 How can jurors rationally apply a standard of proof to the evidence when they have such divergent interpretations of that standard and when so many courts refuse to clarify it for them? If judges can tell jurors no more than that a guilty verdict requires BARD, it is likely – given numbers such as those just cited – that almost every jury will contain jurors who bring inappropriate construals of that standard to the table. In such circumstances, failure to clarify the meaning of BARD to jurors is unconscionable. It is difficult to resist the conclusion that what really drives the growing reluctance to define BARD is not so much the fiction that its meaning is clear to all. It is instead the judiciary’s worry that the coexistence of conflicting definitions of BARD – of the sort that would vividly show up if judges were free to explain BARD to jurors – would bring to public notice the fact that no univocal sense can be attached to the phrase “beyond a reasonable doubt.” Rather than concede that fact and do something to remedy it, many courts – especially at the federal level – seem disposed to paper over the real differences that exist by pretending that the conception requires no further explication or clarification. A system founded on the idea that justice can be secured simply by having the judge, with a wink and a nod, tell jurors, “you already know what reasonable doubt means,” is in urgent need of overhaul. There is doubtless another motive driving the growing disinclination of courts to instruct jurors in the subtleties of BARD. I refer to the fact that, if a judge says something to the jury about BARD with which a higher court might disagree, and if the accused is subsequently convicted and the case is appealed, then the guilty verdict may be overturned. It is better, apparently, to say nothing – leaving the problem unresolved, the guilty verdict in place, and the jury perplexed – than to say something that might trigger a successful appeal. A vacuous BARD instruction is not likely to be reversed, while one with content may well be. Whatever the precise combination of motives for discouraging explanations of BARD, we have every ground for suspecting that there is no uniformstandard for determining guilt in contemporary American criminal law courts. As the justices of the U.S. Ninth Circuit sagely noted: [t]he term [sic] beyond a reasonable doubt . . . may be in common usage by the populace of this nation, but there is no demonstrable or reliable evidence . . . that a reasonably appropriate definition is in common usage or well understood by prospective citizen jurors. 48 47 David Strawn and Raymond Buchanan, Jury Confusion: A Threat to Justice, 59 Judica- ture 478, at 480–1 (1976). 48 U.S. v. Witt, 648 F.2d 608, at 612 (9th Cir. 1981) (Anderson, J., concurring). P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 The Unraveling of Reasonable Doubt 51 Where the standard of proof is concerned, our criminal justice system is blind, not in the customary and commendable sense of being impartial, but in the more literal sense of not knowing what it is doing or where it is going. The wheels of justice keep grinding, juries keep producing verdicts, but the deliberate obscurity and obfuscation surrounding the standard for conviction do little to inspire confidence in the fairness of the system. Taking a Step Back from the Abyss Most critics familiar with the many twists and turns in the recent handling of BARD by the courts seek the remedy in a newly crafted set of jury instructions that will, in plain language, explain to jurors what the idea of (beyond a) rea- sonable doubt means. I have to part ways with my fellow critics at this point. I believe that the proper moral to draw from this sordid story is that no canonical notion of reasonable doubt is on offer. The sad fact is that we have been expecting one doctrine to do multiple tasks. To begin with, we want it to stress to jurors that the burden of proof in criminal proceedings falls on the state, not the defendant. Then, we have used it to warn jurors that they must not let exaggerated, hyperbolic doubts stand in the way of conviction. We have used it to make clear that guilty verdicts in criminal trials must depend on much higher levels of proof than those associated with civil actions or practical life. We have used it to impress on jurors just how much is at stake and how somber must be the decision when they decide to send people to prison, depriving them of their liberty and blackening their good name. Finally, we have used it to secure – insofar as possible – uniformity of standards, making sure that every criminal verdict employs the same bar for conviction. On most of these scores, BARD is failing. It is time to try to diagnose the causes of that failure. It will probably not have escaped your notice that almost all the familiar glosses on BARD(as moral certainty, as firm or abiding conviction, as a belief on which you would base important actions, and so on) define it in terms of the target mental state of the juror. He or she must be “firmlyconvinced,” “almost certain,” “fullypersuaded,” witha “satisfiedconscience” about guilt before a vote for convictionis indicated. It is no accident that these glosses focus on the subjective state of the juror. On several occasions, the Supreme Court has underscored its belief that the right way to characterize the presence or absence of reasonable doubt is in terms of the “subjective state of mind” that jurors should be in if they are to condemn or acquit the accused. 49 This focus on the juror’s mental state is more than a little curious. Suppose we tried to teach young scientists howto judge when a theory was acceptable by 49 See especially Jackson v. Virginia, 443 U.S. 307, at 315 (1979), and In re Winship, 397 U.S. 358, at 364 (1970), which both speak of BARD as denoting a “subjective state.” P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 52 the distribution of error telling themwhat their mental state should be when they should accept a theory. Such advice would be seen as ludicrous. Try to imagine a mathematician saying that he has proved a new theorem and that the proof consists in the fact that he believes the theorem without the slightest hesitation. His colleagues would be aghast since what establishes a mathematical theorem as a theorem is the robustness of its proof, not the confidence of its discoverer. Putting the point a bit more philosophically, we can note that, in most of the areas of life where inquiry after the truth is at stake, the advice offered to ensure that such inquiry is rational specifies the kinds of evidence, tests, or proofs necessary for a well-founded belief. Such advice explains (for instance) when the evidence supports a hypothesis well or poorly or when it is proper to infer a certain conclusion from given premises. Yet, during the last two cen- turies, Anglo-American criminal law has assiduously avoided saying anything specific about the kind of evidence needed for a conviction. (Contrast this with European, Talmudic, or canon law, which, for hundreds of years, specified precisely how much and what kind of evidence was needed to convict.) One reason that the various characterizations of BARD we have surveyed seem so vacuous or unsatisfactory is that they studiously avoid talking about the structure of proof or about the kind of case the prosecution must present. Instead, they focus single-mindedly on the mental state of the juror. True, every jury is enjoined to attend carefully to the evidence and to weigh it, but that process is regarded as a black box, something that happens behind the closed door of the jury room, as if that territory ought not be invaded by a judge telling the jury what a compelling case would look like. Instead, the judge in effect simply says, “Open your mind to all the evidence and then see if you are fully persuaded (or firmly convinced, and such) at the end of the trial.” This formula fails to acknowledge that persuasion is a process of reasoning through the evidence. Judges gives jurors few hints as to how they should do that. This would be forgivable if we could suppose that all, or even most, com- petent jurors already knew how to engage in what philosophers call ampliative reasoning, which leads from the disparate, purported facts presented as evi- dence to a conclusion about the guilt of the accused. If everyone had a “natural” disposition to draw legitimate inferences about complex matters of fact, such information would be superfluous. But there is nothing innate about this. Sci- entists took about two thousand years to figure out how to test their theories correctly. The logic governing clinical trials of drugs emerged only in the twen- tieth century, after eons of the use of medications on only the flimsiest empirical basis. Lawyers, too, have learned much about what distinguishes a powerful case for the prosecution from a weak one. I submit and will argue in detail in the next chapter that the bar for conviction would be better defined in terms of the features of the case needed to convict, not in terms of the jurors’ inner P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 The Unraveling of Reasonable Doubt 53 states of mind.This is especially important since the latter – if not disciplined by certain guidelines about the appropriate logical connections between evidence and verdict – are apt to be ill founded, prejudicial, and irrational, even when they lead to a firm belief in guilt. There is a different way of approaching the same point. Whatever else BARD might mean, it insists that juries should not convict the accused if they have a rational doubt about his guilt. Subjective criteria of the sort that appellate courts favor get at only half of this demand. They tell the juror that he must have no doubts. But, precisely because they are purely subjective criteria, they cannot address the question about the rationality of the juror’s confidence, or lack of confidence, in the guilt of the accused. The required discrimination, between rational and irrational doubts, can never be resolved if one remains focused exclusively at the level of the juror’s degree of conviction. What distinguishes a rational doubt from an irrational one is that the former reacts to a weakness in the case offered by the prosecution, while the latter does not. Let us remember that in a trial by jury, the jurors are supposed to be “the finders of fact.” They are the ones who have to decide whether a crime was committed and whether the defendant committed it. The prosecution will gen- erally lay out a “theory” about the crime, that is, a narrative of supposed events in which the defendant participated. The prosecution, if it is savvy, will adduce evidence in the form of testimony or documents or physical evidence for every key part of its theory. The aim of presenting the evidence is to corroborate the prosecution’s theory of the crime. For its part, the defense may (but is not obliged to) present an alternative theory of the events associated with the crime. At a minimum, the defense will seek to identify weak points or implau- sibilities in the prosecution’s narrative. This critique may involve challeng- ing the evidence or testimony or it may focus on the large inferential leaps required to move from the evidence presented to the thesis that the defendant is guilty. The task of the jury is to evaluate the prosecution’s theory or story, deciding whether it points unambiguously to the conclusion that the defendant is guilty. In making that decision, jurors must decide whether the evidence presented is powerful enough to justify them in concluding that the defendant committed the crime. The principal question is not whether the jurors, individually and collectively, are convinced by the prosecution. The issue is whether the evidence they have seen and heard should be convincing in terms of the level of support it offers to the prosecution’s hypothesis that the defendant is guilty. The key point is that the question, “How strongly does the evidence support the theory of guilt?” is a question about relations between (statements describ- ing) events, not merely or primarily a question about the subjective state of the jurors’ minds. The issue at trial should not be whether, as a contingent matter of fact, the twelve jurors are all fully convinced of the guilt of the accused. P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 54 the distribution of error Instead, it should be, “Does this evidence strongly support the theory that the defendant is guilty?” Putting it differently, “Would a rational and sober-minded person, confronted with this evidence, find that it made a compelling case for guilt?” Better still: “Is it remotely likely that an innocent person would have an evidence profile that looked like the one presented at trial?” For jurors to be able to answer such questions, it is of little or no avail to tell them that they must be firmly convinced or have an abiding conviction of guilt. Rather, they must be able to say to themselves and to one another precisely why the evi- dence at hand does or does not point unambiguously to the guilt of the accused. Doing so requires them to make a whole series of decisions about whether the evidence supports each key claim of the prosecution’s theory to the requisite degree. What the jurors may well be unclear about is howto tell whether the evidence provides powerful or only mild support for the prosecution. For instance, is one credible witness to a crime sufficient to convict? Is circumstantial evidence potentially as powerful as eyewitness testimony? Are some sorts of physical evidence more damning than others? And so on. Given the practical demands of the criminal justice system, the judge cannot give a mini-course in inductive inference to every impaneled jury. Nevertheless, I think that some things that could be said briefly that would provide models or paradigms of powerful forms of argument. Armed with these paradigmatic examples, jurors could go to the jury room ready to struggle with the question of whether the case in hand has the sorts of features that make up powerful proof. I have belabored the question of the standard for conviction at such length because the ambiguities surrounding it are among the largest impediments to ensuring that the criminal justice systemacts fairly toward all defendants and in accordance with society’s views about how high we wish to set the bar before we send someone to prison. “Beyond reasonable doubt” has become a mantra rather than the well-defined standard of proof that it once was. Divested of its entire substantive content by a series of well-meaning but ill-advised judicial findings, it now serves at best as an admonition to jurors not to take their tasks lightly. What it does not do in its present form is to provide the slightest clue to jurors as to what a convincing proof of guilt should look like. Would an Improved BARD Suffice? Let us suppose, however, that this problem could be solved by some judicious characterization of the kind of evidence and argument necessary to convict someone of a crime. Such an instruction (echoing the doctrine of moral cer- tainty) might speak of “independent lines of argument leading to the same conclusion,” or, following Ron Allen’s recent suggestions, of “the absence of any plausible story reconciling defendant’s innocence with the evidence P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 The Unraveling of Reasonable Doubt 55 presented at trial.” 50 In the next chapter, I propose several other ways of defin- ing the standard of proof. But even supposing that some such solution would clean up the worst of the current morass, we still need to ask ourselves whether a more precise version of BARD, or something similar to it, would serve the aims of criminal justice. In particular, I want to ask (without definitively resolving the issue) whether it makes sense to utilize the same standard – whether BARDor something else – for all crimes. Where standards of proof are concerned, I am not convinced that one size fits all. To begin the discussion of this heretical question, it would be well to remind ourselves of two salient historical facts: a) When BARD came into widespread use, all felonies carried the same punishment, and b) that punishment was death. One standard made sense because the cost of a mistaken conviction was the same in every trial for every felony offense. With the abandonment of capital punishment as the customary sentence, the sensible policy emerged of varying the punishment, according to the severity of the crime. Nowadays, the idea that the punishment should fit the crime is universally accepted. This should give us some pause about BARD for two reasons: First, many crimes and their punishments are so minor that the BARD standard seems inappropriately exacting; second, even among serious crimes, BARD – not admitting of degrees (since one either has a reasonable doubt or one doesn’t) – seems a very blunt instrument for determining guilt. I want to look at these points in turn: 1. It is common knowledge that a great many crimes now carry punishments that are no more than fines. Others involve nothing more than unsupervised probation or parole or relatively brief times of incarceration. Are the costs of mistaken guilty verdicts in these cases so steep that we still want to insist on acquitting ten innocent defendants for every false conviction? Hunches about the relative costs of mistakes, perhaps crystal clear in the case of capital crimes, become clouded when we turn to much milder punishments. This difference invites the proposal that the standard for conviction, instead of being the same for every crime from homicide to shoplifting, might – like the punishment – vary with the severity of the crime. After all, many crimes now carry punishments less harsh than one’s potential liability in civil cases, where the standard of proof is simply “more probable than not.” In a civil court, you can be sued for your life savings if you are shown to be probably liable for harm to some third party. You can be denied your parental rights, committed indefinitely to a mental institution, and deprived of your citizenship. Are these outcomes less severe than being found guilty 50 See especially Ron Allen, Constitutional Adjudication, the Demands of Knowledge, and Epistemological Modesty, 88 Nw. U. L. Rev. 436 (1993); Ron Allen, Factual Ambiguity and a Theory of Evidence, 88 Nw. U. L. Rev. 604 (1994). P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 56 the distribution of error Figure 2.1. Familiar Standards of Proof as Probabilities of felony drunk driving and facing a year’s probation? If they are not, what is the sense in holding to such different standards of proof in the two cases? 2. It is also worth asking whether it is an efficient use of money and other resources to require the state to mount the same sort of proof to send an embezzler to jail for a year as it invests to convict a serial killer. Of course, it is not easy to see how to construct a graded scale of standards of conviction that might be mapped onto a scale of criminal severity. Essentially, the U.S. court system officially recognizes only three standards of proof: BARD, proof by the preponderance of the evidence (hereinafter, POE), and proof by clear and convincing evidence (hereinafter, CACE). If courts were disposed to think of standards of proof as probabilities (which they are not), you could imagine at least in principle having a graded series of probabilities associated with increasingly serious crimes, perhaps reserving BARDfor only the most serious. Since that proposal would probably meet fierce opposition – given the courts’ aversion to thinking of standards of proof as degrees of probability – a more realistic possibility might be to use the “intermediate” standard of proof, CACE, for lesser crimes such as misdemeanors and also the less serious felonies. To keep the differences between these three standards firmly in mind, it helps to represent them diagrammatically, as shown in Figure 2.1. If you like, you can explain this scale to yourself as a probabilistic one, ranging from zero to one. On that interpretation, BARD would probably fall somewhere around 0.9 or 0.95 and higher; CACE would kick in at about 0.75; and PoE would be anything above 0.5. If, like me, you find it hard to think of standards of proof as probabilistic degrees of belief, you might want to construe this diagram as representing different weights of evidence and proof. As the diagram shows, CACE, like BARD, is acquittal-friendly in that it requires the state to establish much more than the bare probability of guilt. It has the added virtue of presumably being intelligible to juries and familiar to judges, since civil trials already make frequent use of it. Likewise, CACE is fully compatible with the presumption of innocence. Aproposal to use CACEas the standard in lesser felony cases would require the Supreme Court to reverse its famous opinion, In re Winship, that all criminal defendants are entitled to be tried according to BARD; however, as we have already seen, there are powerful, independent reasons to believe that the framers of the Constitution P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 The Unraveling of Reasonable Doubt 57 never contemplated laying down a specific standard for criminal conviction, let alone using BARD – not the least of these being that BARD was not a part of the common law when the Constitution was adopted. The Chilling Effect of Death Penalty Jurisprudence on a Candid Discussion of the Standard of Proof The history of Anglo-American criminal law and of the BARD standard is inexorably bound up with capital punishment. Any suggestion that the stan- dard of proof might appropriately be less demanding than BARD instantly and appropriately draws the riposte that, where the death penalty is concerned, noth- ing less than the most exacting standard will do. Since it is likewise supposed (without argument) that there should be only one criminal standard of proof, arguments about BARD inevitably fixate on capital cases. This is unfortunate, not least because capital cases are extraordinarily rare. While capital punishment still exists in the United States, most felonies are no longer punishable by death, murder being the only significant exception. 51 Even here, the vast majority of murders are not punishable by death. Even among those that are, the death sentence is still highly unusual since prosecutors need not request it, and juries (or judges in bench trials) may hesitate to impose it, often preferring imprisonment to the ultimate sanction. Numbers probably speak louder than these generalizations. In 2001, according to the Department of Justice: Some 16,000 murders were committed in the United States. Some 150 felons received a death sentence. All are appealed, usually at three different levels. Most cases are reversed or sent back for retrial. Fewer (oftensubstantiallyfewer) than100convictedmurderers were actually executed. Clearly, if the argument for the death sentence is based on its supposed deter- rence of heinous crimes, it hardly constitutes powerful deterrence if would-be murderers know that their chances of actually suffering that fate hover around 1 in 250, and that the penalty will not be exacted, if at all, for at least a decade, often much longer. 52 Likewise, if you see the death penalty as an appropriate form of retribution for especially dastardly crimes committed against society, 51 In certain states, treason and airplane hijacking are also capital crimes. 52 This is about the same risk of death as that faced by an average motorist in twenty years of normal driving. Since the time lapse between committing a murder and being executed can be fifteen years or more, we should not expect the death penalty to deter significantly more would-be murderers than the prospect of dying behind the wheel of a car in the next two decades deters ordinary citizens from driving automobiles (see Larry Laudan, The Book of Risks 170 [1994]). P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 58 the distribution of error the very low rates of capital punishment make it clear that little tit-for-tat retri- bution is being exacted. Similar conclusions seemto followif one considers the economic costs to society of dealing with someone sentenced to death versus someone sentenced to life in prison. 53 What must interest us here is not the debate about the morality of capital punishment but the ways in which the existence of the death penalty warps and thwarts finding out the truth about crimes in general. Two issues, in particular, will concern us: One is the curious epistemology of capital crime prosecution; the other is the probably deleterious impact of the death penalty on the prose- cution of the much larger class of noncapital crimes. I will deal with them in that order. Almost everyone shares the intuition that there is a huge difference between condemning someone to death and committing him or her to life in prison without possibility of parole. (It could be argued that the second is a fate worse than death for many prisoners, but let us leave that possibility to one side.) The principal difference, from an epistemic point of view, is the irrevocability of the death sentence once carried out. If someone is erroneously sent to prison for life, there is always the prospect that the conviction, if false, can be undone by pardon or retrial, if suitably exculpatory evidence emerges. If someone is executed, it cannot be undone. It is a clich´ e in the law, but no less true for that, that the most powerful argument against the death penalty is the ever-present possibility of false convictions. Because of the irreversibility of mistaken executions, special machinery is in place for reviewing death sentences. Apart from the direct appeal to a higher court available to everyone convicted of a crime, there is a second layer of review in most states for capital cases called “post-conviction review.” For cap- ital convictions that survive these two layers of review, there is a third review at the federal level, known as habeas corpus review. At each of these levels, courts may (and frequently do) find sufficiently serious errors that the ver- dict is overruled or a new trial is ordered. Recent studies suggest that two of every three capital cases (68 percent) that pass thorough this appellate sieve are overturned. On its face, this would seem to mean that most capital tri- als are riddled with serious error; otherwise, they would not be reversed or remanded at such a high rate. A more plausible explanation (since nothing close to this rate of error shows up in appeals of noncapital murder cases) is that the judicial system properly regards the costs of a mistake in a capital case as so much dearer than the costs of mistakes in noncapital cases that every 53 “When posttrial review costs are factored in, the cost comparison between capital and noncapital cases is something like $24 million per executed prisoner, compared to $1 million for each inmate serving a life sentence without parole” (James S. Liebman et al., Symposium: Restructuring Federal Courts: Habeas: Capital Attrition: Error Rates in Capital Cases, 1973–1995, 78 Tex. L. Rev. 1839, at 1865 [2000]). P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 The Unraveling of Reasonable Doubt 59 conceivable benefit of the doubt is given to those condemned to die. The Supreme Court itself has acknowledged that death penalty cases call for special epistemic protections: The qualitative difference of death fromall other punishments requires a correspondingly greater degree of scrutiny of the capital sentencing determination. 54 Trial errors that probably would not sway the appellate court in an armed rob- bery case become straws at which justices, confronted with a capital case, grasp in their efforts to be as sure as humanly possible that the condemned defendant is guilty as charged. This meticulous process is extraordinarily costly, both to society and to defendants. The indirect epistemic implications of this situation are not hard to work out. The justice system has neither an unlimited amount of money nor an indefinitely large pool of legal talent to draw on. The scarcity of funds and the time of judges and prosecutors invested in death row appeals means that fewer resources available for investigating and prosecuting other crimes. The punctilious review of capital cases and the high rates of reversal speak of a system unsure that it can catch all the serious errors that may occur. The fact, of course, is that no systemcan catch all its errors. However, despite the recent hue and cry resulting from the discovery (often through new forensic techniques such as DNA analysis) that some death row inmates are genuinely innocent, the system is probably functioning better than we have any right to expect it should. According to the Death Penalty Information Center (DPIC) (not exactly a hotbed of advocates of capital punishment), between 1973 and 2005, some 119 U.S. death row inmates were exonerated. 55 During this same period, well over six thousand defendants received a death sentence. Even if we interpret all these “exonerations” as false convictions (and many of them probablywere not 56 ), this suggests a ratioof false convictions totrue convictions of slightly less than 2 percent. Even if the error rate were double what it appears to be, that would still mean that the system was right in convicting murderers about 96 percent of the time – more often than one would expect, given the frailty of human knowledge. Of course, we would like to do better than that, but, where human affairs are concerned, this is doing about as well as we know how to do. Holding out 54 California v. Ramos, 463 U.S. 992, at 998–9 (1983). 55 For the most current data on exoneration rates, see the DPIC website at www. 56 Consider one sample case of “exoneration” according to the quasi-official list. In 1981, Robert Cruz was convicted of two murders in Phoenix in 1980. This conviction, on appeal, was overturned on a technicality. He was tried again. The jury was hung. He was retried, producing a second hung jury. At his fourth trial, Cruz was convicted, but that verdict was overturned because there were no Hispanics on his jury. Finally, at his fifth trial, he was acquitted. This sordid story (two convictions, two mistrials, and one acquittal) does not provide impressive evidence for the claim that the long-suffering Cruz is probably innocent m . (For details, see State v. Cruz, 857 P.2d 1249 [Ariz. 1993].) P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 60 the distribution of error hope that the errors can be reduced without limit is a fool’s game. Either we acknowledge that we are prepared to tolerate some realistic degree of error in capital cases or else we eliminate capital punishment. The current charade – that if we simply pass a case through enough filters, we will catch all the errors – should be an embarrassment to all who are propounding it. Since the criminal standard of proof, BARD, emerged at a time when guilty felons uniformly faced the death penalty, the standard for conviction was prop- erly set very high. Because capital punishment remains, at least in principle, a sanction open to American courts, and because it informs the popular stereotype of the risks that criminals face before the bar (despite its rarity), it is difficult to find a hearing for the idea that the criminal standard of proof might be less heavily skewed in the defendant’s favor than it is. If the death penalty were to be eliminated overnight, one weighty argument against serious reconsideration of BARDwould vanish. Absent the irrevocable punishment, the debate about how the calculation of costs of judicial errors should go would, I suspect, change its complexion entirely. There is one further facet worthy of mention concerning the epistemology of capital punishment. When someone faces a trial in which the state is asking for the death penalty, the prosecution in most jurisdictions is permitted to ask prospective jurors whether, if the evidence points strongly to the guilt of the defendant, they would be willing to condemn the defendant to death. Those panelists who cannot answer this question affirmatively are struck for cause and cannot serve on the jury. 57 This inevitably means that no one on the jury in a capital trial has overriding moral qualms about the death penalty. The fact is, however, that our society is one deeply divided about capital punishment. If a sizable proportion of the jury pool is excluded on principle from juries in capital cases, this raises some doubts about how representative the jury is of community sentiment. More than that, those who favor capital punishment are more likely to be “law-and-order” types than “bleeding hearts.” To the extent that this is so, they are probably more likely to find the defendant guilty than a randomly selected jury would. In that sense, the defendant’s prospects for acquittal are less than they would probably be with a jury selected blindly. This is a genuine conundrum. It is easy enough to see why the prosecution should be able to exclude from the jury those who have principled objections to the punishment the state is seeking. Equally, it takes no imagination to see why a defendant might fear that a jury full of those comfortable with capital punishment may give him something less than a fair hearing. 57 There is an excellent discussion of some of the epistemic consequences of the death- qualification rule nowused to exclude prospective jurors who have grave misgivings about the death sentence in Erik Lillquist, Absolute Certainty and the Death Penalty, 42 Am. Crim. L. Rev. 45 (2005). P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 The Unraveling of Reasonable Doubt 61 The existence of capital punishment, as well as the safeguards set up to avoid wrongful capital punishment, cast a very long epistemic shadow over the whole of the criminal law. Its elimination would contribute mightily to an environment in which a reasonably dispassionate discussion about the relative costs of different kinds of legal mistakes could take place and where a reasoned discussion about the possibility of different standards for different crimes could occur. Conclusion Clarification of the meaning of the criminal standard of proof is urgently required. Various confused and mutually conflicting versions of BARD abound in the justice system. Pretending that the problem doesn’t exist because BARD is self-explanatory is a ruse that the courts should be ashamed of perpetuat- ing. Still, more is needed than a clear statement of what BARD means. The fetish during the last century and a half for defining BARD in terms of jurors’ subjective states of mind should give way to definitions and illustrations of BARD that focus on the kind of evidence needed to make a rational person rea- sonably certain of someone’s guilt. Jurors (and judges as well) need to under- stand the logical and epistemic features that the prosecution’s case must exhibit before conviction is justified. So long as definitions of BARDfail to address the robustness of the evidence and so long as they remain fixated purely on the strength of the juror’s belief, BARD will remain open to the devastating criti- cism that it confuses mere strength of belief (which may be wildly irrational) withwarrantedbelief. Likewise obfuscatoryis the court’s unwillingness toinsist that jurors be able to give (at least to and among themselves) reasons for such doubts as they may entertain about guilt. As we have seen, an inarticulable, or even an unarticulated, doubt cannot claim to be a rational one. Yet current practice allows acquittal on precisely such inexpressible worries. As recently as 2002, the Second Federal Circuit was still insisting that “some doubts may be reasonable, but not articulable.” 58 This muddled thinking does not augur well for the prospects of the judiciary cleaning up its own act, since it evidently does not yet even see the problem. Historical research shows that BARD was introduced around the turn of the nineteenth century in order to encourage jurors to distinguish between trivial, elusive, and ill-founded doubts, which ought not to block conviction, and rea- sonable doubts, which should. Current practice makes it next to impossible for judges to say anything to jurors about how to draw that crucial distinction. Without it, reasonable doubt simply reduces to doubt per se, and, since any con- tingent statement can be doubted if we are so minded, the currently accepted instructions are tantamount to demanding certainty in order to convict. The 58 Fluellen v. Walker, 41 Fed. Appx. 497, at 502 (2002). P1: JZZ 0521861667c02 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:31 62 the distribution of error profound Enlightenment lesson that every contingent belief is open to skeptical doubt, a lesson that BARD was designed to accommodate, has been lost in the shuffle. The systemic failure to address the question of the structure of the proof of guilt demanded from the prosecution invites invalid convictions and invalid acquittals alike. Simply identifying levels of confidence required from jurors does nothing to enhance verdict validity, nor does it assist juries in attempting to distinguish reasonable doubts from unreasonable ones. Short of some form of radical surgery, BARD’s day has come and gone. P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 3 Fixing the Standard of Proof ‘Tis much more Prudence to acquit two Persons, tho’ actually guilty, than to pass Sentence of Condemnation on one that is virtuous and innocent. – Voltaire (1749) 1 It is better that five guilty persons should escape punishment than one person should die. – Matthew Hale (1678) 2 It is better that ten guilty persons escape [punishment] than that one innocent suffer. – William Blackstone (eighteenth century) 3 I should, indeed, prefer twenty guilty men to escape death through mercy, than one innocent to be condemned unjustly. – John Fortesque (1471) 4 It is better a hundred guilty persons should escape than one innocent person should suffer. – Benjamin Franklin (1785) 5 It is better . . . to acquit a thousand guilty persons than to put a single innocent man to death. – Moses Maimonides (twelfth century) 6 1 Voltaire, Zadig 53 (photo. reprint 1974) (1749). 2 2 Hale P.C. 290 (1678). 3 2 Sir William Blackstone, Commentaries on the Laws of England, c. 27, margin p. 358. 4 J. Fortescue, De Laudibus Legum Angliae 65 [1471] (S. Chrimes, ed. and trans., 1942). 5 Letter from Benjamin Franklin to Benjamin Vaughan (Mar. 14, 1785), in 11 The Works of Benjamin Franklin 11, at 13 (John Bigelow, ed., fed. ed., 1904). 6 2 Moses Maimonides, The Commandments (Charles B. Chavel, trans., 1967), at 270. 63 P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 64 the distribution of error The standard of proof (hereinafter, SoP) in criminal trials in the United States is, I have just claimed, a mess. It is not only ill defined, but it smacks of the arbitrary. Why, one is moved to ask, should it be set at the level of BARD – assuming we knew where that was – rather than still higher or much lower? As we saw in the last chapter, the Supreme Court saidin Winshipthat due process requires BARD, nothing more and nothing less. That is daft. What due process (understood as fairness and the right not to be subjected to capricious or unreasonable demands or restrictions) implies is that the same SoP – whatever it may be – be brought to bear in all similar cases. Neither the Constitution in general nor the due process clause in particular does anything to spell out what would be a fair and just standard of proof. Even if it can be argued that due process requires a standard more defendant-friendly than the preponderance of the evidence, proof by CACE would appear to meet the constitutional requirements of a SoP every bit as much as BARD does. So, we have two interrelated problems: Is there a better way than BARD for characterizing the appropriate SoP for a criminal trial? And how could we identify a nonarbitrary level for that standard that was not simply pulled out of a hat? It is time to ask what, if anything, epistemology can do to help put this situation right. The answer, this chapter will argue, is quite a lot. Specifically, I will suggest a) that there is a way to set the height of the SoP that is nonarbitrary, b) that a standard so generated will incorporate all the benefit of the doubt that we consider appropriate to give to the defendant, c) that this standard will guarantee over the long run that no more innocent defendants will be falsely convicted than we find acceptable, and d) that this standard, unlike BARD, can be cashed out in nonsubjectivist terms. The BoD and the SoP Let us begin by getting clear, in a general way, about what a standard of proof is and how it works. Basically, whether in the law or elsewhere, a SoP specifies a minimum threshold for asserting as proven some hypothesis. The hypothesis in question may be a scientific one (smoking causes lung cancer) or a legal one (Smith murdered Jones). There is nothing nature- or God-given about the height of a standard of proof. Rather, its height reflects a collective decision on our part to place the threshold at one point rather than another. That decision may be whimsical or it might be well thought out. The practical question, of course, is how demanding to make it. This requires us to have some principled basis for setting it there, rather than somewhere else. The appropriate answer will vary from one context to another. In the civil law, that threshold is set at 50+ percent, or “more likely than not.” In the testing of statistical hypotheses in science and medicine, the threshold is usually set at 95 (or sometimes 90 or even 99) percent. P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 Fixing the Standard of Proof 65 Begin with the obvious question: Why would one ever set the threshold at a value higher than the preponderance of the evidence? After all, if a hypoth- esis has a probability greater than half, the proverbial reasonable man should accept it as probably true. Basically, we set a standard well above the midpoint only when we believe that one sort of mistake is costlier, and thus more to be avoided, than another. Medical researchers plausibly think that certifying a drug as safe, when it is not, is more dangerous than denying it is safe, when it is. So, they do not approve a drug as safe unless it has passed trials that establish that it is about 95 percent sure to be safe. In the criminal context, when we set the bar of proof in a criminal trial higher than the preponder- ance of the evidence, we are saying that we intend to give the defendant the benefit of the doubt (hereinafter, BoD), because we think that falsely convict- ing the innocent is worse than falsely acquitting the guilty. If, after hearing all the evidence in a case, the jury thinks the defendant is probably guilty but still acknowledges that the prosecution’s case includes some less than fully convincing elements, then it is enjoined to give the defendant the BoD and acquit him. Setting the SoP significantly higher than the PoE is how we do that. The logic is crucial to grasp. Granting a defendant the BoD is not something that we do over and above specifying the SoP. It is the SoP that incorporates such benefit of the doubt as we consider appropriate. I repeat: The way we give the defendant the BoD is by adopting a SoP significantly higher than 0.5. What this clearly implies is that the question about how high to set the SoP is simply another way of asking how much BoD we should give to the def- endant. The more of the latter we consider appropriate, the higher the former must be. Some thinkers, especially those in Roman law countries, traditionally held that the defendant is entitled to every BoD. If one took this notion seriously, then the SoP would have to be set to 1.0 (or 100 percent), thus excluding all possible doubts. But, of course, we cannot take that proposal seriously. To begin with, it implies that we can have fully certain, secondhand knowledge about past events. The empiricist philosophers decisively undermined that possibility more than two centuries ago. As they showed, there is always room for some doubt in any criminal trial, as there is likewise scope for doubt about almost any assertion about the past. Perhaps key prosecution witnesses were lying; perhaps the police planted incriminatory evidence; perhaps a seemingly voluntary confession was false. If one is of a mind to credit every possible sort of doubt, inductively reasoning to a conclusion of guilt about a crime becomes impossible. Clearly, this notion of giving the defendant every BoDimplies that whenever any sort of doubt arises about guilt, acquittal is necessary. This will not do. There must be an upper bound on the doubt the accused should receive, a lesson that the U.S. Supreme Court tended to unlearn in its recent dicta on BARD, which suggested (as we saw in Chapter 2) that to convict, a jury needed virtual certainty about guilt. P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 66 the distribution of error That is why we know – even without yet having in hand a precise definition of the common law SoP – that it must be less than 1.0, since even unreasonable doubts indicate some residual uncertainty. Defendants in our legal system, the courts say, are not entitled to the benefit of unreasonable doubts, always suppos- ing that the courts were able and willing to tell us what an unreasonable doubt encompassed. But our question still remains: To how much BoD is a criminal defendant entitled? Having decided that we can set the height of the SoP once we know how much BoD society wants to give the defendant, we seem up against a new problem no less inscrutable than the SoP itself: the degree of BoD to which a defendant can properly lay claim. n, the SoP, and the BoD To make precise either the appropriate quantity of the benefit of the doubt a defendant deserves or the height of the SoP, we need to become clearer about certain logical features of any SoP, whether in the lawor elsewhere. Consider, to begin with, Figure 3.1. 7 Along the horizontal axis, we have a scale representing the apparent likelihood of guilt as evaluated by a jury (or by a judge in a bench trial). Along the vertical axis, we represent the frequency of cases that exhibit the relevant apparent probability, that is, howfrequently someone with the relevant apparent probability of guilt appears in a given distribution. The curve itself is a representation of all those truly innocent persons who come to trial, distributed with respect to their apparent guilt, once all the evidence has been weighed. 8 As the location of the peak of the curve indicates, most of those persons will have an apparent guilt below 0.5, that is, their apparent guilt in light of the evidence presented will make it unlikely that they are guilty. However, at the right tail of this curve, we see that some small fraction of those who are in fact innocent will have an apparent guilt significantly greater than 0.5. They look guilty even though, in fact, they are innocent. Now, suppose that we draw a vertical line just to the right of 0.5, which represents the civil SoP (more likely than not). If that standard were used in a criminal trial, we have every reason to 7 The use of diagrams of the sort that follow derives originally from a series of technical applications of signal detection theory to the law beginning (I believe) with Richard Bell, Decision Theory and Due Process, 78 J. Crim. L. & Crim. 557 (1987). Other notable contributions to this literature are Michael DeKay, The Difference between Blackstone- Like Error Ratios and Probabilistic Standards of Proof, 21 Law & So. Inquiry 95 (1996); Erik Lillquist, Recasting Reasonable Doubt: Decision Theory and the Virtues of Variability, 36 U.C. Davis L. Rev. 85 (2002). None of these authors considers the hypothesis, explored here, that the SoP is a measure of the benefit of the doubt. 8 The precise shape of the curve is not important for our purposes. Its peak might be higher, its tails shorter, its dispersion less, its precise position along the x-axis different from the representation. The distribution need not even be “normal.” All that I am about to say will apply to the whole family of curves that might represent the distribution of apparent guilt among genuinely innocent defendants. P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 Fixing the Standard of Proof 67 Figure 3.1. Sample Guilt Distribution among the Innocent expect that numerous truly innocent people would be convicted because their apparent guilt puts them on the wrong side of this standard of proof. These would be false convictions. Obviously, if we want to reduce the frequency of false convictions, we can slide the SoP to the right, that is, toward a higher standard of proof. Other things being equal, the farther to the right that we move the standard, the less likely it is that we will be convicting someone who is truly innocent. We can draw the same sort of diagram for the truly guilty. As Figure 3.2 shows, we have every reason to expect that their apparent guilt will cluster to the right of 0.5, even though many of them will have an apparent guilt that falls below the civil standard. If we were to use the civil standard of proof, we would be falsely acquitting some small fraction of the truly guilty (supposing that the distribution of apparent guilt is as I have depicted it). However, if we do here what we just did with Figure 3.1 – namely, if we imagine moving the SoP farther and farther to the right (in the name of reducing the number of false convictions) – we are going to increase the likelihood of false acquittals dramatically, other things being equal. Indeed, if we put the SoP somewhere close to 0.9, we might end up acquitting more of the truly guilty than we are convicting. Now, things are nowhere near as simple as I have suggested because the actual determination of the SoP depends on knowing more than we now do about the structure of the two curves (corresponding to apparent guilt of the P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 68 the distribution of error Figure 3.2. Sample Guilt Distribution among the Guilty truly guilty and the truly innocent). The point of these figures is not to set the SoP but to provide a vivid sense of the relations between the location of the SoP and the distribution of the sorts of errors likely to occur. If we superimpose these two diagrams, as in Figure 3.3, we begin to see some important interrelations between the two sorts of errors. Supposing that the distributions of apparent guilt and innocence of defendants in real trials were to be as I have depicted them in these figures, we can see that with every increase in the SoP above 0.5, we are both decreasing the frequency of false convictions and increasing the number of false acquittals. As Michael DeKay has put it, “Higher standards of proof lead to more erroneous acquittals and fewer erroneous convictions, all else being equal.” 9 The upshot of all this is that a standard of proof is best conceived as a mechanism for distributing errors. That, in turn, suggests that, if we could fig- ure out the relative costs to society of false convictions and false acquittals, we might be able to use the ratio of those costs as a mechanism for determin- ing the height of the SoP. To that end, let us define n as the socially accept- able ratio of false acquittals to false convictions. Two questions immediately arise: Is there any reason to believe that consensus about the value of n could be achieved (as there is clearly little consensus about n among our epigraph authors)? And, if there were such consensus, would that be sufficient to define a SoP? 9 DeKay, at 97. P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 Fixing the Standard of Proof 69 Figure 3.3. The Standard of Proof and the Distribution of Error The Relative Costs of False Acquittals and False Convictions There are ample grounds for accepting the hunch that false convictions are much more costly than false acquittals. The principal cost of a false acquittal is that a guilty felon goes free, unpunished, perhaps to commit other crimes; justice fails to be done; the victimof the crime is given no closure and comes away fromthe experience embittered because justice has not been done. The message is sent to prospective criminals that they, too, may be able to avoid prosecution for their crimes. Crimes rates will presumably soar if false acquittals are seen to occur frequently. Such costs are far from trivial. But very similar costs are imposed by every false conviction. There, the truly guilty felon is not even apprehended, let alone locked away. Criminally inclined folks who know the identity of the real felon will conclude that they too can commit crimes and escape justice. So far, the costs of the two errors more or less wash one another out (save for the faux sense of closure offered to a crime victim by a false conviction). But false convictions carry other, much weightier costs. They imply that the good name of an innocent person is permanently stained, that he is denied his liberty for the time of his incarceration, and that he may permanently lose other key benefits of citizenship (sometimes, for instance, the right to vote). These costs are awesome and are clearly greater than those associated with a false acquittal. Because they are, we want to make it harder to convict an innocent person than to acquit a guilty one. Having said that, however, I do not want to leave the impression that false acquittals are little more than a minor nuisance. Indeed, one of the key theses P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 70 the distribution of error of this book is that the current legal system allows, even encourages, far more false acquittals than are either necessary or desirable. Their very occurrence shows that the justice system is failing as a device for finding out the truth and for ensuring that those who commit crimes pay for their misdeeds. As Carleton Allen noted a long time ago: We must never forget that . . . the acquittal of ten guilty persons is exactly ten times as great a failure of justice as the conviction of one innocent person. 10 There may be those who, mindful of the fact that most felony trials end in conviction, will suppose that we ought not to blow the problem of false acquittals out of proportion. That is sound advice. Some numbers may help here. Every year in felony trials at the state and federal level in the United States, there are some 14,000 acquittals. In addition, 300,000 felony defendants have the charges against themdismissed by judges, grand juries, or prosecutors. To a first approximation, legal statistics suggest that roughly half of all those initially arrested for a crime walk away, untried and relatively unscathed. 11 If even a modest proportion of these suspects and defendants are guilty (and the proportion is probably not so small, given the skewing of the standard in favor of false acquittals), we are talking about tens, perhaps hundreds, of thousands of guilty defendants and suspects beating the system annually. 12 It would be one thing, and a desirable one, if those against whom charges were dropped prior to trial were innocent. But Floyd Feeney and others report empirical research showing that “most suspects who are arrested but not convicted are thought by police and prosecutors to be guilty.” 13 The police could, of course, be badly wrong about this; but, for reasons we will explore later in this chapter, I doubt it. Turning to data of a different sort, Harry Kalven and Hans Zeisel, in their classic study of judges and juries in the 1960s, found judges reporting that (in their judgment) juries falsely acquitted defendants about 20 percent of the time and falsely convicted defendants 3 percent of the time. 14 A second survey of criminal lawyers and judges reports that these players reckon that, among cases 10 Carleton Kemp Allen, Legal Duties and Other Essays in Jurisprudence 286–7 (1931). 11 Y. Kamisar et al., Modern Criminal Procedure 25 (1994). 12 These numbers come from Andrew D. Leipold, The Problem of the Innocent, Acquitted Defendant, 94 Nw. U. L. Rev. 1297 (2000). 13 Floyd Feeney et al., Arrests without Conviction: How Often They Occur and Why 243 (1983). 14 Harry Kalven and Hans Zeisel, The American Jury 68ff. (1966). It is essential to under- stand what is meant by “false acquittals” and ”false convictions” in the study in question. What Kalven and Zeisel found was not that two of every ten acquittals were of someone who was guilty but that two of every ten acquittals, in the opinion of the judges, justified a conviction, that is, satisfied the criminal SoP. Since that standard is quite high, we have every reason to expect that false acquittals, as we are using that term– namely, the acquittal of truly guilty defendants – happen far more often than the 20 percent figure would imply. P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 Fixing the Standard of Proof 71 leading to conviction, false convictions occur about 0.5 percent of the time. 15 While this is only anecdotal evidence, it is impressive for all that, suggesting that (if judges are right) one in five acquittals at trial is an invalid verdict. It is also worthy of note that, by judges’ estimates in the first study, the value of n in those cases that actually come to trial is around 7:1, and this figure must drastically understate the actual value of that ratio since, as I have said, the ratio estimate by the judges is not of false acquittals to false convictions but of invalid acquittals to invalid convictions. Despite such numbers, some scholars are still inclined to the view that false acquittals must be rare. Their thinking is that the vast majority of clearly guilty defendants will accept a plea bargain and plead guilty rather than face a jury and a potentially harsher sentence. That, these scholars say, means that most of those who elect a trial are probably innocent. One legal scholar, Donald Dripps, frames the argument this way: If we recall, however, that the majority of convictions result from pleas, we can see that the real hole in the system is not in the trial, but in the pretrial stage. Once it becomes clear that the government has a strong case, the factually guilty plead in large numbers. This tendency in turn suggests that jury trials do not run a terribly high risk of false acquittal, or far more guilty defendants would opt for trial. 16 What we know is this: Roughly one-third of those who go to trial are acquit- ted. Let us also suppose that someone who elects trial and is convicted receives a sentence 50 percent longer than the sentence proffered in the plea bargain. Under these circumstances, defendants, whether guilty or innocent, face the same choice: Do they accept two-to-one odds of a conviction and a longer sen- tence or do they accept the certainty of (lessened) jail time and lose any chance to get out of jail free? Different defendants will answer that question differently. But that is no reason to believe that most of those who choose to go to trial are innocent m of the crime. The fact that the case against a defendant is on the borderline between proof BARD and proof weaker than that – and that is the calculation that the shrewd defendant makes when he decides whether to plea bargain – does not entail or even make probable that most acquittals are true acquittals. 17 What it does make plausible is that those who elect a jury trial fancy that the case against them may not meet the high standard required for conviction. That, however, 15 Ronald Huff et al., Convicted but Innocent: Wrongful Conviction and Public Policy 61 (1996). Huff and his colleagues also discovered that more than 70 percent of the criminal justice officials who provided an estimate believed that false convictions occurred in fewer than 1 percent of all cases (Ronald Huff et al., Guilty until Proven Innocent: Wrongful Conviction and Public Policy, 32 Crime & Delinq. 518, at 522–3 [1986]). 16 Donald Dripps, Miscarriages of Justice and the Constitution, 2 Buff. Crim. L. Rev., 635, at 646 (1999). 17 Using language that I explained in Chapter 1, most of these acquittals may be valid ones but they are not thereby likely to be true acquittals. P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 72 the distribution of error is a very different thing from saying that most of those who choose to go to trial are innocent. Put in stark terms, what most determines whether a rational defendant elects a jurytrial is her assessment of the likelihoodof anacquittal, not the objective fact of her guilt or innocence. Arguing that truly guilty defendants will all cop a plea rather than face trial ignores the fact that the defendant’s calculation depends less on his sense of guilt or innocence than on his (and his attorney’s) perception of the conclusiveness of the case against him. In sum, we have no reason to think that false acquittals are only very occasional aberrations. On the contrary, they probably happen with high frequency, especially if we include among the class of acquittals all those pretrial decisions that shunt vast numbers of people (many of them probably guilty) back out onto the streets. In sum, false convictions are much more serious than false acquittals, but that is not to gainsay that the latter also carry heavy costs. So, while false convictions are clearly the more egregious error, false acquit- tals bear nontrivial costs. Given that, it does not seem wildly incredible to suggest that some societywide consensus could be forged about the value of n. We could put in place some mechanism (either by plebiscite or legislative action) for determining the sentiment of the people or their representatives about the respective costs of these two sorts of mistakes. Some citizens, when asked what they consider the appropriate ratio of false acquittals to false convictions, might glaze over at the apparent abstractness of the question. But I daresay few would claim to have no opinion if the wording (but not the substance) were changed, such as in the question, “How many murders or rapes that go unpun- ished are you willing to accept for every case of someone wrongfully convicted of murder or rape?” Once we have hammered out what a socially acceptable value for n would be, this could constitute a kind of social compact. We can now return to the task of asking ourselves what, if anything, a social consensus about the value of the ratio n might do to help us determine the height of the SoP. Using standard expected utility theory, several scholars have derived a value for the SoP that expresses it in relation to n as follows: SoP = 1/(1 +1/n). 18 18 The usual derivation of the formula goes along these lines: It is customary to define the disutility of an action as the product of its probability times its costs. (Variations of this derivation can be found in the articles cited in footnote 7.) A reasonable juror will vote to convict only when the expected disutility of an acquittal is greater than the expected disutility of a conviction. That is, when: C FA ×prob(acq) > C FC ×prob(conv), (1) (where C FA is the cost of a false acquittal, C FC is the cost of a false conviction, prob (acq) is the probability of acquittal, and prob (conv) is the probability of a conviction). What we want to know is the minimum probability or threshold where the inequality expressed in (1) obtains, for only here would a conviction be justified. Standards lower than this one would give society a vested interest in seeing an acquittal. Higher standards would tilt P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 Fixing the Standard of Proof 73 This formula – they claim – offers us a nonarbitrary way to determine the height of the SoP and, at the same time, it seems to provide us a tempting mechanism for ensuring that the ratio of errors that we regard as desirable will be exemplified in any sufficiently long run of real trials. Still, the temptation must be resisted. It turns out that a SoP by itself is quite insufficient to fix the ratio of errors that actually occur in criminal trials. This is because (as Ron Allen and Michael DeKay have cogently argued) the ratio of errors in real trials depends not only on how demanding the SoP is but on at least two other factors: 1. The robustness of the evidence and the validity of the jury’s inferences from that evidence. 2. The distribution of truly innocent m and truly guilty m defendants who go to trial. 19 The first of these problems is at least partially remediable. If, as I argue later in this book, we put in place rules of evidence more likely than the current ones to inform jurors of the important and relevant facts of the case, then we could expect jurors’ estimates of apparent guilt and apparent innocence to be reasonably accurate. But the second of these problems is fatal for this form of analysis. Judges have limited say and jurors have no say, over the proportion of truly guilty and truly innocent defendants who come to trial. That is a matter for prosecutorial discretion. These prosecutorial decisions will dramatically impact the ratio of false acquittals to false convictions that actually appears, since the number of false acquittals and false convictions depends upon the mix of truly innocent and truly guilty among the defendants. To see this concretely, imagine a run of fifty trials. Suppose that forty defendants are truly innocent and ten are truly society’s interest in the direction of favoring convictions. That threshold will define the SoP. Since the probabilities on both sides of the inequality must sumto 1.0, we can rewrite (1) as C FA ×SoP = C FC ×(1 −SoP) (2) (The inequality in (1) becomes an equality here because we are identifying a minimal threshold of apparent guilt.) Solving (2) for the value of the SoP, we obtain SoP = 1/[1 +(C FA /C FC )] (3) We will recognize the ratio in the denominator as nothing other than the inverse of our old friend, n. So, we can rewrite (3) as SoP = 1/(1 +1/n). (4) 19 Allenwrites, “Without knowingthe distributionof [apparent] guilt probabilities of factually innocent and guilty defendants, we cannot know the actual effect [on the distribution of errors] of choosing one standard of proof over another” (Ron Allen, The Restoration of In Re Winship, 76 Mich. L. Rev. 30, at 47 [1977]). See also DeKay: “[The] standard of proof employed by the jury does not, by itself, determine the ratio of judicial errors . . . [which] also depends on the prior odds of guilt and the accuracy of the jury” (DeKay, at 126). P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 74 the distribution of error guilty. Using a SoP of 90 percent, we can predict that the jurors will wrongly convict about 10 percent of the truly innocent, that is, four of them. Now, if a SoP in the neighborhood of 0.9 manages to capture the Blackstonian estimate of 10 as the value of n, then we should expect ten false acquittals for every false conviction. As there were four of the latter, we can expect forty false acquittals. Or can we? Remember that, by hypothesis, there are only ten truly guilty defendants in our sample. Even if the jury falsely acquits all of them, the resulting ratio,10:4, is nowhere close to Blackstone’s 10:1. In short, defining the SoP in this fashion offers no guarantee that the desired ratio of errors will be satisfied. Indeed, it is arguable that there is no machinery for generating a SoP that will capture that ratio in question for every conceivable distribution of guilty and innocent defendants. This disappointing result does not make n trivial or insignificant. To the contrary, it retains its importance as an expression of society’s attitudes toward the respective costs of false acquittals and false convictions. What n will not do, however, is fix the height of a SoP in a way that guarantees that real trials will exhibit this ratio of errors, or anything close to it. A Fresh Start at Setting the SoP So, we still have no nonarbitrary mechanismfor fixing the height of the SoP that would also guarantee that any series of real trials will come close to generating the desired ratio of errors. But before we resign ourselves to having to pull the SoP blindly out of a hat, it is worth exploring whether there isn’t a different ratio of interest that might serve to ground the SoP. The obvious alternative candidate is the ratio of true acquittals to false convictions. Let us suppose, counterfactually, that the ratios in the epigraph referred not to the ratio of errors, but to something relevantly similar. Specifically, let us assume for the sake of argument that Blackstone’s 10:1 ratio (which I seemto have settled on as probably representing our best guess as to what the social consensus would be) were a proposal about the acceptable ratio of true acquittals to false convictions. Read this way, Blackstone and the rest would be saying: “We don’t like false convictions at all. But we understand that, from time to time, the innocent will be found guilty. But we don’t want this error to occur more often than one time in ten (or one in twenty, or one in one thousand) cases.” I will call this ratio of true acquittals to false convictions m. Like the earlier formulation of the problem, this way of putting the point stresses that we must take pains to reduce the frequency of false convictions to an acceptable minimum. But now, that minimum is defined in terms of our successes in acquitting the innocent rather than, as before, our failures in acquitting the guilty. Fortunately, the SoP offers us a way to capture mprecisely. Indeed, havinga SoPof (for example) 90percent, over the longrun, will generate convictions of innocent defendants no more than 10 percent of the time and, P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 Fixing the Standard of Proof 75 obviously, result in acquittals for at least 90 percent of the innocent persons who come to trial. 20 Hence a SoP of 0.9 will give us a ratio of 9:1 for m. Alter it slightly, to about 91 percent, and we have Blackstone’s ratio of 10:1. But, it is natural to ask: Isn’t this way of defining the SoP tarred by the same brush as n was, rendered unworkable by problems about the unknowable character of, and changes in, the distribution of the innocent and guilty who go to trial? The short answer is: no. Over the long run, a SoP in the region of 90 percent will generate about ten true acquittals for every false conviction, regardless of how many truly innocent and truly guilty defendants stand trial. So, a SoP based on a social consensus about m guarantees – subject to the admission of all relevant evidence and reasonably clear-headed jurors – that m innocent defendants will be acquitted for every innocent defendant wrongly convicted. But, one might argue, why not set m even higher than 10 (or whatever value of m society would adopt)? After all, since we all believe that false convictions are terribly costly, reducing them still further would surely be a good thing. Sadly, it would not. We understand that, other things being equal, every increase in the SoP makes it harder to convict some of the truly guilty. Unless the over- whelming majority of those who come to trial are genuinely innocent, raising the SoP will make false convictions more likely than they would otherwise be. We grasp that a very high SoP is bad news in terms of both deterring crime and exacting retribution for egregious wrongs committed by guilty defendants. In short, our decision about an acceptable ratio of true acquittals to false convic- tions (m) must be supplemented by a side constraint, reflecting the fact that we are mindful of the importance of not making it too easy for the guilty to get off the hook. Specifically, the socially settled value of n acts as a side constraint on further jiggering with m. We acknowledge that a high value for m will give us a SoP that makes some false acquittals inevitable. Still, we should insist that the system commit as many false acquittals as are necessary in order to preserve m, but no more than that. Given that increases in m tend to push n higher, we should resist calls to drive down the frequency of false convictions even further below that value we have already agreed upon and incorporated into the SoP. Once we have settled on a socially acceptable ratio of true acquittals to false convictions, we know what we are looking for in a SoP. Specifically, we seek a SoP that will, over the long run, generate m true acquittals for every false conviction. If we can identify such a standard, we will find ourselves in a situation where society should be utterly indifferent as to whether trials end in acquittals or convictions. If the SoP were less exacting than this, society would have a vested interest in seeing to it that trials ended in acquittals, for that lower standard would be falsely convicting a higher proportion of truly innocent 20 For those technical aficionados to whom it matters, I should stress that this SoP is under- stood as an error probability, not as a Bayesian posterior probability. P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 76 the distribution of error defendants than society is willing to accept. If it were any more exacting, society would prefer trials to end in convictions, since trials would then be prone to produce more false acquittals than society’s acceptance of m necessitates. This result, which I will call the principle of indifference between acquittals and convictions, is extremely important for the themes treated in the rest of this book. It means that, if we can once fix on a SoP that incorporates society’s views about the appropriate value of m, it will be a mistake to argue that other aspects of trial procedure (such as the rules of evidence) should give generous additional dollops of the BoDto the accused. To adopt such acquittal-biased and acquittal-enhancing procedures would, in fact, violate the social contract about the value of m, since every rule of evidence or procedure that tilts the scale further toward acquittal will be equivalent to increasing the value of m. In those circumstances, justice would no longer be indifferent to the outcome of a trial. Nor should we. An additional bonus of this way of proceeding is that it enables us to crack the seemingly intractable problem of the appropriate size of the BoD. The amount of the BoD to which a defendant is entitled depends, as the SoP does, on the value of the ratio of true acquittals to false convictions, m. That is, once we fix the height of the bar for conviction, we have simultaneously decided how much BoD the defendant deserves and have incorporated that BoD into the standard of proof. If I have seemed to belabor this point, it is because both the popular and the academic legal literature on the law tend to assume that a criminal defendant should receive the BoD at every stage of a trial. As we will see in Chapter 5, many, probably most, scholars and jurists have argued at one time or other that the rules of evidence should err toward producing acquittals. Others have insisted that trial procedures should likewise be acquittal-friendly. Both points of view ignore the fact that, having set the bar for conviction very high, we have already given the defendant an enormous advantage at trial, one that in principle incorporates society’s view about how much BoD is appropriate. We will pursue this argument in greater detail in subsequent chapters, where we will see how deeply acquittal-enhancing measures have penetrated into trial procedure; but it is crucial to bear it in mind even at this stage. Replacing a Subjective Standard with an Objective One: The Real Reform I keep saying that if we can find a SoP that incorporates and safeguards m, we are on the road to solving the problem of identifying a suitable SoP. But I have said nothing yet about how we move from a settled value of m to a SoP. One obvious strategy would be to seek the solution to our problem in terms of a probabilistic standard of proof. It could be argued, for instance, that if mis about 10, then a SoP defined as an apparent guilt of 91 percent or greater would ensure P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 Fixing the Standard of Proof 77 that there were at least ten true acquittals for every false conviction. Plausible though it is, I want to resist this quick fix for the problem. My resistance is due to the fact that, in the final analysis, any probabilistic SoP will suffer from the same problems of subjectivity that, as we sawin Chapter 2, proved to be among the principal causes of BARD’s undoing. This warning may come as a surprise since I have made frequent use in this chapter of the language of probabilities with respect to the SoP. My reasons for doing so were chiefly heuristic; we can learn a great deal about how a SoP behaves, and especially about its role in the distribution of errors, by drawing on well-established results deriving from the analysis of probability and statistics. But like Ludwig Wittgenstein’s famous ladder that we use for climbing a wall and then – having reached the top – discard, the technical discourse of probabilities is now best laid aside as a tool for formulating a standard of proof. There are two reasons that drive me to this conclusion. Either, taken singly, is sufficient to raise serious doubts about the “probabilification” of the SoP. Taken jointly, they thoroughly undermine that project. I will consider them in turn. Problems with Estimating the Probabilities of One’s Beliefs The best, even the paradigmatic, case of assigning probabilities to one’s beliefs arises in contexts of specifically statistical events. If I know that the local American Legion chapter is raffling off a World War II tank, and I also know that the raffle will consist of one thousand tickets, I might decide to buy nine hundred of them. Supposing the raffle to be fair, I will then believe that I am likely to win the tank. More than that, I will be able to report that I am90 percent confident of the truth of the hypothesis that I will soon be the owner of a vin- tage tank. Here, the quantitative determination of my degree of confidence is a straightforward matter of a quick mental calculation. But move away from events of this sort, and it becomes vastly more difficult to assign a specific probability to beliefs. Suppose that I am a juror sitting through a trial for armed robbery. I have heard much inculpatory evidence and some exculpatory material. Having listened carefully to the evidence and thought very hard about it, I have come to the conclusion that the accused committed the crime. I now have to decide whether to vote to convict or acquit. I knowthat my belief that the accused did it is not by itself sufficient issue since the judge has instructed me that it is not sufficient merely to believe that the defendant probably committed the crime. It may occur to me to try to figure out how probable I think the hypothesis of the accused’s guilt really is. Indeed, if the criminal SoP were defined explicitly in terms of probabilities, I would be obliged to make such a determination. This, in general, is not something that I or most jurors could do with any degree of reliability. While I could perhaps tell you that I believe strongly in the guilt of the accused, that I found P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 78 the distribution of error the prosecution’s case powerful and that of the defense weak, I would be hard pressed to tell you whether I would assign a probability of 95, 85, or 75 percent to the prosecution’s hypothesis that the defendant committed the crime. And even if I did pull some specific number out of a hat, it would be a stretch to imagine that it accurately represented something like my “degree of belief ” in the defendant’s guilt. I would not be alone in this. If we look outside the area of the criminal law and to other forms of empirical investigation, we find a similar reticence of investigators to assign specific degrees of probability to the beliefs that they hold. For instance, almost every paleontologist believes that there was a massive, earthwide event some 65 million years ago that caused the extinction of dinosaurs. I venture, however, that no life scientist could tell you what specific probability he assigns to the abrupt extinction hypothesis. More than that, he would probably be very hard pressed even to indicate a small interval in which that probability falls. For instance, if you asked him: “Is the probability that this extinction was abruptly caused greater than 95 percent?” the vast majority of paleontologists would shrug their shoulders and say that they are not sure. If such a widely supported theory – one that scientists have been thinking very hard about and strenuously testing for more than a quarter century – enjoys broad but still not quantifiable support, what reason have we for thinking that a jury, having investigated a given hypothesis for a few hours (in the typical trial) or for a few weeks (in an atypical one), is better placed than the biology community to decide what specific probability they would assign to the guilt of the accused in a trial? The eminent jurist Learned Hand famously claimed that he could not even tell the difference between BARD and the PoE. 21 It is not hard to imagine his reaction had he been pressed to assign a specific probability to a trial verdict. Aficionados of probability theory will tell us that such difficulties can be solved handily enough. Simply ask a juror how much he would be willing to wager and at what odds that the hypothesis of guilt is true. If, the experts tell us, he would wager $20 or more with the prospect of winning $21 if he is right (and losing the $20 if he is wrong), then he assigns a probability to the guilt hypothesis of about 95 percent. While plausible enough, this gimmick simply displaces the problem one level. If I am genuinely unable to assign a probability to my belief that the accused is guilty, then I will be equally at a loss when asked precisely how much I would wager on his guilt. In short, if the SoP were couched in terms of a specific probability, such as 90 or 95 percent, most jurors who had decided that the accused committed the crime would be hard pressed to determine nonarbitrarily whether their confidence in guilt did or did not satisfy the demands imposed by this SoP. 21 See Barbara D. Underwood, The Thumb on the Scales of Justice: Burdens of Persuasion in Criminal Cases, 86 Yale L. J. 1299, at 1309 (1977). P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 Fixing the Standard of Proof 79 The Subjectivity Question But evenif that awkwardproblemwere soluble, a more serious one lurks beneath the surface. It goes to the heart of the question of what a SoP should be and what we intend it to accomplish. It arises not only with specifically probabilistic standards of proof but likewise with many nonprobabilistic standards such as BARD. The problem is that all such standards of proof are fundamentally subjective. They say the following to jurors: Listen carefully to the testimony and other evidence in this trial. At the end of the trial, you will probably have formed some belief about whether the defendant did or did not commit the crime. If you believe he did not commit the crime, then of course you must acquit him. If you have come to believe that he did commit the crime, you must now ask yourself a further question. If BARD is the standard in place, the question (as we saw in Chapter 2) comes down to this: Are you fully persuaded that he committed it? If 90 percent were to be the SoP, the question becomes this: Is your assessment of the likelihood of his guilt greater than 90 percent? If the juror can answer either question affirmatively, he must convict the accused. Otherwise, an acquittal is in order. The easy familiarity of this picture makes it difficult to take a couple of steps back and recognize just how strange this story is. And nothing about it is more peculiar about it than the fact that it offers no free-standing standard of proof. Instead, the systemtells the juror the following: At the end of the trial, think carefully about the evidence; note what level of confidence you have reached about the defendant’s guilt; if you are really, really confident that he committed the crime, then you should convict; if you are less than deeply persuaded, acquit. Not to mince words, this is a travesty of a system of proof. A SoP – in every area in which proof is called for outside the law (including natural science, clinical trials in medicine, mathematics, epidemiological studies, and so on) – is meant to tell the investigator or inquirer when she is entitled to regard something as proved, that is, when the relation of the evidence or the premises to the sought conclusion warrants the acceptance of the conclusion as proven for the purposes at hand. In the criminal law, by contrast, that issue is either wholly ignored or shamelessly finessed. Instead of specifying that the juror’s level of confidence in guilt should depend on whether a robust proof has been offered, the criminal law makes the SoP parasitic on the inquirer’s (that is, the juror’s) level of confidence in the defendant’s guilt. We have a proof, says the law, so long as jurors are strongly persuaded of the guilt of the accused (or so long as they will assign it a probability greater than x, where x is the probabilistic SoP). Never mind how they arrived at their high confidence, we have a proof. This gets things precisely backwards. In case it is not already clear why this perverts the logical order of things, imagine that we were to say to mathematicians that henceforth they have a legitimate proof of a theorem so long as they are firmly convinced the theorem P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 80 the distribution of error is true. Or, suppose that we told epidemiologists that they have a proof of the existence of a causal link between A and B provided that they are highly confident that such a link exists. Such a proof policy would be a laughingstock. One ought not to say to any trier of fact, “You have a proof of Aprovided you are firmly convinced of A” (nor, “You have a proof of A so long as your subjective confidence is higher than x”). To the contrary, we say, “You have no entitlement to be strongly convinced of A unless and until you have a proof of A,” adding for good measure that “your firm convictions about A count for nothing absent an acceptable proof of A.” And we then tell them what a proof of A would look like. That is what it is to have a standard of proof. Aproper SoP does not depend on one’s subjective confidence in a hypothesis; on the contrary, the standard tells us whether our subjective confidence is justified. In marked contrast, proof practices in the criminal law make the very exis- tence of a proof of guilt parasitic on the prior existence of a firm conviction about guilt. The language used to capture this subjective conviction varies from one era to another and from one jurisdiction to another. Sometimes, it is an “abiding belief” in guilt that is called for. Other times, it is a “moral certainty” or a “settled belief” about guilt, or a “satisfied conscience” or a “firm convic- tion.” It might be something like “confidence greater than 90 percent in guilt.” In Roman law courts, the corresponding word games involve things like “an intimate conviction” about the guilt of the accused, or “a sense of certainty, freely arrived at.” What these and other familiar variants share is the use of the existence of high personal confidence in the prosecution’s story as the touch- stone for deciding whether guilt has been proven. 22 The Supreme Court has recognized numerous times that the current rule for deciding guilt or inno- cence is an entirely subjective criterion (without, I might add, objecting to the fact that it is so). Recall from Chapter 2 that, in Jackson v. Virginia, the high court insisted that jurors, in order to convict someone, must achieve “a sub- jective state of near certitude.” 23 Outside the law, standards of proof are never couched in terms of the subjective confidence of investigators but in terms of the inferential links that must exist between the evidence available and the hypoth- esis in question in order to regard said hypothesis as proven. Outside the law, rational confidence in a conjecture follows on its proof, it does not precede it. Inside the law, such confidence precedes, certifies, and even constitutes the “proof.” 22 The great nineteenth-century legal theorist Thomas Starkie captured the widespread sen- timent I am criticizing when he insisted that, “What circumstances will amount to proof can never be [a] matter of general definition; the legal test is the sufficiency of the evidence to satisfy the understanding and conscience of the jury” (Thomas Starkie, A Practical Treatise of the Law of Evidence 513 [Benjamin Gerhard, ed., 6th ed., 1842] [1837]). 23 Jackson v. Virginia, supra, 443 U.S. at 315. Similar sentiments can be found inter alia in Accord Johnson v. Louisiana, 406 U.S. 356, at 360, (1972), and In re Winship, 397 U.S., at 364 (1970). P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 Fixing the Standard of Proof 81 “Well,” one might reply, “doesn’t the BARDrule specify a genuine standard of proof?” The answer, as a little reflection will show, is a resounding no. As I have already argued in detail, 24 current practice in the United States and other common law countries leaves reasonable doubt either wholly undefined or defined so vaguely as to be entirely unhelpful. I have shown that, de facto, the current SoP in an American criminal trial is firm conviction of guilt and nothing more. The system offers the juror no neutral or objective standard of proof, saying instead that the intensity of her subjective confidence in guilt determines whether she should convict or acquit the defendant. Making matters worse, the system puts no checks on how the juror goes about arriving at that subjective level of confidence. She is given wholly free rein to make of the evidence what she will and is required at the end only to affirm, if she votes to convict, that she is genuinely persuaded that the accused committed the crime. True, if she acquits the accused, she is bound to have a reasonable doubt about his guilt, but since the courts go to great lengths to insist that jurors need not even be able to articulate their doubts, let alone justify them to their fellow jurors, acquittals remain as arbitrary and subjective as convictions. What we face here is not a SoP but a pretext – and a flimsy one at that – for a conviction or for an acquittal. The Core Reform: An Objective Standard But suppose, for the sake of argument, that the criminal law had a genuine standard of proof, one that did not depend on the jurors’ subjective evaluation of guilt but on the establishment by the prosecutor of a powerful inferential link between the evidence presented and the guilt of the accused. In such circum- stances, guilt would not depend on jurors’ introspection of their confidence in guilt but on a determination by themof whether the standard had been satisfied. It is fair to ask what would it mean to have a legal SoPthat was not parasitic on subjective juror assignments of high confidence or high probability to the guilt of the accused. For starters, let us remind ourselves of the SoP that dominated Romano-German law from the late middle ages through the Enlightenment. Put simply, and ignoring the nuances, the formula for securing a conviction in criminal trials in those days was two-pronged: the production of two credible eyewitnesses to the crime or a plausible confession fromthe defendant. (I hasten to add that I am not endorsing this standard; its use here is simply illustrative.) With this standard in place, the subjective belief of the judge (obviously, juries were not used as the trier of fact) about the innocence or guilt of the accused was neither here nor there. The SoP specified what was needed to condemn the defendant. While this was a sorry SoP (in the sense that the absence of two 24 See Chapter 2 as well as Larry Laudan, Is Reasonable Doubt Reasonable? 9Legal Theory 295 (2003). P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 82 the distribution of error witnesses or a confession was not a very a reliable sign of innocence nor was a confession exacted by torture a good indicator of guilt), it must be said that this was, at least, a SoP in the strict sense of the term. The question we have to ask ourselves is whether we can fashion something like it out of contemporary practices. The fact is that there are already several interesting possibilities on offer that might do the job. Consider, for instance, a SoP along the following lines: a. If there is credible, inculpatory evidence or testimony that would be very hard to explain if the defendant were innocent, and no credible, exculpatory evidence or testimony that would be very difficult to explain if the defendant were guilty, then convict. Otherwise, acquit. Or, taking a cue from some of Ron Allen’s interesting proposals 25 : b. If the prosecutor’s story about the crime is plausible and you can conceive of no plausible story that leaves the defendant innocent, then convict. Oth- erwise, acquit. For present purposes, I am endorsing neither of these proposed standards. I cite them because they both exhibit one key virtue lacking in existing practice in criminal trials; to wit, they tell the juror what features he should be looking for in the evidence at hand that would justify him in convicting the accused. They do not ask the juror to engage in a meta-analysis of his own degree of confidence in the guilt or innocence of the accused. (It may be, of course, that many cases that would satisfy standard a or b would also leave most jurors with a firm conviction about the guilt of the accused. One would even hope that this was the case. But it is not necessary.) Neither a nor b would represent a radical break with historical practices in the common law. For a very long time, specifically until 1954, when the Supreme Court eliminated it from federal trials, a standard jury instruction was given in cases involving circumstantial evidence. It followed the lines of the circum- stantial evidence instruction still used in California state courts. It reads, in part: However, a finding of guilt as to any crime may not be based on circumstantial evidence unless the proved circumstances are not only (1) consistent with the theory that the defendant is guilty of the crime, but (2) cannot be reconciled with any other rational conclusion. . . . Also, if the circumstantial evidence permits two reasonable interpreta- tions, one of which points to the defendant’s guilt and the other to innocence, you must adopt that interpretation that points to the defendant’s innocence, and reject that interpretation that points to guilt. 26 25 See especially Ron Allen, Constitutional Adjudication, the Demands of Knowledge, and Epistemological Modesty. 88 Nw. U. L. Rev. 436 (1993); Ron Allen, Factual Ambiguity and a Theory of Evidence, 88 Nw. U. L. Rw. 604 (1994). 26 CALJIC 2.01. P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 Fixing the Standard of Proof 83 Although the specific language of this rule is both redundant (since the second sentence is entailed by the first) and infelicitous (since condition (2) should read something like “is inconsistent with the theory that the defendant is innocent of the crime”), the point is clear. This rule says that, in cases involving circumstantial evidence, a conviction is in order only if the evidence rules out every reasonable hypothesis except that of the guilt m of the defendant. It goes directly to the question of the nature of the evidence and avoids talking about how jurors may feel about the matter. It says to the jurors: c. Figure out whether the facts established by the prosecution rule out every rea- sonable hypothesis you can think of that would leave the defendant innocent. If they do, convict; otherwise, acquit. We should note that c does not ask jurors to assess their confidence about guilt any more than a or b does. Rather, it asks them whether the credible evidence presented at trial excludes every prima facie plausible hypothesis of innocence that they can imagine. Allen’s proposed standard, b, and a, involving the element of surprising evidence, are clearly in the same family. But instead of endorsing this promising instruction, the U.S. Supreme Court incredibly decided to eliminate it and ordered judges to tell jurors that they could convict only if they were “firmly convinced” of the guilt of the accused! 27 While such a decision was in keeping with that Court’s belief that something akin to subjective certainty is a sine qua non for a guilty verdict, it was nonetheless a retrograde move, further obscuring and mystifying an already murky process. None of the three standards under discussiondepends uponthe juror’s subjec- tive estimate of the likelihood of guilt. Indeed, all would allow for a situation in which a juror is obliged to convict, even if she (perhaps perversely) still believes in the innocence of the accused, despite the fact that the standard has been satisfied. One might imagine that jurors would bridle at the suggestion that they could be obliged to convict someone whom they regard as innocent. But this dilemma is not different in kind from the current situation, in which jurors often find themselves obliged to declare not guilty someone whom they believe to be guilty. If jurors can already grasp (as they obviously do) that the demands of proof must trump their private beliefs about guilt, they can surely learn that their private beliefs about innocence may likewise have to give way before a plausible standard of proof. Proposed standards of proof such as a, b, or c might seem overly simple and formulaic. We are dealing, after all, with highly complex human events and (in the case of jury deliberation) obscure thought processes. Standards a through c might sometimes fail to acquit the guilty or, worse, they might condemn the innocent. But that is beside the point. We are not dealing with deductive relations 27 Holland v. U.S., 348 U.S. 121 (1954). The court opined that “the instruction on circum- stantial evidence is confusing and incorrect.” Ibid., at 139–40. P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 84 the distribution of error between premises and conclusions but with uncertainty and fallible inferences. It is in the nature of ampliative rules of inference that they will sometimes lead to erroneous conclusions. No SoP can fully eliminate the uncertainty or render the kind of abductive inferences used in legal reasoning truth preserving. The germane question is whether standards such as a through c would lead to more or fewer mistakes than occur when (as now) the verdict relies on undis- ciplined juror instincts about guilt and innocence or on intrinsically arbitrary assignments of probability by jurors. Unfortunately, no one has collected the relevant empirical information to settle this question definitively. But it could be settled by empirical research. In the meantime, there are some plausible grounds to believe that rules like a through c would fare better than the subjective criteria. Consider, for instance, rule a. It asks whether the prosecution has managed to present credible, relevant testimony or physical evidence that would be very hard to explain if the defendant were innocent. It is a general rule of ordinary reasoning about the world that any historical hypothesis is best tested by seeing whether there are well-established, relevant facts that it should explain but cannot. If the answer to that question is affirmative, then the hypothesis in question has been discredited. This strategy of hypothesis evaluation occurs constantly in the so- called historical sciences, whether we look to human history or earth history. If, for instance, we want to evaluate the hypothesis that a collision with a celestial body was responsible for wiping out the dinosaurs, we ask ourselves whether there is any credible evidence dating from the relevant period that would be entirely explicable by the collision hypothesis but not intelligible if (for example) dinosaurs became extinct due to poor body ventilation. We likewise ask whether there are well-established facts about dinosaur extinction that do not fit the collision hypothesis. Since there is such evidence (in the famous “iridium spike”) explicable by the collision hypothesis, and no salient facts not explicable by it, we have powerful reasons to credit the extraterrestrial hypothesis. Scientific or historical inquiry is not different in kind from what goes on in courtrooms on a daily basis. Hypotheses about guilt and innocence get bandied about and one looks to the evidence to find particularly probing tests of those hypotheses. A standard like a makes fully explicit the kind of proof that jurors need for a conviction. In short, the kind of inference that a invites is not new or foreign to a courtroom, nor would it be wholly unfamiliar to jurors in their ordinary lives. Up to now, however, courts have not been willing to say to jurors that, if a rule like a is satisfied by the prosecution’s case, then jurors have a proof of guilt, and that otherwise they do not. To the contrary, various appellate courts have specifically said that standards such as those on offer here are too demanding. For instance, the First Federal Appeals Court has argued that “The prosecution may prove its case by circumstantial evidence, and it need not exclude every reasonable hypothesis of innocence so long as the total evidence permits a P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 Fixing the Standard of Proof 85 conclusion of guilt beyond a reasonable doubt.” 28 Quite howthe Court imagines that a juror can be sure BARD of the guilt of the accused when he likewise believes that there are “reasonable hypotheses of innocence” that have not been excluded by the evidence is one of the many mysteries surrounding BARD. That same court had argued eight years earlier that “The trier of fact is free to choose among various reasonable constructions of the evidence.” 29 But how can a juror be sure BARD of the guilt of the defendant when there are “various reasonable constructions of the evidence,” some of which are exculpatory? If BARD is as permissive as this, its replacement by a more demanding standard such as a through c becomes all the more urgent. Rules such as these, if adopted as standards of proof, would solve the two problems we have posed here. They are genuine, nonsubjective standards of proof and, unlike both BARD and a specific probability value, they leave little room for doubt about whether they have been satisfied in any particular case. That said, a serious conundrum remains. To put the challenge directly, what reason have we to believe that a, b, or c manages to capture the ratio of true acquittals to false convictions expressed by m? After all, these standards that I have been exploring are not framed in terms of probabilities. If (as I have argued) the function of a SoP is to distribute trial outcomes in accord with our wishes and to safeguard innocent defendants against more false convictions than we consider acceptable, then we need to be able to ascertain whether any candidate SoP succeeds in capturing our views about this issue. This is not an easy challenge, but neither is it insurmountable. I think it could be met in the following way: Let us proceed, as I argued earlier, to determine a socially acceptable value of m. What we then want in the criminal SoP is a decision rule that will yield more or less m acquittals of the innocent for every conviction of an innocent defendant. Consider, again, the proposed rule a. What we need to know about a is how often, if implemented, it would find guilty those who are in fact innocent. If that relative frequency is close to 1/m, then a passes muster. If it is not close to that value, then we have discovered that a would be an inappropriate SoP. But of course, we do not yet know the answer to that hypothetical question about the success/failure rates of a with respect to innocent defendants. So we have to find out by empirical research. That task appears very difficult. After all, it is not usually an easy matter to figure out, in any given trial, whether someone convicted of a crime (using any standard) is truly innocent or truly guilty. Exacerbating the problem further, courts are generally very reluctant to permit such probing by outside parties. Fortunately, criminal trials are not our only point of empirical access to such information. As I have already said, rules like a through c are widely used outside the law in testing empirical hypotheses about historical events. It 28 U.S. v. Gabriner, 571 F.2d 48, at 50 (1st Cir., 1978). 29 U.S. v. Thornley, 707 F.2d 622, at 625 (1st Cir., 1983). P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 86 the distribution of error would not be particularly difficult to conceive a series of experiments that could reveal the frequency with which each of these rules produced false positives (which correspond, of course, to false convictions). Even limiting our purview to criminal cases, it would not be so difficult to find out how well a might fare. Consider the following situation: The prosecution charges Jones with a crime. It presents much inculpatory evidence. Jones, as it happens, has an iron-clad alibi: Numerous reliable witnesses place him far from the crime scene at the time of the crime. Here, it is reasonable to suppose that Jones is highly likely to be truly innocent. But after the trial, as an intellectual exercise, we can apply standard a to all the nonalibi evidence in the trial, asking whether a was satisfied. Then, we repeat this test with several dozen other cases of solid alibis. If it should turn out, over a long run of cases, that a, absent the alibi evidence, would convict clearly innocent defendants like Jones much more than 10 percent of the time, then a fails the relevant test. Otherwise, it seems to have passed muster. While we will not be confident that we have an adequate SoP until we have both ascertained the socially acceptable value of m and found a standard that will reproduce m in practice, that is no reason for clinging onto BARD and the status quo until such time as both those projects are completed. As I already suggested, BARD is both ambiguous and subjective, weaknesses exhibited by neither a, b, nor c. While we don’t yet know whether the latter would replicate m, neither do we know that of BARD. Multiple Standards? Throughout this chapter and the previous one, I have often spoken of the crim- inal standard of proof, as if there were only one of them. While this is the conventional way of thinking about these matters, it is crucial to remind our- selves that there are, in fact, a plethora of different proof standards in play in any criminal proceeding. Indeed, at virtually every point where a judge or jury is called upon to make a determination of lawor of fact, a SoP will be associated with that decision. At the points of arrest and arraignment, the standard is prob- able cause. For a grand jury indictment, the standard is proof of guilt sufficient to convict if not contradicted by defense evidence. For a decision about denial of bail, the usual standard is CACE that the accused poses a risk of flight or of danger to others. If the defendant offers what is called an affirmative defense, he will often have to prove it to a PoE. If there is to be a change of venue, there must be proof by CACE that the accused cannot get a fair trial in the jurisdic- tion where his trial is scheduled. At the preliminary evidence hearing, multiple decisions have to be made about the admissibility of contested evidence. The standard here is usually the PoE, although (as we will see in the next chapter) many legal experts would like to see BARD as the standard for such decisions. Once the defendant is convicted and goes to a sentencing hearing, the standard transmogrifies into proof by a PoE. P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 Fixing the Standard of Proof 87 In appellate review, if the appeals court decides that a constitutional error occurred in the trial, it then faces the question of whether to order a new trial. It will do so unless the judges can persuade themselves BARD that the error did not affect the outcome of the trial. If the appellate court is asked to determine whether there was sufficient evidence to justify a conviction, it will not ask itself whether a reasonable person should or would have convicted the defendant but whether a reasonable person, confronted with the evidence presented at trial, could have decided that the defendant was guilty BARD. While none of these standards is as key as “the” SoP used by the trier of fact to determine guilt or innocence, they are all important epistemic links in the process and warrant much more serious scrutiny than they have received. It seems that a good general rule of thumb in the law is that wherever there is a decision to be made, there is, implicitly or (more commonly) explicitly, a SoP associated with it. And, as this chapter has argued, wherever there is a SoP, there is implicitly is an expression of the relative costs of the successes and failures that may ensue from that decision. If the general machinery for evaluating a SoP that I have described in this chapter is viable, it could be used for the analysis of all the different standards that are in play, inquiring in each case whether the standard in question captures our beliefs about the respective costs of the outcomes associated with each of these decisions. Conclusion I have argued that, in light of the patent failure of BARD to discharge the functions that we expect of a criminal standard of proof, we need a replacement. The challenge, I have suggested, is to find a SoP that simultaneously: r is clear and concise enough for jurors to understand and to apply; r is objective, directing itself to the structure of the proofs offered by the parties rather than resting on the subjective hunches of the trier of fact: and r embodies the social contract about the acceptable ratio of true acquittals to false convictions. While we do not yet know whether there is any SoP that will meet all these demands, we have considered several possibilities. BARD fails on the first two counts, and possibly on the third. Probabilistic standards fail the first two tests. I described several prima facie plausible alternatives. These satisfy the first two demands of clarity and objectivity. Still, none can be deemed adequate until significantly more empirical research has been done on their frequency of error generation. But at least we are now clearer about how to proceed. During the eleventh-century reign of King Ethelred, judges obliged English grand jurors to take an oath, requiring them to “swear, with their hands on a holy thing, that they will condemn no man that is innocent, nor acquit any that P1: JZZ 0521861667c03 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:5 88 the distribution of error is guilty.” 30 In good conscience, we could not ask modern jurors to take such an oath since we have deliberately designed our system with a high tolerance for false acquittals. An intellectually honest modern equivalent would have jurors swear to condemn few men who are innocent and, perhaps, to acquit many who are guilty. 31 But even that oath would remain a hollow one until such time as we put in place a SoP that makes it likely that we are acquitting the innocent most of the time (in line with m) and convicting the guilty as frequently as we can, compatibly with our views about the relative costs of false acquittals and false convictions (n). 30 See 4 Blackstone, Commentaries on the Laws of England, 302. 31 A typical oath in contemporary federal courts requires jurors to “solemnly swear that you will well and truly try the issues herein . . . and a true verdict render according to the evidence and the charge of the Court, so help you God.” (See, for instance, U.S. v. Pinero, 948 F.2d 698–9 [11th Cir., 1991].) Essentially, this oath aims at producing valid verdicts (“true . . . according to the evidence”), not true ones. P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 4 Innocence, the Burden of Proof, and the Puzzle of Affirmative Defenses In all cases of penal procedure, the declared supposition is, that the party accused is innocent; and for this supposition, mighty is the laud bestowed upon one another by judges and law-writers. This supposition is at once contrary to fact, and belied by their own practice. . . . The defendant is not in fact treated as if he were innocent, and it would be absurd to deal by him as if he were. The state he is in is a dubious one, betwixt non-delinquency and delinquency: supposing him non-delinquent, the[n] immediately should that procedure against him drop; everything that fol- lows is oppression and injustice. – Jeremy Bentham (1829) 1 If the presumption of innocence is not sacred in our systemthen what is to become of us? – Judge Reta Strubhar (Oklahoma Court of Criminal Appeals) 2 In the last two chapters, we have looked at the SoP and the BoD. It is time to turn our attention to the other two principal ingredients of what I have been calling error distribution doctrine: the burden of proof and the presumption of innocence. 3 On their face, these might seemto be comparatively simple matters. Basically, the thesis of the burden of proof says that the defendant need prove nothing, while the state carries the full obligation to establish the guilt of the accused to the appropriate standard of proof. The presumption of innocence instructs jurors that they are to suppose the defendant innocent until and unless the prosecutor has discharged that burden. Hidden behind these seemingly uncontroversial doctrines lie several tanta- lizing issues of legal epistemology. I will deal with both in the same chap- ter because, although generally considered distinct concepts, it will probably 1 Jeremy Bentham, Principles of Judicial Procedure 169 (1829). 2 Flores v. State, 1995 OK CR 31 (Okla. Crim. App., 1995). 3 I am grateful to Ron Allen for his encouragement in bringing this chapter to fruition and for his critical comments once it had been. A version of this chapter was discussed at the Tenth Annual Conference on Analytic Philosophy of Law (Austin, 2005). 89 P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 90 the distribution of error already have occurred to you that they seemto be little more than opposite sides of the same coin. If the state bears the full burden of proof, then, of course, one might say, the defendant is presumed innocent. Contrariwise, if the defendant is genuinely presumed innocent, then it naturally follows that the state must defeat that presumption by proving his guilt. 4 Among the questions we will be asking is whether these two traditional doctrines are simply redundant restatements of the same underlying theme or whether they are autonomous one from the other. Before we can proceed to that issue, however, we need to get clearer about what each entails. The Presumption of Innocence: Material or Probatory? The presumption of innocence (hereinafter PI) is among the small handful of doctrines in criminal law that are ubiquitous across a very broad spectrum of legal systems. Enshrined in the constitutions of countries as diverse as France 5 and Argentina, 6 it is as much a commonplace in Roman law countries as in common law jurisdictions. Its historical pedigree has been traced as far back as that of any doctrine now current in the criminal law, even to Deuteronomy (if we are to believe the U.S. Supreme Court in Coffin 7 ). On various occasions, the Supreme Court has seen in the PI the wellspring of American criminal procedure. In 1979, Justices Stewart, Brennan, and Marshall insisted that: “No principle is more firmly established in our system of criminal justice than the presumption of innocence that is accorded to the defendant in every criminal trial.” 8 As long ago as 1895, the Court unanimously held that “The principle that there is a presumption of innocence in favor of the accused is the undoubted law, axiomatic and elementary, and its enforcement lies at the foundation of the administration of our criminal law.” 9 This is heady stuff: the PI as “undoubted law,” “axiomatic,” and at the very “foundation” of the criminal justice system. The Court added, in Estelle v. Williams, that the PI “is a basic component of a fair trial under our system of criminal justice.” 10 4 Here is California’s model jury instruction on the presumption of innocence: “A defendant in a criminal action is presumed to be innocent until the contrary is proved, and in case of a reasonable doubt whether [his or her] guilt is satisfactorily shown, [he or she] is entitled to a verdict of not guilty. This presumption places upon the People the burden of proving [him or her] guilty beyond a reasonable doubt” (CALJIC 2.90). Typical of U.S. model instructions, it runs together the presumption of innocence and the burden of proof. 5 The Declaration of the Rights of Man and of the Citizen says, “Every man is supposed innocent until having been declared guilty.” 6 Article 11 of the National Constitution of Argentina insists that “Every person accused of a crime has the right to be presumed innocent.” 7 Coffin v. U.S., 156 U.S. 432 (U.S., 1895). 8 Kentucky v. Whorton, 441 U.S. 786, at 790 (U.S., 1979). 9 Coffin v. U.S., 156 U.S. 432, at 453 (1895). 10 Estelle v. Williams, 425 U.S. 501, at 503 (1976). P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 Innocence, BARD, and Affirmative Defenses 91 Under such circumstances, one expects it to be reasonably clear what the PI is, when and where it applies in criminal proceedings, who is bound by it, and what relation it has to other key doctrines, such as reasonable doubt, the burden of proof, the BoD, and due process generally. It is thus mildly disconcerting to discover that there is little consensus about precisely what the PI means, that there is ardent debate about to whom and when it applies, and that courts and legal scholars disagree about whether it stands on its own legs doctrinally or is simply an obvious, if nontrivial, consequence of the standard (or perhaps the burden) of proof. Many see its scope as very broad. One scholar of the law, William Laufer, claims that the privilege against self-incrimination, the right to silence, the dis- covery rule, even the rights to counsel and to confront one’s accusers all rest on, and are “reflection[s] of, the presumption of innocence.” 11 Even if that is a bit of a stretch, the presumption is widely described as grounding the claim that the burden of proof falls exclusively on the prosecution in a criminal trial, and sometimes as providing a rationale, or at least a motive, for the BARD standard. 12 On occasion, the U.S. Supreme Court has even gone so far as to insist that the PI counts as “evidence in favor of the accused.” 13 In the same vein, a model federal jury instruction propounds the staggering idea that the PI, standing alone, may be enough to produce reasonable doubt about a defendant’s guilt. The instruction says specifically, “The presumption of innocence alone may be sufficient to raise a reasonable doubt and to require the acquittal of a 11 William Laufer, The Rhetoric of Innocence, 70 Wash. L. Rev., 329, at 333–4 (1995): Evidence of this value, as a reflection of the presumption of innocence, may be seen in the reasonable doubt rule, as well as a series of substantive and procedural safeguards that arguably presuppose legal innocence, e.g., the privilege against self-incrimination and the right to remain silent while in police custody and during trial; the duty of the state to disclose exculpatory evidence; the right to compulsory evidence; the right to confront adverse witnesses; and the right to effective assistance of counsel. 12 In Lego v. Twomey, the Supreme Court stressed that “a high standard of proof is necessary, we said [in Winship], to ensure against unjust convictions by giving substance to the presumption of innocence” (404 U.S. 477, at 486 [U.S., 1972]). 13 Emphasis added. In 1895, Justice White wrote for a unanimous court: The fact that the presumption of innocence is recognized as a presumption of law and is characterized by the civilians as a presumptio juris, demonstrates that it is evidence in favor of the accused. For in all systems of law legal presumptions are treated as evidence giving rise to resulting proof to the full extent of their legal efficacy. . . . the presumption of innocence is evidence in favor of the accused introduced by the law in his behalf ” (Coffin v. U.S., 156 U.S. 432, at 460 [U.S., 1895]). This is preposterous since the presumption is patently not exculpatory evidence but an assumption. Even if we were to suppose that the PI, like exculpatory evidence, provides some tipping of the scales of justice in the defendant’s favor, this analysis leaves completely open the question as to how much evidence the presumption provides. Without knowing that, a juror would be hard pressed to know when, if ever, the inculpatory evidence offered by the prosecutor could prevail. P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 92 the distribution of error defendant.” 14 If the PI in itself can generate a reasonable doubt about guilt, and if reasonable doubt demands an acquittal, then jurors who took this instruction lit- erallytoheart wouldhave a rationale for acquittingany criminal defendant, how- ever inculpatory the evidence against him. (Do we really want jurors saying to themselves, “The evidence against the defendant is really quite overwhelming, but I take the presumption of innocence seriously and it gives me a reasonable doubt”?) Apparently bearing out this view of the doctrine’s centrality, the Supreme Court has overturned convictions in cases where the trial judge failed to instruct the jury about the PI, arguing that this is reversible error. 15 On other occasions, however, that same court has held it to be harmless error to fail to instruct the jury concerning the PI since, in the Court’s (occasional) view, a reasonable doubt instruction suffices. 16 Indeed, members of the high court have been known to argue that the PI is derivative and redundant – rather than “axiomatic” or “foundational” – in that it follows logically from the standard of proof. 17 The high courts have even gone so far as to suggest that judges give an instruction on the PI to jurors, not because the PI adds anything to the principle of the prosecu- torial burden of proof, but simply because it reminds jurors of that burden: While the legal scholar may understand that the presumption of innocence and the prosecutor’s burden of proof are logically similar, the ordinary citizen well may draw significant guidance from an instruction on the presumption of innocence. 18 It is my hope here to shed some light on, and perhaps even sort out, some of these and related puzzles. A good place to start is by reverting to an important similarity I have already explored for other purposes between the testing of hypotheses in science and the hypothesis test that is a criminal trial. The most obvious similarity is that both science and the lawdeliberately make it very hard to disprove the so-called 14 STAR-0-3 Modern Federal Jury Instructions–Criminal P 3.02. Interestingly, the Supreme Court has seemingly overruled that idea that the PI alone is sufficient to acquit a defendant. See U.S. v. Ibara-Alcarez, 830 F.2d 968 (9th Cir. 1987). 15 U.S. v. Fernandez, 496 F.2d 1294 (5th Cir. 1974); U.S. v. Dilg, 700 F.2d 620 (11th Cir. 1983). 16 Kentucky v. Whorton, 441 U.S. 786 (1979). The Court noted explicitly that “the failure to give a requested instruction on the presumption of innocence does not in and of itself violate the Constitution” (ibid. at 789). Justice Byron White was to put the point even more forcefully a decade later: “A jury instruction on the presumption of innocence is not constitutionally required in every case to satisfy due process, because such an instruction merely offers an additional safeguard beyond that provided by the constitutionally required instruction on reasonable doubt” (Arizona v. Fulminante, 499 U.S. 279, 111 S. Ct. 1246, 113 L. Ed. 2d 302, at 319 [1991]). 17 Here is how Justice John Paul Stevens, dissenting, put the point in 1978: “The function of the instruction [on the presumption of innocence] is to make it clear that the burden of persuasion rests entirely on the prosecutor. The same function is performed by the instruction requiring proof beyond a reasonable doubt” (Taylor v. Kentucky, 436 U.S. 478, at 491 [1978]). 18 Ibid., at 485. P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 Innocence, BARD, and Affirmative Defenses 93 null hypothesis. In science, the null or default assumption is (for example) that there is no causal link between one thing and another. In law, the default is the claim that the defendant is innocent. There is a less obvious, but no less important, analogy between the two forms of inquiry. It amounts to this: Just as, in the clinical trial of a drug, the failure to prove a drug’s efficacy is not proof that the drug is nonefficacious, so the failure to prove a defendant’s guilt is not proof that he did not commit the crime. In each case, the evidence shows only that the pertinent SoP was not satisfied. Put in its simplest terms, failure to prove X is never a proof of not-X. What this means for the law, I will argue, is that innocence (as in “‘John is innocent’ means he did not commit the crime”) figures scarcely at all in a criminal trial, nor among the plausible inferences that can be drawn from an acquittal. And if innocence does not figure, neither should a PI (if by that we mean that the defendant should be presumed not to have committed the crime). Our route to these conclusions will be slightly tortuous, so I will try to lay out the steps of the argument as clearly as I can. When and to Whom Does the PI Apply? I begin with the easiest of the puzzles we will have to address: At what points along the path marking the sequence of events in the criminal justice system, and to which actors in that sequence, does the PI apply? One might have supposed, naively as it turns out, that a person is presumed innocent from the moment police first suspect him until his trial ends and a verdict has been pronounced (supposing things progress that far). One might have supposed, equally erron- eously, that all those charged to make decisions about the defendant ( judges, grand juries, prosecutors, and so on) are under some duty to presume his innocence. While such a broad construal of the meaning of PI is common in several countries (at least in theory), the Supreme Court has made it very clear that in American criminal procedure, the PI applies only to the trial itself and, within the trial, only to the jury or other trier of fact. In 1979, the Supreme Court considered a class-action suit from various persons in New York State who had been denied bail and jailed pending trial. Their suit claimed that the con- ditions in the correctional facilities housing them violated their constitutional rights, among them the right to be presumed innocent until proven guilty. The Supreme Court, in a majority opinion written by then–associate justice William Rehnquist, found that the PI simply did not apply to extratrial events: Without question, the presumption of innocence plays an important role in our criminal justice system. . . . But it has no application to a determination of the rights of a pretrial detainee during confinement before his trial has even begun. 19 19 Bell v. Wolfish, 441 U.S. 520, at 533 (U.S., 1979). This was a restatement of the argument originally made in dissent by Justice Felix Frankfurter a quarter-century earlier: “[I]f the ‘presumption of innocence’ is read literally to apply to all pretrial procedures, it is P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 94 the distribution of error The Court’s conclusion about pretrial confinement was derived from a gen- eral analysis of the scope of the PI. In the view of the Court, the PI is not an omnibus claimof the innocence of everyone charged with a crime. It is, rather, a mechanismfor localizing the burden of proof during a criminal trial itself. “The presumption of innocence,” the majority wrote, “is a doctrine that allocates the burden of proof in criminal trials; it also may serve as an admonishment to the jury to judge an accused’s guilt or innocence solely on the evidence adduced at trial and not on the basis of suspicions that may arise from the fact of his arrest, indictment, or custody, or from other matters not introduced as proof at trial.” 20 The Louisiana Supreme Court put the same argument in blunter terms: The presumption of innocence is a guide to the jury. If it were absolute and operative at every stage of a prosecution, the defendant could never be jailed until conviction. It does not prevent arrest. . . . There is little relationship between the right to bail and the presumption of innocence. The presumption of innocence is operative and protects against conviction, not against arrest (which is taking into custody). 21 If the PI is but a mechanism for allocating the burden of proof and warn- ing jurors against drawing adverse inferences from the indictment against the defendant, then there is not the slightest warrant for supposing that the PI gov- erns events beyond the trial itself. The courts took a similar line when Congress passed a law making it illegal for someone under indictment for a serious crime to buy a firearm. When the law was challenged on the grounds that it violated the PI to which all persons are entitled, the courts insisted that the law was constitutionally sound because the PI did not apply outside of the trial itself, adding, for good measure, that “the indictment of an individual for a crime pun- ishable by imprisonment for a term exceeding one year is so often indicative of a propensity for violence. . . . ” 22 In a similar case, Thoreson v. U.S., the Ninth Circuit Court of Appeals considered the constitutionality of a federal law that forbade those under indictment from carrying firearms across state lines, 23 an otherwise lawful activity. The court concluded, “We think it not unreasonable for Congress to conclude that there is considerable likelihood that one indicted for such an offense has a propensity to misuse firearms.” 24 These judgments impossible to justify bail or pretrial detention, both of which are restraints imposed upon an accused despite the presumption” (Leland v. Oregon, 343 U.S. 790, at 802–3 [1952]). 20 Ibid. 21 State v. Green, 275 So.2d 184, at 186 (La., 1973). 22 U.S. v. Craven, 478 F.2d 1329, at 1339 (6th Cir., 1973). 23 The law in question is 15 USC. §902(e), which states: “It shall be unlawful for any person who is under indictment or who has been convicted of a crime punishable by imprisonment for a term exceeding one year or who is a fugitive from justice to ship, transport, or cause to be shipped or transported in interstate or foreign commerce any firearmor ammunition.” 24 U.S. v. Thoresen, 428 F.2d 654, at 662 (9th Cir., 1970). P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 Innocence, BARD, and Affirmative Defenses 95 do not exactly read as ringing endorsements of the pretrial applicability of the PI. While there are many critics (including some on the high court itself) of this restricted construal of the scope of the PI – critics who fancy that the PI follows the defendant right through the sequence of decisions made by the justice system – their bit of wishful thinking simply cannot be right. Running through the events in a criminal proceeding are a host of decisions that cannot be squared with a broad construal of the meaning of the PI. Consider a fewof them: The police who arrest Jones clearly do not regard him as innocent, because probable cause precludes the arrest of persons thought likely to be innocent of a crime. The grand jury that indicts Jones likewise no longer presumes his innocence since the indictment indicates that the grand jury is satisfied that he probably committed the crime. When the now-arrested and -charged Jones requests a bail hearing, the presiding judge cannot be presuming Jones’s innocence in any robust sense since, if he did, he would not be demanding that Jones plunk down a tidy sum to ensure his appearance at trial nor, even more vividly, would the judge have the authority to lock away the presumedly “innocent” Jones until the completion of his trial. Indeed, in many jurisdictions, bail can be denied to Jones – even if he is facing a murder charge – only when the prosecution can persuade the judge that the evidence creates “a strong presumption” of Jones’s guilt. 25 If such a showing is made and bail is denied, do we really want to say that the systemis still presuming Jones to be innocent? As a sagacious California Supreme Court opined a century and a half ago, “If [an indictment] furnished no such presumption [of guilt,] it would not justify the exaction of bail or the detention of the defendant.” 26 Once Jones’s trial gets under way, the PI still has a quite limited scope in terms of those who are bound by it. The prosecutor, obviously, is not constrained by the PI; to the contrary, it is his responsibility to depict Jones convincingly as the perpetrator of the crime. The trial judge is likewise forced to reveal his hand as the trial proceeds in ways that cannot easily be squared with the notion that the judge presumes Jones’s factual innocence throughout the ordeal. Specifically, once the prosecution has presented its case-in-chief, a defense motion frequently calls on the judge to dismiss the case against Jones for lack of evidence, thereby producing a directed verdict of not guilty. If the judge permits the trial to proceed, as he usually will, he can do so by law only if he believes that the prosecutor’s case could persuade a reasonable person of Jones’s guilt BARD. That is, the trial judge who allows a case at this stage to continue has already determined – well before the jury adjourns for deliberation – that it would be entirely reasonable to believe firmly that Jones committed the crime and therefore unreasonable to believe that he did not. Of course, the judge’s 25 See especially State v. Konigsberg, 33 NJ 366, at 377 (1960). 26 People v. Tinder, 19 Cal. 539, at 543 (Cal., 1862). P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 96 the distribution of error ruling at this point is still compatible with a jury’s finding of not guilty, but that does not warrant the claim that the judge is still presuming that Jones did not commit the crime. In short, the only judicial proceeding in which the PI unequivocally operates is the trial itself; even there, the only participants in the legal process bound to believe in Jones’s innocence until jury deliberation begins are the jurors themselves. And, if the analysis in the next section is correct, not even jurors are obliged to assume that Jones is innocent in the sense that he did not commit the crime with which he is charged. One might argue, and some courts have, that an attenuated sense of the PI kicks in well before a criminal trial begins. 27 If we construe the PI as implying only that the state has the obligation to prove something before it can move a suspect further along the criminal justice track, while the suspect need prove nothing, we could then argue that the PI provides the rationale for requiring probable cause for the issuance of search and arrest warrants and for the formal charging at arraignment. In these situations, the state is not obliged to show that the defendant is guilty BARD, but it is obliged to show a plausible reason for searching the suspect’s home, tapping his telephone, formally charging him with a crime via information or grand jury indictment, and so on. The trouble with regarding these practices as instances of a robust PI in play prior to trial is that each stage may terminate in a judgment (to arrest, to search, to deny bail, to commit for trial) that is impossible to square with the idea that a defendant enjoys a continuing right to be presumed innocent until a jury votes to convict. Kinds of Innocence and Guilt Having narrowed our focus to the trial and, within it, to the jurors, we can now begin to explore precisely what it is that jurors are supposed to assume or presume. It will prove profitable to remind ourselves of the two distinct notions of innocence and guilt spelled out in Chapter 1. We there distinguished between material guilt (guilt m ) and innocence (innocence m ) and probatory guilty (guilt p ) and innocence (innocence p ). There is a salient asymmetry between the two pairs of distinctions. It consists in the fact that a) while a finding of guilt p sustains (fallibly) the assertion of guilt m (that is, the legal sys- tem justifiably assumes that someone proved to be guilty is genuinely guilty), b) a finding of innocence p (a “not-guilty” verdict) warrants no inference about innocence m . That, indeed, is why juries do not make findings of innocence but of not guilty, that is, insufficient proof to warrant conviction. There can be little doubt that lay people (and hence jurors) construe “inno- cence” as innocence m . When a judge instructs jurors that they should presume 27 A Texas appeals court has argued that “implicit in the probable cause requirement is the presumption of innocence” (Lanes v. State, 767 S.W.2d 789, at 796 [Tex. Crim. App. 1989]). P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 Innocence, BARD, and Affirmative Defenses 97 the defendant to be innocent until proven otherwise, they will almost certainly interpret that as committing them to believing that the defendant did not per- petrate the crime. The judge will rarely gloss the meaning of the PI, supposing (as more than one court has put it) that “The term [sic] presumed innocent has a self-evident meaning comprehensible to the lay juror.” 28 (I note in passing how convenient it is that so many core legal concepts – BARD and the PI among them – are self-evident to jurors.) Not only will jurors construe the instruction about the PI as requiring them to believe in the innocence m of the accused; that is precisely what the legal system intends them to presume (as we will see in detail later in this discussion). Interestingly, however, the lay person’s notion of innocence, innocence m , is an almost empty category in Anglo-Saxon law. From the moment of arrest forward, virtually all the judicial decisions about the accused fail to include a finding of innocence m among their possible outcomes. As we have seen, at no decision point in the process – a prosecutor’s decision not to press charges, a grand jury’s finding of no bill, a judge’s decision to dismiss a case at arraign- ment, a jury’s acquittal, a judge’s directed verdict, or an appellate reversal of a conviction – can we find an outcome that would enable us to infer that the accused was found or believed by the relevant decision maker to be innocent m , that is, that he (probably) did not commit the crime. Instead, every one of these decision points involves asking whether the case against the accused rises to the appropriate SoP. At no point does the failure of the state to satisfy the relevant SoP justify an assertion of innocence m . 29 Curiously, almost the only way a criminal defendant can secure a decision that actually implies his innocence m is first to win an acquittal in a criminal case and then to bring and win a suit for malicious prosecution. No other judgment in the justice systemrepresents a finding of innocence m . 30 The judicial irrelevance of innocence m is borne out not only by the absence of any judicial findings that imply that the accused did not in fact commit the crime; it likewise, and importantly, shows up in his initial plea. A defendant charged with a crime cannot plead innocent, however much the defendant may otherwise assert his innocence m . He or she can only plead guilty or not guilty. Indeed – and the irony should not go unnoted, for it is significant – the only time in a typical criminal proceeding where the innocence m of the defendant is ever alluded to, by law and by tradition, is in the judge’s opening instruction on 28 Flores, 1995 OKCR9 (Okla. Crim. App., 1995). For similar sentiments, see State v. Perce, 260 Kan. 859 (Kan., 1996). 29 Insofar as the courts ever grapple with innocence m , hints of it can be found in habeas corpus review of convictions and in civil and criminal cases alleging false arrest or false prosecution. Still, their relative rarity is powerful testimony to the thesis that the system is largely indifferent to questions about the innocence m of those passing through it. 30 In California, there is a legal category of innocence m . Upon acquittal, any defendant may petition the judge for a ruling of “factual innocence.” The burden then falls on the petitioner to show that “no reasonable cause” exists to believe that she committed the crime of which she has been acquitted. P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 98 the distribution of error the PI. That instruction typically is couched in such terms as “The defendant is presumed to be innocent” (Virginia model instruction) or “The defendant stands before you presumed to be innocent of any crime” (Massachusetts). 31 The “innocence” at issue in such instructions cannot plausibly be said to be innocence p , that is, lack of proof of guilt. As I have said, lay jurors would not construe “innocence” as anything other than innocence m , since “innocence,” to anyone who is not a lawyer or a judge, simply means that one did not do what one is alleged to have done. There are many voices, both from the bench and from the academy, that likewise insist that the “innocence” inquestioninthe PI is innocence m . Laurence Tribe, for instance, has written that the PI represents far more than a rule of evidence. It represents a commitment to the proposition that a man who stands accused of a crime is no less entitled than his accuser to freedom and respect as an innocent member of the community. 32 Numerous scholars agree with Tribe that jurors should presume material inno- cence. 33 That notwithstanding, from the moment when the jury receives the PI instruction forward, references to innocence m disappear from the trial, rarely if ever to resurface, unless the judge repeats the PI instruction in his final charge to the jury. In its place, we see such surrogates as “not guilty” or “acquittal,” which entail nothing about innocence m . Later in this chapter, I will offer a conjecture as to why so many courts and scholars continue to insist that jurors must assume the innocence m of the defendant, even when nothing in the trial hangs on this assumption. In the meantime, it is sufficient to note again that innocence m is an idle wheel in most of the machinery of justice. On those rare occasions when a judge slips up and suggests to jurors that their task is to settle whether the accused is guilty or innocent (as opposed to guilty or not guilty), higher courts invariably cry foul, insisting (as the First 31 It is not clear why Massachusetts thinks it necessary that jurors assume the defendant to be “innocent of any crime.” Surely what is germane is that they presume the defendant to be innocent of the crime(s) of which he is charged in the indictment. 32 Emphasis added. Laurence H. Tribe, An Ounce of Detention: Preventive Justice in the World of John Mitchell, 56 Va. L. Rev. 371, at 404 (1970). Tribe’s analysis of innocence has been endorsed by U.S. District Courts (Augustus v. Roemer, 771 F. Supp. 1458, at 1464 [D. La., 1991]), among others. It is hard for me to conceive what Tribe is thinking here. Can he possibly regard detention, arrest, and pretrial incarceration as instances of the sort of “freedom and respect” that we accord to someone genuinely considered “as an innocent member of the community”? Do we give the same “freedom” to those accused of a crime as we give to their accusers? 33 See, for instance, Rinat Kitai, Presuming Innocence, 55 Okla. L. Rev. 257 (2002); William S. Laufer, The Rhetoric of Innocence, 70 Wash. L. Rev. 329 (1995); Steve Shep- pard, The Metamorphoses of Reasonable Doubt: How Changes in the Burden of Proof Have Weakened the Presumption of Innocence, 78 Notre Dame L. Rev. 1165 (2003); Scott Sundby, The Reasonable Doubt Rule and the Meaning of Innocence, 40 Hastings L. J. 457 (1989); Jeff Thaler, Punishing the Innocent: The Need for Due Process and the Presumption of Innocence Prior to Trial, 1978 Wis. L. Rev., 441 (1978). P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 Innocence, BARD, and Affirmative Defenses 99 Federal Circuit has on several occasions) that “jurors are called upon only to decide whether the prosecution has proven the defendant guilty beyond a reasonable doubt, not whether the defendant is innocent.” 34 Higher courts get even more nervous when judges suggest to jurors that they should presume that the defendant is “not guilty” as opposed to being “innocent.” When Oklahoma judge Clifford Hopper instructed jurors in Flores v. Oklahoma that they should presume that the defendant was “not guilty,” 35 appellate courts were furious, referring to Hopper’s instruction as “infamous.” 36 To explain to the jury why he was instructing themon the presumption of not guilty rather than innocent, Hopper said: “There is a difference between being innocent and not guilty. A person can be found not guilty and still not be inno- cent of the crime charged.” 37 By uttering this truism(taught in every lawschool in the world), the trial judge, in the opinion of the appellate courts, committed “an error which violates both the letter and spirit of Oklahoma statutes as well as the United States Constitution.” 38 The conviction was reversed because of the faulty instruction. (The sole dissenting voice on the Flores appeal court, Judge Gary Lumpkin, insisted that the difference between “presumed inno- cent” and “presumed not guilty,” “from the standpoint of burden of proof and impacting the trial is no more than a gnat’s hair difference to the aver- age juror.” 39 ) The reversal, appealed to the U.S. Supreme Court, was allowed to stand. It is worth noting the tensions implicit in the cases just mentioned. In the first trial, U.S. v. Andujar, we learned that judges must tell jurors that they have to decide whether the defendant is guilty or not guilty, not whether he is innocent or guilty. But when, in Flores, a judge tells jurors that they should presume the defendant is not guilty (rather than innocent), his chain is firmly yanked and higher courts insist that it is innocence m , not lackof guilt, that must be presumed. What juries have to decide (whether the defendant is not guilty) is something wholly different from what they have to presume (the defendant is innocent m ). In brief, the defendant himself has pleaded not guilty (that is, innocent p ), not innocent m . If the jury acquits him, they will be affirming his innocence p , not 34 U.S. v. Andujar, 49 F.3d 16, at 24 (1st Cir., 1995). They added, “We repeat here that, due to the risks of misleading the jury, district courts should refrain wherever possible from using a “‘guilt or innocence’ comparison in their jury instructions” (ibid.). 35 The trial judge’s actual instruction was this: “You are instructed that the defendant is presumed to be not guilty of the crime charged against him. . . unless his guilt is estab- lished by evidence beyond a reasonable doubtand that presumption of being not guilty continues . . . unless every material allegation . . . is proven by evidence beyond a reason- able doubt” (Flores v. State, 1995 OK CR 9 [Okla. Crim. App., 1995]). 36 The appellate court referred to the Flores judge’s “infamous ‘presumed not guilty’ instruc- tion” (Flores v. State, 1995 OK CR 31 [Okla. Crim. App., 1995]). 37 Ibid. 38 Ibid. 39 Ibid. P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 100 the distribution of error his innocence m . Why, then, should the jury presume his innocence m ? We will return to this tension more than once. Much of this is familiar terrain. Still, I doubt whether either the courts or academic lawyers have thought through what is entailed by the fact that the judicial systemnowhere makes a finding about the innocence m of the defendant. The implications of this odd state of affairs showup vividly in our understanding of what the PI means, what work we expect it to do for us, and what protections it can offer to the defendant. problems with the supposition of material innocence I will begin with a critique of the view that the trier of fact must believe, at the outset of a criminal trial, that the defendant is innocent m , that is, that he did not commit the crime of which he is charged. On this view of things, it falls entirely to the state – if it is to secure a conviction – to persuade the trier of fact to change his belief from one of innocence m to one of high confidence in the guilt m of the accused. This premise raises severe conceptual problems. For instance, how strong must the juror’s initial confidence in the innocence m of the defendant be in order to satisfy the demands of the PI? Ajuror cannot begin the trial assigning a probability of 1.0 to the defendant’s innocence (and thus 0.0 to the defendant’s guilt), for it would then be impossible on technical grounds for any evidence to rationally shift the juror’s opinion. 40 Some authors have argued that a juror would be in compliance with the PI as long as, going into the trial, she assigned a probability greater than 0.5 to the defendant’s material innocence. Other scholars have argued that anything close to 0.5 would be far too weak to satisfy the PI. They suggest that the proper value should be found by asking a question like, “What is the probability that a citizen selected at random is innocent m of the crime in question?” 41 In this case, the initial probability of innocence, while marginally below 1.0, would usually still be extraordinarily high. The 40 Some courts have had the temerity to suggest that the PI may require that the juror assume that the initial probability of guilt of a defendant is zero. The Supreme Court of Connecticut opined: If we assume that the presumption of innocence standard would require the prior probability of guilt to be zero, the probability of paternity in a criminal case would always be zero because Bayes’ Theorem requires the paternity index to be multiplied by a positive prior probability in order to have any utility. In other words, Bayes’ Theorem can only work if the presumption of innocence disappears from consideration (State v. Skipper, 637 A.2d 1101, at 1107 [Conn., 1994]). The court might have concluded, instead, that the PI is viable only if it permits assigning some positive magnitude to the probability of guilt m . 41 Richard Friedman, for instance, writes, “A conscientious juror would start with prior odds [of guilt] not of 1:1 but of 1:X, where Xis a large number perhaps on the order of the entire population of people who might have committed the crime” (APresumption of Innocence, Not of Even Odds, 52 Stanford L. Rev. 873, at 885 [2000]). P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 Innocence, BARD, and Affirmative Defenses 101 evident lack of consensus among those who gloss the PI as requiring triers of fact to adopt a hypothesis about the probability of innocence m (falling in the enormous range between 1.0 and 0.5) already suggests that we are less clear than we ought to be about what a presumption of innocence m should or could mean. Leaving technical probabilities to one side, there are other curious features to a presumption of innocence m . In essence, the juror is told that he must adopt an assumption – presumably that means a belief – that the defendant did not commit the crime. He is given no evidence for this proposition. How realistic is it to suppose that jurors – or anyone else, for that matter – can appropriate their beliefs in this fashion? Can we will ourselves to believe X, absent any evidence or arguments for the truth of X, simply because a judge has said that such is our duty? I doubt it. It would be one thing, and a familiar one, if the judge were to say, “For the sake of argument, let us suppose that Jones is innocent m of the charges against him, until and unless we discover otherwise.” All of us know how this conceptual ritual works and we have plenty of practice at suspending disbelief for the sake of seeing where a certain line of inference might lead. But suspension of disbelief is less than what the PI enjoins. It demands, not merely that we approach the case as if the defendant were innocent m , but that we accept the viewthat he is innocent m until such time, if ever, as the prosecutor persuades us otherwise BARD. The salience of the conundrumof believing in the innocence m of the accused shows up more vividly if we consider a common-enough sort of situation, in which jurors have acquitted a defendant but claim that they nonetheless believe that he probably committed the crime. Obviously, this is a perfectly coherent position to take: The prosecutor persuaded them that the defendant actually committed the crime, but the proof did not rise to the level of BARD. Can this be squared with the presumption of innocence m ? As we have seen, jury instructions invariably insist that jurors must presume the defendant to be innocent (and remember that innocence m is what is at stake here) until and unless they are persuaded BARD of the defendant’s guilt m . Here is how one model federal jury instruction puts the point: The lawpresumes the defendant to be innocent of all the charges against him. I therefore instruct you that the defendant is to be presumed by you to be innocent throughout your deliberations until such time, if ever, you as a jury are satisfied that the government has proven him guilty beyond a reasonable doubt. 42 This instruction precludes the possibility that a juror might say to himself, toward the end of jury deliberations, “I believe Jones committed the crime (and is thus guilty m ), but I am not yet persuaded of that BARD.” Clearly, there is 42 This instruction was proposed by the Fifth Circuit in U.S. v. Walker, 861 F.2d 810, at 813 (5th Cir. 1988) P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 102 the distribution of error something fishy going on when the rules of the system preclude this possibility. What the courts want is that jurors should not vote to convict Jones unless his guilt has been shown to the relevant standard. Fair enough. But when they attempt to micromanage this outcome via an instruction to the effect that jurors should believe or suppose Jones innocent m until and unless they are convinced that his guilt has been shown BARD, the courts are confusing the logic of guilt m with that of guilt p . The point is as important as it is subtle. In the case of any jury deliberations that terminate in a conviction, it is natural to suppose that the jurors reach a point where they have decided that Jones is not innocent m , but where they have not yet decided that their confidence in his guilt reaches BARD. That is, they believe that Jones committed the crime, but they are not yet sure that there is enough proof of that to convict him. If that is so, why on earth should jurors be instructed to believe that Jones is innocent m “until such time, if ever, as you as a jury are satisfied that the government has proven him guilty beyond a reasonable doubt”? This instruction simply doesn’t reckon with the fact that the path from believing Jones is innocent m to believing Jones is guilty p typically passes through the stage of believing Jones is guilty m but not believing he is guilty p . On the material innocence model, it is as if at one moment a juror must believe that Jones did not commit the crime and at the next moment she not only believes that Jones committed the crime but is firmly convinced of it BARD. I doubt very much that is how rational agents reason. If, as seems fair to suppose, a typical juror gradually moves from believing that Jones did not commit the crime (innocent m ), to believing that he probably did commit the crime (guilt m ), to finally being morally certain that he committed the crime (guilt p ), then there must be a time when jurors have abandoned their belief in Jones’s innocence m but have not yet embraced the belief that the likelihood of Jones’s guilt m is strong enough to convict him. This model instruction precludes such a possibility, insisting that Jones must be presumed innocent m until such time, if ever, as jurors are persuaded of his guilt m BARD. Moreover, and picking another irritating nit, if jurors are really supposed to believe Jones innocent (either materially or probatorily) until the moment when the jury collectively has reached a verdict, no juror could ever vote to convict in a straw poll, for that would violate the instruction to “presume Jones innocent until the jury is satisfied that he is guilty beyond a reasonable doubt.” 43 When 43 A Texas appellate court has previously noted this problem: “[I]f every juror, by law, had to believe the defendant is actually innocent prior to reaching a verdict, then every juror would have to believe the defendant is innocent when voting on a verdict. And if every juror believed the defendant is innocent at the time he or she votes, what must the verdict inevitably be?” (Miles v. State, 2004 Tex. App. Lexis 9788 [Tex. App., 2004]) (emphasis in the original text). P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 Innocence, BARD, and Affirmative Defenses 103 model jury instructions are as garbled as this, it is small wonder that the status of the PI is unclear. A third problem with the presumption of innocence m is the unclarity about the content of the belief meant to be conjured up in the juror’s mind by the PI. Imagine that Jones is on trial for aggravated assault with a deadly weapon against Smith. If the juror accepts the judge’s admonition to presume Jones’s innocence m , what specifically should she believe? Is it enough to believe that Jones assaulted Smith but without a deadly weapon? Is it sufficient to believe that Jones probably shot at Smith, but did so unwittingly? Would it do for the juror to believe that Jones shot at Smith and intended to harm him, but did so in self-defense? Any one of these beliefs, if true, would be sufficient to win a not-guilty verdict for Jones (and, indeed, enough to imply his innocence m ). Or, does the PI require the juror to believe that Jones never assaulted Smith at all, never aimed and fired a gun at him, did not intend to harm him, and was acting in self-defense and with diminished mental capacity and without negligence? That is, must the juror begin the trial disbelieving every element of the case that the prosecution must prove or is it sufficient to disbelieve at least one? The conventional answer, of course, is the former. But if we take that seriously, then the PI actually requires the juror to believe much more than a literal construal of “presuming Jones to be innocent m ” would demand. After all, Jones is innocent m so long as any key part of the charge against him is false. Worse, it is impossible to believe all of these things at once (for example, shooting at Smith in self-defense obviously rules out not shooting at himat all). Interestingly, model jury instructions address none of these three concerns. They do not tell the juror how strongly he should believe in the defendant’s innocence m , or how the juror is to adopt a belief for which he has no evidence, or howmany of the elements of the indictment he must disbelieve. As intriguing as these puzzles may be, they do not yet bring us to the root problem with PI. This goes deeper than worries about accepting beliefs on no evidence, arbitrarily assigning probabilistic degrees of belief to assertions of innocence m, or parsing innocence across the elements of the crime. The core mistake in the promiscuous uses of “innocence” that we see alike in jury instructions and in academic treatments of the PI is the supposition that the PI per se has anything to do with hypotheses about the innocence m of the defendant. I will argue that the PI, properly understood, should require of jurors not a belief about innocence m at all but rather a belief about the innocence p of criminal defendants at the outset of a trial. Like the judge in Flores, I will maintain that the PI is a presumption of innocence p (that is, guilt not yet established), not one of innocence m . This is a controversial thesis, so we must approach it cautiously. To begin with, let us remind ourselves that an acquittal is not an assertion of innocence m ; to the contrary, it is fully compatible with guilt m . As appellate courts have repeatedly emphasized, an acquittal is simply an assertion that the case offered P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 104 the distribution of error by the prosecution failed to satisfy the relevant standard of proof. 44 An acquittal is agnostic about the guilt m or innocence m of the accused. If we want a fair trial for a defendant, we expect the trier of fact to begin the trial by assuming that, short of hearing powerful proofs fromthe prosecutor, she will end up acquitting the defendant. We want her to assume at the outset that she has no conclusive proof of guilt. But that is not enough; she must assume that, at the outset, there is no proof at all. The fact that the accused is in the dock, the juror must suppose, provides no evidence in and of itself pertinent to questions of guilt or innocence. Nor do any snippets of information that the juror may have come by through other sources provide evidence of guilt or innocence. If the prosecutor can shift the juror’s belief that the defendant’s guilt is unproven by presenting compelling inculpatory evidence (not discredited by the defense), then the defendant should be convicted. If the prosecutor fails to do so, the juror ends the trial where she began, believing that the defendant’s guilt is unproven, although, unlike at the beginning, she need not suppose that there is no evidence of guilt at all. That is, a fair and conscientious juror should begin the trial with a presumption of full innocence p (no inculpatory evidence). To argue that, in addition to or instead of supposing innocence p , the unbiased juror in a fair trial must likewise suppose the innocence m of the defendant is neither necessary nor appropriate. We have already seen why a belief in innocence m is not necessary: It is not presupposed by even the most defendant-favorable outcome of a trial, an acquittal. It is pointless and gratuitous to insist that the juror at the beginning of a trial adopt a belief about innocence m much stronger than what he will later assert if he votes to acquit. Yet that is precisely what the presumption of innocence m requires. By contrast, the presumption of innocence p makes the same assumption about the defendant’s innocence at the opening of a trial that an acquittal implies at its end. The assumption of innocence m is no more appropriate than it is necessary. After all, a juror, at the outset, has no basis for either affirming or denying the innocence m of the defendant. She does not know the defendant, nor has she any idea what sorts of exculpatory and inculpatory evidence will arise in the course of the trial, and so on. Worse, the assumption of innocence m requires the juror to deny what she already knows: to wit, that persons do not get to the trial stage in a criminal proceeding if the probability of their guilt m is vanishingly small. Insisting that the juror must go into a trial utterly convinced that the defendant is innocent m belies the elaborate, winnowing structure of the legal system; it acts as if the decision to put someone on trial were conducted by some sort of randomizing mechanism; and it insults the intelligence of jurors. 44 The Fourth Federal Circuit put the point this way: “Averdict of acquittal demonstrates only a lack of proof beyond a reasonable doubt; it does not necessarily establish the defendant’s innocence“ (U.S. v. Isom, 886 F.2d 736, at 738 (4th Cir., 1989). P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 Innocence, BARD, and Affirmative Defenses 105 By contrast, instructing the juror to assume that she has seen as yet no evidence of the defendant’s guilt correctly mirrors the juror’s own epistemic situation and thus is an injunction that she can act on in good conscience. We can note straightaway that this way of construing the PI avoids several awkward problems, already mentioned, associated with a presumption of inno- cence m . In the latter case, for instance, someone has to decide what probability or degree of confidence the juror should initially assign to the hypothesis that the defendant is innocent m . There is no nonarbitrary solution to that problem. By contrast, the presumption of probatory innocence can be mute on this point. Indeed, it implies that the ideal juror is one who has no beliefs about which side is likely to prevail. Some scholars, most notably Judge Richard Posner, think otherwise. Specif- ically, and in the context of a discussion of possible juror bias, Posner has argued that “Ideally we want the trier of fact to work from prior odds of 1 to 1 that the plaintiff or prosecutor has a meritorious case.” 45 Obviously, with- out using my terminology, Posner is referring to probatory rather than material guilt and innocence, a focus unambiguously signaled by his reference to the quality of the case to be presented rather than the material innocence or guilt of the accused. According to Posner, the ideal juror will be one who, going into a trial, supposes it equally likely that the prosecution or the defense will prevail. Is such an assumption entailed by, or even compatible with, the PI? I think not. Here is why: We all suppose that the criminal SoP is very high. For the sake of argument, let us suppose that it weighs in at about 90 percent. To win a case, the prosecutor must persuade the jury that the apparent guilt of the defendant reaches this value; anythingless demands anacquittal. Nowif, followingPosner, a juror supposes that the prosecutor is as likely to win the case as he is to lose it, that must mean that the juror believes that the apparent guilt of the accused is as likely to be above 90 percent as it is likely to be less than 90 percent. This, in turn, implies – unless we have a fiendishly bizarre distribution of degrees of apparent guilt among a large class of defendants – that most defendants are guilty m (even if only half of them are assumed to be guilty p ). That may be true as a statistical generalization about the cases that actually go to trial, but surely we do not want jurors going into a trial believing that most defendants are guilty m , especially not when (as now) the SoP is defined in terms of subjective conviction about guilt. In sum, Posner’s stipulation of the ideal juror as assigning equal initial odds to guilt p and innocence p is not only unmotivated but is a clear violation of the spirit, and probably the letter, of the PI. It would be far better to say that the fair-minded juror is one who is open- minded about the eventual outcome of the trial and feels under no obligation, 45 Richard Posner, An Economic Approach to the Law of Evidence, 51 Stan. L. Rev. 1477, at 1514 (1999). P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 106 the distribution of error nor even in a position, to assign odds – even or otherwise – to the trial’s out- come. 46 What is important is that the juror concedes that she has no proof now about guilt and that, therefore, she lacks any clue about which side will eventually prevail. This is a patently true description of her situation and any juror should accede to its truth. She can simply take it, as indeed she should, as an obviously correct characterization of the epistemic situation in which she finds herself. If we take this route, we need not ask the juror to believe some- thing for which she has no evidence (such as, for instance, that the defendant is innocent m ). To be clear, I am not claiming that we can be utterly indifferent to the views that jurors may harbor concerning the innocence m of the defendant. High confidence, prior to trial, that the defendant is guilty m is clearly a sign of bias and should be grounds for juror exclusion for cause. But, at the risk of committing heresy, I submit that high confidence in the innocence m of the defendant – of precisely the sort enjoined by the standard reading of the PI – is just as troublesome a source of bias. Anyone who goes into a trial believing that she already has enough information about a case to have a firm opinion about guilt or innocence is suspect as a trier of fact. This is not to gainsay that rational and suitable jurors, on the strength of pretrial publicity and the like, may have tentative hunches about the innocence m or guilt m of the defendant. So long as such persons will grant that they have no evidence of guilt and that the verdict must depend only on the proof offered at trial, they are qualified as jurors. The crucial point is that the juror must be willing to accept, without reservation, the thesis of innocence p, regardless of what tentative hypotheses she may have entertained about guilt m and innocence m . The Core Failing: Confusing Juror Confidence with a Standard of Proof I have been railing in this chapter against the view that jurors must begin a trial with a strong belief in the innocence m of the accused. I have suggested, rather, that what they must believe is that the defendant’s guilt has not yet been proved and that his appearance as defendant constitutes no evidence of his guilt. Since jurors are never called upon to affirmthe defendant’s innocence m , I have argued that presuming such innocence is, at best, gratuitous. If, in retrospect, this seems fairly obvious, we need to try to figure out why it has seemed so natural for so long to think that a presumption of innocence m was essential to a fair trial. I conjecture that the principal factor operating here has been the ubiquity, especially in Anglo-Saxon criminal law (but not without its counterpart in 46 We must disabuse ourselves of the Bayesian-inspired notion that one does or one should assign probabilities here and now to a vast collection of future states of affairs. Just as I have no opinion about whether it is likely to rain next September 23 at noon in Kuala Lumpur, so a reasonable juror can – and should –be unwilling to assign odds at the opening of a trial to its possible outcomes. P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 Innocence, BARD, and Affirmative Defenses 107 Roman law countries, as we will see later), of a certain model of juror/judge deliberation and belief change that I called in Chapter 3 the theory of the subjective standard of proof. Recall that the received view, simultaneously normative and descriptive, of juror decision making goes something like this: The trier of fact begins (and should begin) a trial assigning a high probability to the defendant’s innocence m (and, accordingly, assigning a very low probability to his guilt m) . As the trial proceeds, each juror is thought to be constantly revising her estimate of apparent innocence m in light of evidence presented by both defense and prosecution. As final deliberation begins, the only question facing the juror is whether her by- now multiply revised estimate of the probability of guilt m reaches or exceeds the standard of proof. If she is firmly convinced about guilt m , if she harbors no reasonable doubts about guilt m , if her belief in guilt m is the sort of belief upon which she would base important life decisions, then – the judge will have instructed her – she should convict the defendant. Otherwise, she should acquit him. I have already argued that we must reject this subjective account of the standard of proof. We need a standard that really functions as a standard – that is, a standard that tells one when one is entitled to regard someone’s guilt as genuinely proven, for the purposes at hand. If we had such a standard in place (and I proposed some plausible candidates in Chapter 3), then we would not be beguiled into believing that what goes on in a criminal trial issuing in a conviction is that the juror’s subjective confidence increases step by step, until that level of confidence exceeds some preordained level, p. With a proper standard of proof, the focus of everyone’s epistemic attention would not be on the shifting probabilities that jurors assign to the hypothesis of guilt m. Instead, it would be on the point at which jurors decide that the prosecution’s case exhibits those features demanded by the SoP. The PI that is appropriate in such a context is not one of innocence m but of innocence p. Before quitting the subject of the PI, it is worth briefly setting ourselves an empirical question, as opposed to the normative ones that have occupied us thus far in this chapter. Specifically, what evidence do we have concerning the actual ratio of truly guilty and truly innocent defendants who go to trial? Are most of those defendants who stand trial innocent or are most guilty? The second seems the far more likely hypothesis. There is ample anecdotal evidence to this effect, much of it coming from defense attorneys – who are likely to have a pretty good idea whether their clients are guilty or innocent. Here is how one distinguished member of that fraternity, Allen Dershowitz, sees it: Rule I. Almost all criminal defendants are, in fact, guilty. Rule II. All criminal defense lawyers, prosecutors and judges understand and believe Rule I. 47 47 Alan M. Dershowitz, The Best Defense xxi–xxii (1982). P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 108 the distribution of error Still, it would be preferable to have something better than anecdotal evidence to go on. Where direct evidence is concerned, we obviously have very little since true guilt and true innocence are largely inscrutable in the vast majority of cases. But there are some plausible inferences we can draw from what data we have at hand. Begin with the fact that the acquittal rate in criminal trials across the United States falls in the range of 30–40 percent. (Although there are variations among jurisdictions and among crimes, the overall figures have held steady in this range for decades.) Let us take the midpoint of this range, 35 percent, and see what we can plausibly infer from it. The first and most obvious (but not necessarily the most telling) inference is that, in the opinion of the jurors (and judges), most defendants are guilty. Indeed, their apparent guilt is so striking that a clear majority have been found to be guilty BARD. So, if about 65 percent of those who went to trial were adjudged guilty BARD, there were probably numerous others whose apparent guilt was well above 0.5 and who, there- fore, may well have been guilty. Still, juries, even with the best of intentions, will sometimes make mistakes; so it would be precarious to infer that at least 65 percent of the defendants were truly guilty. The SoPenables us to tighten up this inference a bit. In particular, it gives us a mechanism for putting a plausible upper bound on the proportion of defendants who are innocent m, supposing that the jurors actually utilize a SoPthat produces about ten true acquittals for every false conviction and that jurors’ inferences from the evidence are generally reasonable ones. Taking the most charitable possible construal of the data – charitable in the sense that it gives maximum latitude tothe possibilityof innocence –suppose that the juries make nomistakes when they acquit defendants. So, on this assumption, at least thirty-five of every hundred defendants are innocent m . This hypothesis – that there are no false acquittals – is wildly implausible, of course, but let’s adopt it since we want to find out the maximum plausible proportion of innocent m defendants among those who go to trial. We can infer from the SoP that another 10 percent (here about four) of those who were innocent m were convicted. Hence, under the most favorable possible circumstances, it is plausible to believe that fewer than 40 percent of criminal defendants are innocent m . 48 (In fact, the actual proportion is probably far lower than this since we have powerful reasons to expect that at least some of those acquitted were guilty m .) It will be worth bearing this back-of-the-envelope calculation in mind in later sections of the book, when we turn to discuss rules of evidence and procedure. What will be at stake there is whether we should structure the evidence rules to make important additional concessions to the defendants in a trial, over and 48 Alan Gelfand and Herbert Soloman have estimated, using a different sort of calculation frommine, that the ratio of truly guilty to truly innocent defendants who go to trial is about 2:1 (Modeling Jury Verdicts in the American Legal System, 69 J. Am. Stat. Assn. 32, at 35–6 [1974]). P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 Innocence, BARD, and Affirmative Defenses 109 above those already embodied in the SoP and the PI. We will then need to remind ourselves that, since most defendants are guilty m , any such advantages we offer are more likely to be sops to the guilty m than lifelines to the innocent m . The Burden of Proof The onus probandi in a criminal case is the obligation of one party to persuade the trier of fact that the evidence presented by that party proves, to the level specified by the SoP, a hypothesis of interest. Typically, the hypotheses of interest will be assertions about the defendant’s conduct that exemplify the “elements” of the crime, as defined by the relevant statutes. For simplicity’s sake, we can say that the burden of proof involves the duty to prove the two hypotheses: a) Crime x was committed, and b) defendant y committed x. Officially, everyone agrees that, in criminal cases, the burden of proof falls explicitly and exclusively on the state. (I will explain the “officially” shortly.) The criminal defendant, so the theory goes, need prove nothing, nor is he even under any obligation to offer evidence, should he choose not to do so. Even in such circumstances of a wholly inert defendant, the jury is instructed to acquit unless the prosecution’s case reaches or exceeds the SoP. Like the presumption of innocence, the standard of proof, and the benefit of the doubt, the principle of a prosecutorial burden of proof is designed to favor the defendant. That we immunize the defendant fromany obligation to prove anything makes acquittals much easier to secure than they would be if he were under an obligation to prove his innocence. If we were not interested in promoting false acquittals over false convictions, we would do what civil courts do: Use an undemanding SoP and allow the burden of proof to shift back and forth between the parties as the trial progresses. If we understand the idea of a burden of proof this way, it becomes clear that there is considerable conceptual redundancy between it, the PI, and the SoP. The SoP can readily enough be formulated as follows: “The defendant must be acquitted unless the state establishes that the evidence of the defendant’s guilt exceeds BARD (or whatever other standard may be in play).” That is, the standard of proof specifies a level of proof, it gives a decision rule for conviction or acquittal, and (in the version I have just stated) it spells out that the burden of proof falls on the prosecutor. On this analysis, a separate burden of proof principle is clearly unnecessary. Alternatively, one could formulate the SoP more narrowly, along these lines: “The defendant must be acquitted unless the evidence of the defendant’s guilt exceeds a certain level.” This formulation leaves it less clear where the burden of proof falls and would probably need supplementation by a separate rule that the state bears such a burden. But if we accept both the SoP and the PI, then the burden of proof thesis becomes superfluous. By contrast, conjoining the SoP with the burden of proof would not quite render the PI unnecessary since neither the SoP nor the prosecutorial P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 110 the distribution of error burden thesis makes it entirely explicit that the defendant begins the trial without prejudice. That is, it is the “clean slate” element of the PI that is not already explicitly present in the SoP or in the burden of proof. 49 So conceived, the PI is an important prophylactic device, cautioning jurors to attach no probatory significance to the fact that the defendant sits (metaphorically) in the dock. The same cannot be said, under normal circumstances, for the thesis of the burden of proof, which is fully dispensable provided that a demanding SoP and a presumption of innocence p are in place. The Puzzling Case of Affirmative Defenses Unfortunately, circumstances are not always “normal.” In state courts through- out the United States, and in courts in many other countries, there are numerous situations where the defendant is obliged not only to present some relevant evidence (called the burden of production) but also to prove certain exculpatory assertions that he has made. These situations are generally called affirmative defenses, and they pose a serious conundrum for legal epistemology. 50 We can see why affirmative defenses pose a challenge if we remind ourselves of some striking differences between civil and criminal trials. In the former, the burdens of proof and production of evidence can routinely move back and forth between the contending parties, rather like a ball in a tennis match. By contrast, as we have seen, conventional wisdom has it that in a criminal trial the burden of proof is fixed squarely on the prosecution’s shoulders. As the Model Federal Jury Instructions phrase the point: As a result of the defendant’s plea of not guilty, the burden is on the prosecution to prove guilt beyond a reasonable doubt. This burden never shifts to a defendant for the simple reason that the lawnever imposes upon a defendant in a criminal case the burden or duty of calling any witness or producing any evidence. 51 Sadly, in the criminal law, “never” almost never really means never. Defendants in state courts (where the great majority of criminal prosecutions occur) are 49 Indeed, there is a standard model jury instruction saying as much: I remind you that an indictment itself is not evidence. It merely describes the charges made against the defendant. It is an accusation. It may not be considered by you as any evidence of the guilt of the defendant. In reaching your determination of whether the government has proved the defendant guilty beyond a reasonable doubt, you may consider only the evidence introduced or lack of evidence (1–3 Modern Federal Jury Instructions–Criminal P 3.01). 50 Another important, but unobjectionable, set of circumstances in which the defendant bears a burden of proof occurs at the preliminary or evidential hearing. If the defendant proposes to introduce a piece of evidence, e, to which the prosecutor objects, then it falls to the defendant to show, by a preponderanceof the evidence, that e complies with the Rules of Evidence. I discuss the logic of this situation in Chapter 5. 51 1–4 Modern Federal Jury Instructions–Criminal P 4.01. (Emphasis added.) P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 Innocence, BARD, and Affirmative Defenses 111 frequently required to assume a significant burden of proof in order to win an acquittal. I will describe some of the factors that lead to the creation of a defendant-based burden of proof and then ask what these practices imply not only for the PI but likewise for our attitude toward the preferred ratio of true acquittals to false convictions. Affirmative defenses are not some recherch´ e or arcane area of the law. 52 Such defenses can arise whenever the defendant concedes (or fails to dispute) the prosecution’s claim that he committed the acts of which he stands accused but argues that he is nonetheless not guilty because the lawspecifically excludes punishing someone in his predicament. Consider a few of the many kinds of situations that fall under the rubric of an affirmative defense: r If Jones harms Smith because Smith is about to harm Jones or someone else or to destroy Jones’s property. r If Jones harms Smith while Jones is temporarily insane or involuntarily intoxicated. r If Jones harms Smith with Smith’s permission (perhaps Jones is performing surgery with Smith’s consent or having consensual sex with Smith). r If Jones harms Smith while properly enforcing the law. r If Jones harms Smith while enjoying diplomatic immunity. r If Jones harmed Smith long ago and the statute of limitations has expired. In these and dozens of other situations (by my count, some sixty in all), many jurisdictions will require that, if Jones is to win an acquittal, he must prove his story, usually to the PoE standard but sometimes even to the BARD standard (the latter is not unknown where insanity defenses are involved). At least eleven states 53 have statutes requiring the defendant to prove his defense to a PoE. In Delaware, Georgia, and North Carolina, the defendant must prove certain defenses “to the satisfaction of the jury,” whatever that means. In Kentucky, the evidence for an affirmative defense must be “convincing,” whatever that means. Delaware, Georgia, and Oregon have required defendants to prove an insanity defense BARD, a requirement astonishingly upheld by the U.S. Supreme Court. 54 However these unconventional standards are glossed, it is clear that all require of the defendant substantially more than that he establish a reasonable possibility that his defense is true, which I take to be the functional equivalent of raising a reasonable doubt about his guilt. Such practices generate problems both for our interpretation of the PI and for the prosecutorial burden of proof. For instance, how can we say that jurors 52 A very good account of the philosophical nuances of affirmative defenses can be found in Mitchell Berman, Justification and Excuse, Law and Morality, 53 Duke L. J. 1 (2003). 53 Alaska, Delaware, Illinois, Louisiana, Maryland, Ohio, Pennsylvania, Rhode Island, South Carolina, Texas, and West Virginia. 54 See Rivera v. Delaware, 429 U.S. 877 (1976), and Leland v. Oregon, 343 U.S. 790 (1952). P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 112 the distribution of error are supposing Jones to be innocent p if he has to prove something to secure an acquittal? Howcan we say that the criminal defendant “never” bears the burden of proof, when a whole class of situations explicitly imposes a substantial burden on the defendant? More importantly, affirmative defenses undermine the rationale for setting a SoP at all. As I argued at length in Chapter 3, criminal trials impose a very high SoP on the prosecution for a very good reason – to wit, the broad consensus that we want a high ratio of true acquittals to false convictions. As we saw, the sharedperceptionof howfrequently we are prepared to accept a false conviction is what enables us to define the height of the bar for conviction nonarbitrarily. The problem with affirmative defenses is that the policy of imposing a significant burden of proof on the defendant who (for example) claims he acted in self-defense flies in the face of that social contract. The question before us involves asking specifically what this social contract implies for affirmative defenses that require the defendant to establish more than a reasonable doubt about his innocence. The answer is clear: If a state requires that the defendant establish a certain defense (such as self-defense or consent) to a PoE (or even higher level), that state is saying that erroneously convicting someone who genuinely acted in self-defense or with the consent of the victimis no greater an injustice than acquitting someone who falsely alleges self-defense. If a state says that a defendant claiming insanity must prove his insanity by CACE (let alone BARD), that state is saying that we would much prefer a false conviction to a true acquittal. These, I will argue, are curious and ultimately incoherent judgments of value. It is bad enough that they fly in the face of the Blackstonian thesis that false acquittals are less costly than false convictions. It is still worse that they will lead to false convictions of the innocent far more often than one time in m. They add insult to those injuries by undermining the PI and the prosecutorial burden of proof when they insist that a defendant can win an acquittal only if he can prove his innocence to a relatively high standard. 55 It might be helpful to consider this hypothetical pair of examples: Jones and Smith are both on trial for first-degree murder. Jones offers an alibi (which is not an affirmative defense in the strict sense), presenting witnesses who claim to have been with himelsewhere at the time of the crime. The judge will instruct the jury that for Jones to win an acquittal, he need only raise a reasonable doubt that he was not at the crime scene. Smith, by contrast, claims to have been acting in self-defense. In many jurisdictions (but not in federal trials), Smith 55 Jeffries and Stephan cannot understand why anyone might make a fuss of the sort I am raising here since, they say, “These devices [presumptions and affirmative defenses] have existed for a long time and have not been widely perceived or popularly condemned as invasions of the presumption of innocence” (John Jeffries and Paul Stephan, Defenses, Presumptions, and Burden of Proof in the Criminal Law, 88 Yale L. J., 1325, at 1347 [1978–9]). The absence of a public hue and cry seems an extraordinarily jejune criterion to apply the evaluation of any public policy. P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 Innocence, BARD, and Affirmative Defenses 113 must present enough evidence to make it more likely than not that he was indeed acting in self-defense. Unless he does so, his evidence may be excluded or, even if it is admitted, the judge may elect not to give the jury a self-defense instruction. Jones and Smith, charged with the same crime, obviously enjoy quite different prospects. Jones will be acquitted if he can raise a reasonable doubt about his guilt. Smith, by contrast, is bound to be convicted if all he can do is to raise a reasonable doubt; he must prove it more likely than not that he acted in self-defense. The crucial issue for us involves the messages that the justice system is sending in this pair of cases. In Jones’s trial, the message – implicit in BARD – is that it is far better to acquit the guilty than to convict the innocent. In Smith’s trial, by contrast, the inescapable message is that convicting the innocent and acquitting the guilty are equally undesirable. Trials like Jones’s guarantee that most innocent defendants will be acquitted. Trials like Smith’s do no such thing; they ensure that about half of the innocent defendants will be convicted. The pertinent question is a simple one: What is the principled difference between these cases that would justify such discrepant assessments of the rela- tive costs of errors and such divergent assessments of the importance of acquit- ting the truly innocent? Jones, it is true, is denying that he committed an act that the state is obliged to prove, while Smith is conceding that he committed the act but insisting that his behavior was justified and can point to a state statute on self-defense that stipulates self-defense to be a full justification for acting as he did. If Smith’s action was genuinely one of self-defense, then he is every bit as innocent – legally and morally – of the crime as Jones is, if his alibi is true. But when the state insists that Smith must prove it to be more likely than not that he acted in self-defense, it is saying that convicting an innocent Smith would be a much less weighty error than convicting an innocent Jones. This is daft. If “innocence” is to have a univocal meaning, and only chaos can ensue if it does not, then we must hew to the line that convicting a person innocent m of a given crime brings the same costs, independently of the specific attributes that render innocent m those who are wrongly convicted. Likewise, acquitting a guilty m person arguably generally brings the same costs, whether he adopts an affirmative defense or simply denies his guilt. To hold that convicting the innocent is sometimes much worse than acquitting the guilty while other times saying that convicting the innocent is no worse (or perhaps even better) than acquitting the guilty is to fall into babbling incoherence; even worse, it is unjust. It may be true, as it is often said to be, that state legislatures are consti- tutionally at liberty to allow or disallow certain excuses or justifications, for they write the substantive law. Perhaps some states will decide that insanity, self-defense, or consent is not to be recognized as a legitimate defense. But if a state does recognize that acting in self-defense or with consent makes an oth- erwise criminal act blameless, then that state can have no reason to oblige the defendant to prove that he probably so acted. BARD, or whatever criminal SoP P1: JZZ 0521861667c04 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 15:10 114 the distribution of error is actually in play, must govern such trials if there is to be even a semblance of epistemic coherence in the criminal justice system. 56 Similar arguments apply to most of the other defenses. 57 The Supreme Court decided a long time ago that the Constitution provides no obstacles to states defining whatever SoP they like in cases involving affirmative defenses. Not being a constitutional lawyer, I do not intend to argue that case. But it is one thing to say that such a policy would be constitutionally permissible and quite another to say that it would be epistemically coherent or sound public policy. I have tried to suggest that, what- ever the Constitution allows here, there is no principled rationale for imposing a positive burden of proof on the defendant in cases of insanity, self-defense, or consent. It is sometimes suggested that we should not kick up too much fuss about this anomaly since it offers more protection to a defendant to have the option of claiming self-defense – even if he must prove it himself – than if self-defense were not even recognized as an exculpatory factor. 58 Many legislatures, so the argument goes, agree to allowaffirmative defenses only because they impose a significant burden of proof for the defense. That may well be so; still, it does not gainsay the claim that such an attitude on the part of legislators, or anyone else, reflects a poor grasp of why we set the various standards of proof where we do. 56 As Barbara Underwood rightly puts it, “The costs of erroneous convictions and erroneous acquittals are not different by virtue of the gratuitous character of the defense” (Barbara Underwood, The Thumb on the Scales of Justice, 86 Yale L. J., 1299, at 1322 [1977]). 57 I say “most” because I believe that there are a few defenses where the calculation of costs of errors and the ratio of true acquittals to false convictions should be different from the usual one. Consider, for instance, the affirmative defense of the expiry of the statute of limitations. Here, defendant concedes he committed the crime, offers no exculpatory justification for doing so, but simply claims that the clock has run out on punishment for the crime. In such cases, BARD, or some other high prosecutorial standard, would be inappropriate since cases like this do not involve putting the truly innocent at risk (since the defendant is evidently truly guilty m ). Similar arguments might apply to the affirmative defense of diplomatic immunity. 58 For instance, Jeffries and Stephan have argued that “shifting the burden of proof [to the defendant] is often politically necessary to secure legislative reform. It seems quite possible, therefore, that disallowance of this procedural device would work to inhibit reform and induce retrogression in the penal law” (Jeffries and Stephan, at 1356). P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 part i i FLAWED RULES OF EVI DENCE AND PROCEDURE Due process does not require that every conceivable step be taken, at whatever cost, to eliminate the possibility of convicting an innocent person. – Justice Byron White for the Supreme Court 1 1 Patterson v. N.Y., 432 U.S. 197, at 208 (1977). 115 P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 116 P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 5 Evaluating Evidence and Procedures The great body of the law of evidence consists of rules that operate to exclude relevant evidence. – McCormick’s Handbook of the Law of Evidence 2 [A] criminal trial is anything but a pure search for truth. When defense attorneys represent guilty clients – as most do, most of the time – their responsibility is to try, by all fair and ethical means, to prevent the truth about their client’s guilt from emerging. – Alan M. Dershowitz 3 None of the concepts studied thus far – the standard of proof, the burden of proof, the BoD, or the PI – does anything to reduce erroneous verdicts in the aggregate. All are targeted at making it likely that, when errors do occur, they are more likely to be false acquittals than false convictions. We must now move away from issues of error distribution – what we might call soft-core legal epistemology – and toward its hard-core: error reduction. The following four chapters will examine a variety of specific rules, practices, and procedures that serve as major obstacles to the twin tasks of finding truth and avoiding error. In this chapter, our remit will be broader and more theoretical. It is a long-standing tradition in law schools and law texts to draw a sharp boundary between the rules of evidence and the laws of criminal procedure. Indeed, evidence law and criminal procedures are generally taught in dif- ferent classes and by different specialists. By contrast, my focus will be on both. Having argued that the key distinction in legal epistemology is between principles of error reduction and those of error distribution, rather than between evidence and procedures, I suggest that a failure to grasp the full import of this 2 Charles T. McCormick, McCormick’s Handbook of the Lawof Evidence, at 121, § 53 (Edward W. Cleary, ed., 2nd ed., 1972). 3 Alan M. Dershowitz, Reasonable Doubts 166 (1996). 117 P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 118 flawed rules of evidence and procedure distinction explains why we see a blurring – sometimes even a fusing – of error reduction and error distribution concerns in both the theory of evidence and the theory of procedures. This is, I believe, a fundamental mistake, because the bulk of both procedural and evidence law should form a part of hard-core, not soft-core, legal epistemology. A Meta-Rule for Evaluating Evidence The key diagram in Chapter 3 (Figure 3.3) showed how moving the SoP can impact the distribution of errors. That same figure can be put to good use when we turn to a discussion of rules of evidence and procedure. It showed two curves representing respectively the distribution of apparent guilt among the truly innocent and of apparent guilt among the truly guilty. Because those curves partially overlap, we face the problem of false acquittals and false convictions. Some guiltypersons will appear more innocent thansome trulyinnocent persons do, and, of course, some innocent persons will appear guiltier than some truly guilty persons do. So long as those curves partially overlap, jurors – however well intentioned and sagacious – are going to reach erroneous conclusions about the guilt and innocence of some defendants. Such inferences do not represent failings on the part of jurors, who, we will suppose, are fully attentive and rational in their evaluations. Rather, they reflect the fact that the available evidence does not always offer a true picture of guilt and innocence. While the problem of some degree of overlap between the two curves is insoluble – in the sense that a certain minimal overlap will always exist so long as evidence is incomplete and human judgment is fallible – the magnitude of the problem can be vastly decreased if, by an act of imagination, we were to “pull” the two distributions farther away from one another. In Figure 3.3, the peaks of the two distributions were relatively close. As we have seen, shifting the SoP either to the right or to the left will do nothing to alter the overlap between the two distributions. To be quite specific, shifting the SoP toward the right does nothing to improve the accuracy of verdicts. To the contrary, we already saw in Chapter 3 that its principal effect is to give us a different ratio of true acquittals to false convictions than a less demanding standard would. But suppose we could move the peak of the truly innocent curve to the left and that of the truly guilty to the right. Overlap between the tails of the two curves would not vanish entirely, but there would be far fewer cases of truly innocent persons who looked as guilty as many of the truly guilty do and vice versa. Even if we do not move the midpoints of the two curves apart, we can imagine that the spread (or variance) associated with each curve becomes smaller. Again, the result would be less overlap between the innocent and the guilty. The real-life counterpart to these spatial gymnastics is to improve our eviden- tial practices. The more robust and the more complete the evidence presented to a jury, the greater will become the gap between the midpoints of the two P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 Evaluating Evidence and Procedure 119 distributions, thereby reducing the area of overlap between them. As Michael DeKay has remarked, “Anything that makes truly guilty defendants appear more guilty or truly innocent defendants appear less guilty moves the curves apart and increases accuracy.” 4 By contrast, if we adopt epistemically suspect proce- dures or sloppy rules of evidence, the overlapping areas under the curves will intermingle even further than illustrated in Figure 3.3. Juries will find it harder and harder to distinguish the truly guilty from the truly innocent. We can put the same fundamental point in a different way: If we are interested in reducing the frequency of false convictions, there are precisely three ways to do that. One way is to put fewer truly innocent people on trial. That sounds fine in theory but is wholly impractical, since we are never sure who is truly innocent and who is truly guilty. A more viable solution is to shift the SoP even farther toward the right than we have contemplated thus far. We pay two costs by solving the problem in this fashion: The ratio of true acquittals to false convictions dramatically increases (above that already sanctioned by the social contract), and we may well increase both the frequency of false acquittals and the overall rate of errors. Fortunately, there is a third way of reducing false convictions – improving the discriminatory power of the evidence – that exacts neither of these sacrifices. It consists of two strategies: a) enhance the quality and quantity of the evidence that jurors see, and b) put in place mechanisms likely to catch mistaken verdicts when they occur. What sounds straightforward in theory might be devilishly difficult in prac- tice, of course. But it is not as difficult as might be imagined, provided we are prepared to insist on using rules of evidence and procedure that will enhance the likelihood that jurors can distinguish between the truly guilty and the truly innocent. Such a policy will involve costs, as we will see in later chapters. The costs involved are not epistemic ones. (Specifically, the adoption of rules that make it more likely that jurors would be able to separate the truly guilty from the truly innocent correctly will sometimes require us to reconceptualize certain rights belonging to the defendant. It also implies that the courts should get out of the dubious business of using the laws of evidence to police the police.) What we are talking about here is the epistemologist’s promised land: error reduction. Anything we can do to make it more likely that the truly guilty are seen to be guilty and that the truly innocent appear to be innocent reduces the overall likelihood of erroneous verdicts. It should be the principal function of the rules of evidence and procedure to ensure that a defendant’s apparent guilt is as good an indicator of his true guilt or innocence as we can make it. Our crucial experiment will seek out a set of rules for conducting a trial that would reduce the overlap between these two distributions as far as our current investigative and forensic techniques allow. 4 Michael DeKay, The Difference between Blackstone-Like Error Ratios and Probabilistic Standards of Proof, 21 Law & So. Inquiry 95, at 97 (1996). P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 120 flawed rules of evidence and procedure I just said that the chief aim of the rules of evidence should be to lessen the overlap in question. Patently, many current rules do not do that. Instead, as we will see in detail in the next four chapters, we are lumbered with a set of rules and procedures that do far less than they could, and should, to help jurors mark off the truly guilty from the truly innocent. It is notorious that the bulk of evidence law in common law countries (especially the United States) consists of scores of rules that exclude relevant evidence from the jury’s attention. Occasionally, these exclusions make epistemic sense. If, for instance, we were to discover that there is a certain kind of relevant evidence (hearsay, for example) whose importance juries are apt tooverestimate, thenexcludingit might be appropriate, since giving evidence more weight than it deserves further blurs the boundaries between the two distributions. But the vast majority of current exclusions of evidence are not based on the fear that the excluded evidence might confuse or obfuscate the issues at stake in the trial. Instead, highly germane evidence – evidence that would indisputably reduce the overall rate of erroneous verdicts – is excluded for reasons having nothing to do with the search for the truth. So, inthe spirit of our epistemic thought experiment, I propose toexplore how to formulate rules of evidence and procedure that would increase the likelihood that jurors’ judgments about apparent guilt and innocence map onto the true state of affairs. It is customary, and entirely appropriate, to parse this question in terms of two key concepts: relevance and reliability. As we saw in Chapter 1, physical evidence or testimony is relevant provided that, if true, it increases or decreases the likelihood of guilt of the accused. Evidence or testimony is reliable just in case there is a solid basis for believing that it is true. The most important work in a criminal trial is the assessment of the relevance and the reliability of the various exhibits and testimony that the contending parties enter into evidence. The trial judge handles some of this work. Most importantly, it falls to her to decide whether the evidence offered by the two parties is genuinely relevant. If it is not, she will properly exclude it as time wasting and distracting. Until recently, jurors were expected to make most of the decisions about the reliability of the evidence, because such decisions are closely bound up with that most traditional of jury functions: determining the facts. In recent years, however, judges have been entrusted with making at least some of the reliability decisions as well. For instance, if one party or other wants to introduce expert testimony, judges decide whether the methods used by the expert are sufficiently reliable that jurors should hear his testimony. Even so, it is ultimately the jurors who decide whether what the expert – if they are permitted to hear one – has to say is really worthy of belief. In that sense, the jurors remain the ultimate deciders of questions of the reliability of such evidence as they hear, even when the judge acts as preliminary gatekeeper. For epistemic purposes, it is not very important who makes the decision about the reliability of proffered evidence, so long as it is made in a rational way (of which, more later). P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 Evaluating Evidence and Procedure 121 Obviously, one cannot control what the jury makes of the evidence; that happens behind the closed doors of the jury room and will have to be treated largely as a black box in our analysis. Although epistemology has much to say about what counts as reliable evidence, there is not much point exploring that here since jury decisions about reliability are beyond the reach of epistemic control. 5 True, judges can instruct jurors and counsel can try to persuade them to reason this way rather than that. But nothing guarantees that the jury will comply. The best that can be done is to raise the likelihood that the judgment reached by the jury about apparent guilt is as well informed as it can be. To that end, I propose a simple rule for guiding our decisions about the admission of evidence: The triers of fact – whether jurors or judges in a bench trial – should see all (and only) the reliable, nonredundant evidence that is relevant to the events associated with the alleged crime. An aside: For those like myself keen to give less prominence to the judge’s role as gatekeeper, one could delete the word “reliable,” since there is no epistemic reason not to leave reliability decisions entirely in the hands of the jury itself, providedthat arguments fromcounsel andinstructions fromthe judge informthe jury’s assessments of reliability. Indeed, the rule would be much easier to apply with the suggested deletion since reliability is a great deal more problematic than relevance. For one thing, relevance does not admit readily of degrees: Evidence either is relevant or it isn’t. By contrast, reliability notoriously admits of enormous variations from barely to fully. If evidence is only marginally reliable, and if the judge is charged with determinations of reliability, she might reasonably exclude it, even if it is relevant. For this reason, I prefer the rule in its simpler formulation. It is my hunch that no further rule of evidence is necessary in a truth-driven system of judicial investigation. As David Hume might have said, “Let us commit the massive tomes on evidence law to the flames since their principal function is to articulate hundreds of exceptions, qualifications, and truncations of this core rule of evidential relevance.” 6 Ponder this: The hearsay rule is only one among dozens of exceptions to the principle that relevant evidence is to be admitted. The Federal Rules of Evidence contain twenty-nine specific exceptions to the hearsay rule. To be clear, these are exceptions to the exception. It then adds an all-purpose “residual exception,” admitting almost any hearsay evidence that will serve “the interests of justice,” whatever that might mean. 5 One obvious exception is the adoption of an objective SoP of the sort proposed in Chapter 3, which would give the system a bit more control over how jurors weight the facts. 6 The hornbook version of the classic evidence lawtext, Charles T. McCormick, McCormick on Evidence (John Strong, ed., 5th ed., 1999), devotes fewer than 5 of its 540 pages to a treatment of relevance. P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 122 flawed rules of evidence and procedure I conjecture that every failure to comply with the rule of relevance – that is, everyexceptiontoit –increases the odds that the jurywill err whenit assesses the apparent guilt or innocence of the accused. Of course, I could be wrong. What would count as being wrong would be the discovery of forms of genuinely relevant evidence to which jurors were incapable of giving the appropriate weight. The current rules of evidence abound with situations where the courts fear that jurors may evaluate the evidence wrongly. As we saw in Chapter 1, some evidence is regarded as “unfairly prejudicial” and thus excluded. Hearsay testimony, while relevant, is excluded for fear jurors will not recognize how weak (that is, unreliable) it sometimes is. The defendant’s criminal history is excluded from the prosecution’s case-in-chief for fear that jurors may attach too much significance to the defendant’s shady past or will convict himbecause he is a bad person, not because they are convinced that he committed the crime. In these and many other cases, the rules of evidence are predicated on claims about juror psychology. These generally fail to rise above the level of folk mythology. I have no quarrel with the suggestion that, if we should learn from a well-designed empirical study that jurors are apt to misconstrue the value of a particular sort of relevant evidence, then steps must be taken, either to exclude it or to devise a judge’s instruction that cautions the jury against misinterpreting it. Indeed, I think that any legal epistemology worth its salt must be open to learning from experience in this way. Having said that, I have yet to see robust evidence that any of the just mentioned exclusions are necessary. But should we come to discover such problems, I am quite happy to add a codicil to this effect to the meta-rule of evidence that I just proposed: Where there is compelling evidence that a certain type of relevant and reliable evidence is likely to be misconstrued by jurors and where there is likewise evidence that such misconstruals are not readily susceptible of correction by judge’s instructions and the arguments of opposing counsel, such evidence should be excluded. Given the gap between an epistemologically robust evidence policy and the set of rules currentlyprevailing, there is ample roomfor mistakenverdicts. Some of these mistakes are simply failures of human judgment. But many mistakes – among them the most readily remediable – are due to arcane rules of evidence that offer exceptions and amendments to the rule of relevance. Although the courts frequently pay lip service to relevance as the most important factor in assessing evidence, most of so-called American evidence law consists in a long rehearsal of exceptions to the rule of relevance. While the Federal Rules of Evidence do recognize the relevance principle (Article IV, Rule 402 confidently asserts, “All relevant evidence is admissible. . . . Evidence which is not relevant is inadmissible”), subsequent rules introduce so many exceptions to this rule that it is honored far more in the breach than in the observance. Part of the problem stems from that fact that the drafters of the Federal Rules of Evidence failed to grasp that the principles of relevance and reliability P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 Evaluating Evidence and Procedure 123 are not merely two rules of evidence that coexist alongside many other rules. Relevance and reliability are meta-principles that should be used for evaluating all other proposed rules of evidence. A so-called rule of evidence that denies the jury access to genuinely relevant evidence (and most do) fails the basic test of adequacy as a rule of evidence, which should be to enhance the likelihood that the apparent guilt of a defendant is a good indicator of his true guilt or innocence. There are several reasons why the existing rules of evidence have not been grounded on these epistemic home truths. Many rules attempt to safeguard various nonepistemic values, chiefly having to do with the rights of defendants, the importance of the court not resorting to – or seeming to endorse – illegality, and the role of the courts in policing the behavior of the police. We will consider those worries in detail in the last chapter. But they have no place in our thought experiment, which is concerned with investigating the counterfactual question: What would a criminal trial look like if its fundamental and overriding concern were to find out the truth about a crime? Still, the discrepancy between the rules as they might be if truth were the principal aim and the rules as we actually find them is not due entirely to the intervention of extra-epistemic values. Many of the current evidential practices are there because it is hoped that they will further reduce the likelihood of a false conviction. In other words, the explicit or latent function of many rules of evidence and of procedure is to alter the balance between false convictions and false acquittals, even if that means that the overall accuracy of verdicts (which is what drives the preference for relevance and reliability) is compromised. To be clear, I am asserting that many existing rules of evidence are clumsy efforts at, or have the effect of, increasing the BoDalready given to the defendant by the SoP. This means, in turn, that they shift the values of n and of m, and thereby the SoP, to even higher levels than those already settled in what I have been calling the social contract. Such rules are in place not to increase the likelihood of a true verdict but with a view to decreasing still further the likelihood of a false conviction. They are, in other words, driven by concerns with error distribution rather than error reduction. They are seen as instruments for reducing one sort of error, the false conviction, while their proponents remain blissfully indifferent to their nugatory impact on error reduction in general. Importing distribution issues into the rules of evidence and procedure is a grave error. Much of the rest of this chapter will be devoted to documenting how far the preoccupation with distribution has penetrated into the rules of evidence and procedure. But first, I want to explain why, if many rules of evidence are motivated by distributionist motives, this should be a source of serious concern. As I argued in Chapter 3, the only way to define a nonarbi- trary SoPis to build it upon a determination of the acceptable ratio of true acquit- tals and false convictions (leavened by a knowledge of the impact of a change P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 124 flawed rules of evidence and procedure of that ratio on the respective costs of false acquittals and false convictions). We saw that the SoP should be defined as that point at which society could be genuinely indifferent about whether a trial ends in acquittal or in conviction. A SoP higher than that point would generate a rational preference for convictions over acquittals; a lower SoP would tilt the balance the other way, with society preferring an acquittal over a conviction. It follows that, once such a SoP has been settled, all our subsequent delib- erations should be driven by a concern with reducing errors rather than further distributing them. Having selected a SoP that should make us indifferent to whether someone is acquitted or convicted, we have no conceivable incentive for trying to reduce still further the frequency of false convictions, if the price we thereby pay is increasing the frequency of false acquittals. Indifferent to whether a trial ends in acquittal or conviction, our only remaining concern should be with error reduction, not with error distribution. We reduce errors, I have argued, by seeing to it that the jury hears as much relevant evidence as the parties have in hand. We do not promote that goal if we then say that we should make it easier for the defendant to present exculpatory evidence than for the state to present such evidence. Nor do we promote it if we then say that certain sorts of evidence should be excluded because they might lead to a false conviction (if such exclusions are even more apt to lead to still more false acquittals). Nor do we promote it if we put in place procedures, such as the asymmetric discovery rule, that then give a specific evidentiary advantage to the defendant. Probatory even-handedness – which is another phrase for admitting all relevant evidence – is what should be driving the selection of rules of evidence and procedure. A failure on the part of judges, legal scholars, and legislators to grasp that the SoP alone should incorporate distributionist sentiments has created an intel- lectual milieu in which it is considered entirely appropriate to commend a rule of evidence or procedure on the grounds that the rule will decrease the chance of false acquittals or that it will give the defendant a supposedly well-deserved BoD. If we are to have a coherent theory of evidence and procedure, we need to insist that such concerns are entirely out of place in this arena, having already been fully catered to in the setting of the standard of proof. In the rest of this chapter, I will showhowpervasive distributionist sentiment is in the literature of the law of evidence and how many existing rules and procedures derive their rationale not from the rule of relevance but from a desire to inject further distributionist sensibilities into the conduct of a trial. Reining In Distributionist Excesses Contemplate for a moment the situation of the innocent defendant facing trial. He knows (if he has followed our argument this far) that, thanks to the SoP, he P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 Evaluating Evidence and Procedure 125 runs only a small to slight chance of being wrongly convicted, supposing the SoP to be in the neighborhood of 90–95 percent. He knows, further, that even if falsely convicted, he has the prospect of getting the conviction reversed on appeal. He likewise knows, thanks to the PI, that the jury must ignore the fact that several earlier actors in his drama (judges, police, and prosecutors) have found significant evidence against him. He knows, too, that (unless he adopts an affirmative defense) he is not obliged to say or do anything to win an acquittal. He desperately wants, we may suppose, a true verdict. Thus far, his interests coincide with society’s. Still, he is understandably worried by the slim but nonnegligible chance a) that he will be convicted and b) that his false conviction will be allowed to stand on appeal. So, he (and more importantly his advocates in the defense bar) would like to reduce further the prospects of his being falsely convicted. There is some incriminatory evidence against him; otherwise, his case would not have come this far. Nothing would please himmore than to discover a body of exclusionary rules that will keep the jury from hearing and seeing some of that. At this point, his interests in winning an acquittal and the interests of guilty defendants in doing the same begin to coalesce, while our interests as truth seekers begin to diverge fromthe fondest hopes of either. Although we would certainly like to see the innocent defendant acquitted, we understand full well that any evidentiary concessions made to the innocent defendant will benefit those defendants who are guilty as much as, and sometimes more than, the innocent. Having already made life relatively low-risk for the innocent defendant via the SoP, we should be reluctant to see further concessions ladled out, since we know that most of the recipients of this largesse are guilty, and that their guilt will become harder and harder to discern as we erect more and more barriers to the admission of relevant evidence. Still, many legal experts do not see it this way. They believe that pro- defendant bias should be incorporated in all the elements of a trials, heaping benefit upon benefit, with a view to narrowing still further the risks of convict- ing the innocent. They do this even while conceding that such concessions will help far more guilty defendants than innocent ones. Consider a hypothetical example. Suppose someone were to argue as follows: We all concede that a few retracted confessions are false. We likewise grant that from time to time a jury might take such a false confession as dispositive, thereby convicting an inno- cent defendant. Accordingly, we could avoid the undesirable false convictions generated by permitting the admission of retracted confessions if we were to exclude all retracted confessions. Such a rule would clearly extend the BoDwell beyond the standard of proof. It says that if there is any doubt about whether a confession was voluntary (doubt of the sort that would be implied by the defendant’s retraction of his confession), then we should exclude it. If, as I believe and most commentators concede, a P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 126 flawed rules of evidence and procedure clear majority of retracted confessions are certainly relevant and probably true, then such a rule implies a minor increase in m(the proportion of true acquittals to false convictions) and a drastic increase in n (the ratio of false acquittals to false convictions). Several independent studies indicate that the resolution of around 50–60 percent of criminal cases depend, in part, on evidence fromconfessions. 7 If defendants, having once confessed to the police, suddenly learned that they could get their confessions automatically thrown out simply by retracting them, presumably large numbers of cases now resolvable as true convictions would end up as false acquittals. Note that the exclusion of all retracted confessions would not only change the ratio of true acquittals to false convictions from that ratio already incorporated into the SoP; almost certainly it would likewise increase the aggregate number of false verdicts in comparison with current practice, which admits many (but not all) retracted confessions. But let us leave to one side for the moment the fact that this rule would break the social agreement (already incorporated into the SoP). Let us likewise ignore its negative impact on the frequency of false verdicts. Further, let us not belabor the fact that acquittal-conducive rules such as this create a situation in which defendants, whose cases turn on the rules in question, receive more BoD than those whose cases do not, raising significant issues of fairness and due process. Let us focus instead on the fact that, although this proposed rule would increase the BoD enjoyed by the defendant, it is wholly unclear by how much. Since we don’t know, not even approximately, either how many retracted confessions are false nor how often juries regard retracted confessions as dispositive, it is impossible to estimate how large an effect the implementation of this rule would have on n and on m. All we can say is that it would increase the former ratio by some indeterminate amount and the latter by an even greater, but still indeterminate, amount. In sum, were we to accept such rules as this silly one (“exclude all retracted confessions”), we would be giving an indeterminate extra dollop of BoD to the defendant. As I have already argued, any incremental increase in the BoD above that already incorporated into a properly designed standard of proof would renege on the social contract. This policy would also make it impossible to ascertain how much additional BoD any given defendant was receiving. Why does all this matter? What is at stake is whether the criminal justice system effectively implements society’s values concerning the kinds and quan- tity of mistakes and successes that should occur in a criminal trial. When we 7 See, for instance, Paul G. Cassell, Miranda’s Social Costs: An Empirical Reassessment, 90 Nw. U. L. Rev. 387, at 394–418, 437–8 (1996) (57 percent confession rate); Richard A. Leo, The Impact of Miranda Revisited, 86 J. Crim. L. & Criminology 621, at 659 (1996) (64 percent rate); George C. Thomas, III, Plain Talk about the Miranda Empirical Debate: A“Steady-State” Theory of Confessions, 43 UCLAL. Rev. 933, at 955 (1996) (confession rate of 55 percent). P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 Evaluating Evidence and Procedure 127 have no clue about how much BoD any given defendant is receiving (which is the current situation in American law), we cannot decide whether the courts are being too lenient or too tough on those who face trial. The only way to ensure the court’s compliance with sound public policy is by localizing all benefits of the doubt to those areas of trial procedure where we can calibrate how heavily the scales of justice are tilted in the defendant’s favor. I have already argued that the SoP is the only appropriate place for incorporating BoD concerns, because that standardmakes it clear for all whowant tosee just what concessions have been built into the system. If we permit such distributionist instincts to shape rules of evidence and procedure over and above the SoP, it becomes impossible for anyone to figure out whether criminal defendants are receiving more or less BoD than we consider appropriate. Concerns such as these are germane whenever distributionist sensibilities drive the choice of a rule of evidence or procedure. Indeed, if (as I have argued) a properly defined SoPalready incorporates society’s judgment about the appro- priate amount of BoD to be granted to the accused, then the adoption of any acquittal-prone (as opposed to an error-reducing) rule of evidence or procedure implies that the defendant is receiving more BoD than that to which he or she is entitled. (Likewise, a conviction-prone rule would alter mand n unacceptably in the direction of giving the defendant less BoD than that to which he is entitled. But there are relatively few of those. 8 ) A critic of this analysis might voice the perfectly legitimate worry that per- haps the existing SoP fails to incorporate the full BoD regarded as socially acceptable. Wouldn’t it then be appropriate to opt for defendant-friendly rules? The problem, as I’ve already hinted, is that adding additional dollops of the BoD by adopting defendant-friendly rules may produce a BoD that is mas- sively greater than the acceptable one. I reiterate a key corollary of the argu- ment from Chapter 3: The only way to ensure that a criminal trial gives the defendant something close to the level of the BoD that is socially acceptable is to incorporate that quantum of doubt directly into the SoP and to exclude error-distributionist concerns from the debate about the appropriate rules of evidence. It follows that, if we discover that the current SoP confers less BoD than society regards as desirable, the only appropriate remedy is to raise the existing standard of proof, not to skew the rules of evidence in favor of the defendant. 8 I have already discussed at some length in Chapter 4 one instance where the rules of evi- dence and procedure skew the trial outcome in a prosecutor-friendly direction (the case of affirmative defenses). I will mention here two other rules that are, in my view, unacceptably prosecutor-congenial: 1) the rule allowing the prosecutor to oblige a reluctant witness to testify by granting himimmunity, while the defendant sometimes lacks such powers, and 2) the rule not requiring the prosecutor to divulge to the defense during discovery information that she regards as only mildly exculpatory. P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 128 flawed rules of evidence and procedure Insofar as there is a coherent rationale behind the idea that the rules of evidence should favor acquittals, it would seem to be this: In deciding whether to admit a given sort of evidence, a rational person must assess the utilities of admitting it and of excluding it. That assessment will depend on two sorts of things: the cost of the error likely to be produced by the rule and the probability that the error in question will occur. Consider our hypothetical rule: Exclude all retracted confessions. Any proponent of this rule knows that it is likely to exclude more true confessions than false ones. Thus, he knows that it will produce more false acquittals than would occur absent the rule. Still, he might say, that is not an end to the matter. Since false convictions are (for example) ten times much more costly than false acquittals, it is reasonable to accept a rule that is likely to produce (for example) five times more false acquittals than false convictions because the multiplication of each of these costs by their respective probabilities will still produce a positive utility for the rule. Where this otherwise plausible strategy goes awry is in neglecting the fact that the SoP was set as high as it was precisely because at that point, and only at that point, the utilities of an acquittal and a conviction are the same. Having set the SoP at that point, coherence requires us to be utterly indifferent as to whether an acquittal or a conviction occurs in trials governed by such a standard. Accordingly, we should favor rules that are neutral between acquittals and convictions (or better, rules that reduce the likelihood of error in general rather than rules that favor one sort of error over another), which is precisely the policy that the rule of relevance independently urges on us. Those who advocate rules that are more likely to produce acquittals than convictions ignore these facts at their peril and to the detriment of a coherent theory of evidence. In effect, they are engaged in a game of double counting, where the SoP is first raised to protect the innocent defendant and then the rules of evidence are further skewed to favor acquittals, as if the requisite concessions to the defendant were not already reflected in the SoP. This is about as rational as it would be if someone, having bought comprehensive health insurance to cover her medical expenses, then continued to set aside substantial savings in the event she had expensive surgery. Just as the point of buying insurance is to avoid having to salt away money for medical emergencies, the reason for adopting a high SoP is to avoid having to shave the rules of evidence so that they favor the defendant. So far, the argument I’ve sketched is fairly abstract. I’ve offered only one example of an epistemically offensive rule of evidence (the proposal to exclude all retracted confessions) and that, you may imagine, is so absurd as to constitute little more than a straw man. So, I need to put that situation right. The first step involves pointing out that my example of a silly distributionist rule was not my own invention, designed to paint the distributionists as mildly demented. Far from being a figment of my imagination, the rule in question was proposed by Corey Ayling in the Wisconsin Law Review in 1984. Ayling’s argument is, P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 Evaluating Evidence and Procedure 129 as we will see repeatedly, typical of a whole genre of rule justifications and criticisms that warrant our vigilance. In making the case for his proposed rule, Ayling writes: Since the prevailing ethos of American criminal law is that it is better to free a guilty person than to convict an innocent person, outlawing confessions would be justified even if the net result were to free more guilty persons than innocent ones. 9 Clearly, this is an unmitigatedly distributionist argument. A proposed rule of evidence is to be preferred over an existing one, Ayling says, not because it will decrease the overall error rate but because, while increasing the overall inaccuracy of the system, it is also likely to produce fewer false convictions than the current rule! Katherine Goldwasser, Professor of Law at Washington University, makes a similar argument to justify a rule that would permit the defendant to intro- duce evidence that fails the reliability test associated with hearsay and expert testimony. Such acquittal-augmenting rules as she offers are acceptable, she wrote in 1998, because “producing fewer wrong convictions at the cost of also producing a few extra wrong acquittals is fully consistent with (indeed, rein- forces) reasonable doubt values.” She continues, “to exclude defense evidence (and thereby increase the risk of an erroneous conviction) solely out of concern about the risk of an erroneous acquittal is flatly unacceptable.” 10 Other examples of this tendency to use the rules of evidence to alter the distribution of errors abound in the scholarly literature. I will cite just a few. In 2003, Richard Friedman argued, in a vein similar to Goldwasser’s, that standards for the admission of evidence should be lower for the defense than for the prosecution because of the “shared perception that the social cost of an inaccurate judgment, given that the defendant is in fact innocent, is many times greater than the social cost of an inaccurate judgment given that the defendant is in fact guilty.” 11 He holds that the defendant should be able to introduce into evidence kinds of testimony and physical evidence that, if offered by the prosecution, would be excluded. I amall in favor of making it as easy as possible for parties to introduce relevant evidence, but I can see no reason to apply that principle differently to the prosecution than to the defense. Another pair of authors, Daniel J. Seidmann and Alex Stein, have used the same strategy to justify what is known as the Griffin rule (to be discussed at length in Chapter 7), which holds that the jury cannot use the silence of the defendant as a basis for drawing any inferences about the defendant’s guilt 9 Corey J. Ayling, Corroborating Confessions: An Empirical Analysis of Legal Safeguards against False Confessions. Wis. L. Rev. 1121, at 1199 (1984). 10 Katherine Goldwasser, Vindicating the Right to Trial by Jury and the Requirement of Proof beyond a Reasonable Doubt: A Critique of the Conventional Wisdom about Excluding Defense Evidence, 86 Geo. L. J. 621, at 635 (1998). 11 Green v. U.S., 355 U.S. 184, at 187–8 (1957). P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 130 flawed rules of evidence and procedure or innocence. No one, least of all Seidmann and Stein, disputes that this rule produces far more false acquittals than false convictions and far more false verdicts than would be likely if the Griffin rule were abandoned. But these scholars evidently prefer it that way. They argue: The abolition of Griffin would produce obvious benefits in all three situations; it would force guilty suspects to plead guilty, to lie and to face rebuttal, or to risk adverse infer- ences from silence, thereby increasing the rate of correct convictions. The costs of eliminating Griffin, however, would undermine these benefits. The abolition of Griffin would . . . increase the rate of wrongful convictions. If the prevention of wrongful convic- tions is of immensely greater value to society than the prevention of wrongful acquittals, then the retention of the Griffin doctrine . . . would be the socially optimal choice. 12 It is not only the truth-thwarting Griffin rule that Seidmann and Stein are out to defend. They insist that it is never sound to prefer aggregate error reduction to the reduction of false convictions: The wrongful conviction of an innocent defendant (a “false positive”) is much costlier than the wrongful acquittal of a criminal (a “false negative”). Therefore, in contrast to civil case errors, a reduction in criminal false negatives cannot offset false positives. 13 Even if one supposed (which I patently do not) that it is appropriate to bring worries about error distribution into the evaluation of rules of evidence and procedure, the assertion they have just made – to the effect that a decrease in false acquittals can never offset any increase in false convictions – would still not be valid. Suppose, for example, we could show that changing some evidence rule, R, would eliminate fifty false acquittals for every false conviction that it generated. Suppose, too, that like Blackstone, they believed that false convictions were ten times worse than false acquittals. In these circumstances, even the staunchest advocate of distributionism in evidence law should favor the emendation of R, even though it increases the rate of false convictions. The common thread running through all these arguments – if we put it in its most general terms – is that any rule of evidence apt to produce fewer false convictions than a rival is preferable to that rival, even if the rule in question produces overall more false verdicts thanits rival. Some authors might hesitate to put it that way, of course, since stated so baldly, this meta-rule is patently absurd (though this fact seems to have gone unnoticed by Seidmann and Stein). But that is precisely the principle driving their arguments. As Goldwasser formulates this perspective: It is “flatly unacceptable” to reject a rule that would lower the risk of false convictions simply because of the impact of that rule on the frequency 12 Daniel J. Seidmann and Alex Stein, The Right to Silence Helps the Innocent, 114 Harv. L. Rev. 430, at 494 (2000). 13 Ibid., at 487. (Emphasis added.) P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 Evaluating Evidence and Procedure 131 of false acquittals. The doctrine that I am describing probably received its most general and explicit formulation in an argument of Stephen Saltzburg: Since our society has chosen to give criminal defendants the benefit of all reasonable factual doubts – a benefit not usually conferred upon civil litigants – rules of evidence [in the criminal law] may be tailored in a principled way to reflect and support this choice. 14 This is the position, and the meta-rule, that I have in mind when I refer to “error distributionism” in the law of evidence. It is, as Saltzburg says, a “principled” approach to evidence law. The problem is that it is grounded on the wrong principle. I submit that such scholars are offering us a very slippery slope, one that we don’t want to venture down, for there’s no point along it where traction can be had. There is virtually no plausible rule of evidence or procedure that will never produce the occasional false conviction. There is virtually no existing rule that couldnot be replacedbya weaker, evenmore acquittal-prone, versionthat would decrease still further the likelihood of a false conviction. After all, eyewitness testimony sometimes leads to wrong identification and false conviction. So do expert testimony, forensic evidence, confessions, and so on. If we once accept that it is appropriate in the evaluation of rules of evidence to ignore aggregate errors and to valorize the avoidance of false convictions over everything else, the only acceptable rule for the conduct of a criminal trial would be to “acquit every defendant.” No other rule will so effectively avoid false convictions. I wish I could report that it is only a handful of academic lawyers who slip down this slope. Unfortunately, the same BoDcreep that we see in the arguments of many academic lawyers shows up in arguments offered by appellate courts for favoring rules of evidence that are defendant-friendly rather than rules that promote true verdicts. Consider three examples, all taken from Supreme Court rulings: r In 1957, Justice Hugo Black set out to defend the patently defendant-friendly attitude embodied in the current American practice of allowing appeals of convictions but not of acquittals. No serious student of the lawwould dispute that this appellate asymmetry distorts the ratio of true acquittals to false con- victions already incorporated in the SoP, since higher courts can reverse convictions, some of which may indeed be false, while it does nothing to catch false acquittals. No one denies that many erroneous acquittals might be corrected on appeal, if it were allowed. That notwithstanding, Black rejected appeals of acquittals – not because they would violate double jeopardy, but because such appeals would “enhance the possibility that even though inno- cent, [the defendant] may be found guilty [in a second trial].” 15 Black is 14 Stephen Saltzburg, Standards of Proof and Preliminary Questions of Fact, 27 Stan L. Rev. 271, at 304 (1974–5). 15 Green v. U.S., 355 U.S. 184, at 188 (U.S., 1957). P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 132 flawed rules of evidence and procedure surely right that if acquittals could be appealed, an innocent person, cor- rectly acquitted in his first trial, might occasionally find that result reversed by an appellate court and end up falsely convicted by a second jury. This would, of course, be a gross miscarriage of justice. But what Black fails to say is that moving to a system of symmetric appeals would vastly decrease the overall frequency of erroneous verdicts. Further, it would do so without altering one whit the BoD concessions embodied in the SoP. Black’s distri- butionist worries (“the truly innocent might occasionally be convicted”) pale incomparisonwiththe error reducers’ worries growingout of the asymmetric nature of appeals. r Seven years later, Justice Arthur J. Goldberg mounted a similar argument in Murphy concerning the privilege against self-incrimination. While grant- ing that this privilege is a “shelter to the guilty,” Goldberg insists – as the Supreme Court had in Twining half a century earlier – that this privilege should be maintained because it offers “a protection [against false convic- tion] to the innocent.” 16 There may be powerful moral arguments in favor of this privilege (though I doubt it, as I argue in Chapter 6). But Goldberg is purporting to offer an epistemic argument for the silence rule, saying that it reduces the rate of false convictions, even while conceding that it dramatically increases the number of false acquittals. 17 r An even more overt attempt to import the BARDstandard and BoDconcerns into debates about the rules of evidence emerged in a 1975 case, Lego v. Twomey. Here, the defendant had offered a confession that he subsequently sought to retract, claiming it was involuntary. The judge in the case heard the arguments on both sides and decided, by a PoE, that the confession had been freely given and so was admissible. Convicted, the defendant appealed, arguing that his confession should have been admitted only if the evidence supporting the hypothesis that it was voluntary had been proved to a BARD standard. Fortunately, the court rejected this argument on a split vote. But three of the then-justices, WilliamBrennan, WilliamDouglas, and Thurgood Marshall, insisted that confessions should be admitted only if there were no reasonable doubts about their being voluntary. Here is how Brennan framed the argument (with Douglas and Marshall concurring): It seems to me that the same considerations that demand the reasonable-doubt stan- dard when guilt or innocence is at stake also demand that standard when the question is the admissibility of an allegedly involuntary confession. If we permit the prose- cution to prove by a preponderance of the evidence that a confession was voluntary, 16 Murphy v. Waterfront Comm’n of New York Harbor, 378 U.S. 52, at 55 (U.S., 1964). 17 Perhaps sensing that Murphy had tilted too far toward distributionism, the Court two years later went on record saying pointedly that “the basic purposes that lie behind the privi- lege against self-incrimination do not relate to protecting the innocent from conviction” (Tehan v. U.S., 382 U.S. 406, at 415 [U.S., 1966]). P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 Evaluating Evidence and Procedure 133 then, to paraphrase Mr. Justice Harlan, we must be prepared to justify the view that it is no more serious in general to admit involuntary confessions than it is to exclude voluntary confessions. I am not prepared to justify that view. Compelled self-incrimination is so alien to the American sense of justice that I see no way that such a view could ever be justified. If we are to provide “concrete substance” for the command of the Fifth Amendment that no person shall be compelled to condemn himself, we must insist, as we do at the trial of guilt or innocence, that the pros- ecution prove that the defendant’s confession was voluntary beyond a reasonable doubt. . . . In my judgment, to paraphrase Mr. Justice Harlan again, the command of the Fifth Amendment reflects the determination of our society that it is worse to permit involuntary self-condemnation than it is to deprive a jury of probative evi- dence. Just as we do not convict when there is a reasonable doubt of guilt, we should not permit the prosecution to introduce into evidence a defendant’s confession when there is a reasonable doubt that it was the product of his free and rational choice. 18 It is crucial that we understand what this strategy would entail. Brennan and his colleagues are arguing that, when the defendant alleges that the admission of a certain type of evidence would violate his constitutional rights, it falls to the prosecutor to prove BARD that the admission of the evidence in question would not compromise those rights. If this strategy were accepted, it would have implications far beyond confessions. If, for instance, a defendant claimed that some incriminatory evidence was seized from his home without probable cause, the state – in order even to present such evidence to the jury – would have to prove BARD that there had been probable cause for police to enter the defendant’s premises and seize the items in question. The argument of the minority in Lego v. Twomey was no aberration. In numerous other cases, courts have insisted that a disputed confession can be admitted only if the state has proved BARDor by CACEthat it was voluntary. 19 Numerous appellate court rulings similarly have insisted that decisions about admissibility of nonconfessional evidence should be driven by concerns about error distribution. Some courts have held, for instance, that evidence obtained by searches must be “proven by clear and convincing evidence” 20 to have been based on legitimate warrants. Other rulings have held that in-court eyewitness identifications of the accused, in order to be admissible, must “establish by clear and convincing evidence that the in-court identifications were based upon observations of the suspect” rather than observations made in a police lineup. 21 Still others have insisted that evidence obtained by electronic surveillance can 18 Lego v. Twomey, 404 U.S. 477, at 493–5 (U.S., 1972). 19 Among the many examples, see Clifton v. U.S., 125 US App. D.C. 257, 371 F.2d 354, 360–4 (1966); Mullins v. U.S., 382 F.2d 258, at 261 (1967); U.S. v. Inman, 352 F.2d 954, at 956 (1965); U.S. v. Smith, 308 F.2d 657, at 663 (1962). 20 See ibid. 21 U.S. v. Wade, 388 U.S. 218, at 240 (1967). P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 134 flawed rules of evidence and procedure be admitted only if it has been shown to a “high standard” that it was constitu- tionally secured. 22 All such proposals are profoundly confused and epistemically subversive. They are confused because they specify a high quantum of proof, not for deci- sions about the defendant’s guilt (where it would be appropriate) but for the admission of evidence (where it patently is not). They confuse relevance and admissibility with proof. No serious court in the history of the common law ever imagined that every piece of evidence or testimony presented to a jury must be shown beyond a reasonable doubt to have been legally obtained. Such proposals subvert truth seeking because they deny the trier of fact access to information that is indisputably relevant and has been certified by a judge as probably compatible with the constitutional rights of the defendant. If no retracted confession could be admitted unless it could be shown BARD to have violated no rights of the accused, then far fewer confessions could be certified as admissible, since BARD is a very demanding standard. That would land us in a position virtually indistinguishable from the one that Ayling advocated earlier, in which all retracted confessions are withheld from the jury. We need to keep reminding ourselves that a high SoP is appropriate only when we want to distribute errors in a particular fashion. If we focus, for the moment, on decisions about the admissibility of evidence, we have, as usual, two sorts of errors that may arise. But the germane errors here are not ones of false acquittal and false conviction. A preliminary hearing on the evidence, after all, does not issue in a finding of guilt or innocence. Instead, the errors in question concern themselves with the decisions that this tribunal is capable of making. One such error would be the admission of relevant evidence (such as the result of a seizure) that violated the defendant’s right against illegal search and seizure. The counterpart error would be that of excluding some relevant evidence that was legally seized. Even if we believe that it is ten times worse to convict an innocent defendant than to acquit a guilty one, there cannot be any clear-headed analyst who believes that this Blackstonian maximentails that a) the admission of relevant but illegally seized evidence would be ten times more egregious than b) the exclusion of relevant and legally seized evidence. Insisting on the BARD standard for the admissibility of evidence is tanta- mount to saying that it is ten times worse to violate a defendant’s rights than it is to keep legally seized, relevant evidence out of the hands of the jury. In my view, this sentiment is preposterous, entailing as it does that the interests of truth seeking are trivial in comparison with the protection of defendants’ rights. To sum up, the only sorts of errors to which Blackstonian, distributionist sentiments properly apply are erroneous verdicts, that is, false convictions or false acquittals. No clear-headed person holds that the admission of illegally 22 U.S. v. Taglianetti, 274 F. Supp. 220, at 226 (D.R.I. 1967). P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 Evaluating Evidence and Procedure 135 seized but relevant evidence (which is the pertinent error in question here) is likely to lead to false convictions. To the contrary, relevant evidence – however it was obtained (short of outright fabrication by the police) – is likely to produce fewer false convictions and fewer false acquittals. By insisting (Lego notwith- standing) that evidence admissibility decisions should rest on a PoE standard, courts have indicated that outcomes a) and b) are equally undesirable. This seems to me the right call. The use of BARD to govern admissibility decisions would violate the principle of indifference. Some legal analysts accept that a mere PoE is the appropriate standard for admitting most sorts of evidence but still hold out for a much tougher standard (usually BARD) where the admission of confessions is concerned. They seem to think that to admit a confession is, de facto, to convict the accused. As one legal scholar asks rhetorically, “In short, isn’t it likely that the existence of a confession, no matter how obtained, facilitates conviction by a jury?” 23 The answer to that question surely is yes. But one is minded to add: So what? Every piece of relevant, incriminating evidence, if admitted, “is likely to facilitate conviction by a jury.” And so it should. That is what we mean when we call it incriminating or inculpatory. The question here should not be this one but rather, “Do juries, given a confession of whatever provenance, regard a conviction as a done deal?” We have reason to think not. In most jurisdictions, even if a judge has ruled a confession admissible, the jury is expected to make its own determination as to whether it was voluntary. Moreover, defense counsel can introduce any relevant evidence that he has tending to show that the confession was not given willingly. Likewise, the jury can – and should – attempt to ascertain whether the content of a confession is independently corroborated by other evidence offered by the prosecution. If it is not, they should, and probably will, discount it. None of this is to deny that jurors may frequently give a corroborated confession considerable weight in their deliberations. But that is only proper; a corroborated confession, I will argue in Chapter 7, is powerful inculpatory evidence. After this digression about the logic of the admission of evidence and con- fessions, we must not lose sight of the larger issue before us. We have seen in the last several pages several thinkers arguing that we should raise the bar for the admission of relevant evidence proffered by the prosecution. Others have insisted that we should lower the bar for the admission of evidence proffered by the defendant. Both policies entail that inculpatory evidence must meet a tougher standardthanexculpatoryevidence. Either of these policies wouldmake it more difficult to convict the truly guilty and easier to acquit both the innocent and the guilty. We have no incentive to accept such distributionist trade-offs if we are genuinely indifferent to the outcome of a trial. 23 Saltzburg, at 305. P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 136 flawed rules of evidence and procedure To this point, I have been trying to illustrate the extent to which a familiar but flawed argumentative strategy for the appraisal of evidence rules aims to extend distributional concerns well beyond the scope of the standard of proof. We can characterize the implicit meta-rule in play here, without undeserved caricature, as amounting to the claimthat “error distributionist values should always trump error reduction values.” This meta-rule is thoroughly undermining of the search for truth in verdicts. If I have belabored it, that is because it undergirds a great many debates about the rules of evidence. Implicit Distributionist Creep The problembefore us, however, goes well beyondsuchexplicit instances of this meta-rule in play. What confronts us, if we clearheadedly examine (for example) the U.S. Federal Rules of Evidence or the Federal Rules of Criminal Procedure, is a host of rules governing the conduct of a criminal trial that, whatever their original or current rationale, de facto result in a skewing of m and n further in the direction of bestowing additional BoD on the defendant well beyond that already incorporated into the SoP. Whether these rules were originally put in place specifically because they were intended as acquittal-conducive concessions or for other reasons, the fact is that these rules violate the principle that rules of evidence and procedure should be driven by error reduction values (basically, relevance and reliability) rather than error distributionist ones. Some Rules of Evidence That Offend against the Antidistributionist Policy I offer here a partial list of some federal rules of evidence that exhibit such skewing: 1. The exclusion of non-Mirandized confessions, even when their contents reveal a knowledge of the details of the crime likely to be known only to the perpetrator. 2. The exclusionary rule prohibiting the use by the prosecution of relevant, but illegally obtained, evidence. 3. The so-called poison-fruit corollary of the exclusionary rule, which likewise rules as inadmissible any evidence the police may have uncovered by fol- lowing up on information obtained from an illegal search or seizure or an illegal confession. 4. The creation of a class of privileged witnesses not obliged to testify concern- ing what the defendant may have told them about the crime. 5. The Griffin-rule insistence that jurors must be instructed to draw no adverse inferences from the silence of the defendant. P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 Evaluating Evidence and Procedure 137 6. The exclusion of character evidence from the prosecution’s case-in-chief (in all cases except sex crimes, where it can be presented). 7. The obligation of the prosecution to prove it likely that a retracted confes- sion was voluntary before it can be admitted. 8. The rejection of relevant evidence if it might “unfairly prejudice” the jury. 9. The exclusion of confessions, however well corroborated, if obtained after more than six or eight hours of detention of the defendant without seeing a magistrate. 10. The failure to informjuries when a defendant has refused to allowtestimony from a privileged witness. 11. The exclusion of a voluntary and corroborated confession, proffered after an arrest without probable cause. 12. The failure to inform juries if there were witnesses with knowledge of the crime who invoked their Fifth Amendment right not to testify. 13. The obstacles that discourage jurors from asking questions of witnesses; the jury’s inability to subpoena those who appear to have a knowledge of the crime, and their inability to ask for independent expert witnesses to address issues they find perplexing. 14. The inadmissibility of hearsay evidence. While hearsay is considered suf- ficient to establish probable cause for arresting someone or for issuing a search warrant, it is held to be inadmissible in a criminal trial. (Ponder the paradox: Evidence regarded in one context as making it more likely than not that the defendant committed a crime cannot then be introduced into his trial. 24 ) 15. The rule against informing the jury when two or more people are on trial for the same crime and one of them has given a voluntary and admissible confession, implicating the other defendant. 16. The Doyle rule insistence that jurors cannot be informed if the defendant refused to answer questions posed by police interrogators. Many of these evidentiary practices hinder the ability of the jury to come to a correct verdict because they block the jury’s access to relevant, often powerfully relevant, evidence. Others preclude the jury from making plausible inferences (5, 10, 12, and 16). Most of these rules increase the likelihood of more false verdicts overall than would be the case under a set of less acquittal-conducive rules. Even though we lack hard data on the impact each of these rules has on the distribution of errors, it seems plausible to believe that it is significant. For instance, in the case of the exclusionary rule (2), a study conducted shortly 24 See Federal Rules of Procedure (FROP): “The finding of probable cause may be based upon hearsay evidence in whole or in part” (FROP, Rule 4, and repeated again at Rule 41(c)(1)). P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 138 flawed rules of evidence and procedure after Mapp v. Ohio extended the rule to state courts found that guilty pleas on narcotics charges in New York City dropped from more than 85 percent of the cases to fewer than 30 percent. 25 Astudy a decade later reported by the National Institute of Justice found that the prosecutor’s office dropped about one-third of all the cases against drug suspects in Los Angeles in 1981 because of fears about search and seizure problems. 26 If such figures have any validity, this rule alone represents, in all those cases where it is implicated, a significant, de facto shift upward in the SoP. To make matters worse, the distorting effects of these rules can aggregate cumulatively. While virtually no criminal trial involves all of these sixteen concessions to the defendant, it is not uncommon to see a case in which half a dozen or more of these rules are in play. If each confers an additional but indeterminable additional BoD on the defendant over and above that already incorporated into the SoP, then many defendants are receiving massively more probatory concessions than are either necessary or sanctioned by the social compact about the relative costs of false acquittals and false convictions. In later chapters, I will analyze more fully many of the previously listed rules of evidence with a view to showing that they have unmistakably acquittal-enhancing tendencies and to exploring how we might change such rules to make them more conducive to the ends of finding out the truth in a trial. But, for now, it might be useful toconsider one example of howdistributionist values showup in the current rules of evidence. Consider item6 of the preceding list. It refers to Rule 404 of the Federal Rules of Evidence, which addresses the admissibility of evidence concerning the character of the accused. While admitting certain exceptions, this rule (in clause 404(a)) generally prohibits the prosecution from introducing into its case-in-chief any evidence of the bad character of the defendant. Nothing can be said about his previous legal convictions, his known associates, his former brushes with the law, and so on. However, the same rule (clause 404(a)1) permits the defendant to present evidence of his good character, should he choose to do so. The asymmetry here is stark. The defendant’s right to present evidence of his good character clearly increases the likelihood of an acquittal. The preclusion of the prosecution from presenting evidence of the defendant’s bad character diminishes the likelihood of a conviction. Whatever else is driving the logic of this rule, it cannot be concerns with rele- vance and reliability. Aperson’s past actions, good or bad, provide some rational basis for assessing whether he is likely to have committed the crimes with which he is currently charged. If that is so, then both defense and prosecution should 25 Comment, Effect of Mapp v. Ohio on Police Search-and-Seizure Practices in Narcotics Cases, 4 Colum. J .L. & Soc. Probs. 87, at 97 (1968). 26 National Institute of Justice, The Effects of the Exclusionary Rule: A Study in California 2 (1982). P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 Evaluating Evidence and Procedure 139 be free to present to the jury whatever evidence they like about the defendant’s past. Skeptics might argue that a defendant’s past – whether good or bad – has little bearing on the likelihood that he committed this crime in particu- lar. But that argument, even if sound, will not make sense of Rule 404. If the defendant’s past history really has no bearing on his guilt or innocence in the present case, then the rule should exclude all character evidence, whether prof- fered by the prosecution or the defense. The only conceivable circumstance in which this rule would make sense is if we had empirical evidence a) that people with a good past character are generally unlikely to commit crimes and b) that people with a bad past character are neither more nor less likely to com- mit crimes. Only if both of those generalizations are true would we be justified in including good character evidence and excluding evidence of bad character. But virtually no one believes both a and b, nor is there any empirical evidence to support either of them. Maybe what is going on here reflects a worry that, if jurors learn that the defendant has a criminal past, they will automatically convict him, regardless of whether the evidence supports his guilt in the case in hand. Fortunately, we have ample evidence that juries do not rush to that sort of judgment. 27 Accordingly, we must look for some other rationale for Rule 404. The only possible motivation for it is that it is a vehicle for givingthe defendant, especially the defendant with a checkered past, a strong dose of the BoD. The rule says that, even while we grant the relevance of a bad past to the current investigation, we are not going to let the jury know anything about that past since doing so would make it more likely that the defendant will be convicted. Sometimes, such convictions would be erroneous. To foreclose that possibility, we will exclude such relevant evidence. Given this asymmetry, if an erroneous verdict ensues in a case to which character evidence would have been germane, it is clearly more likely to be a false acquittal than a false conviction. Worse, because of this skewing of the evidence in favor of the defendant, there will be more erroneous verdicts overall than would occur if both sides could freely present relevant evidence of character. This is unconscionably bad epistemic policy. One key twist to the character rule is important to mention: If the defendant chooses to present evidence of his good character, then the prosecution is freed to present evidence of his bad character in rebuttal. Giving the prosecution the right to rebut claims of an upright character might seem to nullify whatever 27 According to a study reported by Harry Kalven, Jr., and Hans Zeisel, in situations where the prosecution’s case is not rated as strong, jurors who are informed of the defendant’s criminal past still acquit in two cases out of five (38 percent). This hardly suggests that jurors – once informed that the defendant has a criminal record – rush blindly to convict simply on the strength of a criminal past. (See Harry Kalven, Jr., and Hans Zeisel, The American Jury 160, Table 52 [1966].) P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 140 flawed rules of evidence and procedure advantages this rule confers on the defendant. But a bit of analysis should dispel that hunch. Since the defendant knows that any evidence he presents of his sterling character will enable the prosecution to present whatever negative evidence it has, a rational defendant will not present evidence of his good character unless he reckons that the prosecution has nothing to counter that evidence. The upshot is that character evidence will generally be introduced only by those defendants who have an upright history. That is unobjectionable, as far as it goes. A person with a clean record should be allowed to offer it on his behalf, since it is relevant exculpatory evidence. The problem is that defendants with an unsavory past will opt to present no character evidence, thereby blocking the prosecution from informing the jury about their past. The exclusion of this clearly relevant evidence makes it more likely that the defendant with a nasty past will be acquitted than it would be if both sides were free to present evidence about the defendant’s past. Insofar as that evidence of a shady past makes it objectively more likely that the defendant is guilty, this policy encourages false acquittals much more than false convictions. It likewise produces more false verdicts in aggregate than would a policy that permitted both sides to present relevant character evidence. It was for such reasons that a famous nineteenth-century English jurist, Chief Justice Sir George Cockburn, argued that the asymmetric character rule is incoherent: The truth is, this part of our law is an anomaly. Although, logically speaking, it is quite clear that an antecedent bad character would form quite as reasonable a ground for the presumption and probability of guilt as previous good character lays the foundation of innocence, yet you cannot, on the part of the prosecution, go into evidence as to bad character. 28 The U.S. Supreme Court recognizes the same problem, characterizing the law about character evidence as “archaic, paradoxical and full of compromises and compensations by which an irrational advantage to one side is offset by a poorly reasoned counter-privilege to the other.” 29 Fewdeny that bad character evidence is relevant to, without being dispositive of, the question of the guilt of the defendant. It is excluded not because of its irrelevance but for fear that jurors may attach more significance to it than it reasonably deserves. If that is the fear, then why should the prosecution be free to present such evidence when the defendant takes the stand in his own defense? The official explanation is that if the defendant testifies, the evidence of previous convictions and bad behavior can be used not to argue his guilt but to impeach his testimony by suggesting that, as a convicted felon, his testimony is not to be taken at face value. But ponder the logic of this position for a 28 Reg. v. Rowton, 10 Cox’s Criminal Cases 25, at 29–30. 29 Michelson v. U.S., 335 U.S. 469, at 486 (1948). P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 Evaluating Evidence and Procedure 141 moment. If Jones’s having committed a crime in the past (say, bank robbery) makes it more likely that he is lying now (and that is what this exception supposes), why doesn’t Jones’s criminal record likewise make it more likely that Jones committed the crime with which he is currently charged? If a history of robbing banks is reason to believe that Jones is a perjurer, why isn’t it likewise a reason to believe that Jones passed fake twenty dollar bills? The whole policy about when character evidence can and when it cannot be used is a tissue of implausibilities. In the course of later chapters, we will see several other important rules of evidence that function similarly, each conferring an additional BoD on the accused, over and above the massive one already provided by BARD. Assessing Procedures The course and eventual outcome of a trial depend not only on what evidence a jury manages to see and hear but also on the sequence of events that con- stitute the trial itself, its prelude and its aftermath – what lawyers usually call “procedures.” Just as there are elaborate rules of evidence, so are there rules of procedure that determine the ebb and flow of events. The law recognizes no hard and fast dividing line between rules of evidence and rules of proce- dure. Indeed, at the federal level, the rules of evidence form a subset of the broader rules of criminal procedure. Still, epistemically they function differ- ently. As I will be using the terms, the “rules of evidence” determine what putative facts the jury comes to learn about the crime, while the “rules of procedure” determine when and how those facts are learned. But procedural rules go much further than simply settling the agenda for a trial. They deter- mine, for instance, how a jury is selected, what sorts of verdicts are subject to appeal, who can interrogate whom, what instructions the judge gives to jurors, what standards the judge must use for his various rulings, and sundry related matters. Obviously, such procedures can profoundly influence the outcome of a trial. If we ask ourselves what sorts of epistemic yardsticks can be used for the evaluation of procedural rules, our previous analytic crutches – relevance and reliability – are of little avail, because procedural rules do not determine, at least not directly, what evidence is admitted and what is excluded. Nor can we hope to reduce the rules of procedure, as we did with the rules of evidence, to a single overarching principle. There are, however, at least two epistemic demands that we should make of any rule of procedure: a. Rules of Procedure should be designed to optimize the likelihood that the triers of fact, typically jurors, receive their information in a way that enables them to draw valid inferences from the evidence about the guilt of the P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 142 flawed rules of evidence and procedure accused. That is, procedures should be chosen to reduce the likelihood of an invalid verdict. 30 b. Rules of procedure, taken as a whole, should create a self-correcting system of checks so that if serious errors are made along the way, they are likely to be discovered and rectified. Neither of these suggested meta-rules would seem to be very controversial. They are precisely the sorts of principles we would fall back on in designing any system of inquiry, legal or otherwise. That notwithstanding, we will see in later chapters that several procedures currently in place fail to pass these simple tests. Among them are these: 1. The inability of the prosecution to appeal an acquittal, even when a trial is riddled with serious errors. 2. The inability of the prosecution to demand pretrial “discovery” from the defense, unless the defense agrees. 3. The prohibition against jurors testifying about what went on in jury deliber- ations, even if there are suspicions of grave irregularities or misconduct. 4. The stipulation that the defendant, if he testifies at all, does so only after he has seen and heard the evidence and witnesses against him. 5. The widespread tendency of courts to refuse to define for jurors what BARD means. 6. The ability of both defense and prosecution to invoke a set of so-called peremptory challenges, which they can use to disqualify prospective jurors without cause. (As if these arbitrary acts were not bad enough, distributionist concerns are present here too; in felony cases, the defense is given more challenges than the state.) 7. The fact that the jury is under no obligation either to explain or to justify its verdict (thereby making it extremely difficult to identify errors in the conclusions that jurors have drawn). I will defer serious discussion of procedures until a later chapter, but I will offer one brief example to illustrate the kind of problem that we will be up against. Consider the role of “discovery” in a criminal trial (procedure 2 in the preceding list). Discovery itself, in the legal sense, is a relatively new notion. Traditionally, both parties went into the courtroomlargely ignorant of the sort of case that the opposing side might present. The defendant had the indictment, of course, spelling out the crimes he was charged with, and the prosecution knew how the defendant had pled. But beyond that, there was little foreknowledge. This meant that either party could be blind-sided by the other. Revelations of 30 The Supreme Court suggested as much when it insisted that, “We must be mindful that the function of legal process is to minimize the risk of erroneous decisions” (Addington v. Texas, 441 U.S. 418, at 425 [1979]). P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 Evaluating Evidence and Procedure 143 surprise witnesses, damning evidence, or ironclad alibis could often halt a trial in its tracks, or at least divert it from its expected direction. More importantly, from an epistemic point of view, this system meant that one party or the other in a trial was sometimes caught flat-footed, having to improvise frantically and ex tempore in order to cope with the unforeseen situation. With the growing recognition that this surprise element was not conducive to finding out the truth, the rules of procedure were modified to put in place pretrial machinery for preventing such grandstanding. Now, under Rule 16(a) of the Federal Rules of Procedure, the defendant has the right to demand, at a pretrial hearing, that the state provide to him copies of all statements by the defendant in the state’s possession and by intended witnesses (including expert witnesses), copies of all documents the state has that are relevant to the crime, and access to all the physical evidence that the state intends to present, among other things. Should the defendant make this request, the state can issue a similar request to the defense. So far, so good. The sharing of information before a trial begins makes it more likely that the in-trial discussion of testimony and other evidence will be as well informed as we can make it. That is clearly in keeping with the aforementioned principle a. But this reform, while overdue, was drastically ill-thought-out because it gives the defendant unilateral control over triggering the discovery process. That is, if the defendant does not ask the state to reveal what evidence it has in hand, then the state is precluded from demanding “discovery” from the defendant. (There are certain exceptions here, specifically when the defendant intends to offer a so-called affirmative defense.) The obvious question is why, if discovery is unequivocally a boon to truth seeking (as it would seem to be), should it take place only when and if the defendant desires? If discovery ensures that each side can have advance access to evidence necessary to prepare its case thoroughly, then discovery should not be made hostage to the self-serving calculations of either party. And whose interests are served by setting up an asymmetrical discovery rule of this sort? The current rule allows the defendant to spring surprises on the prosecution if he chooses not to request discovery, while the prosecution has no comparable mechanism for suppressing advance knowledge of its own would-be surprises. The predictable consequence of the discovery rule in its current form is that defendants with surprises to deliver – surprises that they fear might not stand up to the sort of advance scrutiny offered by discovery – can keep their powder dry, hoping to catch the prosecution off balance. That is no way to discover the truth about a crime. If discovery has any epistemic merit (and I think it does), then its exercise should neither be discretionary nor left in the hands of just one party. In its current form, the discovery rule gives a decided and unprincipled advantage to the defendant that is neither grounded in, nor compatible with, maximizing the likelihood that the jury will hear the best case that the two parties can present. It is one more instance of a judicial system attempting to P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 144 flawed rules of evidence and procedure jigger with the distribution of errors when it should be concerning itself with their reduction. In Chapter 8, we will analyze some other epistemically dubious procedures. Conclusion Before closing this discussion of some of the meta-rules of the theory of evi- dence and procedure, I should probably say something to allay the worry that, in my rejection of rampant distributionism with respect to evidence and pro- cedures, my arguments betray an indifference, perhaps even hostility, to the defendant in a criminal trial. I flatly deny the charge. To begin with, it cannot be hostile to the innocent defendant to propose that the rules governing a trial should be those most likely to lead to a true verdict. Above all else, the innocent defendant seeks a true verdict. If my stance is less than indulgent to anyone, it is to guilty defendants, for which I make no profuse apologies. But the more important response to the charge of downplaying the impor- tance of tilting the system toward the defendant is this: Nothing I have said precludes skewing the SoP as far as society likes in the defendant’s favor. My advice has been to build explicitly into that standard every bit of acquittal- friendly bias that society deems appropriate in view of its appraisal of the proportion of innocent defendants it is prepared to see convicted and its assess- ment of the relative costs of mistaken convictions and mistaken acquittals. (For my part as an epistemologist, I have proposed, and will propose, no upper bound for it, as that falls well outside my brief. As a citizen, my disposition would be to put it in the 0.9 range, but I could live contentedly with values mildly lower or higher than that. I have to confess that I would not particularly want to live in a society that, like Maimonides, believed that false convictions were a thousand times worse than false acquittals or that the justice system could convict no more than one in one thousand innocent defendants. 31 Under such a regime, there would be precious few convictions of the guilty and virtually no deterrence of crime.) Having settled on the appropriate level of bias, however, we should let the rest of the system function as an epistemically respectable engine, that is, as a viable, error-reducing, distributionally neutral tool for investigation that tries unstintingly to find the truth. Only in that way can we ensure that a trial honors the prior social compact about the relative costs of mistakes and successes. Only in that way can we ensure that the jury’s estimate of the apparent guilt of the accused is as reliable as we can make it. 31 “It is better . . . to acquit a thousand guilty persons than to put a single innocent man to death” (2 Moses Maimonides, The Commandments [Charles B. Chavel, trans., 1967], at 270). P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 Evaluating Evidence and Procedure 145 WhenI make this argument tofriends andcolleagues, theysometimes remind me that, in the standard criminal trial, an isolated defendant – often indigent – faces the awesome apparatus of the police and the prosecutor’s office. He is typically represented by a public defender who, in some jurisdictions, has fewer thansevenhours inall todevote toeachcase. 32 Theysuggest that it is necessary– and I can’t count how many times I have heard the phrase repeated in this context – “to level the playing field.” Adopting acquittal-enhancing rules of evidence or procedure is seen as a way of restoring parity to a situation where a hobbled David is up against such an imposing Goliath. Aficionados of this slogan would have us believe that, unless we shave the evidence rules and procedures in the defendant’s favor, he starts the “game” at an enormous handicap. This form of special pleading should be recognized for the sham that it is. The playing field – to adopt this misleading metaphor for a moment – already begins tilted heavily toward the defendant by virtue of the SoP. By announcing our insistence on (for example) ten or twenty true acquittals for every false conviction, and by incorporating that decision into the SoP, we have already given an enormous advantage to every criminal defendant, whether guilty or innocent. Common sense makes clear that the actual handicapping in a criminal trial is precisely the opposite of what the distributionists would have us believe. From the outset of a criminal proceeding, the average guilty defendant is assured that, though guilty, he is unlikely to be convicted of a crime. The innocent defendant, for his part, knows that it is extraordinarily unlikely, given the standard of proof, that he will be falsely convicted. If we allowthe rules of evidence and procedure to skew the trial further in acquittal-enhancing ways, then the risk we run is not that of leveling the playing field but of perching it precariously on its edge. It is true, of course, that some prosecutors have vast financial and investiga- tive resources at their disposal, which the (indigent) defendant does not. But does that tilt the playing field so far that the defendant is unfairly disadvan- taged? To determine whether this is the case, answer this question: If you were an innocent defendant in a criminal trial and you had to choose between one of these options, would you pick a) a PoE standard and a prosecutor’s office on hard financial times or b) a very high SoP and a resource-rich prosecutor’s office? If, like me, you would – without the slightest hesitation – prefer b over a, then that settles the question. Above all, this chapter has insisted that, if our principal interest is in finding out the truth, then we must resist the temptation to replace relevance- and reliability-based rules of evidence by defendant-friendly rules of exclusion. The latter inevitably blur the boundaries between the truly guilty and the truly 32 Office of the Missouri State Public Defender, Fiscal Year 2004 Annual Report, October 1, 2004. P1: JZZ 0521861667c05 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:38 146 flawed rules of evidence and procedure innocent, making it likely that the total number of mistaken verdicts will be vastly greater than they would be if the excluded evidence were admitted. We want evidentiary practices that will generally make the truly innocent look innocent and the truly guilty appear guilty. What we don’t want are practices that make a few more of the truly innocent look innocent while making many more of the truly guilty look innocent as well. P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 6 Silent Defendants, Silent Witnesses, and Lobotomized Jurors Today as in the past there are students of our penal system who look upon immunity [that is, the defendant’s privilege against self-incrimination] as a mis- chief rather than a benefit, and who would limit its scope, or destroy it altogether. No doubt there would remain the need to give protection against torture, physical or mental. . . . Justice, however, would not perish if the accused were subject to a duty to respond to orderly inquiry. – Justice Benjamin Cardozo (1937) 1 If all the criminals of every class had assembled, and framed a system after their own wishes, is not this rule [of silence] the very first which they would have established for their security? – Jeremy Bentham 2 The Fifth Amendment’s privilege against self-incrimination . . . is not designed to enhance the reliability of a fact-finding determination but stands in the Constitu- tion for entirely independent reasons. – U.S. Supreme Court (1986) 3 Silence and Truth Seeking In the following four chapters, I am going to examine a few selected rules of evidence and procedure to see whether they are truth promoting or truth thwarting. Those I have chosen to look at are particularly problematic. When I find rules wanting from this point of view, my thought experiment will require that I explore alternatives. I will begin by discussing a broad family of evidence rules that address the questions of who is obliged to give testimony and what inferences the jury should be entitled to make when such testimony is not forthcoming. Depending on whose testimony is at stake, these fall into three 1 Palko v. Connecticut, 302 U.S. 319, at 320 (U.S., 1937). 2 J. Bentham, A Treatise on Judicial Evidence 241 (1825). 3 Allen v. Illinois, 478 U.S. 364, at 375 (1986). 147 P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 148 flawed rules of evidence and procedure classes: the accused, the witness with something to hide, and the privileged witness. The Accused When a suspect is arrested and charged with a crime, she is – from the police point of view – a potentially potent source of information. If she committed the offense, she may confess. Whether she committed it or not, she may well have information about what took place and who was involved. From the moment of her arrest forward, however, the law shields the accused from having to reveal what she knows. She need say nothing to the investigating detectives or the arresting officers. When she eventually comes to trial, if she does, she is once again shielded from having to say anything in her defense. More than that, nothing can be made of her silence. Follwing what is known as the Doyle rule, the jury cannot be informed that she refused to answer questions from police interrogators. 4 The prosecutor cannot refer to the silence of the accused during the trial without risking a mistrial. The judge will instruct the jury that the defendant’s failure to make statements during pretrial interviews or during the trial itself cannot under any circumstances be taken as evidence, however partial, of her complicity in the crime. Here is a typical model jury instruction used in federal courts: You may not attach any significance to the fact that the defendant did not testify. No adverse inference against him may be drawn by you because he did not take the witness stand. You may not consider this against the defendant in any way in your deliberations in the jury room. . . . In other words, you should simply put the whole matter of the defendant not testifying out of your mind. There may be any number of reasons which are perfectly innocent for him not testifying. He may be nervous, he may stammer when he speaks, he may have an unattractive voice. There are hundreds of possible reasons. It is simply a matter you should not consider and I am instructing you, as a matter of law, not to. 5 In some jurisdictions, jurors are further admonished that it would be a violation of their oath to attach any meaning to the defendant’s silence. 6 In a word, the courts deem defendant silence irrelevant. 4 Ever since 1976, if not before, courts have held that federal judges may not admit evidence that a defendant under arrest refused to respond to questions (Doyle v. Ohio, 426 U.S. 610 [1976]). See also Miranda v. Arizona, 384 U.S. 436, at 468 (1966). 5 Model Federal Jury Instructions–Criminal P 5.07. 6 The First Circuit warns jurors: “[The defendant] has a constitutional right not to testify and no inference of guilt, or of anything else, may be drawn from the fact that [the defendant] did not testify. For any of you to draw such an inference would be wrong; indeed, it would be a violation of your oath as a juror” (ibid., P 3.03). P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 Silent Defendants and Silent Witnesses 149 The Witness with Something to Hide Something similar happens when a subpoenaed witness asserts his right not to incriminate himself and thereby avoids testifying. U.S. v. Johnson established that a judge may refuse to allow a witness to take the stand when it appears that the witness intends to claimthe privilege against self-incrimination. 7 If this happens, the prosecution and defense counsel will know that, if they mention that a prospective witness “claimed the Fifth,” the trial may be aborted. If the recalcitrant witness does not assert this right until he is actually on the stand, the jury will usually be admonished to attach no significance to what they have just witnessed. The Privileged Witness The third time when silence triggers special treatment occurs when a subpoe- naed witness declines to testify, not on grounds of self-incrimination, but by claiming exemption from the obligation to testify by virtue of a special rela- tionship to the accused (for instance, his spouse, lawyer, psychiatrist, social worker, or priest). If these assertions of the privilege occur outside the hearing of the jury, neither the defense nor the prosecution may mention them. If within earshot, the assertion of the privilege will provoke a stern warning to jurors to ignore what they have just seen and heard. Each of these silences has two troubling aspects from an epistemic point of view: first, that the law allows those with relevant information to avoid divulging it; and second, that jurors are either kept in the dark about, or urged to pretend they do not know, the fact that relevant evidence was suppressed. Our general meta-rule about relevance and reliability in Chapter 4 warned us to be wary whenever it is deemed impermissible to draw an inference from relevant evidence. Decisions to do so, I’ve argued, are appropriate only if the sequestered information is clearly irrelevant or so confusing that we have powerful reason to believe that the jury will probably draw the wrong conclusions from it. In thinking our way through cases of this sort, the principal issues will be the following: a. Does it further the interests of truth seeking to allow defendants, guilty witnesses, and privileged witnesses to refuse to tell what they know about a crime? b. When these silences do occur, does it promote the search for the truth to instruct jurors to attach no meaning to such events? In each case, finding the answer will require us to decide whether it is relevant to a jury’s deliberations to know that someone with knowledge of the crime 7 U.S. v. Johnson, 488 F.2d 1206, at 1211 (1st Cir. 1973). P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 150 flawed rules of evidence and procedure refused to reveal what she knows. In a word, is silence significant? I will deal with the three cases separately. Silent Defendants: The Problem At the risk of belaboring the obvious, it remains important to say that police investigative workis clearlyhamperedif anyone withinformationabout a crime, let alone the suspect, refuses to tell what she knows. Likewise, the silence of the defendant thwarts the jury’s fact-finding duty, except perhaps in those rare cases where the defendant knows literally nothing whatever about the crime, its prelude, or its aftermath. Consider a simple case: The police find stolen goods in defendant Jones’s warehouse. Several witnesses testify that these were stolen fromthem. Innocent or guilty, Jones is probably in a good position to shed important light on the question of how he came to have those purloined items. But he is not obliged to say anything. Clearly, this policy (what I will call “Silent Defendant ”) works to increase the likelihood that the state will have insufficient evidence to convince a jury of Jones’s guilt BARD. The key question for us, however, is not whether Silent Defendant works to increase the rate of acquittals (since acquittals are not a bad thing per se) but rather what impact this policy has on the frequency of false acquittals and false convictions. Compared to a hypothetical system in which defendants helped the police with their inquiries, we can expect that Silent Defendant will lead to far more false acquittals. If genuinely guilty parties told what they know about a crime, it is hard to imagine that many would be acquitted. Likewise, there would be fewer false convictions since criminal charges would be less likely to be brought against the innocent if the guilty always confessed. But this is not a very useful comparison since truly guilty defendants who have resisted confessing to or commenting on their crimes prior to trial are unlikely to tell the truth if they do testify, even under oath. A more useful comparison is with a system in which defendants are obliged to testify but where there is no guarantee of their veracity. Obviously, requiring a defendant to testify is not the same thing as making himtell the truth. Still, every attorney knows that lies can be nearly as revealing as the truth. A guilty defen- dant who testifies on his behalf can be subjected to probing cross-examination. Inconsistencies or implausibilities in his account of events can be drawn to the jury’s attention. Jurors themselves can drawinferences fromthe behavior of the defendant on the stand that may be epistemically more telling than what they can infer from his mute behavior sitting next to counsel. Accordingly, if, absent the doctrine of Silent Defendant, defendants could be made to take the stand, we can expect that false acquittals would probably decrease (even if most guilty defendants lied while on the stand). Of course, even absent a constitutional right to silence, a defendant may choose to remain mute when called to the stand. Short of using thumbscrews P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 Silent Defendants and Silent Witnesses 151 or the rack or putting her in jail for contempt, nothing can force a reluctant defendant to give testimony. Whether the system gives defendants the right to remain silent or not, it is within defendant’s power to do so. In that sense, the Silent Defendant is potentially an element of any set of criminal procedures, whether we like it or not. Even so, the systemcould include built-in incentives to encourage defendant testimony. Indeed, through most of the history of the common law, there was just such an incentive: Jurors were at liberty to draw whatever conclusions they thought appropriate from a defendant’s silence. Defendants who declined to testifyranthe riskthat sucha choice might leadthe jurors toadverse conclusions. Byholdingthat the jurywouldbe violatinga constitutional right of the defendant if it attributed any significance to his silence, the famous Griffin case put an end to that practice in 1965. Reversing Griffin would restore one powerful motive for the defendant to explain his side of the story. The problemis not solely lack of incentive; further exacerbating the problem is a series of strong disincentives to testify, created, like Griffin, by the courts. Consider a few of the things that can happen to a defendant who chooses to testify. The moment he takes the stand, the prosecution can introduce into its cross-examination a long list of incriminatory items that were previously excluded from its case-in-chief. For instance, if the defendant has a criminal history, the prosecution can now use it to challenge the defendant’s testimony. If the defendant gave the police a confession that he subsequently retracted and the judge excluded, that too is now fair game. If the judge excluded illegally seized but relevant evidence from the trial, it can suddenly be introduced to impeach the defendant’s testimony. In short, current rules create a situation in which the typical defendant, especially the guilty one, has overwhelming reasons not to testify. This does not augur well for finding out the truth. If there were no Silent Defendant policy, jurors could draw appropriate inferences from the refusal of the defendant to address the evidence presented against him. That policy would tend to reduce the number of false acquit- tals. So, there seems little doubt, as Bentham argued in this chapter’s epi- graph, that the Silent Defendant policy serves the interests of the truly guilty defendant. Does it likewise promote the interests of the innocent? It is possible to imagine circumstances in which a truly innocent defendant might believe that offering evidence to the police or testimony to the court would make her erro- neous conviction more likely. For instance, perhaps she has information about the crime that appears highly inculpatory, even though someone else committed it. Perhaps she fears that her demeanor on the witness stand would unreason- ably prejudice the jury against her. Or perhaps she worries that her truthful testimony would implicate someone precious to her, whom she wants to shield from suspicion. Let us grant straightaway that Silent Defendant – understood now as both the right to remain silent and the “right” not to have that silence factored into P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 152 flawed rules of evidence and procedure the jury’s calculations – may protect some small subset of innocent defen- dants from being convicted. Is that sufficient to establish its legitimacy? As I argued in the previous chapter, answering such questions requires us to esti- mate and then compare both the false convictions averted and the false acquittals generated, a comparison that Silent Defendant’s advocates conveniently never make. Should a defendant’s refusal to tell what she knows (either to the police or to the jury) be an element in a jury’s fact-finding deliberations? That is an easy call. In ordinary life, when there is plausible evidence of someone’s wrongdo- ing, and when, confronted with that evidence, he refuses to say anything, it is natural to take that as an indication (not a proof, mind you, but still a relevant indicator) that he has something to hide. In sum, the defendant’s silence, in the face of plausible evidence of his guilt, is relevant to the question of that guilt. So it is clear that Silent Defendant, by relieving the defendant of the obligation to say anything and then precluding the jury from attaching any sig- nificance to his silence, is far more likely to shield the guilty from conviction than to protect the innocent. There simply is no epistemic justification for Silent Defendant. That notwithstanding, the prospect of moving to a system in which defen- dants are required to give testimony is slim. A large body of theory – or better, of ideology – holds that Silent Defendant is the very bulwark, the palla- dium, of our freedoms. Here is a typical bit of hype from Justice Arthur J. Goldberg (writing for a unanimous Supreme Court) in a classic case from 1964: The privilege against self-incrimination “registers an important advance in the devel- opment of our liberty – ‘one of the great landmarks in man’s struggle to make himself civilized’.” It reflects many of our fundamental values and most noble aspirations: our unwillingness tosubject those suspectedof crime tothe cruel trilemma of self-accusation, perjury or contempt; our preference for an accusatorial rather than an inquisitorial sys- tem of criminal justice; our fear that self-incriminating statements will be elicited by inhumane treatment and abuses; our sense of fair play which dictates “a fair state– individual balance by requiring the government to leave the individual alone until good cause is shown for disturbing him and by requiring the government in its contest with the individual to shoulder the entire load,”; our respect for the inviolability of the human personality and of the right of each individual “to a private enclave where he may lead a private life;” our distrust of self-deprecatory statements; and our realization that the privilege, while sometimes “a shelter to the guilty,” is often “a protection to the innocent.” 8 8 Murphy v. Waterfront, 378 U.S. 52 (1964). (Footnotes omitted.) Judge Henry Friendly pointedly remarked on “the extent to which eloquent phrases have been accepted as a substitute for thorough thought” in discussions of the silence issue (Henry Friendly, The Fifth Amendment Tomorrow, 37 U. Cin. L. Rev. 671, at 677 [1968]). Goldberg’s purple prose is typical of its genre. P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 Silent Defendants and Silent Witnesses 153 Let us suppress the suspicion that when clever justices like Goldberg resort to this sort of rhetorical grandstanding, they are signaling that serious arguments for their position are in short supply. Instead, let us take this passage at face value and see whether it makes any sense. Clearly, only the last of Goldberg’s claims for Silent Defendant has anything to do with getting at the truth. So let’s begin there and then turn to the nonepistemic arguments. Goldberg’s last clause rhetorically plays off the adverbs “sometimes” and “often,” misleadingly inviting us to believe that Silent Defendant is more likely to protect the innocent from conviction than it is to secure acquittal for the guilty. Of course, as I have conceded, we can conceive of cases where exercising one’s “right” to silence might protect the innocent from conviction. But such cases are swamped by the number of situations in which we can imagine Silent Defendant leading to an acquittal for the guilty. Had Goldberg not been involved in special pleading, he might less misleadingly have said that Silent Defendant “often” offers shelter to the guilty and “sometimes” to the innocent. Such candor would, of course, have undermined the conclusion he wants us to draw. But, it might be said, even if Silent Defendant protects only the rare innocent defendant from conviction, those victories are precious and should be encour- aged. So as not to be taken in by the general logic of Goldberg’s argument here, remember that many patently silly policies would “offer a protection to the innocent.” For instance, policies that said that all female defendants should be automatically acquitted or that all left-handed black defendants get to sup- press any evidence against them would “often” protect the innocent. Have we any reason to think that being female, a left-handed black person, or a Silent Defendant is more likely to indicate innocence than guilt? Unless we do, there is no justification for the privilege. Let us consider, briefly, some of Goldberg’s other points. He refers to a “cruel trilemma”: If defendants were obliged to testify under oath, the nasty choices facing them, he says, would be a) admitting their guilt, b) lying under oath and by that committing perjury, or c) remaining silent and facing charges of contempt of court or an adverse inference fromtheir silence by the jury. Note that the innocent defendant does not face this trilemma. She can tell what she knows without either admitting her guilt or perjuring herself. The individual facing Goldberg’s “cruel” set of choices is the guilty defendant. And are these choices really so cruel? Granted, no reasonable person would welcome choices a through c. But if the defendant is guilty, it is she who brought the trilemma upon herself; no one obliged her to commit the crime. Having committed it, and without a right to silence, she might indeed have to choose between confessing, lying, or obdurately remaining silent and suffering whatever hostile inferences the jury might make about her testimonial intransigence. Nothing about this conundrum is any more “cruel” than the situation of the philandering husband who, when challenged by his wife with evidence of his infidelity, must decide whether to confess, to lie about his affairs, or to tough P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 154 flawed rules of evidence and procedure it out by ignoring the question. What Goldberg seems to be arguing is that the guilty defendant should be shielded from some of the consequences of his having broken the law. This is not only sentimentality but it is directed at a source who has forfeited any strong claim to such sentiments. Another part of Goldberg’s rhapsody to which I want to drawyour attention – because it represents a point of viewcommon in the legal profession, especially among members of the defense bar – is his conception of a trial as a process that must satisfy our sense of “fair play” and in which a “balance” needs to be struck between the state and the accused. Such a view of a criminal trial invites us to think of it as a game of sport. As in a game, neither side must be given any advantage. Akhil Amar and Renee Lettow have aptly described the situation in these terms: “At times, the fair balance idea collapses into . . . the idea that we should boost the odds for criminals just to keep the game interesting.” 9 In effect, Goldberg proposes “handicapping” the prosecution by insisting that it can neither oblige the defendant to testify nor use her silence to argue her guilt to the jury. But a trial is neither a game nor (as Goldberg puts it) a “contest.” It is a serious and weighty effort to find out the truth about a crime. It demeans the gravity of the event to demand that the location of the goal posts be fine-tuned so that each side has an equal chance of “winning.” Only if precisely half of those who came to trial were innocent would this be appropriate. We have no grounds to believe that and plenty, rehearsed in earlier chapters, to doubt it. What is important is not that every defendant has a fifty-fifty chance of winning, but that the system is set up to make it very unlikely that the innocent are convicted and at least mildly unlikely that the guilty are acquitted. 10 For sure, the deck should not be so stacked that the innocent defender is likely to be convicted; BARD already ensures that the opposite bias obtains. As Richard Uviller has rightly observed, “Handicapping has no place in the determination of criminality by trial.” 11 The ultimate riposte to those in the defense bar who insist on a “fair fight” is to remind themthat, if they really want fair play between prosecution and defense, then BARD should be replaced by a standard of proof that awards victory to whoever has the more plausible case. Goldberg’s other point, and it too is commonly voiced, is that the right to silence accordedbythe FifthAmendment is part of a larger regardfor the privacy of the person, whether guilty or innocent. Since this argument has nothing to do with truth seeking, I will postpone a detailed analysis of it until Chapter 9. 9 Akhil Amar and Renee Lettow, Fifth Amendment First Principles, 93 Mich. L. Rev. 857, at 892 (1995). 10 “Equality between contestants makes for good sports, but in a criminal investigation we should be seeking truth rather than entertainment” (J. Grano, Selling the Idea to Tell the Truth, 84 Mich. L. Rev. 662, at 677 [1986]). 11 H. Richard Uviller, Evidence from the Mind of the Criminal Suspect, 87 Colum. L. Rev. 1137, at 1174 (1987). P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 Silent Defendants and Silent Witnesses 155 The Shaky Epistemology of Lobotomized Jurors In an ideal world, then, defendants wouldbe obligedtotell what theyknewabout the crimes of which they were accused. But even if the prevailing (mis)construal of the Fifth Amendment is so deeply entrenched that it cannot be overturned by mere worries about finding the truth, current policies about defendant silence could be tweaked to make them much more likely to reduce aggregate verdict error. The key, as we have seen, is that Silent Defendant has two component parts. One says that the defendant needn’t testify. The second precludes the prosecutor and judge from remarking on the defendant’s refusal to mount a defense or on his failure to cooperate with the police in their inquiries. It likewise insists that the judge – on request from defense counsel or his own initiative – instruct the jurors that they are to draw no adverse inferences from the defendant’s silence. They must, as the court says, pretend it never happened. For obvious reasons, I will call this part of Silent Defendant doctrine, which blocks juries fromdrawing plausible inferences fromdefendant’s silence, “Lobotomized Juror.” By universal admission, it has no constitutional mandate. There is nothing in the Fifth Amendment or anywhere else in the Constitution concerning this issue. That is scarcely surprising since it was routine, from the early days of the Republic until the middle of the twentieth century, for jurors to be left free to drawwhatever inferences seemed to them appropriate from the silence of a defendant. As many studies of criminal practice in the first century of the Republic make clear, no one then understood the Fifth Amendment to sanction without penalty a defendant’s refusal to respond to questions from police or prosecutors. On the contrary, it was a part of the common law rules of evidence, and a doctrine enshrined in many state constitutions, that if the defendant said nothing in his trial, the jury could choose to take that silence as partial evidence of guilt. Here, for instance, is howthis principle was formulated in the California State constitution: [Defendant’s] failure to explain or to deny by his testimony any evidence or facts in the case against him may be commented upon by the court and by counsel, and may be considered by the court or the jury. 12 In several other states of the Union, constitutions or legislatures made similar provisions. Justice Felix Frankfurter clearly laid out why courts and legislators insisted that the jury be allowed to drawinferences fromthe defendant’s silence: Sensible and just-minded men, in important affairs of life, deem it significant that a man remains silent when confronted with serious and responsible evidence against himself which it is within his power to contradict. The notion that to allow jurors to do that which sensible and right-minded men do every day violates the “immutable principles 12 Cal. Const. art. I, § 13. P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 156 flawed rules of evidence and procedure of justice” as conceived by a civilized society is to trivialize the importance of “due process.” Nor does it. 13 As recently as 1947, the U.S. Supreme Court itself agreed with California that a defendant’s failure to testify is not without significance. Here is what it said in Adamson v. California: It seems quite natural that when a defendant has opportunity to deny or explain facts and determines not to do so, the prosecution should bring out the strength of the evidence by commenting upon defendant’s failure to explain or deny it. 14 Only in the last half-century, through a series of rulings by the Court begin- ning in the 1960s, was this practice deemed unconstitutional. What had seemed “quite natural” to the Supreme Court in the 1940s became, in less than two decades, a practice that violated a defendant’s constitutional rights. By 1965, the Supreme Court – in a flagrant case of historical myth making – was argu- ing that permitting the prosecution to comment on the silence of the defendant would be reverting to that bogeyman of all Anglo-Saxon systems of justice, the Inquisition: “Comment on the [defendant’s] refusal to testify,” wrote the Supreme Court, “is a remnant of the ‘inquisitorial system of criminal jus- tice’.” 15 How, one wonders, can such comment be a remnant of that foreign system when Anglo-American common law, and the Supreme Court itself, had routinely endorsed such comment until the 1960s? Not content with denying judges or prosecutors the traditional right to com- ment on the defendant’s silence, the Court went one step further. In 1981, it insisted that, if so requested by defense counsel, the judge must instruct jurors that they are to attach no significance to the defendant’s silence. Here is how the Court saw things then: The Griffin case stands for the proposition that a defendant must pay no court-imposed price for the exercise of his constitutional privilege not to testify. The penalty was exacted in Griffin by adverse comment on the defendant’s silence; the penalty may be just as severe when there is no adverse comment, but when the jury is left to roam at large with only its untutored instincts to guide it, to draw from the defendant’s silence broad inferences of guilt. Even without adverse comment, the members of a jury, unless instructed otherwise, may well draw adverse inferences from a defendant’s silence. 16 By 1981, all the key elements of Lobotomized Juror were in place: Neither judge nor prosecutor could mention that the defendant hadn’t testified and (in case the jury had noted that fact) jurors were instructed to drawno conclusions whatever 13 Adamson v. California, 332 U.S. 46, at 60 (1947). 14 Ibid., at 56. 15 Griffin v. California, 380 U.S. 609, at 614 (1965). 16 Carter v. Kentucky, 450 U.S. 288, at 301 (1981). P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 Silent Defendants and Silent Witnesses 157 from his nontestimony. Making the story complete, the Supreme Court ruled in Miranda and in Doyle that jurors could not be informed if a defendant had refused to answer questions from police in the pretrial proceedings. 17 Lobotomized Juror is thus a relatively new policy, fashioned out of whole cloth by the Supreme Court in the mid-twentieth century. Like the Miranda rights invented by the same court in the same era, Lobotomized Jury lacks significant historical and constitutional precedent. The question for us is whether it is conceptually coherent. As the previously quoted passage makes clear, Lobotomized Juror rests on two quite distinct arguments. The first argument is that the defendant’s right to silence would not really be a right if defendants were penalized for exercising it by allowing the jury to draw any inferences from it. In the view of the Court, permitting jurors to attach any significance to silence “cuts down on the privilege [of silence] by making its assertion costly.” 18 As one scholar has summarized the current attitude of the Court, “If the right to silence is to have any meaning, it must lie in restrictions on the adverse consequences flowing fromits assertion.” 19 Whether having a right entitles one to penalty-free exercise of that right is an issue we will take up in Chapter 9, since it falls well outside epistemology per se. The second and much more pertinent argument is the claim that the court must stifle the drawing of “adverse inferences” from a defendant’s failure to testify since such inferences could be misleading or invalid. There are two ways in which this claim can be understood. One of them holds that the decision of the defendant to say nothing is literally irrelevant (in the strict sense discussed in Chapter 4) to his guilt or innocence. The other concedes the relevance of the defendant’s silence but supposes that jurors – if not prevented from drawing any inferences – will give that silence more weight than it can reasonably bear. Silence, it is feared, might be regarded by juries as so dispositive that they will convict simply because the defendant was silent – unless the judge explicitly tells them that they can’t do that. An epistemic appraisal of the merits of Lobotomized Juror must try to resolve these two questions: a) First, is it plausible that defendant’s silence objectively has no bearing on the question as to whether he is guilty or innocent? We have already acknowl- edged that there can be all sorts of circumstances in which an innocent defendant might choose not to take the stand. This should already be sufficient to persuade us (and jurors) that any simplistic inference from “the defendant is silent” to “the defendant is guilty” would be fallacious. But current court practice is not 17 The Court wrote: “It is impermissible to penalize an individual for exercising his Fifth Amendment privilege when he is under police custodial interrogation. The prosecution may not, therefore, use at trial the fact that he stood mute or claimed his privilege in the face of accusation” (Miranda v. Arizona, 384 U.S. 436, at 468 [1966]). 18 From Murphy v. Waterfront Comm’n, 378 U.S. 52 (1964). 19 Gordon van Kessel, The Suspect as a Source of Testimonial Evidence, 38 Hastings L. J. 1, at 11 (1986). P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 158 flawed rules of evidence and procedure one that advises juries to be cautious about linking silence and guilt. It tells them rather that under no circumstances are they to take the defendant’s silence as making his guilt one whit more likely. Worse, courts fail to give jurors a plausi- ble story about why an adverse inference fromsilence would be an unreasonable inference for them to make. 20 Indeed, ignoring the defendant’s silence seems to be a bad mistake, for it is easy to imagine circumstances in which a defendant’s silence should have a powerful bearing on the jury’s deliberations. Consider this hypothetical case: John is accused of stealing Joan’s car. Several seemingly reliable witnesses place John in Joan’s neighborhood when the theft occurred. The police located the stolen car in a locked garage belonging to John. The steering wheel bore John’s fingerprints. It is still conceivable that John is innocent. The question paramount in the jurors’ minds, having heard the prosecution’s case, will be whether John can offer a plausible story that will render these inculpatory facts innocuous. If he is innocent, the jurors probably reason, he will be able to explain these facts away. If John refuses to testify, his refusal does not make the state’s case more plausible than it already was. What his failure to testify may do is to leave the state’s case as the only plausible one on the table. The importance that a rational juror would attach to John’s silence arises from the apparent lack of an alternative story to be told. By default, his silence may seal his conviction, provided that the prosecution’s case was reasonably powerful in the first place. Legal experts sympathetic to Lobotomized Juror will remind us that the defendant is under no legal obligation to give the jury his version of events. That is true. But if he fails to offer any exculpatory story, and if the prosecution’s case is plausible and powerful, then it falls to the jurors to think up one. Whatever story they devise will inevitably have less credibility than if John himself had offered it in testimony. Jurists may not like the fact, but it is a fact that rational jurors will find a theory of the crime that makes John innocent more plausible if John himself offers testimony that bears out the theory in question. John needn’t offer such testimony (that is what the right to silence means), but if he does not, we can hardly blame the jurors if they conclude that John has something to hide and that any theory that exonerates him seems less plausible than it would have done had his own testimony supported it. There is another way to formulate the point. Basically, the Supreme Court’s epistemic argument for Lobotomized Juror is that innocent defendants will 20 As Richard Posner has observed, “Judges who want jurors to take seriously the princi- ple that guilt should not be inferred from a refusal to waive the privilege against self- incrimination will have to come up with a credible explanation for why an innocent person might fear the consequences of testifying” (An Economic Approach to the Law of Evi- dence, 51 Stan. L. Rev. 1477, at 1534–5 [1999]). P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 Silent Defendants and Silent Witnesses 159 sometimes elect silence. What that establishes, and all that it establishes, is that it silence is not an infallible indicator of guilt. It does not show that silence has no relevance to guilt, and that is what the law must establish if it is to say with any epistemic authority that jurors should ignore silence. To lobotomize the jury with respect to defendant silence, on the grounds that juror inferences of guilt from this datum might sometimes be mistaken, is exactly like insisting that prosecutors should not be able to present evidence that a defendant had a motive or an opportunity because sometimes people who are innocent have both motive and opportunity. It is like saying that the prosecution cannot produce eyewitness testimony against the defendant because witnesses sometimes fin- ger innocent defendants. To establish the irrelevance of silence, superior courts would have to point to evidence that innocent defendants are as likely as guilty ones to say nothing at their trials. That, after all, is what irrelevance means. But no one has ever produced empirical studies that point to this counterintuitive result. Moreover, it is doubtful whether even the courts would seriously main- tain that silence was irrelevant to a decision about guilt or innocence. Indeed, in 1976, the Supreme Court conceded that the Griffin rule “has little to do with a fair trial and derogates rather than improves the chances for accurate decisions.” 21 b) Even grantingthat one cansometimes drawappropriate adverse inferences from a defendant’s failure to testify, do juries tend – if permitted to assess the meaning of a defendant’s silence – to attach more significance to that silence than they should? The advocates of Lobotomized Juror insinuate that jurors are ready to pounce on every instance of Silent Defendant and to use it to justify a conviction. Here is how one legal scholar, Maria Berger, frames the argument: For many people, no innocent man would stand silent in the face of his accusers. The prejudicial nature of this evidence, combined with its questionable reliability, supports its exclusion to support the truth-seeking function of criminal procedure. 22 This argument is a shambles. Jurors already know that the defendant has said nothing. If it were true that jurors were scandalized by that fact, then no instruc- tions from a judge would prevent them from using that fact in their private calculations of guilt or innocence. But it is simply false that juries will always or generally convict any defen- dant who keeps his mouth shut. Remember that until the 1960s, criminal courts at the state level explicitly allowed comment on the defendant’s silence and allowed the jury to make of that silence what they would. Literally tens of thousands of defendants who gave no testimony were acquitted. Berger offers 21 Baxter v. Palmigiano, 425 U.S. 308, at 319 (U.S., 1976). 22 Maria Berger, Defining the Scope of the Privilege against Self-Incrimination, 1999 U. Ill. L. Rev. 1015, at 1040 (1999). P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 160 flawed rules of evidence and procedure no evidence for her claim that the typical trial jury slides effortlessly from the defendant’s silence to the defendant’s guilt, independent of the evidence available. Through a couple of centuries of American law, juries showed them- selves perfectly capable of seeing that guilt was often not proven even when the defendant said nothing in his own behalf. Berger’s argument is even more seriously misleading when she touts Lobotomized Juror as a principle that fur- thers the “truth-seeking functions” of the law. Quite the reverse: It thwarts them. Hers is Justice Goldberg’s flawed epistemic argument in a different guise. It is still possible that, although jurors do not inevitably rush to a verdict of guilty when confronted by a defendant’s silence, they attribute a greater weight to that silence than it rationally deserves. Lacking telling empirical data on this matter, proponents and opponents of Lobotomized Juror will continue bickering about it. But note the point we have reached. Indisputably, in many criminal cases, the defendant’s refusal to testify increases the likelihood that a rational jury would convict him if Lobotomized Juror did not block all such inferences. If Lobotomized Juror were eliminated, there is a possibility (whose magnitude we don’t know) that juries would convict when, rationally, the evidence is insufficient to warrant it, although suitably crafted judge’s instructions should warn themoff that path. Still, we are comparing a policy – the current one – that is bound to produce many erroneous verdicts with a policy (the pre-1960s one) that may or may not do so. The errors of the current policy will generally be false acquittals; the errors of the previous system were apt to be false convictions. As argued in earlier chapters, our choice between the two should be guided by a concern to reduce overall error rates rather than by a desire to distribute errors further in an acquittal-enhancing way. On those grounds, Lobotomized Juror should be eliminated, the system returned to its pre-Griffin state, and jurors warned not to give silence more significance that it merits in the case under consideration. Such a policy is already the norm in English courts. Here is what a judge there says to the jury when a defendant has exercised his right to silence: The defendant has not given evidence. That is his right. But, as he has been told, the law is that you may draw such inferences as appear proper from his failure to do so. Failure to give evidence on its own cannot prove guilt but depending on the circumstances, you may hold his failure against him when deciding whether he is guilty. . . . [Specifically,] if the evidence he relies on presents no adequate explanation for his absence from the witness box then you may hold his failure to give evidence against him. You do not have to do so. What proper inferences can you draw from the defendant’s decision not to give evidence before you? If you conclude that there is a case for him to answer, you may think that the defendant would have gone into the witness box to give you an explanation for, or an answer to, the case against him. If the only sensible explanation for his decision not to give evidence is that he has no answer to the case against him, or none that could have stood up to cross-examination, then it would be open to you to P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 Silent Defendants and Silent Witnesses 161 hold against him his failure to give evidence. It is for you to decide whether it is fair to do so. 23 If we compare this nugget of good sense with its American counterpart (recall the U.S. jury instruction previously cited: “you should simply put the whole matter of the defendant not testifying out of your mind”), it is not hard to divine which is more likely to produce a sound verdict. Interestingly, prior to Griffin, American judges used to give an instruction to jurors similar to the English one. Here, in fact, is the instruction that the California judge gave to the jury in the Griffin case: It is a constitutional right of a defendant in a criminal trial that he may not be compelled to testify. Thus, whether or not he does testify rests entirely in his own decision. As to any evidence or facts against him which the defendant can reasonably be expected to deny or explain because of facts within his knowledge, if he does not testify, or if, though he does testify, he fails to deny or explain such evidence, the jury may take that failure into consideration as tending to indicate the truth of such evidence and as indicating that among the inferences that may be reasonably drawn therefrom those unfavorable to the defendant are the more probable. In this connection, however, it should be noted that if a defendant does not have the knowledge that he would need to deny or to explain any certain evidence against him, it would be unreasonable to drawan inference unfavorable to him because of his failure to deny or to explain such evidence. The failure of a defendant to deny or explain evidence against him does not create a presumption of guilt or by itself warrant an inference of guilt, nor does it relieve the prosecution of its burden of proving every essential element of the crime and the guilt of the defendant beyond a reasonable doubt. 24 This instruction issues all the appropriate warnings against giving excess weight to the defendant’s silence. It is carefully crafted to stress a) that the defendant’s silence may (depending on the circumstance) be of minimal relevance, and b) that even in the most powerful circumstances, silence alone is insufficient grounds for convicting the defendant. Notwithstanding, the Supreme Court held this instruction to violate the constitutional rights of the defendant. In a civil trial, even in the United States, defendants usually do not enjoy a Griffin-type right. In Baxter v. Palmigiano, the Supreme Court held that the Fifth Amendment did not forbid “adverse inferences against parties to civil actions when they refuse to testify in response to probative evidence offered against them.” 25 In short, the constitutional right to silence in a civil trial carries none of the inferential prohibitions that that same principle carries in a criminal trial. Are we to conclude, using the logic of the Griffin case, that civil defendants 23 Quoted from Mark Berger, Reforming Confession Law British Style: A Decade of Expe- rience with Adverse Inferences from Silence, 31 Colum. Human Rights L. Rev. 243 (2000). 24 Griffin v. California, 380 U.S. 609, at 618 (1965). (Emphasis added.) 25 Baxter v. Palmigiano, 425 U.S. 308 (U.S., 1976). P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 162 flawed rules of evidence and procedure really have no right to silence at all (since the exercise of that right carries a penalty), or might we better infer that the right to silence implies nothing with respect to inferences that juries, whether criminal or civil, might draw from a defendant’s silence? Defenders of Lobotomized Juror see in proposals such as mine the undo- ing of everything from Western justice to the American way of life. As one member of the defense bar recently put it, enabling the jury to draw its own conclusions from a defendant’s silence would “shift the burden of proof to the accused[,] . . . weaken the prosecution’s burden of proof[,] . . . undermine the accusatorial systemof justice[,] . . . [and threaten] open and democratic society.” He adds, for good measure, that it might, as well, “compromise individual dig- nity, privacy, and autonomy.” 26 I leave you to draw your own conclusions as to whether life as we know it would end with the undoing of Lobotomized Juror or whether, as in the epigraph, Justice Cardozo is right that justice would not be compromised and that the interests of securing true and valid verdicts would be significantly enhanced. Silent Witness If the defendant is the only participant who can lay claimto an unrestricted right to silence, certain others can, under special circumstances, decline to tell what they know. One important class consists of those witnesses whose testimony might expose themto criminal liability. When witnesses refuse to testify for this reason, Lobotomized Juror kicks in once again: If the refusal occurs outside the jury’s presence, jurors cannot be informed; if it occurs before a jury, the jurors will be instructed to pretend it never happened. Ironically, Silent Witness, as I will call this policy, often works against the interests of an innocent defendant. It thus exhibits the worst possible com- bination of circumstances, sometimes making it more likely that an innocent defendant will be convicted and less likely that a guilty one will be. If a sub- poenaed witness with knowledge that would support the defendant’s innocence asserts his right not to incriminate himself, then the jury – unaware of the exculpatory information – may well convict an innocent defendant. Imagine the following circumstances: Jones is accused of blackmailing Smith. Key evi- dence against Jones is expert testimony that the blackmail letters demanding money fromSmith were typed on Jones’s typewriter. Jones claims to be (and is, let us suppose) innocent. He strongly suspects that Wilkins, a boarder in Jones’s house, is the true blackmailer. Jones’s defense strategy is to subpoena Wilkins and get his testimony, which Jones expects will shift suspicion from himself to his boarder. However, as soon as Wilkins receives the subpoena, he notifies 26 Gregory O’Reilly, Criminal Law, 85 J. Crim. L. & Criminology 402, at 444 (1994). P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 Silent Defendants and Silent Witnesses 163 the court that he does not intend to testify, citing his right not to incriminate himself. Before the trial, the judge convenes a hearing, obviously out of hearing of the yet-to-be-appointed jury. The judge questions Wilkins about his refusal to testify and Wilkins reiterates his Fifth Amendment right. Wilkins is excused fromtestifying, even though his testimony might have been of inestimable value to Jones. Once the trial gets under way, Jones’s lawyer tries to deflect suspicion from his client to the boarder, Wilkins. The latter, protected by Silent Witness, does not appear, nor can the jury be told that Wilkins refused to testify, even though that information is surely relevant to the case. From the jury’s point of view, Jones’s strategy of attempting to deflect guilt onto a third party, who doesn’t even appear before them, smacks of desperation. But if the jurors knew that someone besides Jones, with access to Jones’s typewriter, had refused to testify because he might thereby incriminate himself, they would probably take a very different view of things. At a minimum, such information would be relevant to their deliberations about Jones’s guilt. Unaware of that fact, however, the jury may find the circumstantial evidence sufficiently compelling to convict the innocent Jones. That would be an example of what I called in Chapter 1 a valid but false conviction. It would be valid because the evidence before the jury requires it; it is false because Jones is innocent. Silent Witness, driven by the same principle of protection against self-incrimination as Silent Defender, can thus make it more difficult than it should be for innocent defendants to secure the acquittals they deserve. Judge Henry Friendly summed up the problem succinctly: “ . . . a man in suspicious circumstances but not in fact guilty is deprived of official interrogation of another whom he knows to be the true culprit. . . . ” 27 A further wrinkle is worthy of brief mention. In principle, any witness – whether for the defense or for the prosecution – can invoke Silent Witness. However, when a witness does so, the prosecutor can compel that witness to testify by offering testimonial immunity, which basically means that her testi- mony cannot be used against her in subsequent legal proceedings. Granted such immunity, witnesses must testify. By contrast, the defense has no authority to grant immunity. 28 It can petition the prosecutor or the judge to grant immunity to an uncooperative witness whomthe defense has subpoenaed, hoping to trump Silent Witness. But such petitions are often rejected. The prosecutor may reject them because he doesn’t want his case against the current defendant weakened. The judge may be reluctant to grant thembecause he does not want to undermine 27 Friendly, at 680–1. 28 An exception is the Third Federal Circuit Court, which believes that a judge can order a prosecutor to grant immunity for a defense witness. Even here, however, the grant- ing of the immunity depends on the discretion of the judge. See Amar and Lettow, 863. P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 164 flawed rules of evidence and procedure pending prosecutions against the recalcitrant witness. 29 In such circumstances, the defense is powerless to oblige the uncooperative witness to come forward. This asymmetric practice again reinforces the tendency of the right of silence to make it easier to convict the innocent. This is another potent reason to reject it. Insum, the rule of silence makes it more likelythat the guiltywill be acquitted and more likely that the innocent will be convicted. That, of course, is exactly the reverse of the way a truth-seeking or error-reducing systemshould function. No system dedicated to finding out the truth should be so compromised unless there are compelling reasons, apart from truth seeking, for doing so. Even if such reasons exist, and I have already voiced skepticism about that, they do not reach so far as to justify keeping the jury in the dark about exercise of the privilege of silence. Privileged Testimony So far, our focus has been on the silence of the defendant or the witness with somethingtohide. We turnnowtoa thirdcategoryof silent nonparticipants inthe trial. These are individuals with information that the lawtreats as privileged and thus not susceptible of being elicited by interrogation or testimony. I conclude this chapter by considering the impact of these privileges on the truth-seeking process. As a general rule, the criminal lawholds that “every man’s evidence” is like- wise the state’s evidence. 30 If someone knows something relevant to a crime, unless he is the defendant or otherwise implicated in the crime itself (in which case he can claim the Fifth Amendment), he can be sworn as a witness and obliged to reveal what he knows, on pain of being found in contempt of court. However, certain kinds of information enjoy special legal status. Those who possess such information cannot be obliged to testify about it nor punished for not testifying. Most of this information arises from special relationships between persons. Thus, husbands and wives cannot be made to testify against one another. A lawyer may not be forced to testify about his conversations with his client, a psychiatrist about the content of sessions with his patient, a social worker about her communications with those on her caseload, nor a clergy- man about revelations fromhis penitent. Certain states have created other privi- leged relationships. In Massachusetts, for instance, a rape crisis counselor, even 29 As Amar and Lettow have written, “Most courts, however, have refused to grant immunity even if the testimonyis crucial to a defendant’s case” (Amar and Lettow, at 863 n. 15). 30 As John Wigmore put it, “For more than three centuries it has nowbeen recognized as a fun- damental maxim that the public . . . has a right to every man’s evidence.” 8 John Wigmore, Treatise on the Anglo-American System of Evidence in Trials at Common Law, 2192 (3rd ed., 1940). P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 Silent Defendants and Silent Witnesses 165 if unlicensed and without professional credentials, is privileged. 31 In others, “jailhouse lawyers” – that is, fellow convicts who offer legal advice – are like- wise privileged. Occasionally, the privilege focuses not on a relation between the defendant and some professional but on the information itself. Someone who knows state secrets cannot be made to divulge them in court. The identity of police informants does not have to be revealed. No one can be obliged to divulge how he voted in a political election. The rationale for the relationship privileges (as they are sometimes called) is that, if confidential information exchanged between someone and her spouse, lawyer, analyst, or priest could be elicited as testimony, then the integrity and viability of those relationships would be severely threatened. 32 As Charles McCormick’s classic textbook on evidence explains it: Rules of privilege are not without a rationale. Their warrant is the protection of interests and relationships which . . . are regarded as of sufficient social importance to justify some sacrifice of availability of evidence relevant to the administration of justice. 33 As for state secrets, voting behavior, and informant identity, their rationale is sufficiently obvious to require no elaboration here. The first curiosity to note about the relationship privileges currently recog- nized in federal law is what they fail to privilege. For instance: r Children can generally be made to testify against parents and vice versa. Is this because the parent-child relation is less threatened by the revelation of confidences than a spousal one is? Or is it supposed that children and parents never reveal confidences to one another? On either assumption, the distinction seems dubious. r Although psychiatrists cannot be made to testify in federal courts about what their patients have told them, other physicians can. Is it clear that the doctor-patient relationship is less dependent on patient candor than the analyst-patient or social worker–client one? 31 U.S. v. Lowe, 948 F. Supp 97, at 99 (D. Mass. 1996). 32 In granting privileged status to psychiatrists and social workers in 1996, the Supreme Court held that “the psychotherapist privilege serves the public interest by facilitating the provision of appropriate treatment for individuals suffering the effects of a mental or emotional problem,” adding that “effective psychotherapy . . . depends on an atmosphere of confidence and trust” (Jaffee v. Redmond, 518 U.S. 1, at 10–11 [1996]). While that may all be true – while it is even conceivable that psychotherapy is not the pseudoscience its numerous critics take it to be – the Court did not showthat the interests of mental health are sufficient to trump the opposing interests of truth seeking in the criminal justice system. In his dissent in this case, Justice Antonin Scalia asked rhetorically, “How likely is it that a person will be deterred from seeking psychological counseling, or from being completely truthful in the course of such counseling, because of fear of later disclosure in litigation?” (ibid., at 22). 33 Charles T. McCormick, McCormick on Evidence 72, at 171 (Edward W. Cleary, ed., 3rd ed., 1984). P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 166 flawed rules of evidence and procedure r Althoughone’s conversations withone’s lawyer are privileged, conversations covering the same subjects with one’s accountant usually are not. r Journalists have to reveal their sources if such information is relevant to a criminal proceeding. r Scientists conducting research on sensitive topics (such as sexual or crim- inal behavior) cannot promise full confidentiality to those who participate in their experiments, since federal courts recognize no researcher-subject privilege. One might pursue these contrasts in considerable detail if our interests were not primarily epistemological. Suffice it to say that the distinction between those relationships that are privileged and many that are not appears arbitrary. Because the privilege exclusions are driven, not by the content of the commu- nication but by the specific personal relationships involved, curious situations can arise. Consider a few: If John tells his wife, Sue, “I robbed the bank,” Sue’s information is privileged. If John says the same thing to Sue while ten-year-old John, Jr., is in the room, the information is no longer privileged. If Sue tells her mother, “John says that he robbed the bank,” that information – even though it is technically hearsay – is not privileged and must be divulged. If John tells a police informant, “I robbed the bank,” it can be used as testimony. If, however, the police plant a speaker in a confessional and have a recording of John telling his priest that he robbed the bank, the recording is off limits because it violates the privilege. All the privileges, with the possible exception of how you voted, pose obsta- cles to truth seeking. Although often criticized for making it more difficult to convict the guilty (whichtheyfrequentlydo), the privileges likewise worksome- times to convict the innocent. For instance, if Jones is on trial for raping Smith and a priest or analyst learns in a confession or therapy session that Wilson is the rapist, the inability to elicit the privileged information from Wilson’s priest may lead to an incorrect conviction of Jones. In essence, the relationship privileges say to those guilty of a crime, “You can reveal what you did, however horrible it was, to certain persons in the full knowledge that they will not be allowed to pass along those revelations in ways that will be harmful to you.” The courts say that they privilege such communications because it is “in the broader public interest” to do so. Is this a viable claim? Does it seem plausible that, if I reveal to my social worker that I just robbed the local liquor store, she will be better able to solve my problems with (for example) domestic violence or a slum landlord? Indeed, if I have just robbed the local liquor store, do I even have a legitimate claimon the unstinting confidentiality of my social worker? That seems doubtful. To be charitable, perhaps the argument in favor of these privileges has less to do with protecting the guilty and more to do with protecting the relationships P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 Silent Defendants and Silent Witnesses 167 that innocent people have with their social workers. But an innocent person should have nothing to fear from telling the truth to his social worker, even if the privilege didn’t exist, since nothing he revealed to her would be the sort of thing that would land him in trouble if repeated in a criminal trial. I flatly deny that a social worker can do her work in a way that promotes the public interest only if she can tell her clients that everything they say to her, however revelatory of criminal activity, will be shielded from legal scrutiny. (Indeed, social work flourished for more than a century before 1996, when the Supreme Court invented this privileged category.) Some may grant that the social worker privilege is dubious but say that matters are different with respect to one of the oldest of the privileges and the one most revered by the legal profession, the attorney-client privilege. Here is how the usual story goes: A lawyer has an obligation to represent the interests of his client. In an adversarial trial, this means constructing the most powerful defense that can be mounted. Therefore, the lawyer needs as much relevant information as possible, to anticipate the case that the prosecution will present against his client. Unless the accused knows that he can tell all to his lawyer in strictest confidence, he may well withhold information and weaken the defense that his attorney can mount. If attorneys could be subpoenaed to testify about the content of communications with their clients, they might be transformed from indefatigable defenders of defendants’ interests into unwitting enemies of those interests. The natural consequence of this argument for creating circumstances that allow fully candid communication between defendants and their counsel is that full candor with one’s attorney would include a confession of guilt, if the defendant were guilty. That would hardly serve the interests of the guilty client, however. Defense attorneys are officers of the court. Their professional code of ethics forbids that (in the words of the American Bar Association) “the relation of attorney and client should be used to conceal wrongdoing on the part of his client.” 34 Furthermore, lawyers are duty bound to informthe police if they know that their clients are about to commit a crime. No savvy guilty defendant, about to commit perjury by testifying falsely as to his innocence, is going to confide all to his attorney. Clearly, the system is not designed to encourage rational, guilty defendants to be wholly candid with their counsel. Revealing too much will mean that counsel, at a minimum, must refuse to carry on representing the accused. It follows that “full disclosure” to one’s attorney makes sense only for the innocent. But the innocent are precisely the ones who would have little to fear from the disappearance of the attorney-client privilege. So it must be something short of full candor that is protected by the attorney- client privilege. Presumably, the idea is that an attorney should be privileged 34 Canon 4 of the ABA Code of Professional Conduct. P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 168 flawed rules of evidence and procedure to learn lots of apparently inculpatory facts from those whom she serves while never learning those things that would seal her client’s guilt. This subterfuge may give a guilty defendant the most robust defense that ethics will allow and money can buy, but it does nothing to further the ends of true acquittals and true convictions. All the privileges mean that the jury is denied access to relevant informa- tion. Worse, in many jurisdictions, jurors are not allowed to be informed when a potential witness has invoked the privilege. In other jurisdictions, which per- mit the invocation of the privilege in front of the jury, jurors are instructed that they are to draw no adverse inferences from the invocation of the priv- ilege. Just as we saw in our discussion of Silent Defender, jurors are told to ignore the fact that someone with relevant but privileged information did not divulge it. Even if there were a justification for recognizing certain classes of privileged relationships (and I am not fully persuaded of that), no compelling evidential rationale exists for failing to inform jurors when a witness has invoked a privi- lege or for obliging themto repress any memory of its occurrence. The only hint of an argument relating such privileges to truth finding involves the claim that no legitimate adverse inference could ever be drawn froma witness choosing to invoke one of these privileges. If, for instance, a psychiatrist steadfastly refuses to answer all questions about the content of his conversations with his patient, or if a priest refuses to say anything about what happened in the confessional, what inference could the jury legitimately make concerning the guilt or inno- cence of the patient or the penitent? We cannot blame the defendant, after all, for the testimonial recalcitrance of some third party. The antidote to this form of self-deception is to remind ourselves who owns the privilege in question. Except in the case of the spousal privilege, it belongs not to the psychiatrist but to his patient, the defendant. It belongs not to the priest but to the penitent. If it belonged to the psychiatrist or the priest, then its exercise could sustain no plausible, adverse inferences against the defen- dant. But, as it belongs to the defendant, he can waive the privilege, allow- ing his psychiatrist, priest, or attorney to respond freely to the prosecution’s questions. If a defendant chooses not to remove that muzzle, and jurors are informed of that fact, they may well conclude that this is because the defendant wants to hide something that he fears his analyst or priest will reveal. This is why a jury, in certain circumstances, might be justified in drawing adverse inferences from the assertion of a testimonial privilege. That inference will often be a rational one to make, which is why juries should be both informed if defendants have asserted the privilege and allowed to make of that what they will. It is hard to fault Jeremy Bentham’s observation that the belief that relevant evidence can be legitimately excluded if it might create “unpleasant” conse- quences for various human relationships is “one of the most pernicious and P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 Silent Defendants and Silent Witnesses 169 most irrational notions that ever found its way into the human mind.” 35 In the balancing act between society’s interest in justice being done and its interest in fostering certain interpersonal relationships, courts have fairly consistently sac- rificed the interest in truth and justice to the “larger social good,” even when (as in the case of the social worker–client relation) they have little or no empirical evidence that the relationship in question would be undermined if the privilege were to vanish. Perhaps the last word belongs to McCormick, who, in his classic text on evidence, remarks apropos the marital privilege: We must conclude that, while the danger of injustice from suppression of relevant proof is clear and certain, the probable benefits of the rule of privilege in encouraging marital confidences and wedded harmony is at best doubtful and marginal. 36 That is, privileging certain forms of testimony exacts an undeniable epistemic cost in the name of possibly conferring certain social benefits. Occasionally it has been argued that eliminating the relationship privileges would make little difference to the outcome of a trial since, if these privileges were removed, wrongdoers would simply stop talking to their spouses, lawyers, priests, and psychiatrists about their nefarious behavior. 37 Under such circum- stances, defendants would no longer need to invoke the privilege to keep their confidants off the stand, since they would, in effect, have no confidants. This seems doubtful since, except in the case of the lawyer-client privilege, most wrongdoers are likely to be unaware of or – even if knowledgeable about – indifferent to the arcana of the law of testamentary privilege. What proportion of Americans would know, for example, that the social worker–client relation is privileged? How many would care, even if they knew? Conclusion There may be situations in which silence is golden. A criminal proceeding is not among them. If the accused, a suspect, or a witness knows something about a crime, truth seeking is hobbled when obstacles are put in the way of securing his testimony. Courts have to find ways to encourage those with relevant information to come forward. In the case of recalcitrant witnesses, that vehicle is a finding of contempt if they do not testify and vigorous prosecution for perjury if they lie on the stand. Where reluctant defendants are concerned, that mechanism should involve allowing triers of fact to make what they will of a defendant’s silence. Finally, concerning privileged witnesses, if they refuse to testify, that is usually because the defendant does not want jurors to hear their 35 J. Bentham, Rationale of Judicial Evidence 193–4 (J. S. Mill, ed. 1827). 36 McCormick, §86. 37 Richard Posner has argued, for instance, that “If the [spousal] privilege were abolished, and this were widely known, spouses would be much less likely to make damaging admissions to each other; so abolition . . . would not create a cornucopia of evidence” (Posner, 1532). P1: JZZ 0521861667c06 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:5 170 flawed rules of evidence and procedure testimony. When a witness invokes the privilege, the jury should be informed to whom that privilege belongs and left at liberty to attach whatever significance to that act seems appropriate. The rights to silence and to privileged relationships give an enormous epis- temic advantage to the guilty defendant and create a major epistemic obstacle for the state. A way to restore the epistemic balance, commonplace until the last quarter of the twentieth century, was to allow jurors to make what they would of such actions by the defendant. The newly coined “rights” preventing jurors drawing any conclusions from a defendant’s silence or from privileged nontestimony undermine that delicate equilibrium and make the silence rules serious obstacles to finding out the truth. Given the judicial practices and attitudes we have surveyed in this chapter, it becomes increasingly hard to suppress the conclusion that the courts basically do not want defendants to speak. They make silence a painless decision for the defendant – shielding him from its usual and natural consequences by jury instructions to make nothing of it. They penalize him if he opens his mouth. The courts have turned a right of the defendant (the right to say nothing) into the preferred state, thereby making a virtue out of a possibility. Instead of breaking down the barriers to the defendant’s participation in his trial by telling his own story, they have built those walls ever higher, piling disincentive upon disincentive to the defendant telling what he knows. This is patently contrary to the pursuit of accuracy in verdicts. P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 7 Confessions, Poison Fruit, and Other Exclusions Men are not to be exploited for the information necessary to condemn thembefore the law. . . the state which proposes to convict and punish an individual [should] produce the evidence against him by the independent labor of its officers. – U.S. Supreme Court (1961) 1 Admissions of guilt are more than “desirable,” . . . they are essential to society’s compelling interest in finding, convicting and punishing those who violate the law. – U.S. Supreme Court (1986) 2 Evidence is the basis of justice: to exclude evidence is to exclude justice. – Jeremy Bentham (1827) 3 The Tortured Logic of the Admission of Confessions Our focus in the last chapter was on the Silent Defendant. Here, we look at the case of the defendant who wasn’t silent but later wishes he had been. In short, confessions – especially retracted ones – will be the center of attention. Numerous empirical studies make it clear that confessions play a significant role in American criminal procedure. 4 Making allowances for discrepancies between various jurisdictions, it seems a fair summary of that research to say that more than a third of resolved criminal cases involve a confession from the accused. Given their frequency, no study of the epistemology of the law would be even modestly adequate without a discussion of the reliability of such self-incriminatory activity. 1 Culombe v. Ct., 367 U.S. 568, at 581–2 (1961). 2 Moran v. Burbine, 106 S. Ct. 1135, at 1144 (1986). 3 Jeremy Bentham, A Rationale of Judicial Evidence, bk. 9, ch. 3, at 490 (1827). 4 There is an excellent summary of dozens of these studies in Gordon van Kessel, The Suspect as a Source of Testimonial Evidence, 38 Hastings L. J., 1 (1986). 171 P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 172 flawed rules of evidence and procedure Unhappily, fewparts of the contemporary lawof evidence are more epistem- ically challenged than the rules governing the admissibility of confessions. A conflicted Supreme Court, vividly revealed in the epigraph, is but one indicator of the problem. In the first quotation, the justices suggest that a robust system of criminal justice would have little need for confessions. A quarter-century later, they insist that it requires confessions. Two factors drive the preoccupa- tion with confessions, especially those later repudiated by the defendant: a) a suspicion that some of those confessions are false; and b) a worry that juries will regard a confession, even a false one, as tantamount to a full proof of guilt. For many experts, these add urgency to the view that monumental efforts must be made to ensure that false confessions never get as far as a jury. In due course, this chapter will take up those concerns, but we need to do some conceptual spadework before we enter that fray. It will be useful to begin with a distinction between two principal contexts in which confessions may emerge. The first is while the accused is on the stand in a courtroom. Confessions of this sort are generally treated as ordinary testi- mony. The judge or jury hears them just as it hears whatever any other witness has to say. The legal system regards these as rather unproblematic. In the case of retracted confessions, however, especially those rendered to interrogators while in police custody, the judicial system interposes several hurdles between the confession and the jury. To this end, a whole body of evidence law regu- lates how this second form of confession is handled. It is this context that will chiefly occupy us in this chapter. Whenever I refer in the rest of this chapter to confessions without qualification, I will have in mind those proffered prior to trial, usually to the police, and subsequently retracted. The Triple Test Before a confession can reach the jury, it must pass a triad of tests. The first concerns itself with whether the confessions were offered voluntarily or were coerced. A judge makes this determination, using a tortuous set of benchmarks whose aim is to establish whether the defendant – who has now retracted his confession and is pleading innocent – gave the confession of his own free will. The second test deals with the legality of the circumstances in which the con- fession was obtained. The defense may argue, for instance, that the confession was obtained after an illegal arrest of his client or after an interrogation longer than the law allows. If and only if the confession passes both these tests, the judge then submits it to the third test of corroboration. This involves asking whether other evidence proposed by the prosecutor bears out some of the claims about the crime contained in the confession. If the judge finds that a confes- sion is either involuntary, illegal, or uncorroborated, she may exclude it. All this deliberation occurs absent the jury, which will generally learn about the P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 Confessions, Poison Fruit, and Other Exclusions 173 existence of a confession only if the judge has ruled it to be voluntary, legal, and corroborated. On its face, this elaborate ritual seems extraordinary to those of us who think that juries should see all relevant evidence. For starters, it seems that there can be few sorts of evidence more obviously relevant than a confession from the accused. True, some confessions are false, but we have numerous times made the point that proffered evidence need not be certifiable as true to be admitted as relevant. Just as we leave it to the adversarial process and the good sense of jurors to figure out which eyewitness testimony to believe and which to reject, so we might be inclined to leave it to jurors to sort out which confessions are true and which are false. Jurors, after all, are supposed to be the principal trier of fact in a criminal trial. Whether a confession is voluntary, legal, and corroborated is at least as much a matter of the facts of the case as it is a matter of law. Why, then, do we encounter all these barriers standing between a signed or taped confession, on the one side, and the jury, on the other? While the question, as posed, may seem rhetorical, it is a genuine one. We need to figure out whether, from an epistemic point of view, confessional evidence should face so many obstacles before a jury is allowed to learn about it. Resolving that question will require us to look at several elaborate doctrines, principally developed in the common law, about these issues of voluntariness, legality, and corroboration. Voluntariness Suppose that Gomez has just been arrested for robbery. He is being questioned at the police station and, after some initial intransigence, finally admits that he is the culprit. He dictates a detailed confession, signs it, and is then removed to a cell. Afterward, let us suppose, Gomez has a change of heart and, through his lawyer, seeks to retract his confession. He intends to plead innocent and wants his confession suppressed so that the jury does not learn about it. Of course, he cannot simply withdraw his confession, once made. Several different sorts of factors can trigger suppression of a confession. It may be, for instance, that Gomez was arrested, absent the police having probable cause for his arrest. 5 Or, the police may have failed to inform him of his Miranda rights to silence and to have an attorney present. Or, even having been informed of his rights, Gomez may have confessed without explicitly indicating that he wished to waive those rights. Worse, the police may have tortured or otherwise intimidated him. He may have confessed because the police offered false inducements, perhaps in terms of more favorable treatment by the jury if he owned up to what he had done. Perhaps the police interrogated him for an unreasonably long time before bringing himbefore a magistrate. Any of these situations would serve as a basis 5 For a classic case that established that confessions pursuant to a false arrest must be excluded, see Brown v. Illinois, 422 U.S. 590, at 599 (1975). P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 174 flawed rules of evidence and procedure in law for excluding Gomez’s confession. Prior to the trial itself, there will be an evidentiary hearing in which the prosecution and defense will debate the merits of admitting or excluding a contested confession. Our initial concerns will center on the kinds of arguments that the two sides will be permitted to make in their quest to introduce or exclude Gomez’s confession. Before we proceed, we need to clear up any possible confusion about what “coerced” means in the law. Those unversed in the law are apt to imagine that “coercion” refers tosituations inwhichsomeone is beingsubjectedtosome form or other of police brutality. Violence and threats of further violence break down one’s defenses. Eventually, we are apt to suspect, one will admit to virtually anything, simply to bring the agony to an end. If one is innocent, a coerced, false confession will be the result. Now, if this scenario is plausible, the fact that one confesses to a crime in such nightmarish circumstances probably has nothing to do with whether one committed the crime. Here, it seems perfectly reasonable to say that a coerced confession is as likely to be false as true and hence it is strictly irrelevant. The law, however, understands these words very differently. And that takes us to the core of the first problem we have to grapple with. American courts have decided that the use of force is not necessary to create a situation in which “coercion” occurs. Likewise, they have decided that the threat of force or violence is not necessary to constitute “coercion” either. They have even decided that psychological trickery is not necessary for “coercion” to have occurred. To put it bluntly, the view of the current Supreme Court (and of its predecessors since the 1960s) is that police custody itself is inherently “coercive.” That is, from the moment one is taken into a police interrogation room, or even taken into custody by an arresting officer on the street, one’s choices – including the choice to confess or not – are no longer “voluntary.” (This is the core premise of the Miranda decision.) It follows from the logic of this curious way of conceiving the world that no confession given by an arrested suspect is ever “voluntary” and, therefore, that no confession tendered by an arrested person should be admissible in court. As the Supreme Court put it in 1966, “no statement obtained from the defendant [while in custody] can truly be the product of free choice.” 6 From the earliest origins of the common law, interrogation by a constable or a justice of the peace (since there were no police forces until the nineteenth century) was regarded as entirely proper and fully compatible with the idea that a confession given in the absence of force and violence is voluntary. As recently as 1953, the Supreme Court itself was explicitly denying that police custody in and of itself removed the suspect’s free will: 6 A Book Named “John Cleland’s Memoirs of a Woman of Pleasure” v. Attorney Gen. of Mass., 383 U.S. 413, at 436 (U.S. 1966). P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 Confessions, Poison Fruit, and Other Exclusions 175 Interrogation is not inherently coercive, as is physical violence. Interrogation does have social value in solving crime, as physical force does not. By their own answers many suspects clear themselves, and the information they give frequently points out another who is guilty. Indeed, interrogation of those who know something about the facts is the chief means to solution of crime. The duty to disclose knowledge of crime rests upon all citizens. . . . This Court never has held that the Fourteenth Amendment prohibits a state from such detention and interrogation of a suspect as under the circumstances appears reasonable and not coercive. 7 Within a decade, this commonsensical view had given way to the idea that any custodial confession was prima facie “coerced.” Because “coercion” (mere police custody) and coercion (use of force or threats of force and intimidation) are such different things, we must lay aside our common-sense intuitions about such cases when we speak about “coercion” as the courts now understand that term. We must ask ourselves anew: Is it plausible that an innocent suspect, sub- jected to neither torture nor the threat of it, offered no particular incentives to confess, and treated reasonably humanely by arresting officers, is likely to tender a false confession simply to get himself out of the interrogation room and into a jail cell? The question answers itself. Innocent people, even when subjected to the humiliation of an arrest, do not routinely confess to crimes they did not commit, especially since they fear, understandably enough, that such confessions may mean several years in prison. In sum, we have precious little reason to doubt the veracity of a “coerced” confession. “Coerced” confessions are potentially relevant evidence in a way that coerced confessions may not be. I had just said that current legal opinion holds that the police station rep- resents an inherently “coercive” environment and that any confession made in police custody must be considered so dubious as to be inadmissible. Does this mean that all confessions obtained by the police are thrown out? No, because the Supreme Court, in the famous Miranda case, invented a legal fiction for extricating itself from the corner into which its voluntarist logic was leading. According to the Court, it is possible to get a legitimate and “uncoerced” confes- sion froma suspect provided that, after his arrest, he is read what we have come to call his Miranda rights, and that he expressly waives his right to silence and to an attorney. Absent that ritual, no confession is acceptable, allegedly because no confession is “voluntary” without it. There are conceptual problems aplenty here. I will mention only the two most important: First, this analysis asks us to assume that mere police custody takes away our free will and makes our behavior and statements “involuntary.” That already seems dubious, but let us accept the point for the sake of argument. Then we have to believe that something scarcely short of miraculous happens. Provided 7 Stein v. NY, 346 U.S. 158, at 184 (1953). P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 176 flawed rules of evidence and procedure that the interrogating officer rehearses the Miranda formula at the outset of the interrogation, and that the suspect agrees to waive his rights, the “coercive” and intimidating environment of the police station vanishes and the suspect reassumes responsibility for his actions. In other words, if, after hearing a ritual rehearsal of his rights, he decides to waive them and confess, the confession will be deemed “voluntary.” Had he given the identical confession one minute earlier (that is, after arrest but prior to hearing his rights), it would have been “coerced” and inadmissible. Had he confessed in the police station prior to being informed of his arrest, that confession, too, would have been “voluntary.” Timing, it seems, is everything. This is absurd on its face. If it is really true that being in police custody is intrinsically “coercive” (and that is the core premise of Miranda), then we cannot believe that a suspect, not even after hearing his Miranda rights laid out, freely chooses to make a confession. Free choices, by definition, cannot be assured in genuinely coercive environments. On the other hand, if custody is not intrinsically coercive, then a confession given in the absence of violence or threats should be a candidate for admission, regardless of whether the police have engaged in the Miranda ritual. I will leave to one side further discussion of the perplexing question of how the repetition of a string of phrases, which everyone who watches television has already heard hundreds of times and that most of us can repeat from memory, can transform an inherently “coercive” situation into a “voluntary” one. Second, we have to ask ourselves whether, even if a “coerced” confession is “involuntary,” that fact has anything to do with its relevance and reliability as evidence. The voluntariness of a confession is given special significance mainly because we assume that coerced confessions are probably unreliable. Is there the slightest reason to think that “coerced” confessions raise similar doubts? If the police forget to Mirandize the arrestee, but otherwise behave reasonably toward the suspect, have we grounds to believe that a confession proffered in such circumstances is so likely to be false as to justify us in preventing the jury hearing it? Is the ambiance of the interrogation room – absent Miranda – so “coercive” that a confession elicited in such circumstances is as likely to be false as to be true? In sum, there is no epistemic motive for excluding non-Miranda confessions as such. They are indisputably relevant evidence, particularly if they have been corroborated. The Supreme Court thinks otherwise. In 1966, it insisted that “Miranda guard[s] against the possibility of unreliable statements in every instance of in- custody interrogation.” 8 Note the structure of the argument. By excluding all confessions that fail to meet the Miranda demands, says the Court, we protect against the false confessions that might ensue from police interrogation. With 8 Johnson v. New Jersey, 384 U.S. 719, at 730 (1966). P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 Confessions, Poison Fruit, and Other Exclusions 177 as much plausibility, one could say that by excluding all confessions (whether “coerced” or “uncoerced”), we would further “guard against the possibility of unreliable statements.” The Court patently fails to give any weight to the fact that, in excluding all non-Miranda confessions, it is doubtless throwing out many confessions that are true. Insteadof askingthe right question(towit, “Wouldthe exclusionof Miranda- violating confessions be likely to produce more true verdicts than their admis- sion would?”), the Court asks, “Could we reduce the number of false convic- tions by rejecting confessions proffered without Miranda?” While that question admits of an affirmative answer, so do questions like, “Could we reduce the possibility of false convictions by excluding all inculpatory evidence?” The balancing act between false convictions and false acquittals required to evalu- ate the bona fides of any proposed rule of evidence simply was not brought into play by the Court in deciding Miranda. If we judge the exclusion of Miranda- violating confessions against that yardstick, we come to a conclusion directly contrary to the Supreme Court’s. While the odd confession obtained by virtue of a Miranda violation may be false, the vast majority of them are likely to be true, even if they were secured in the “coercive” environment of a police sta- tion, and especially if they have also been corroborated. The policy of throwing out all Miranda-violating confessions will predictably decrease the number of false convictions slightly in return for vastly increasing the number of false acquittals. A dissenting member of the Miranda Court, Justice Byron White, pointed out as much when Miranda was adopted: In some unknown number of cases the Court’s rule will return a killer, a rapist or other criminal to the streets and to the environment which produced him, to repeat his crime whenever it pleases him. As a consequence there will not be a gain, but a loss, in human dignity. 9 Once again, we see distributionist worries driving the formulation of rules of evidence, with the inevitable compromises of truth that such concerns always bring. Illegally Obtained Confessions According to current law, the Miranda ritual is necessary to establish the vol- untariness of a confession but it is not sufficient to establish admissibility. If the accused alleges that he was beaten, threatened, or offered unreasonable incen- tives to confess, that, too, will raise doubts about whether his confession was “voluntary,” even if the interviewing officers fully complied with the Miranda demands. Similar problems arise if the accused can show that he was interro- gated by police officers over an extended period of time (the usual cutoff is eight 9 Justice White, dissenting in Miranda v. Arizona, 384 U.S. 436, at 542 (1966). P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 178 flawed rules of evidence and procedure hours) prior to being presented before a judge for arraignment. 10 Likewise, if the accused can showthat, although read his Miranda rights, he was nonetheless denied access to a lawyer when he requested to see one, this provides grounds for excluding the confession. Again, if the accused was arrested when the police had less than probable cause, any ensuing confession is considered illegal and is thus ruled inadmissible. 11 To admit a confession, the judge must assure himself that all such allegations are groundless. Or, to be more precise, the prosecutor is obligated to convince the judge by a preponderance of the evidence that the confession was legally obtained – if the defendant makes a claim about the use of illegal tactics. From an epistemic point of view, this is a confusing grab bag of situations. While it may be reasonable to suppose that a confession induced by violence or threats is as likely to be false as true, it is surely unconvincing to say the same thing about a confession offered when one is refused access to a lawyer or when one has been interrogated for nine hours instead of eight or when one’s arrest was based on less than probable cause. The motive for throwing out such confessions has everything to do with rule breaking by the interrogators and nothing to do with legitimate grounds for doubt about the content of the confession. To encourage full compliance by the police with existing laws about the treatment of those under arrest, the courts have decided that punishing the prosecutor by excluding an indisputably relevant confession is appropriate. This decision to subordinate the importance of getting a correct verdict to a concern with the legality of obtaining the relevant evidence obviously represents a policy decision that goes well beyond our epistemic brief. I do not intend to skirt that issue, but I will postpone serious discussion of it until Chapter 9. Corroboration Having passed tests for voluntariness and legality, a confession is still not yet ready for prime time. In most U.S. jurisdictions, the judge must also satisfy himself that there is independent evidence for the guilt of the accused, apart from that contained in the confession itself. The specific structure of the corroboration requirement varies from one jurisdiction to another. In most, the corroboration requirement insists that the evidence presented by the prosecution, apart from the confession itself, must include proofs of the corpus delicti. Absent any other proofs of guilt, a confession standing alone will, in most jurisdictions, be excluded, even if apparently voluntary and legal. Such exclusion not only 10 On exclusion of confessions obtained after delay, see McNabb v. U.S., 318 U.S. 332, at 344 (1943). 11 Perhaps the two chief cases that established the idea that illegal confessions must be excluded were Miranda v. Arizona, 383 U.S. 436, at 481–2 (1966); Rogers v. Richmond, 365 U.S. 534, at 544 (1961). P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 Confessions, Poison Fruit, and Other Exclusions 179 means that the prosecution cannot include the confession among the evidence it presents, but also that it cannot reveal to the jury in its case-in-chief anything the accused said in his confession, nor even that there was a confession. The confession is treated legally as if it never happened if, in the opinion of the presiding judge, it is not corroborated by other facts in the case. On the face of it, the corroboration requirement has an obvious epistemic rationale. There can be no doubt that a confession becomes more reliable as it is supported by independent lines of evidence. Nonetheless, I will suggest that the corroboration test comes at the wrong place in the judicial proceedings and that, conceived as a test of reliability, the test is far too weak to do the work expected of it. It comes at the wrong place because the judge should be making judgments of relevance, not judgments of reliability. No one disputes that confessions, whether retracted or not, are relevant evidence. If we regard jurors as the triers of fact, it is they and not the judge who should assess the reliability of a confession, just as they are left to assess the reliability of eyewitnesses or much physical evidence. But since, as I said in Chapter 4, epistemology is neutral on the question of whether it is judge or jury who is charged with determining the reliability of evidence, I will not press this point further here. I will simply note that courts seem very worried that jurors will overinterpret the authenticity of a retracted confession and that, therefore, confessions should go to a jury only if they have been precertified as probably authentic. The principal reason that courts are reluctant to trust jurors to handle confes- sions responsibly is the fear that jurors will automatically convict any defendant whohas givena recantedconfession. Legal scholars RichardA. LeoandRichard J. Ofsche state what appears to be the consensus among appellate judges when they write: Because a confession is universally treated as damning and compelling evidence of guilt, it is likely to dominate all other case evidence and lead a trier of fact to convict the defendant. 12 If this claim were true, there might be good reason to exclude as unfairly prejudicial all but the most powerfully corroborated confessions. But there is plenty of empirical evidence that makes it clear that jurors are not as credulous about confessions as some experts think. For instance, in one well-designed study of felony trials in Pittsburgh, the authors found that the conviction rate for defendants who offered no (admissible) confession was 54.5 percent, while those who made an admitted confession were convicted in 78.7 percent of the 12 Richard A. Leo and Richard J. Ofshe, Criminal Law: The Consequences of False Confessions: Deprivations of Liberty and Miscarriages of Justice in the Age of Psycho- logical Interrogation, 88 Crim. L. & Criminology 429 (1998). P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 180 flawed rules of evidence and procedure cases. 13 Although, to judge by these figures, the conviction rate is clearly higher when there is a confession than when there is not, two observations are in order. The first is that, supposing the cases to have been otherwise similar in terms of inculpatory evidence, the availability of a confession should increase the con- viction rate, as should any other highly relevant evidence that is available in one class of cases and not in another. Second, even when the juries had a confession in hand, they voted to acquit in almost one-quarter of the cases. This gives the lie to the myth that jurors regard confessions as a clincher or as “compelling evidence of guilt.” It is worth adding that the confessions heard by Pittsburgh juries were already judge-sanitized by the various tests of voluntariness and legality. This means that many of the most contested and least plausible con- fessions were never heard by the jurors. So, there is reason to think that if, as I have recommended, jurors were to hear all confessions except the genuinely coerced ones, the conviction rates in confession cases would be even lower than the levels found in the Pittsburgh study. Still, the most serious problem with confession law does not hang on the question of who makes the determination of reliability but rather on what the test for reliability is, that is, on how we understand this notion of corroboration. In most jurisdictions, judges are told by higher courts that a confession is corroborated so long as there is any independent evidence that supports the hypothesis that a crime occurred. This requirement is, in my judgment, far too weak for purposes of determining the reliability of a confession. Instead, the judge (if he is the one making the determination of reliability) or the jury should understand that a confession is corroborated just in case it contains information about the crime of a sort that only its perpetrator could know. Corroborating any hypothesis requires more than finding out merely that some of its consequences are true, which is the prevailing legal gloss on “cor- roborated.” The only category of evidence that strongly corroborates a hypoth- esis, H, consists of those facts that seriously probe the truth of H – that is, facts that are highly unlikely if His false. Thus, in the case of a confession, we should be seeking facts that are highly unlikely to be true if the confession is false. A confession replete with fine-grained details concerning the crime would come in this category, since it is obviously unlikely that someone innocent of a crime could describe it accurately and in detail. On the other hand, if the confession does not describe the crime in detail, if it cannot lead the police to evidence they do not already have, if it gives a generic rather than a specific description of the crime, then we still lack strong corroboration of the defendant’s guilt. (Of course, some true confessions may lack these features. Perhaps the crime was committed when the defendant was too drunk or too high to remember the details and all he can recall is committing the crime itself. Although some such 13 R. Seeberger and R. C. Wettick, Miranda in Pittsburgh – A Statistical Study, 29 U. Pitt. L. Rev. 1, at 20 (1967). P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 Confessions, Poison Fruit, and Other Exclusions 181 confessions may be true, the confessions themselves give us no reason to credit them strongly enough to use them as a basis for conviction.) One can already see one potential snag with the “facts known only to the perpetrator” rule of corroboration. The implied contrast class here, of course, is third parties who did not commit the crime. This ignores the fact that police investigators, by the time they have brought charges against a suspect, will prob- ably already have learned a great deal about the crime that only the perpetrator would know– what sort of weapon was used, the “M.O.” of the perpetrator, and so on. If unscrupulous interrogators were to feed this information to the suspect during questioning, the suspect would then be able to make a false confession that could nonetheless be corroborated in this demanding sense. A few simple prophylactic steps would circumvent, or at least minimize, these difficulties. We could, for instance, require the audio- or videotaping of interrogations (as already occurs in much of the English-speaking world and even in many U.S. jurisdictions 14 ). These records would reveal whether the material contained in a confession reflects the firsthand knowledge of the suspect or was provided to him, intentionally or unintentionally, by the police during their interrogation. Provided that the judge manages to communicate to jurors that a bare confes- sion should not be dispositive (since it may be false) and that the jury should seek corroboration for it in other evidence presented by the prosecutor, then a jury should hear any confession that is relevant, that is, any confession not exacted by coercion. (Whether the jury should hear illegally obtained confessions is, as I have said, a policy question that goes beyond the epistemic.) The Supreme Court repudiates the suggestion that relevance rather than voluntariness should be the key to admissibility of confessions. In 1961, for instance, the Court argued that, in deciding whether a confession was voluntary (and thus admissible), it was inappropriate to consider whether the confession was probably true. 15 That is, a confession that offers every internal sign of coherence and that is broadly corroborated will nonetheless be excluded unless the state can prove that it was voluntarily given. This is, to put it mildly, curious. Recall that the original position of the Court – fromthe nineteenth century until the Miranda reforms of the 1960s – was that the voluntariness of a confession was important precisely because it was a marker as to whether the confession was likely to be reliable or true. By the mid-twentieth century, however, volun- tariness had taken on a life of its own, severed from its earlier role as a test of the truth of the confession. What was originally introduced as an aid to truth seeking has now become an impediment to that search, capable of trumping 14 A nice treatment of current interrogation practices can be found in William Geller, Video- taping Interrogations, Nat’l. Inst. of Justice, Research in Brief, at 2 (Mar. 1993). 15 InRogers v. Richmond, the Court arguedthat a challenge tothe voluntariness of a confession could not be resolved using “a legal standard which took into account the circumstances of [`ıts] probable truth or falsity” (365 U.S. 534 [1961]). In other words, apparently true confessions can be held inadmissible if they were “involuntary.” P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 182 flawed rules of evidence and procedure truth itself. The price paid by society for this voluntarism, Miranda or other- wise, is not negligible. Paul Cassell has conducted extensive empirical studies of the effect of Miranda on conviction rates. Among his conclusions: “[E]ach year the number of crimes that go unsolved because of Miranda is between 56,000 and 136,000 violent crimes and 72,000 to 299,000 property crimes.” 16 While some scholars dispute these specific numbers, it is undeniable that many true confessions continue to be ruled inadmissible because of Miranda. Poison Fruit Confessions are useful to police and prosecutors not only, sometimes not even chiefly, because they point to the guilt of the confessing party but also because they provide an invaluable resource for collateral investigation. A confession may well name accomplices or point the police to the location of important clues and other physical evidence related to the crime. Current practice holds that when a confession has been obtained illegally (for instance, after an arrest without probable cause), not only must the confession itself be thrown out, but so too must the courts exclude all evidence obtained as a result of following up the leads provided in the confession. Suppose that poor Mr. Gomez’s confession indicated the current location of the goods that he is alleged to have stolen. Pursuing that lead, the police go where Gomez claims to have hidden the stolen goods, and there they are, replete with Gomez’s fingerprints. On my analysis, this would constitute impor- tant corroboration of Gomez’s confession and be a powerful argument for its admission. In American courts, however, if Gomez were denied access to his attorney before his confession or held in custody for longer than the courts allow before facing a judge, the latter will not only exclude his confession but also all the other evidence found as a result of that confession, no matter how inculpatory that evidence is. Federal courts have gone so far as to suggest the possibility that, when an illegal confession leads the police to the location of the corpse in a murder case, the police and prosecutor must pretend, for purposes of the trial, that they do not know the whereabouts of the body. 17 Indisputably, this practice (often known as excluding the “poison fruit of a coerced confession”) runs directly counter to the court’s avowed interest in finding out the truth. It precludes the jury’s access to evidence that is sometimes overwhelmingly relevant to their decision about guilt or innocence. The ratio- nale for excluding this evidence cannot be epistemic since, unlike the exclusion of a coerced confession (where the coercion ostensibly increases the probability 16 Paul Cassell, Miranda’s Social Costs: An Empirical Reassessment, 90 Nw. U. L. Rev. 387 (1996). 17 The case in which this issue was explicitly raised by the Second Circuit Court of Appeals is Killough v. U.S., 315 2d 241, at 245 (1962). P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 Confessions, Poison Fruit, and Other Exclusions 183 that the confession is false), we are dealing here with the admissibility of evi- dence that (unlike a false confession) is not fabricated by a terrified defendant. Why, then, is it excluded? The official answer is that prosecutors may not use against a defendant evidence that came intotheir possessionillegally. Eventhoughthe “poisonfruit” in question may have been seized with a judge’s court order, it is nonetheless considered illegally seized if knowledge as to its existence or whereabouts arose out of a confession that violated legal procedures. (There is one telling exception to this rule. If the poison fruit arises from a failure by the police to go through the Miranda ritual before obtaining a confession, the courts will allow the police to use evidence obtained from the content of the confession against the accused, even though the confession itself will be excluded. This clearly suggests that the courts themselves entertain doubts as to whether the so-called Miranda rights are full-blown rights in the usual sense.) In most other cases of illegally obtained confessions, the fruit will be excluded along with the confession. There is nothing in the U.S. Constitution that would imply or sanction this rule. Like so many other truth-thwarting practices, it is in place because the courts have decided that police illegality should be discouraged at whatever price. This doctrine, like many other epistemic obstacles current in U.S. criminal law, was a creation of the courts in the twentieth century. Before that, the prevailing view was that evidence arising out of following up leads provided in an illegal confession was fully admissible, even if the confession itself was not. No other common law courts in the world exclude poison fruit as system- atically as American ones do. (As we will see in this section, the poison fruit doctrine excludes much more than evidence found as a result of an illegally obtained confession.) Jurisprudence in other common law countries holds that, while the rejection of an illegal or coerced confession is rooted in a concern with finding the truth, there is no truth-related rationale for excluding indis- putable and relevant facts uncovered as a result of checking out a confession of doubtful legality. Those countries are right, and the sooner the U.S. Supreme Court faces up to the epistemically deleterious effects of the poison fruit doc- trine, the sooner American evidence law will regain a modicum of epistemic integrity. We saw in Chapter 5 that current rules of evidence and procedure go to great lengths to encourage a defendant’s silence. These measures range from innocuous reminders to the defendant that he has the right to say nothing, to serious truth-threatening devices such as prohibiting the jury from drawing any inference from a defendant’s silence, to a barrage of defendant-unfriendly consequences if he does choose to testify. What this chapter has shown so far is that if the defendant chooses to speak and does so in the formof a confession that he later retracts, the lawimposes a series of hoops that the prosecutor must jump through if he is to get that confession before a jury. Only one of these hoops, P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 184 flawed rules of evidence and procedure the existence of corroborating evidence, has a clear epistemic motivation. The others – demanding that, before a jury hears a confession, it must be shown to be “voluntary” (in the idiosyncratic sense of that term now used by the courts) and it must have been legally obtained – represent well-meaning but misguided efforts to reduce false confessions and to keep police investigators on the straight and narrow. Making matters worse, the law, besides excluding a great many probably true confessions, generally prevents the state from using evidence, however reliable and inculpatory it may be, arising out of police follow-up research on excluded confessions. Instead of encouraging the defendant to tell what he knows and leaving the jury to decide whether a confession is plausible and corroborated, the regimen now in place does everything in its power to ensure that the defendant says nothing during his trial and that, should he have spilled the beans prior to trial, those revelations, and their fruit, often fail to reach the jury. If you detect in these policies a whiff of hostility on the part of the judiciary toward the use of confessions as an instrument of proof, you would probably not be far off the mark. As the Supreme Court itself once famously argued in Escobedo: [A] system of criminal law enforcement which comes to depend on the “confession” will, in the long run, be less reliable and more subject to abuses than a system which depends on extrinsic evidence independently secured through skillful investigation. 18 What the Court offers here is a set of false choices. Liberalizing the rules about the admissibility of confessions is no substitute for, but an indispens- able supplement to, independent prosecutorial investigation. Permitting juries to hear relevant confessions is patently not the same thing as instituting a system that “comes to depend on confessions.” There is no reason to believe that if confessions are allowed as one of several forms of proof, those other forms of proof would soon atrophy. Rephrasing the Court’s remark from Escobedo, we can say that a system of law enforcement that often denies jurors access to a relevant confession will, in the long run, be much less reliable than a system that puts most confessions before a jury, instructing the latter to demand cor- roboration of any proffered confession. The Court may insist, as it did in 1949, that “The lawwill not suffer a prisoner to be made the deluded instrument of his own conviction;” 19 still, many of the rest of us will think that a judicial system that often denies itself access to plausible confessions and to their indisputably relevant fruit is badly in need of a quick epistemic fix. 18 Escobedo v. Illinois, 378 U.S. 478, at 488–9 (1964). See also the passage in the epigraph from Culombe v. Ct. 19 Watts v. Indiana, 338 U.S. 49, at 54 (1949). P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 Confessions, Poison Fruit, and Other Exclusions 185 The Exclusion of Nonconfessional Evidence In this section, I want to discuss some of the ramifications that flow from the exclusionary rules that have nothing to do with confessions. Put very simply, if evidence was obtained in a way that violates the courts’ interpretation of the defendant’s Fourth Amendment rights (chiefly having to do with arrests, searches, and seizures), then that evidence will be excluded, regardless of its relevance to the case at hand. So, too, will any of the poison fruit associ- ated with such illegal searches or seizures. Neglecting fine points of doctrine, a search or seizure (including an arrest) is deemed illegal if the police offi- cers involved lacked a warrant or at least “probable cause” to believe that the defendant had committed a crime or, in the case of a search, that they would find evidence of criminal activity. It is important to add that the exclusion- ary rules under discussion here are triggered not only by a violation of the defendant’s constitutional rights but also, sometimes, by a violation of statute law in the acquisition of the evidence in question. Consider a few concrete examples: r If a police officer on duty in a train station sees a passenger about to board a train with what looks to be a concealed weapon under his bulging jacket, the officer may legally stop that passenger and frisk him. If the officer finds a gun in that search, or any other sort of contraband that he can identify by touch, he may seize the relevant evidence and arrest the passenger. If, however, the officer reaches inside the suspect’s pocket and pulls out (for example) a small packet of high explosives, that evidence will be excluded because a frisk, as defined by law, does not allow entry inside someone’s garments. Absent probable cause for searching the subject’s pockets for a prohibited chemical substance, this search was illegal. Evidence so obtained will be excluded. r The police learn from an informant that John keeps the accounts for a local extortion racket. The police take what evidence they have to a judge and get a legal order to search John’s study for documents relevant to this alleged activity. The police show their warrant, enter John’s home, and are led to his desk, which they search thoroughly, to no avail. Before leaving, they ask John about what a closet in his hallway is used for. “Winter clothes,” he replies. The police open the closet and find, amid the coats and raingear, a covered box. They lift the cover on the box and find stashes of drugs in plastic bags. They arrest John for possession. This, however, will be to no avail since the judge, in any trial of John for possession of drugs, will note that the warrant was issued for a search of John’s study for documents, not for a search of John’s house for drugs. r The police pull over Peter’s minivan for speeding. As the officer examines Peter’s license and title, she notices that a long cloth is draped conspicuously over something in the row of seats behind the driver. Recalling a police P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 186 flawed rules of evidence and procedure message from an hour earlier about the holdup of a liquor store with a shotgun, the officer reaches through the open window and lifts up the cloth. Under it, she finds a shotgun, which subsequent laboratory analysis reveals to have been the one used in the crime. However, the gun arguably cannot be used as evidence in a trial of Peter for robbery since the arresting officer did not have a reasonable basis for her suspicion that the cloth concealed a shotgun. What these examples illustrate is that the sort of police illegality that can trigger exclusion is not only flagrant abuse of authority (for instance, arbitrarily frisking someone for no reason whatever, entering someone’s home without a warrant, and so on), but what might seem to most of us to be relatively innocuous and natural actions by the police. Still, it is crucial to stress that the only sort of illegality that will lead to exclusion of evidence involves acts committed specifically against the defen- dant, rather than police lawbreaking in general. This curious premise – that illegality against nondefendants does not warrant exclusion of evidence while illegality against the defendant does – leads to many bizarre situations. Sup- pose, for instance, that the chief of the local vice squad, Sergeant O’Leary, gets an anonymous telephone tip that Big Al is running dope and prostitution out of his office downtown. Knowing that anonymous tips do not reach the level of probable cause, O’Leary nonetheless decides to followup and organizes a team of fellow vice squad members to break into Big Al’s office. Inside, he finds Big Al’s Big Black Book, containing information about drug sales, dates, appoint- ments with prostitutes, and all the other details of Big Al’s operations. O’Leary takes these to the police station and studies them closely. On the strength of what he learns, he proceeds to arrest four of Al’s associates for selling drugs and seven for prostitution. A key bit of evidence in those cases will be Big Al’s book. Although O’Leary indisputably seized the book illegally, the state can use it against Al’s associates since their Fourth Amendment rights were not violated by the police break-in. What the state cannot do is to use that same illegal evidence against Al himself. Still keen on getting the head man, O’Leary guesses that Al’s lieutenant, Schwartz, might have a second copy of the Big Black Book, reckoning that Al would probably not trust all his contacts and accounts to a single copy. Again, this is just a hunch, nothing like enough evidence to persuade a judge to issue a warrant to search Schwartz’s house. Nonetheless, O’Leary forces his way into Schwartz’s house and finds the second copy. Clearly, the state cannot use that second copy to prosecute Schwartz (since the search violated his rights), but it already has him nailed by the material in Al’s original copy. What the state can do is to use the information in Schwartz’s copy to convict Big Al himself. Once again, the use of the illegally seized evidence will meet no judicial hurdle as to its admissibility, provided it is not used to prosecute the person whose P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 Confessions, Poison Fruit, and Other Exclusions 187 rights were violated by its seizure. So, here we have identical copies of the one book, both seized illegally. The first copy can be used against all of Big Al’s accomplices, while the second can be used against the boss himself. The point is that the Exclusionary Rule does nothing to sanction or regulate illegal police behavior directed at someone other than the defendant, while it offers an ironclad guarantee to the defendant that illegal violations of his rights will lead automatically to exclusion of the seized evidence, no matter howincriminatory. In this hypothetical example, all the involved parties were assumed guilty. The workings of the Exclusionary Rule in the case of innocent persons are even more bizarre. So long as the prosecutor never brings you to trial, the remedy incorporated in the Exclusionary Rule provides you no relief or protection from illegal searches and seizures. Unless you are the defendant in a trial, you have no official standing and thus cannot protest against the state using the evidence it has illegally seized from you against whomever else it likes. If this seems to you to be a reasoned and coherent policy for conducting criminal trials and investigations, then the Exclusionary Rule will not strike you as bizarre. For many of us, however, it stands as an insult not only to the search for the truth but to basic standards of fair play and the protection of the innocent. This rule offers very broad protections to the probably guilty while doing little or nothing to safeguard the interests of the probably innocent. Worse, while pretending to be driven by a concern to check police illegality, it does nothing to penalize illegal police behavior toward anyone but the defendant. No one disputes that much of the evidence seized in illegal searches can be highly relevant, that is, that it would increase a reasonable person’s belief in the likelihood that the accused committed the crime with which he is charged. No one disputes that many guilty felons walk free because of the Exclusionary Rule. (It has been estimated that the Exclusionary Rule undermines an otherwise powerful prosecution in roughly 5 percent of general felony cases and in some 30 percent of drug cases. 20 ) The rule’s elimination is thus highly desirable, whether we are concerned narrowly with truth or broadly with justice. If we take seriously the notion of Chapter 5 that the trier of fact should see relevant evidence, then the requirement that evidence must be excluded if it was illegally obtained should be abandoned. There is another twist to exclusion law that is noteworthy. If the defendant takes the stand in his defense, then evidence that was previously excluded because of its illegality is nowfair game and can be admitted in order to impeach the defendant’s testimony. Multiplying the ironies, the prosecution cannot use such evidence to impeach the testimony of any defense witnesses besides the defendant himself. Consider a typical case of appellate reasoning about these 20 National Institute of Justice, U.S. Department of Justice, The Effects of the Exclusionary Rule (1982): “4.8 percent of all felony arrests rejected for prosecution from 1976 through 1979 [in California] . . . were rejected because of search and seizure problems. . . . ” P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 188 flawed rules of evidence and procedure matters. In a 1966 case, U.S. v. Morla-Trinidad, a suspected crack dealer was on trial for sale and possession of a prohibited substance. This was not his first brush with the law. Prior to this trial, he had been tried on a similar charge, but the principal evidence, crack cocaine found in his possession, was held to have been illegally seized and thus was excluded from the previous trial. 21 (He was acquitted.) Some years later, Morla-Trinidad was charged with a new case of possession. The defendant took the stand in his own defense and claimed that he had never had crack in his possession. At that point, the judge permitted the prosecution to introduce the illegal, excluded evidence from the previous trial in order to refute the claim of the defendant. In sum, clearly relevant evidence was initially excluded in the first trial because of its illegality. But that obstacle vanished as soon as the defendant in the second trial declared himself to have never dealt in drugs. The defendant was convicted and, on appeal, the appellate court let the ruling stand, quoting the view of the Supreme Court that: . . . a defendant’s statements made in response to proper cross-examination reasonably suggestedbythe defendant’s direct examinationare subject tootherwise proper impeach- ment, albeit by evidence that has been illegally obtained and that is inadmissible on the government’s direct case, or otherwise, as substantive evidence of guilt. 22 The twisted judicial logic here is that the evidence is not being used in the second case to show the defendant’s guilt but to show that he lied in his testimony. But if evidence of possession of crack is pertinent and admissible to show that the defendant is a liar, why is it not admissible and pertinent to showing that the defendant was probably in possession of the illicit drug? I said earlier that the United States is unique among common law nations in insisting that neither the evidence found in an illegal search nor other evidence to which it led (for example, bank accounts, names of associates) can be used as a part of the case-in-chief against someone charged with a crime. It was not always so. Until the early years of the twentieth century, the predominant theory among the American judiciary was that evidence relevant to the guilt or innocence of an accused person could be presented at trial, regardless of how it came into police possession. In 1841, for instance, the Supreme Judicial Court of Massachusetts was grappling with a case in which the police had seized some incriminating papers of a suspect without a warrant. The defendant argued that such evidence could not be used against him. The judgment of the court was unambiguous: Admitting that the . . . materials were illegally seized, still this is no legal objection to the admission of them in evidence. If the search warrant were illegal, or if the officer serving the warrant exceeded his authority, the party on whose complaint the warrant 21 100 F.3d 1, at 6 (1st Cir. 1996). 22 U.S. v. Havens, 446 U.S. 620 (U.S. 1980). P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 Confessions, Poison Fruit, and Other Exclusions 189 issued, or the officer, would be responsible for the wrong done; but this is no good reason for excluding the papers seized as evidence, if they were pertinent to the issue, as they unquestionably were. When papers are offered in evidence, the court can take no notice how they were obtained, whether lawfully or unlawfully. . . . 23 Granting that the documents should not have been illegally seized, the court pointedout that the defendant alreadyhada remedyfor the wrongdone tohimby the illegal search and seizure: He could sue the police who conducted the illegal search for trespass and wrongful seizure. What the defendant could not do, argued the court, was to block the admission of clearly relevant and inculpatory evidence in the case against him. This doctrine had been the prevailing theory for several centuries of English and American common law. Evidence was evidence, and how the court or the police came to possess it had nothing to do with its admissibility. (The situation with illegally obtained confessions, as opposed to physical evidence, was more complex, but we can ignore that for our purposes.) As recently as 1904, the Supreme Court found a lower court’s admission of illegally seized evidence to be fully constitutional. 24 However, a famous Supreme Court ruling in 1914 changed all this at the federal level. The Court held that the admission of tainted, illegally obtained evidence at trial would be tantamount to a judicial endorsement of police illegality. That case, Weeks v. U.S., was the principal source of the Exclusionary Rule. (In 1961, the Supreme Court extended the application of the Exclusionary Rule to state proceedings as well as federal trials. 25 ) As the justices wrote in Weeks: The tendency of those who execute the criminal laws of the country to obtain conviction by means of unlawful seizures and enforced confessions, the latter often obtained after subjecting accused persons to unwarranted practices destructive of rights secured by the Federal Constitution, should find no sanction in the judgments of the courts which are charged at all times with the support of the Constitution and to which people of all conditions have a right to appeal for the maintenance of such fundamental rights. 26 A decade later, Justice Louis Brandeis staked out the same terrain, arguing the importance of the Exclusionary Rule for preserving judicial integrity: In a government of laws, existence of the government will be imperiled if it fails to observe the law scrupulously. . . . Crime is contagious. If the government becomes a lawbreaker, it breeds contempt for the law; it invites every man to become a law unto himself; it invites anarchy. 27 23 43 Mass. 329, at 337 (1841). (Emphasis added.) 24 Adams v. New York, 192 U.S. 585 (1904). 25 The case in question was Mapp v. Ohio, 364 U.S. 643 (1961). 26 Weeks v. U.S., 232 U.S. 383, at 393 (1914). 27 Olmstead v. U.S., 277 U.S. 438 (1928). P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 190 flawed rules of evidence and procedure We will turn in Chapter 9 to assess these moral and political sentiments. Before we do, it is important toget some epistemic facts straight. The maineffect of the Exclusionary Rule is to prevent relevant evidence, however inculpatory, from being used at trial if it was illegally seized. “Illegal,” as used here, usually means evidence seized without a valid warrant or, more generally, evidence seized from the defendant when there was no probable cause to believe that a police search would yield evidence of criminal activity. From its introduction, the Exclusionary Rule has generated fierce debate between civil libertarians and the tough-on-crime faction. While the first argue that citizens must be protected from arbitrary and unjustified intrusions into their homes, cars, and other private spaces, the second maintains that evidence is evidence, and that the bad guys should not get out of jail free simply because of a constable’s error of judgment. It is perfectly clear that the exclusion of germane, inculpatory evidence, on account of the way in which it was obtained, significantly increases the likelihood of false acquittals. The courts universally acknowledge this fact. The Supreme Court itself minced fewwords on the matter when it remarked in 1973 that the protections against searches and seizures in the Fourth Amendment “have nothing to do with promoting the fair ascertainment of truth at a criminal trial.” 28 Infact, this is toogenerous since those protections are not neutral –as the “nothing to do with” language suggests – with respect to “the fair ascertainment of truth” but in fact undermine it. As the classical evidence text, McCormick on Evidence, puts it: “The fact that the evidence excluded [by exclusionary rules] is not only relevant and competent but also highly reliable increases the difficulty of justifying” such rules. 29 How, then, are such rules currently justified? There is a range of standard defenses for the Exclusionary Rule, some of which have already been hinted at in passages quoted in this chapter. Basically, they fall under three headings: a) that evidence seized by an act violating the rights of the defendant must be excluded because otherwise the courts would not be protecting the rights of the accused; b) that the courts must be seen to be paragons of legal integrity, not complicit in acts that flout the law; and c) that the refusal of courts to admit evidence seized illegally by the police will send a powerful and salu- tary message to police officers that, if they hope to have their evidence used in convicting a felon, they must be scrupulous about how they acquired that evi- dence. 30 Chapter 9 will consider these nonepistemic arguments in considerable detail. 28 Schneckloth v. Bustamonte, 412 U.S. 218, at 242 (1973). 29 Charles T. McCormick, McCormick on Evidence, §165. 30 The Supreme Court wrote inMichiganv. Tucker: “The deterrent purpose of the exclusionary rule necessarily assumes that the police have engaged in willful, or at the very least negligent, conduct which has deprived the defendant of some right. By refusing to admit evidence gained as a result of such conduct, the courts hope to instill in those particular P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 Confessions, Poison Fruit, and Other Exclusions 191 For the purposes of our thought experiment, it is sufficient to note that a sweeping rule of inadmissibility like this one could find no place in a set of evidence rules that were principally designed to promote truth seeking. The existence of such rules gives the lie to the claim that the current system of criminal justice seeks, above all, to find out the truth about a crime. Two final observations about the Exclusionary Rule are in order. The first concerns itself with what might at first seem to be an economic issue rather than an epistemic one. As anyone who has followed one or another celebrity trial knows, attorneys invest a great deal of their time and resources in fighting, during preliminary hearings, about the admission of evidence. There are no rules of evidence that generate more of this sort of controversy than debates about alleged Fourth Amendment irregularities. It is likewise true that the dock- ets of appellate courts are full of appeals of rulings issuing from preliminary hearings about the admission of evidence under this rule. Clearly, if this rule were replaced by a simple rule saying that all relevant evidence was admissible, these delays and expenses would be drastically curtailed. Why does that matter from a truth-seeking perspective? Simply because the time and financial resources of prosecutors, defense counsel, and appel- late courts are acutely limited. Energies currently going into wrangling about the admissibility of clearly relevant evidence could be profitably invested in a more thorough development of the evidential base that each side brings to trial. Seeking additional witnesses, if they exist, or investing more in exploring the forensics of the case would often make a positive contribution to the ability of the jury to find out the truth about a crime. In the case of appellate review, higher court judges – if freed from the onerous task of adjudicating disputed rulings about admissibility – would be able to review far more cases than is now possible and, with any luck, they would be able to catch and correct more mistakes than they now do. In short, reliable verdicts would be more likely to ensue if human and economic resources – currently devoted to largely idle debates about whether the triers of fact should or should not see indisputably relevant and reliable evidence – were devoted to reducing the likelihood of an erroneous verdict. There is one further argument to consider before we leave this subject: While critics of the Exclusionary Rule (like me) point to the costs that are exacted when relevant but illegally seized evidence is excluded and insist that illegal behavior can be remedied via other channels (for instance, civil and criminal charges against the offending officers), defenders of the rule sometimes point out that if those alternative remedies really worked, then illegal evidence would not be coming into the hands of the police and the Exclusionary Rule would never have to be invoked. More important, they say, is the fact that the “costs” attributed investigating officers, or in their future counterparts, a greater degree of care toward the rights of an accused” (417 U.S. 433, at 447 [U.S. 1974]). P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 192 flawed rules of evidence and procedure to the Exclusionary Rule are exactly the same as the “costs” that would be paid in a society with no Exclusionary Rule but scrupulously legal police officers. Here is how Yale Kamisar frames the argument: Doesn’t a societywhose police complywiththe FourthAmendment in the first place “pay the same price” as the society whose law enforcement officials cannot use the evidence they obtained because they violated the Fourth Amendment? Don’t both societies convict fewer criminals? 31 The answer to Kamisar’s first question is, I think, negative. It is one thing if Jones is acquittedbecause the police, havingbrokennolaw, fail tohave sufficient evidence to persuade a jury of his guilt. It is quite another if, having illegally secured very incriminating evidence against Jones, the court refuses to allowits use. In both cases, as Kamisar says, Jones will be acquitted. In that respect, the two societies Kamisar describes are similar. Still, in the second case, it becomes a matter of public knowledge that the state, having committed an illegal act at Jones’s expense, now has powerful evidence of his guilt. Knowing that, the court still refuses to let the jury know what it knows. In the circumstances, the general public perceives a gross miscarriage of justice. The injustice they associate with an illegal search of the guilty Jones’s home is seen as minor in comparison with the injustice of permitting an obviously guilty felon to escape conviction by refusing to let the jury know what the state knows about his guilt. By failing to utilize any sort of balancing test in these circumstances, the courts seemto presume a moral equivalence between (for example) committing an illegal search and committing murder (or whatever other crime with which Jones is charged). The murderer goes free because the police lacked probable cause to search his warehouse, where they found compelling evidence of his bad deeds. In that important sense, the two societies Kamisar describes are paying quite different costs. For such reasons, it seems to me that a liberal government has to walk a narrow line, vigorously discouraging police illegality while not indulging in “acquittal by technicality” when illegally obtained evidence – especially if the illegality was not especially offensive – does come into its hands. There is another important difference that Kamisar ignores. Many of the evidence-acquiring practices that lead to exclusion do not involve acts on the part of the police that are criminal, even though they may violate rights. For instance, it is not a crime if the police fail to read a detainee his Miranda 31 Yale Kamisar, Confessions, Search and Seizure and the Rehnquist Court, 34 Tulsa L. J. 465, at 486n (1999). (Emphasis in the original.) Asimilar argument can be found in Richard Posner, An Economic Approach to the Law of Evidence, 51 Stan. L. Rev. 1477, at 1533 (1999). P1: JZZ 0521861667c07 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:9 Confessions, Poison Fruit, and Other Exclusions 193 rights immediately upon his arrest. It is not a crime if the police, conduct- ing a frisk search, pull a packet of drugs from someone’s pocket. It is not a crime if the police manage inadvertently to overhear a criminal’s confession to his priest. Yet, the courts will refuse to admit such evidence just as surely as they would if the police obtained evidence by committing some egregious criminal act. P1: JZZ 0521861667c08 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:12 8 Double Jeopardy and False Acquittals: Letting Felons and Judges off the Hook? Whenever, and by whatever means, there is an acquittal in a criminal prosecution, the scene is closed and the curtain drops. – U.K. Solicitor General 1 Insulation [fromappealing an acquittal] is an arbitrary windfall to the guilty, not a carefully structured scheme to protect the innocent. . . . A defendant has no vested right to a legal error in his favor. – Akhil Amar 2 The community incurs an incalculable expense when the vast machinery con- structed to bring criminals to justice can be felled by the simple error of a single unreviewable judge. – Scott Shapiro 3 The Core Problem The preceding four chapters have examined some rules of evidence that are epistemically counterfeit. We turn now to look at some procedural rules that pose similar kinds of problems. I will focus on what is probably the single largest source of avoidable error in the judicial system: the asymmetry of the appellate process. As we turn from evidence to procedures, our philosophical criteria of eval- uation must likewise shift. Relevance is no longer the guiding metaprinciple since procedural rules donot concernthemselves withthe admissionof evidence per se. As I indicatedinChapter 5, whenit comes tothe evaluationof procedures, 1 From Duchess of Kingston’s Case 2 Smith’s LC (13th ed.) 644, at 647–8, [1776]. Quoted by Lord Patrick Devlin, Trial by Jury 75 (1956). 2 Akhil Amar, Double Jeopardy Law Made Simple, 106 Yale L. J. 1807, at 1844 (1997). 3 Scott Shapiro, Reviewing the Unreviewable Judge: Federal Prosecution Appeals of Mid-Trial Evidentiary Rulings, 99 Yale L. J. 905, at 906 (1990). 194 P1: JZZ 0521861667c08 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:12 Double Jeopardy and False Acquittals 195 the appropriate meta-rules stress the importance of both maximizing the likeli- hood that a jury will reach a valid verdict and ensuring that the systemis capable of catching and correcting (some of ) the mistakes that it does make. It is clearly the second sort of issue that will concern us here, since appellate procedures typically determine what happens once the trier of fact has already reached a verdict, whether it is valid or not. There are basically two sorts of mistakes in play here. In one case, the jury may draw a conclusion about the innocence or guilt of the defendant that fails to square with the evidence. I earlier called this kind of mistake an invalid verdict. In the other, the judge, who presides over the course of the trial, may have deviated significantly from the rules of evidence and pro- cedure. He may have excluded witnesses or evidence that the rules insist on including, or, much more likely, he may have included evidence that the existing rules require to be excluded. Alternatively, he may have given faulty instructions to the jury concerning anything from the legal definition of the crime in question to the standard of proof. And so on. Any one of these mistakes by the judge might be responsible for the jury coming to a wrong verdict. In an epistemically ideal legal system, machinery would be in place for catching each of these types of mistakes and correcting them. The key question in this chapter will be the following: Howclose does the current American legal system come to this ideal? The system in U.S. federal courts has a variety of ways in which it tries to prevent or to correct such mistakes. One such mechanism belongs to the trial judge. More or less anywhere in the course of a trial, she can decide that the prosecution’s case is too weak to justify a conviction and can dismiss the charges against the defendant. This can even occur after the jury has announced a guilty verdict. The trial judge’s power to vacate a jury’s conviction is a powerful instrument for helping to ensure that false convictions do not go forward – always supposing that the judge’s determination that the prosecution’s case has not reached the level of BARD is sound. What a trial judge cannot be relied on to do, however, is to identify the errors that he himself may have committed in the course of the trial. That is why the bulk of appeals in criminal cases focus on supposed errors made by the judge conducting the case. Where appellate courts believe that there may be merit in such claims, they will review a case, opening the way to setting aside the existing verdict and either ordering a new trial or simply replacing a conviction by an acquittal. If someone convicted of a crime files an unsuccessful appeal, it is generally open to him to carry the appeal at least one step further up the system, hoping that a higher appellate court than the one that rejected his initial appeal will find merit in his argument. These various mechanisms for imposing checks on the system are highly commendable. In principle, they enable the system to catch many of the errors P1: JZZ 0521861667c08 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:12 196 flawed rules of evidence and procedure that would otherwise go undetected and uncorrected. Still, there is one grand flaw in this otherwise rosy scenario. The functioning of the appellate review mechanism is appallingly one-sided. While the trial judge can reverse a jury’s guilty verdict, on the grounds that the evidence does not justify a finding of guilt, she can do nothing to shift a jury’s verdict of acquittal, however persuaded she may be that the evidence of the defendant’s guilt was overwhelming. The same asymmetry ascends right up the hierarchy. If a defendant is convicted, he has the right to appeal the verdict to a higher court. If he is acquitted, the prosecutor is not allowed to seek redress from an appellate court, however foolish the jury’s verdict or however outrageous the errors committed by the trial judge in the course of the trial. In short, acquittals by a jury or by a judge are the final word on the matter. The unmistakable upshot of this asymmetry is that many erroneous con- victions will be put right while erroneous acquittals are untouchable. Instead of using the appellate machinery to reduce whatever type of error occurs (as epistemological interests would demand), the justice system proclaims unam- biguously that it is interested in reviewing and correcting only false convictions, while leaving in place whatever false acquittals may arise from the decisions of incompetent judges or inferentially challenged jurors. One must wonder aloud why this is so. It cannot be because the system believes either that juries never issue invalid acquittals or that judges commit serious mistakes only in trials that issue in a finding of guilt. The usual answer to this question, or at least the initial answer, is that both the history of the common law and the Fifth Amendment to the Constitution make it so. The common law, we are told, has never permitted appeals of acquittals, and the double jeopardy protection afford by the Fifth Amendment to the Constitution precludes a second trial for any criminal defendant. Both explanations can be misleading. While it is true that the common law traditionally did not permit appeals of acquittals, it is likewise true that it did not permit appeals of convictions either. The great eighteenth-century English jurist Sir William Blackstone repeat- edly referred to “this universal maxim of the common law of England, that no man is to be brought into jeopardy of his life, more than once, for the same offence.” 4 This meant absolutely no retrial. If common law tradition were suf- ficient to settle the question, then we should not permit appeal of either sort of verdict. This common law doctrine was enshrined as a part of the Fifth Amendment to the U.S. Constitution. “No person,” reads the relevant text, “shall be subject for the same offense to be twice put in jeopardy of life or limb.” Taken literally, this clause (as Blackstone’s earlier formulation) protects citizens from repeated 4 4 Blackstone, Commentaries on the Laws of England, 335. P1: JZZ 0521861667c08 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:12 Double Jeopardy and False Acquittals 197 prosecution only with respect to capital crimes, since “life or limb” was a standard phrase in the eighteenth century for the death penalty. (The “limb” language was a reference tothe practice of quarteringthe bodyof the condemned as a part of the death penalty.) During the early years of the Republic, the amendment meant what it said: Anyone who had been tried by a federal jury and convicted or acquitted of a capital crime could not be tried a second time. 5 (Double jeopardy did not always apply to trials in state courts.) This principle was not, it needs to be stressed, aimed specifically at prohibiting appeals of acquittals. When the Bill of Rights was passed, the mechanism of judicial appeal as we know it had not been invented, and so there were no appeals as we would understand that term, either of convictions or acquittals. Such mistakes as the system might make were correctable chiefly by the use of pardons or other extrajudicial remedies. In the course of the nineteenth century, however, the doctrine of double jeopardy underwent several transformations. In the 1870s, double jeopardy protection was extended from capital crimes to lesser felonies and even to mis- demeanors, despite apparently clear constitutional language to the contrary. 6 By then, it had become vividly clear to lawmakers and judges that jurors sometimes falselyconvicteddefendants, convictions that were oftennot set aside bythe trial judge (as was his right). To protect against that outcome, appellate machinery was put in place in 1889 letting federal defendants in capital cases appeal con- victions to higher courts, provided that they could identify serious errors made during their trials. If the appellate court agreed that a serious procedural breach had occurred, a new trial would be ordered and the defendant would have a second shot at an acquittal. In 1891, the right to appeal was extended to all federal defendants convicted of major felonies. So far, so good. This was clearly a change that enhanced the truth-seeking capacity of the criminal justice system by enabling it to catch and correct many false convictions that would have gone undetected under the previous no-appeal regime. Unfortunately, lawmakers and judges at the time strenuously resisted the idea that one should allow appeals of acquittals, even when monumental 5 Prominent nineteenth-century cases that argued the prohibition of retrials following either acquittals or convictions are People v. Comstock (8 Wend. 549 [1832]); U.S. v. Gilbert (25 F. Cas. 1287 [1834]). 6 The crucial case extending double jeopardy protection to all federal crimes was Ex parte Lange (1873). Oliver Wendell Holmes regarded the extension of double jeopardy to mis- demeanors as a grave mistake: “If . . . the constitutional prohibition [of double jeopardy] should be extended to misdemeanors, we shall have fastened upon the country a doctrine covering the whole criminal law, which, it seems to me, will have serious and evil conse- quences. At the present time in this country there is more danger that criminals will escape justice than that they will be subjected to tyranny” (Kepner v. U.S., 195 U.S. 100, at 134 [U.S. 1904]). P1: JZZ 0521861667c08 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:12 198 flawed rules of evidence and procedure errors had occurred. 7 This, I will argue, was patently a mistake and for more reasons than the obvious one. It also became the practice that directed acquittals from a judge or verdicts of not guilty from a bench trial, and not merely jury verdicts, were protected against appeal on double jeopardy grounds. 8 Finally, in the 1960s, the Supreme Court insisted that double jeopardy protection governed state as well as federal trials. 9 Since then, double jeopardyhas governedvirtually all criminal trials in the United States. Through these many convolutions, it is easy to lose sight of the fact that the original motive of the principle of double jeopardy – ensuring that trial verdicts were final and definitive – was quietly abandoned. Better, it was abandoned in a decidedly one-sided fashion. What was retained – indeed what became the core of the modern theory of double jeopardy – was the doctrine that acquittals, like all verdicts before the late-nineteenth century, are not subject to judicial review. Convictions, by contrast, came to be seen as provisional way stations, subject to review, rethinking, and reversal by higher courts, when jurists could detect serious irregularities in the trial that worked against the interests of the defendant. (Bear in mind that most felony convictions are now appealed. 10 ) Double jeopardy was thus converted froma doctrine that insisted on the integrity and finality of the trial verdict – whatever that verdict was – into a policy that now specifically excludes the possibility of appealing jury or bench acquittals or bench dismissals. Everyone acknowledges that false acquittals routinely occur and that, if the law permitted it, some and perhaps many of those errors could be corrected upon review, just as incorrect convictions are routinely set aside. The false acquittals now under discussion are not those that arise inevitably from the skewing of BARD in the defendant’s direction. Those are built-in features of the system. What will concern us here, rather, are those false acquittals arising from such factors as inappropriate judicial exclusion or inclusion of evidence 7 The first case in which the Supreme Court suggested that acquittals specifically could not be appealed was U.S. v. Sanges, 144 U.S. 310 (1892), shortly after appeals of convictions were introduced. The case that finally established that doctrine as a constitutional right was Kepner v. U.S. (195 U.S. 100 [1904]). In 1978, the Supreme Court insisted that the ban on appeals of acquittals applied even when the acquittal was due to “egregiously erroneous” rulings by the trial judge (Sanabria v. U.S., 437 U.S. 54, at 59 [1978]). 8 As the Court of Appeals of the Second Circuit noted in Lynch v. U.S., “the Supreme Court has made it equally clear that the rule prohibiting the revisitation of acquittals applies as much to trials in which a judge decides the facts as it does to a prosecution tried to a jury” (162 F.3d 732, at 738 [1998]). 9 The relevant case is Benton v. Maryland, 395 U.S. 784 (1969). The Supreme Court said, “we today find that the double jeopardy prohibition of the Fifth Amendment represents a fundamental ideal in our constitutional heritage, and that it should apply to the States through the Fourteenth Amendment” (at 794). 10 According to the Bureau of Justice Statistics, four out of every five defendants (81 percent) convicted at trial in federal courts appeal their convictions (Bureau of Justice Statistics, Federal Criminal Appeals, 1999 [April 2001], p. 4). P1: JZZ 0521861667c08 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:12 Double Jeopardy and False Acquittals 199 and testimony, irregular trial procedures, inaccurate instructions from bench to jurors, corrupt jurors and witnesses, sloppy juror inferences, and similar errors. 11 The historyof double jeopardylawis the storyof howa party-neutral doctrine that originally allowed no appeals of verdicts has been transformed into the defendant-biased idea that those who have been convicted may be retried while those who have been acquitted cannot. It is some of the contemporary twists and turns implied by this history that will form the focus of this chapter. Appeals of Acquittals: A No-Brainer I will begin by being a bit more concrete and specific about the current state of things. Shortly before a criminal trial begins, there is routinely a hearing concerning the admissibility of evidence. Both prosecution and defense can take exception to the judge’s pretrial rulings about admissibility and can appeal those decisions to higher courts. Both sides are on an equal footing to this preliminary point. As soon as the jury is sworn, however, the situation changes drastically. Jeopardy suddenly “attaches,” in the jargon, and the asymmetries begin to multiply. Decisions from the bench about the admissibility of evidence not presented prior to trial can be appealed by the defense when there is a conviction but not by the prosecution when there is an acquittal. If the judge decides (as is her right) to dismiss the charges against the defendant on the grounds that the state’s case is too weak to convict, this decision cannot be appealed and the defendant cannot be retried on the same charge. If the judge misinstructs the jury about the law or the standard for conviction, the defense can use this as a basis for appeal. The state cannot. (Leaving bribery aside, the only circumstance in which the state may appeal an acquittal, and it is extraordinarily rare, is when the jury has voted to convict the defendant but the judge sets that verdict aside and acquits him. 12 ) The reach of a judge’s potential errors goes well beyond botching jury instructions and making erroneous rulings about evidence. She has the power to penalize the prosecution for a serious breach of legal protocol by immediately dismissing the case against the defendant, however incriminatory it may be. Such a decision cannot be appealed and the defendant cannot be retried. (By contrast, a judge cannot penalize outrageous conduct by defense counsel by declaring the defendant guilty.) An error by the prosecution or defense is not 11 According to one well-known study of appeals, almost one-third (29 percent) of all capital cases reversed on appeal are due to erroneous instructions in the law from the judge to the jury (James Liebman et al., A Broken System: Error Rates in Capital Cases, Columbia Law School, June 12, 2000, section iv). 12 In People v. Aleman, an Illinois court ruled that an acquittal obtained by bribing the trial judge did not place the defendant in jeopardy. Therefore, it ruled, the defendant could be retried for the same crime. P1: JZZ 0521861667c08 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:12 200 flawed rules of evidence and procedure even necessary to trigger a judge’s decision to intervene and dismiss a case, thereby precluding any appeal by the prosecution and further prosecution of the defendant on the same charge. Because this power of the judge is so awesome and absolute (being unreview- able), it is perhaps worth recalling the particulars of the case that established this doctrine. It was Fong Foo v. U.S., and was decided in 1962. 13 The case involved a claim of criminal conspiracy and fraud against the officers of a pri- vate firm providing various types of meteorological equipment on contract to the government. The state alleged that the officers of the company knew the equipment to be defective and substandard. The prosecution began its case by presenting several preliminary witnesses. Halfway through the testimony of the third witness and long before the prosecutor had laid out his case-in-chief, the judge intervened and dismissed the case against Fong Foo. The Circuit Court to whom the prosecution appealed summarized what had happened: [The judge] abruptly terminated the Government’s case . . . long before the Government had had an opportunity to show whether or not it had a case; and, moreover, he did so in ignorance of either the exact nature or the cogency of the specific evidence of guilt which Government’s counsel said he had available and was ready to present. 14 All this notwithstanding, the Supreme Court insisted that the prosecution had no right to appeal the judge’s directed dismissal, even though that dismissal may well have been ill advised and even though it precluded any further trial of Fong Foo on the charges in question. It is bad enough that judges have the power to silence the prosecution in midsentence, as it were; it is incomparably worse that there are no checks on this power. Still, this is the precisely sort of epistemic silliness to which the asymmetry of appeals leads. Why should the opportunity for appeal, for checking that a verdict was legitimately arrived at, be so one-sided? The immediate answer is that the appeal of an acquittal might well require a newtrial (specifically if the appellate court found serious errors in the judge’s conduct of the trial). But a second trial, so the story goes, is precluded by the constitutional protection against double jeopardy. One could be forgiven for thinking that the double jeopardy clause, literally construed, should preclude the appeal of convictions as much as acquittals, since either type of event could issue in what the person on the street (and probably most defendants) would regard as a second trial. The standard response is that, in filing an appeal of her conviction or in demanding that the judge declare a mistrial, the defendant agrees to waive her constitutional right not to have to undergo a second trial. 15 But, so the story continues, an acquitted 13 369 U.S. 141 (1962). 14 286 F.2d, at 562–3 (1962). 15 As the Supreme Court put it, “a motion by the defendant for mistrial is ordinarily assumed to remove any barrier to reprosecution” (U.S. v. Jorn, 400 U.S. 470, at 485 [1971]). To my knowledge, the Supreme Court first endorsed this doctrine that a convicted defendant may P1: JZZ 0521861667c08 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:12 Double Jeopardy and False Acquittals 201 defendant will most certainly not waive that right so that the prosecution can appeal an acquittal. This argument can be readily countered by pointing out that (as Oliver Wendell Holmes famously argued) it is not wholly clear precisely when a trial as such has ended or, come to that, when a “real” trial has taken place. Holmes believed that a trial has not properly ended until all the relevant appeals and, if granted, retrials have been exhausted. Consider a very simple example of the problem: Suppose that Wilson is on trial for rape. The prosecution’s case is very persuasive. Just before sending the jury off to deliberate, the judge explains to the jurors that they may not convict unless they are certain “beyond all doubt” of Wilson’s guilt. The jury acquits. Now, it is perfectly clear that the judge has made a terrible mistake by misdefining the standard for conviction. There is plenty of reason, let us suppose, to suspect that the judge’s mistaken instruction caused the anomalous verdict. Has a trial just occurred? Not if we understand by a “trial” a judicial inquiry in which BARD is the standard for conviction. Even if it is a trial, are we obliged to say that it has ended when the jury delivers its verdict, however flawed the trial and however suspect the verdict may be? Wilson has a right, let us suppose, to be tried fairly only once, but arguably the process that he just suffered through was not a fair trial. If this seems strained and scholastic, remember what happens when a mistrial is declared. A judge, perhaps because she believes that the prosecution or the defense attorney has flagrantly compromised the integrity of the trial by some statement or tactic (perhaps the prosecutor mentioned the prior criminal record of the defendant), can declare a mistrial. Likewise, if the jury cannot reach consensus about a verdict (which happens in about one in nine trials 16 ), the judge will declare a mistrial. When that happens, the jury is excused and the defendant faces a newtrial. This doesn’t violate the principle of double jeopardy because of the legal fiction that the defendant was not tried the first time around but was “mistried.” Now, if it is acceptable to try a defendant a second time when a mistrial has been declared because of misconduct by the prosecutor, defense attorney, or jury, or when the jury cannot agree on a verdict, why is it impossible, when the bench has made gross errors, to try the acquitted defendant a second time? Clearly, if our principal interest were in error reduction, it would not require much ingenuity to redefine the conception of a trial so that it was not considered properly ended until all legal appeals arising from the case had been exhausted by whichever side wished to pursue them. What is it, then, that convinces so many observers (especially in the defense bar) that asymmetrical double jeopardy must be retained, however much it waive his double jeopardy rights by appealing his conviction in Trono v. U.S. (199 U.S. 521 [1905]). 16 In a study based on the Los Angeles courts in the 1990s, jury trials resulted in a hung jury 11 percent of the time (29 U. West. L.A. Rev. 215 [1998]). P1: JZZ 0521861667c08 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:12 202 flawed rules of evidence and procedure thwarts the interests of truth seeking? The U.S. Supreme Court has made it clear on countless occasions that double jeopardy functions to promulgate errors favoring the defendant. Nevertheless, the Court holds the constitutional prohi- bition against second trials to be ironclad. As the justices wrote in a classic double jeopardy case in 1984: The constitutional protection against double jeopardy unequivocally prohibits a sec- ond trial following an acquittal, [because the] public interest in the finality of criminal judgments is so strong that an acquitted defendant may not be retried even though the acquittal was based upon an egregiously erroneous foundation. . . . If the innocence of the accused has been confirmed by a final judgment, the Constitution presumes that a second trial would be unfair. 17 This passage means precisely what it says. Even if a witness were suborned by an accomplice of the defendant, even if the judge made incredible errors, an acquittal will still stand. What, we need to ask, would be “unfair” about a second trial, supposing as the Court does that the defendant was acquitted for “egregiously erroneous” reasons in the first? Why would it be “unfair” to refuse to let a defendant off if he won acquittal in a trial that was wrongfully conducted? To the contrary, it seems unfair to society to let a patently flawed acquittal stand just as it is unfair to let a flawed conviction stand. Apart fromthe fairness issue, defenders of double jeopardy have used other equally dubious arguments. Here is how Martin Friedland, author of a book on double jeopardy, formulates the core case against retrial: [If two trials were permitted,] an innocent person will not have the stamina or resources to fight a second charge. And, knowing that a second proceeding is possible, an innocent person may plead guilty at the first trial. But even if the accused vigorously fights the second charge he may be at a greater disadvantage . . . because he will normally have disclosed his complete defense at the former trial. . . . The prosecutor can study the tran- script and may thereby find apparent defects and inconsistencies in the defense evidence to use at the second trial. 18 This is, of course, a series of non sequiturs. If it were true that innocent per- sons “lack the stamina or resources” to endure a second trial, then we should prohibit judges from declaring mistrials, since those usually require the defen- dant to start over again from scratch. If it were true that many defendants do not have the stomach for a second bout, we should not expect to see more than 80 percent of all major felony convictions appealed, 19 since winning an appeal ordinarily entails retrial. As for the suggestion that the innocent defendant, faced with the prospect of a second trial, might plead guilty immediately to 17 Rodrigues v. Hawaii, 469 U.S. 1078, at 1079 (1984) (emphasis added). 18 Martin Friedland, Double Jeopardy 4 (1969). 19 This figure comes from Kate Stith, The Risk of Legal Error in Criminal Cases, 57 U. Chi. L. Rev. 1, at n. 39 (1990). P1: JZZ 0521861667c08 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:12 Double Jeopardy and False Acquittals 203 avoid such a fate, this seems wildly implausible on its face. As unpleasant as one trial, let alone two, must be, that prospect has to pale in comparison with the unpleasantness associated with most criminal punishments. Most telling of all is the last argument Friedland reprises. It voices the worry that, if retrial were allowed (which of course it already is after a mistrial or reversed conviction), then the prosecution would know the second time around much about the defense strategy and might even be able to detect “inconsis- tencies in the defense evidence.” If we have the slightest inclination to see the purpose of a trial as getting at the truth, we would surely welcome rather than deplore the opportunity for either side to be able to expose incoherencies in the case presented by the other side. That is precisely what the adversarial systemis designed to foster. Only a zealous member of the defense bar, single-mindedly concerned with his client winning rather than with whether truth emerges, could expect the rest of us to take seriously the argument that a prime advantage of a system permitting only one trial is that it allows defense inconsistencies to go undetected and unchallenged. Note, again, what happens when a mistrial is declared or when a conviction is reversed. Typically, a new jury is impaneled and the procedure begins again from scratch. The defendant loses quick finality and is made to bear additional expense. The prosecution nowknows the defense strategy. Despite all that, dou- ble jeopardy concerns do not protect the defendant from retrial after a mistrial. Why should those concerns suddenly kick in when a defendant is acquitted in a trial riddled with errors favorable to him? Another argument commonly heard in defense of double jeopardy involves the idea that if the state were not limited to one shot at a defendant, then it could harass someone indefinitely, bringing him or her to trial over and again on the same charge until it either found a jury that would convict or drove the defendant mad. That is a red herring in this context. The proposal under consideration is not that the state should be able to recharge someone with a crime indefinitely many times. It is simply that an acquittal, like a conviction, should be open to judicial review. If the appeal of an acquittal is unsuccessful – that is, if no serious errors occurred in the original trial – then that would put an end to the state’s power to harass the defendant, as least where the crime in question is concerned. A final, familiar argument in defense of the prohibition against appeals of acquittals concerns the jury’s right to nullify. Although jurors cannot be informed that they have such a right, and although it is probably more a power than a right per se, it is clear that under current law jurors can deliver a verdict of acquittal even when the evidence produced at trial is overwhelmingly incul- patory. 20 That power, it is suggested, would be put at risk if acquittals could be 20 The Supreme Court has described jury nullification as the “assumption of a power” that a jury has “no right to exercise” (Dunn v. U.S., 284 U.S. 390, at 393 [1932]), and as “the P1: JZZ 0521861667c08 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:12 204 flawed rules of evidence and procedure reversed. This defense of double jeopardy likewise misses the point. If judicial review is limited (as it usually is) to catching legal errors made during a trial rather than with second-guessing the motives or reasoning of the jurors, jury nullification would be largely unaffected by permitting appeals of acquittals. Reversal of a jury’s verdict could be ordered only if the appellate court iden- tified serious procedural errors in the trial itself. (Nullification of a verdict is decidedly not a procedural error. It is an invalid verdict.) More generally, no acquittal could be turned into a conviction simply because the appellate bench believed the accused was guilty. The extent of the appellate court’s power would be to order a new jury trial. If the law that the first jury nullified were generally unpopular with the public, the second jury could nullify again. Note that none of the arguments against appeals of acquittals reprised thus far offers even a hint of a concern for issues of truth and falsity. I have been able to find only one that attempts to relate double jeopardy protection to questions of truth finding and error reduction. It comes from Justice Hugo Black in a well-known split decision from the Supreme Court in 1957. He wrote: The State with all its resources and power should not be allowed to make repeated attempts to convict an individual for an alleged offense . . . [thereby] enhancing the pos- sibility that even though innocent he may be found guilty. 21 In our terminology, Black is arguing that permitting retrial after a reversed acquittal would inevitably entail more false convictions than occur under the current regime. He is surely right about that. All it takes is one mistake by an appellate court sending a true acquittal back for retrial and a complementary false conviction by the jury at the new trial. There is every reason to expect that this concatenation of errors would be uncommon, but it would surely occur from time to time. Still, this has by now become a familiar and well-worn stratagem. We have already seen too many examples of it earlier in this book. Someone argues against a particular rule or policy by noting that it has a tendency, however slight, to increase the rate of false convictions. That alone is taken as sufficient to discredit the policyin question. We have also seen, however, that suchan argu- ment, standing alone, is woefully incomplete. When appraising any judicial rule or practice, we need to estimate its aggregate effect on truth seeking and error reduction. Is a given policy likely to vastly reduce the number of false acquittals while bringing only a minor increase in false convictions? Such a policy would surely be wise to adopt. The appeal of acquittals comes, I have argued, squarely in this category. What Justice Black should have asked is whether permitting appeals of acquittals would, in aggregate, increase or decrease the frequency of unreviewable power of a jury to return a verdict of not guilty for impermissible reasons” (U.S. v. Powell, 469 U.S. 57, at 63 [1984]). 21 Green v. U.S., 355 U.S. 184, at 187–8 (1957). P1: JZZ 0521861667c08 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:12 Double Jeopardy and False Acquittals 205 false verdicts. There is no room for dispute about the answer to that one: The overall error rate would drop sharply if acquittals were subject to the same sort of review as convictions. To this point, my quarrel with the appellate asymmetry has been that it catches fewer errors in individual cases than a more even-handed policy would. But there is a larger worry generated by a policy that allows appeals only of convictions and not of acquittals. Unlike the problem we have just examined (where everyone explicitly acknowledges that many false acquittals that could be caught on review slip through the system), this second concern focuses on a situation that is less obvious, and certainly less intentional. It is a clich´ e that, whenever any policy is put in place, it will eventually reveal itself to have multiple unintended, and unforeseen, consequences. The policy of asymmetric appeals is no exception. Now that we have slightly more than a century’s worth of experience with it, we can identify some of those consequences that have epistemic implications. Trial judges, more than anything else, are umpires or referees. Their job is to rule on all the questions and controversies that arise in the course of a jury trial. They must respond to objections that respective counsel make to one another’s partisan efforts. Their task, as well, is to instruct the jury about the relevant law. The defense can appeal almost every ruling and instruction that a judge is asked to make. In a typical appeal, the defense counsel may challenge dozens of rulings or instructions issued by the judge in the course of a trial. It seems reasonable to suppose that, other things being equal, trial judges prefer not to have a higher court reverse too many cases over which they have presided. Ajudge whose cases are regularly overturned on appeal probably will be perceived by his peers as less than fully competent. Given this, it becomes significant that every judge knows that virtually no trial ruling she makes favor- ing the defendant can be appealed and that every ruling she makes favoring the prosecution can be. In the extreme case, a higher court would never reverse a judge who consistently ruled in favor of the defense during a trial, however ill informed and incorrect her rulings. Now, I do not for a minute suppose that judges are less than persons of high integrity. If the law is clear in a given area, I assume that they will apply it neutrally, unmindful of which side it may favor. But many trials involve motions calling for rulings that are close calls or borderline cases, where a plausible argument can be made for ruling one way or the other. In those circumstances, we would expect to find a pattern of judicial rulings in favor of the defendant and, not coincidentally, of the sort that cannot be overruled on appeal. As one astute legal scholar, Kate Stith, has argued: In a system that absolutely bars any prosecution appeals . . . a trial court seeking to mini- mize reversal would rule for the defendant on all contested issues. If the trial then ended in acquittal (and, presumably, ruling for the defendant would increase the likelihood of P1: JZZ 0521861667c08 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:12 206 flawed rules of evidence and procedure this outcome), the legal issues could not be appealed. If the trial ended in conviction, the defendant would have nothing to appeal, since he prevailed on all legal issues. 22 In short, there is an inescapable bias in a system that permits appeals only of convictions, a bias that encourages rulings on controverted questions of law favorable to the defendant. To deny this thesis would require us to believe that judges were wholly indifferent to whether the trials they handled were likely to be reversed by higher courts. Given human nature, that seems unlikely. By contrast, a systemin which there were appeals of both convictions and acquittals would offer no incentives to trial judges to shade their rulings one way or the other. The preliminary evidentiary hearing was adopted, in part, to reduce the potential damage associated with this asymmetry. In principle, the preliminary hearing – whose rulings can be appealed by either party – will settle many questions as to the admissibility of evidence, testimony, and so on. Still, this provides only a partial fix for the problem. During the trial itself, many issues that affect its outcome are likely to arise. One counsel will ask questions of witnesses that the other holds to be inappropriate. Freshly discovered evidence may be introduced, whose admissibility the judge must settle. Motions for striking comments by witnesses or counsel will routinely be made and must be decided then and there. At the last minute, the defendant may choose to testify in his own defense, which can spark multiple controversies about the introduction of previously excluded evidence. Both sides are apt to propose instructions that they will ask the judge to pass along to jurors. And so on. The point is that the dynamics of any complex trial require many decisions from the judge that cannot be anticipated at the preliminary hearing. The pros- ecution is lumbered with a situation in which the judge’s is the last word on almost every acquittal-enhancing ruling and only the first word on every ruling favoring the prosecution. Defendant-friendly trial errors on the part of the judge are both irreversible and cost-free to her while any prosecution-friendly errors she may make are subject to review and correction. It would be surprising if this difference did not work to skew judges’ rulings in the direction of erring on the side of the defendant. Obviously, if the appellate policy were symmetric, such tendencies would vanish. Self-Correction and Learning from Our Mistakes I have been arguing that we face significant epistemic liabilities arising from the fact that the appellate process examines only the ways in which convictions can go wrong. I have claimed that countless false acquittals every year could be undone if review of such cases were allowed. We are making more mistakes 22 Stith, at 37. P1: JZZ 0521861667c08 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:12 Double Jeopardy and False Acquittals 207 than we need to, and those mistakes could be corrected without compromising our commitment to preserving m as the acceptable ratio of true acquittals to false convictions. But there is a deeper problem than this with the asymmetry of appellate review. It has to do with the ability of the courts to learn over the long run when existing practices are failing to promote accurate verdicts. I invite you to think, for a moment, of a criminal justice systemon the model of a system of inquiry. It consists of elaborate rules and baroque procedures creating a highly structured environment whose putative goal is to find out the truth about certain matters of high interest. Conceived in this way, we can see several parallels with that other great cultural system of inquiry: the natural sciences. If science has its methods, the criminal law has its procedures. If science has rules for the evaluation of theories and the design of experiments, the criminal law has its rules of evidence and proof. If science supposes that the null hypothesis is sound until and unless it has been refuted, the law has its PI, which similarly must be defeated if one is to draw positive conclusions about the guilt of the accused. Afundamental feature of any systemof inquiry is that it is a work in progress. The rules that that system has adopted must themselves constantly be subject to review, precisely because the current rules may be suboptimal for reaching their desired ends. Just as the methods of science evolve as scientists discover ways in which existing methods fail to conduce to the truth, so the rules of proof and procedure in the law evolve as judicial authorities learn how those rules sometimes lead us astray or function as obstacles to the realization of the very ends they were meant to promote. It will be helpful to distinguish between two quite different things that happen in the appellate process. On the one hand, the appellate system is charged to identify and rectify specific mistakes made in particular cases, where the term “mistake” generally means a failure to following existing rules of evidence or procedure. We might describe it as a systemlearning of its mistakes. At the same time, the machinery of the appellate process allows the judiciary to learn from its mistakes. By that I mean that courts gradually come to perceive that certain existing rules of procedure and evidence constitute obstacles to finding out the truth. Such discoveries prompt calls for emendations in the rules themselves. These are very different roles and it is crucial to be aware of both. This duality of roles can appear paradoxical. On the one hand, appellate courts are looking to ensure that lower courts scrupulously follow existing rules. On the other hand, those higher courts must be constantly alert to the ways in which existing rules – which they are charged to enforce – are flawed. But this paradox, if paradox it be, is not unique to the criminal justice system. It is at the core of every enterprise and activity seriously committed to the idea that one can and must learn from one’s mistakes. More than a century ago, the American philosopher Charles Sanders Peirce argued that any robust, undogmatic system of inquiry into empirical P1: JZZ 0521861667c08 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:12 208 flawed rules of evidence and procedure questions must contain mechanisms that enable inquirers to discover both sorts of mistakes. When Peirce referred to mistakes, he did not only mean, not even principally, specific conclusions erroneously reached by the application of the methods or rules in question. On the contrary, Peirce thought that what a system of inquiry must allowfor is systematic discovery, over time, of the ways in which the existing rules guiding the system are bad or inadequate, that is, how they thwart finding out the truth about the objects or processes under investigation. In the case of science, which was Peirce’s principal concern, this meant that inquiry should be so constituted that it was capable of finding whether existing rules or methods were conducive to the truth or whether, to the contrary, they were often responsible for generating erroneous beliefs. He maintained that one of the great virtues of science was that it was a system capable of discov- ering when its rules were systematically leading it astray. By discovering such fallacious procedures and modifying them, Peirce argued, scientists, in effect, were continuously learning howto learn. The rules of scientific method, for him, were not commandments cast in Mosaic stone, but a set of fallible guesses about how best to interrogate nature. Because of the existence of various feedback mechanisms in science, we can discover when those guesses are misleading us and then change them for the better. Peirce’s point is that in an ideal system of inquiry we are not only learning about specific mistakes we have made in the acceptance or rejection of particular hypotheses. More importantly, we are learning about how the very rules that constitute the anatomy of inquiry fail to be optimal. This self-correction was, for Peirce, a necessary condition for judg- ing any system of inquiry to be genuinely empirical. We need to ask ourselves whether the legal system exhibits the sort of self-correction that Peirce had in mind. Specifically, I want to focus on the logic of the appellate process with this set of questions in mind. Legal systems like the American one allowfor the appeal of convictions. This means, of course, that superior courts are routinely presented with situations in which persons are claiming to have been wrongfully convicted. A simple- minded version of what goes on in such appellate discussions is that they result in a determination by superior courts of whether the rules currently in place were assiduously followed in particular trials. On this view, higher courts, in reviewing appeals submitted to them, are chiefly concerned to find out whether existing rules of evidence and procedure were assiduously followed in the case in question. But that is not all that is going on; it is not even the most important thing that is going on in the appellate process. As the history of twentieth-century American jurisprudence vividly reveals, higher courts, when they accept an appeal, are not simply deciding whether a given case was conducted and decided according to the existing rules. Courts are likewise, and more importantly, looking for and sometimes discovering ways in which the current rules conduce P1: JZZ 0521861667c08 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:12 Double Jeopardy and False Acquittals 209 to erroneous convictions. On discovering that existing rules persistently lead to the conviction of the apparently innocent, circuit and Supreme Court justices frequently propose modifications of the existing rules and procedures to make them less likely to conduce to errors of the sorts they see coming across their desks on a regular basis. This is precisely the sort of feedback mechanism that Peirce had in mind when he talked about authentic systems of inquiry. One learns fromone’s mistakes not only that the mistakes have occurred but also that their occurrence was made more likely by the use of certain rules or procedures that are themselves open to review and modification. This is genuinely learning how to learn. In the legal case, however, unlike the scientific one, the learning is decidedly one-sided. The systemis continuouslyreviewingpotentiallyflawedconvictions, diagnosing their causes, and tinkering with or adjusting the rules to reduce the likelihood of erroneous convictions in the future. That is doubtless a good thing; but it is not necessarily an unmitigated good, because tinkering with the rules to reduce further the occurrence of false convictions may lead to new rules that produce vastly more false acquittals than the existing system does. A genuinely self-corrective system of inquiry could quickly identify such retrograde changes because they would show up in the form of increased false acquittals. However, when appeals are uniquely one-sided, there is no feedback mechanism for discovering whether a given change in the rules (motivated by a concern to protect the innocent from false conviction) has inadvertently made the prospect of false acquittals much more likely. Obviously, the asymmetric appeals process is not only more likely to iden- tify the causes of false convictions than of false acquittals; it is also likely to introduce new rules or revisions of old rules that, while reducing the frequency of false convictions, increase the frequency of false acquittals. If you doubt this, thinkof the effect of the Mirandarule introducedinthe 1960s. The WarrenCourt was convinced that the earlier rules governing confessions – basically variants on the notion of voluntariness – were leading occasionally to the admission of false confessions and thus producing false convictions fromtime to time. Under the Miranda regime, as we saw in Chapter 7, the frequency of false acquittals jumped significantly, although it is not clear whether Miranda reduced false convictions to any significant degree. There is a second problem, related to, but distinct from, this one. Given the asymmetry of appeals, there is no comparable process with respect to learn- ing about the rules that conduce to mistaken acquittals. Their causes remain shrouded in mystery. Absent judicial review of acquittals, judges have nothing other than folklore to fall back on in deciding whether existing rules could be modified to make false acquittals less likely. Absent a systematic scrutiny of cases resulting in acquittal, no one has anything more than hunches about either the frequency of such errors or about which current rules of evidence may be P1: JZZ 0521861667c08 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:12 210 flawed rules of evidence and procedure responsible for producing such errors as do occur. In sum, we have much less information than we should about what causes false acquittals and no regular mechanism for ascertaining whether recent modifications to those rules have increased their frequency. Error Correction versus “Error” Correction Thus far, our discussion of appellate strategies has been assuming (as virtually everyone who discusses this theme does) that a key function of the review process is to detect errors and that, despite the asymmetry of the current appeals procedure, the system manages to catch many false verdicts, specifically, false convictions. It is time to challenge that assumption itself. As I have been using the term “error” throughout this book, it refers to a verdict – or other concluding result of a criminal investigation – that is either false or invalid. If we could catch such errors of those sorts that the system commits, this would clearly be a good thing. That is what judicial appeal should be about. The courts, however, work – in practice if not in theory – with a very different notion of error. For them, as I pointed out in Chapter 1, an “error” occurs when a rule of evidence or procedure has been violated in a criminal investigation and where that violation appears to have played a role in leading to the conviction of a defendant. In short, when appellate courts are looking for “error,” they are typically looking for instances of rule breaking. If they find them, and regard them as other than harmless, they will throw out a conviction. But it is fair to ponder whether finding evidence of rule breaking is likely to be a good indicator that a guilty verdict is false. As we have seen repeatedly, many of the rules currently governing a criminal trial and investigation are not truth conducive. Some of themare neutral with respect to truth seeking. Others, I have argued, are unambiguously truth thwarting. If the courts decide that the conduct of a particular trial violated either of those sorts of rules and send the case back for retrial, the existence of the “error” will have nothing to do with whether the guilty verdict is true or false. Worse, if someone was convicted (for example) by virtue of the trial judge breaking one of the exclusionary rules that keep relevant evidence away fromthe jury, an appellate court will set this verdict aside as “erroneous,” even though the rule breaking in question probably gave the jurors a much more accurate picture of the events surrounding the crime than they would have had otherwise. The moral is that a verdict that emerges from breaking the rules of evidence and procedure is likely to be a false or invalid verdict only when the rules in question are themselves truth conducive. Since many existing rules fail to be so, an appellate court’s identification of serious “errors” may have little to do with whether the verdicts in question are really erroneous. This has a number of practical implications. If we learn, for instance, that about one in every ten appealed convictions is reversed on appeal, our naive inference might be that P1: JZZ 0521861667c08 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:12 Double Jeopardy and False Acquittals 211 such statistics indicate that one in every ten convictions is false. Yet, given the disconnect I have been describing between “a trial that broke the rules” and “a trial that produced a false or invalid verdict,” there is no plausible inference we can reliably draw from the frequency of verdicts set aside to the frequency of false convictions. Under current circumstances, an overturned or reversed verdict is not ipso facto likely to be an erroneous verdict in the sense in which I have been using the term. By contrast, if the kinds of rules of evidence and procedure that I have been advocating here were in place, then evidence of rule breaking in a trial would warrant a presumption of a genuinely erroneous verdict. Conclusion Given the arguments in this chapter, it is difficult to decide who is the greater recipient of double jeopardy’s largesse: guilty defendants who win irreversible acquittals or sloppy trial judges who find their acquittal-enhancing errors immu- nized from review. Whatever dispute there may be about the beneficiaries of this policy, it is clear that the most serious victim of double jeopardy law is the justice system itself, which is precluded from eliminating many patent errors of judgment by the insulation of an acquittal from any further scrutiny. I have argued that the appellate asymmetry has two predictable consequences: r There are more erroneous verdicts overall than would occur under symmetric appeal. r Over time, evidence law itself comes to contain more and more protections against false convictions, while doinglittle or nothingtobringfalse acquittals under control. Arguably, almost no other error-reduction step that the criminal justice sys- tem could take would correct as many erroneous verdicts as the institution of appeals of acquittals. Apart from enabling the state to convict more of the truly guilty than it now does, opening up the system to appeals of acquittals would halt the drift of evidence law in the direction of becoming ever more acquittal- friendly. If appellate courts were routinely confronting injustices of both sorts, we have every reason to expect that the courts would come up with fixes that worked to reduce the total number of erroneous verdicts instead of producing remedies that focus myopically on false convictions. We should recall that the Supreme Court decision of 1904 that established that convictions cannot be appealed, Kepner v. U.S., was decided by a slim five-to-four majority. 23 Among the dissenters was the cleverest mind then on the Court, Oliver Wendell Holmes. Holmes and the minority argued, as I have 23 Kepner v. U.S., 195 U.S. 100, at 133 (1904). For a detailed critique of the Kepner doctrine, see Amar, 1841–3. P1: JZZ 0521861667c08 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:12 212 flawed rules of evidence and procedure here, that there is no constitutional impediment to appeal of acquittals and that retrial after such a successful appeal would be as appropriate as retrial after the successful appeal of a conviction. Whatever the merits of this constitutional argument, it is clear that truth seeking took a major hit by the slimmest of margins when the Court decided against allowing appeals of acquittals. P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 9 Dubious Motives for Flawed Rules: The Clash between Values The central purpose of a criminal trial is to decide the factual question of the defendant’s guilt or innocence. – U.S. Supreme Court (1986) 1 Rights always trump reliability. – William Pizzi 2 The manifest destiny of evidence law is a progressive lowering of the barriers to truth. – C. McCormick 3 In the previous four chapters, we have seen repeatedly how truth-thwarting rules or procedures have been put in place across a wide spectrum of rules of evidence and procedures. In several cases, I have suggested alternatives to those rules that would be more conducive to finding out the truth about a crime. Now, at the end of this long thought experiment, we have to choose between these genuinely truth-promoting rules and their truth-thwarting counterparts. If epistemic values were the only ones in play, the choice would be a foregone conclusion. But other values – fromthose focused on the rights of participants in a criminal proceeding to questions of public image and efficiency – do intrude, meaning that difficult choices must be made. If we are to decide between existing rules and some of my proposed alterna- tives, we have to face squarely the question of how, if at all, conflicts between these different sorts of values can be resolved. It will probably be helpful if, from the outset, we recognize that the sorts of reasons given for the various 1 Rose v. Clark, 478 U.S. 570, at 577 (1986). 2 William Pizzi, Trials without Truth 167 (1999). 3 Charles T. McCormick, Handbook of the Law of Evidence §18 (1954). 213 P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 214 flawed rules of evidence and procedure truth-thwarting rules I have been scrutinizing fall more or less cleanly into five distinct classes: r Distributionist concerns, seeking to lower still further the likelihood of a false conviction r Worries about the jury’s inability to interpret evidence appropriately r The importance of safeguarding the rights of the defendant r The necessity of policing the police to discourage unacceptable behavior r The desirability of maintaining the moral integrity of the courts, by ensuring that they are untainted by suspicions of complicity in the illegal behavior of other agents of the state. I will have nothing further to say about the first type of argument, which imports distributionist values into the appraisal of rules of evidence and pro- cedure, having already explained in Chapters 4 and 5 why it is unacceptable to bring those concerns to bear in any context except the determination of the SoP. The last four chapters have presented an abundance of instances of error- distributionist arguments cropping up whenever legal scholars or judges hope to justify the adoption of epistemically suboptimal rules. If I have not yet per- suaded you of both the ubiquity and the undesirability of this practice, nothing I can add will change your mind. However, each of the other four sources of conflict with epistemic interests will require separate discussion. Three of the four deal with what are basically moral questions. The remaining one, with which I will begin, concerns itself with the epistemic compromises supposedly demanded by the use of trial by jury. Alleged Jury Incompetence and the Rules of Evidence and Procedure In all the arguments before us, save this one, we face a decision between truth seeking and other values. Here, however, we have a debate that takes place largely within the confines of legal epistemology. As we have seen repeatedly, the idea is put forward that jurors should not be allowed to see or hear this or that bit of relevant information because they are apt to evaluate it inappropriately. Consider a few of the examples we have already encountered: r Jurors should not see very vivid evidence of a crime if it might “unfairly prejudice” them against the defendant. r Jurors should not be allowed to hear a retracted confession that might lack corroboration because they are prone to regard confessions as dispositive and will convict without further ado. r Jurors must be sternly instructed not to drawany conclusions fromthe silence of the defendant because otherwise they might convict him simply because he was silent. P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 Dubious Motives for Flawed Rules 215 r Similarly, jurors must not be informed that witnesses invoked their Fifth Amendment right against self-incrimination or that the defendant refused to permit privileged witnesses to testify. r Jurors must learn nothing from the prosecution about the criminal past of the defendant, not because such information is irrelevant but because they might convict him on the grounds that he is a bad guy. Clearly, undergirding every one of these arguments is a set of empirical hypotheses about juror psychology that may be true or false. In some cases, we have already seen grounds to doubt them. For instance, empirical studies cited earlier have shown that jurors, even given a retracted confession, do not automatically convict the accused. Even when allowed to draw inferences from a defendant’s silence, they do not necessarily convict. Still, it would be prema- ture to suggest that we have any robust evidence from well-designed studies that would either corroborate or undermine the hypotheses that drive these evi- dentiary exclusions. Until such studies are available, our default assumption should be that relevant evidence is admissible. As I have often remarked, the courts do not encourage empirical study of such questions. So, it is not clear whether we will ever have the evidence necessary to decide whether these prophylactic restrictions on the jury’s access to relevant evidence are necessary. I have supposed that a combination of judge’s instructions and argument by respective counsel may be sufficient to warn jurors against the obvious fallacies associated with this sort of evidence. But if empirical research were to show that these measures are insufficient, or if we simply assume them to be insufficient absent evidence to the contrary, then this should give us all pause about the epistemic desirability of trial by jury. If it is really true that trial by jury requires the wholesale exclusion of so much relevant evidence, then we should think again about the advisability of trial by jury. My own preference – absent reliable information about juror psychology – would be for strengthening rather than weakening the powers of the jury. Such efforts would, I think, increase the likelihood that juror decision making would be as rational as we want it to be. What I have in mind can best be highlighted by comparing a modern jury with a traditional one. The English have been using jury trials as the instrument for determining guilt at least since the thirteenth century. Through most of that time, juries were both active and robust. An early English jury would normally draft into its ranks those who had witnessed the crime. It would conduct its own inquiries, often including interviews outside of the courtroom proper. Jurors were simultaneously the chief repository of knowledge about the crime, the interpreters of the relevant law that applied to the case, and the triers of fact, who would settle issues of guilt and innocence. Jurors could interrogate whomever they liked (including the defendant), and put to them whatever questions they thought relevant. P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 216 flawed rules of evidence and procedure Beginning in the eighteenth century, the rot began to set in with the rise of professional prosecutors and defense attorneys. Judges came to insist that it was their role, not the jury’s, to interpret the relevant questions of law that touched on the case before the court. Prosecutors and defense attorneys conspired to take away fromjurors all of their independent investigative functions. The ideal juror came to be conceived as a person who, instead of knowing something about the crime and about the principals in the case, was completely ignorant of such matters. Jurors could no longer freely ask questions of witnesses, neither in nor outside the courtroom. They could no longer subpoena witnesses to appear before them. By the early-nineteenth century, a whole body of law and tradition had evolved concerning the issues of evidence and proof. Those rules were quite explicit about what sort of testimony a jury could hear (no hearsay, for instance), about who could testify (privileges extended to employees and employers of the accused as well as his family members, and the accused himself was not allowed to give sworn testimony on his own behalf), about what sort of physical evidence could be shown to the jury, and so on. Jurors came to be instructed (as they still usually are) that they could not discuss the pending case among themselves until after closing arguments. They could perform no experiments, consult no books of reference, and take no notes during the trial. In sum, the jury was rendered inert, stripped of its investigative and inter- rogatory powers, firmly told that it was to decide questions of fact and not questions of the law (which had become the judge’s territory), and that it was to do all this shielded by the rules of evidence fromseeing and hearing much of the evidence relevant to the case. Jurors came to be subject to elaborate instructions from the judge about the law and about how they were to deliberate. They were given mandatory presumptions to “aid” them in drawing inferences. They were firmly told to ignore certain items of evidence or to suppress memory of certain testimony they had heard. That is substantially where things still stand. Although there has been some loosening of the rules of evidence, which allows modern jurors to learn a little more about the case before them than they would probably have learned in the nineteenth century, it is likewise true (and depressing) that twentieth-century jurisprudence has created many new obstacles (most notably the exclusionary rules, the Miranda rules, and the suppression of jury inferences from silence). If we can briefly step away fromthe legal context, it is easy see howartificial and sclerotic the system has become. The legal system, after all, is not the only social institution whose principal goal is truth seeking. Imagine trying to conduct scientific or engineering research on the legal model. Specifically, since we are focused here on juries, let us imagine telling scientists working on a problem (for example, “Is the trajectory of that asteroid likely to bring it on a collision course with the Earth?”) that they will be given access to only a subset of the available, relevant evidence. Further, they can put no questions P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 Dubious Motives for Flawed Rules 217 to those feeding them relevant information. They cannot collect any evidence themselves, nor request its collection, nor consult any outside sources. Nor can they talk among themselves about the arguments pro and con until they have heard the full cases presented by their colleagues, and so on. No one would expect a system of inquiry constructed along such lines to be efficient, reliable, or long-lived. Indeed, no self-respecting scientist or engineer would agree to participate in such a charade. Without exception, they would say that this is patently not how you would go about determining the asteroid’s trajectory. There are, of course, important disanalogies between the scientific case and the legal one; prominently, jurors, unlike scientists and engineers, have no special expertise in conducting investigations nor much practice in drawing complex inferences from messy data. Granting all that, I remain unpersuaded of the desirability of shielding jurors from relevant facts to the extent that the legal system now does. Instead of leaving a jury to decide whether a retracted confession was coerced or is corroborated, we usually assign that decision to a judge. Instead of allowing the jury to make what they will of a defendant’s silence or a witness’s invocation of the right against self-incrimination or other privilege, we organize things so that jurors are either unaware of these develop- ments or (in the case of the defendant’s silence) explicitly instructed to assign no significance to that fact. Judges and attorneys debate furiously in judges’ chambers or in preliminary hearings concerning what evidence jurors will be allowed to see, what testimony they will hear, and what instructions they will be given. Throughout this process, of course, jurors are treated as simpletons, inca- pable of coping with (for example) a confession that may be suspect or a certain bit of evidence that may be less inculpatory than might appear at first glance. As Mirjan Damaˇ ska has observed, “The complete passivity of the paradig- matic Anglo-American fact finder – the jury – is far from an ideal epistemic arrangement.” 4 It is important to add that, apart from juries not being allowed to hear much relevant evidence, there are likewise many truths that prosecutors are not allowed to utter before a jury. Among other things not already mentioned, the prosecutor is not permitted to tell the jury that the state’s evidence is “unre- futed,” for that might remind the jury that the defendant is saying nothing. 5 He cannot ask whether the defendant or a witness took, or is willing to take, a lie detector test. He cannot utter the truth that a testifying defendant, coming last in the trial proceedings, is in a position to shape his testimony – if he gives it – to fit that of the other witnesses. 6 The prosecutor cannot inform the jury in a murder trial that there will be a review of the conviction and the sentence 4 Mirjan Damaˇ ska, Evidence Law Adrift 96 (1997). (Emphasis in the original text.) 5 U.S. v. Skandier, 758 F.2d 43, at 44 (1985). 6 Connecticut. v. Cassidy, 672 A.2d 899 (1996). P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 218 flawed rules of evidence and procedure by a higher court. 7 Juries, it seems, cannot be trusted with such sensitive facts. Some of these prohibited expressions are not only true but are so obvious that any awake juror could see them for herself. That perhaps undercuts the urgency of the prosecutor being able to say them openly but it does not gainsay the inappropriateness of court suppression of truth telling. You might already have noted the irony of a system that extols the virtues of trial by jury, the “palladium” of our system of justice, all the while arguing that jurors are too stupid, naive, or immature to be trusted with complex arguments, inflammatory or prejudicial evidence, confessions of uncertain provenance, and so on. To quote Damaˇ ska again, “One would expect a legal process that glorifies novice amateurs as fact finders to presume their intellectual and emotional capacity for the job.” 8 What compounds that irony is that, having treated jurors as if they were impressionable fools all through the trial, the legal system then turns around and treats their verdicts as if written on Mosaic tablets. Juries are not obliged to explain or justify their verdicts. Appellate courts will rarely reverse a jury’s judgment with respect to the facts of a case, however invalid their inferences appear to be. More emphatically, a jury’s acquittal is literally the last word, even if it arose from juror incompetence or tampering, or (more commonly) if it derived from rules that drastically restricted jurors’ access to relevant evidence. That, in its way, is every bit as objectionable as the fact that the jury, prior to reaching its verdict, is treated so condescendingly. Safeguarding the Rights of Defendants On multiple occasions, my analysis has been butting its head up against the fact that certain rights, especially rights belonging to the defendant in a criminal trial, pose serious obstacles to getting at the truth. Whenever I could gracefully drop that hot potato, I have done so, promising that I would eventually get around to it. That time has come. The first point to note – because it may not have been obvious thus far – is that many of the rights given to defendants by the Constitution or the courts unquestionably serve epistemic ends. The rights to counsel, to speak in one’s own defense, to confront one’s accusers, to knowwhat exculpatory information the prosecution may have been concealing, to be tried in public, to appeal a conviction, to subpoena witnesses, and to know precisely what crime one is being charged with represent policies that pose no problems for the legal epistemologist. Even if these were not already rights, there would be a powerful epistemic case for including them all in the procedural rules of a criminal trial. 7 Caldwell v. Mississippi, 472 U.S. 320 (1985). 8 Damaˇ ska, at 29. P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 Dubious Motives for Flawed Rules 219 Other rights are largely truth neutral, and so not troubling to the truth seeker. Most obvious here is the right to trial by jury (supposing, as I have just argued, that one does not construe the debilities of the jury so broadly that they become excuses for evidence suppression). Others include the right to serve as one’s own attorney, the right to be tried to the BARDstandard, and the right in federal courts to demand a grand jury indictment instead of a judge’s finding that one should go to trial (often known as an “information”). The much more difficult class of rights is that of those rights that (at least under their orthodox interpretation) constitute serious, sometimes even grave, obstacles to accurate verdicts. They include the right to silence (and to have that silence ignored), the right to invoke the relationship privileges, the right not to have an acquittal appealed, the right to demand exclusion of evidence acquired as a result of violations of one’s search and seizure rights, and the right of a witness to avoid giving incriminating testimony. The question we have to confront is how to resolve the tension between the values associated with finding out the truth and those thought crucial for safe- guarding human dignity, privacy, and our general sense of respect for persons. There is no omnibus solution to this problem, as far as I can see. Still, there are solutions to particular bits of it. Having already discussed in Chapter 8 the right not to have an acquittal appealed, I will focus my discussion on the rights to suppress evidence obtained by illegal searches and seizures and the right to silence. Consider the argument that relevant evidence obtained from rights-violating searches and seizures of the defendant’s property must be excluded from trial because those seizures violated the defendant’s rights. Let us grant, at the outset, that the defendant – like everyone else – has a right not to have his property searchedandseized, absent probable cause or a searchwarrant. The Constitution makes it clear that every citizen enjoys this right, but it says nothing about what the appropriate remedy is when this violated right belongs to someone who also happens to be a criminal defendant. Is the best remedy for this violation of the defendant’s rights the exclusion of relevant evidence from a criminal case against him? Clearly, there are other remedies – the same ones open to every citizen whose property and privacy rights are violated. The defendant can bring a civil suit seeking compensatory damages against the police for violating his rights. If the violation in question involved lawbreaking (such as illegal entry), he can also bring criminal charges and can sue in civil court for punitive as well as compensatory damages. What is unclear is why the courts think that the right (for example) not to have one’s house entered without a search warrant or with- out probable cause entails the additional right to have inculpatory evidence acquired by that right-violating act excluded from trial. These are wholly dis- tinct questions. Remember that the courts allow this right-violating evidence to be used to impeach the defendant should he choose to testify. That much P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 220 flawed rules of evidence and procedure alone establishes that the courts believe that the defendant has no claim to an unconditional exclusion of such evidence. Given that the other remedies avail- able – civil and criminal – are probably much more likely to deter unwelcome police invasions of property and privacy than the exclusion of evidence will, the Exclusionary Rule seems to exact an unnecessarily and unacceptably high epistemic price. Writing in 1954, the Supreme Court itself made the same point: It appears . . . that there is no lack of remedy if an unconstitutional wrong has been done in this instance without upsetting a justifiable conviction of this [defendant]. If the officials have willfully deprived a citizen of the United States of a right or privilege secured to him by the Fourteenth Amendment, that being the right to be secure in his home against unreasonable searches . . . , their conduct may constitute a federal crime under 62 Stat. 696, 18 U.S.C. (Supp. III) §242. 9 The argument is as valid now as it was half a century ago. The righting of this constitutional wrong can be secured without exclusion of relevant evidence. What about the right tosilence andits Griffin-fashionedcorollarytothe effect that the jury must wholly ignore the silence of the defendant? You will perhaps recall a long passage quoted in Chapter 5 from Justice Goldberg, insisting that the right to silence is basically grounded on a more fundamental right, the right to privacy. Goldberg referred to a “private enclave” into which the state should not enter, not even in the interests of solving a crime. He insisted that “our respect for the inviolability of the human personality” demands as much. According to this way of thinking, the contents of an individual’s mind, what he knows, belong exclusively to him, and he has an unconditional right to keep his thoughts to himself. Forcing him to testify, or allowing the jury under any circumstances to infer guilt from his refusal to testify, would – says Justice Goldberg – be an unwarranted invasion of the defendant’s privacy. This is a bizarre constitutional argument, whether we focus on its moral or its constitutional dimension. To begin with, it is almost certainly the case that the original framers of the Constitution did not recognize any generalized right to privacy, at least not where criminal trials were concerned. Moreover, the legal system itself does not generally take the view that individuals can lay claim to a private enclave protected from the search for justice. For instance, when police can showprobable cause for suspicion of someone, they can obtain warrants requiring the suspect to surrender access to his possessions, including private papers, bank statements, personal videotapes, and the contents of his computer’s hard drive. They can record his telephone calls or live conversations with informers. They can intercept his mail. Where probable cause has been 9 Irvine v. California, 347 U.S. 128, at 137–8 (U.S. 1954). The referenced statute provides “that whoever, under color of any law, statute, ordinance, regulation or custom, will- fully subjects any inhabitant of any state to the deprivation of any rights, privileges or immunities secured or protected by the Constitution of the United States shall be fined or imprisoned.” P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 Dubious Motives for Flawed Rules 221 shown, the police can demand samples of a suspect’s blood, his fingerprints, his hair, his wardrobe, and his handwriting. They can make him submit his body and his voice to a line-up. They can insist that a police photographer take pictures of the defendant’s sexual organs, if considered relevant. If a person has committed his private thoughts to paper (such as in correspondence with intimates) or expressed them to friends, who can be subpoenaed, they too are regarded as fair game. As David Dolinko has asked, “Why should we suppose that a person who is legally compelled to reveal incriminating information about himself suffers a loss of privacy different in degree or quality from that which he would experience if anyone else revealed the same information?” 10 There is not much suggestion in any of this of a “private enclave.” What ultimately gives the lie to Goldberg’s privacy argument, however, is the manner in which the justice system treats witnesses, as opposed to defendants. If someone is thought to know something about a crime, she can be subpoe- naed to testify as to what she knows. When she appears in court, she can be questioned about her knowledge of the events surrounding the crime. She is provided no “private enclave.” True, she can always invoke the Fifth Amend- ment herself, declining to testify concerning matters that may tend to implicate her in wrongdoing. But witness silence is a trumpable right. Confronted by an uncooperative silent witness “taking the Fifth,” the prosecutor can grant her immunity and thus oblige her, on pain of contempt of court, to reveal all. If the justice system seriously held that every individual had a right to privacy with respect to what they know, these practices would not be in place. More than that, the basic premise in a criminal trial is that the truth about the crime is being sought. Essential to that endeavor is that the jury come to learn the relevant facts from those who have pertinent information. The logic of a fair-minded trial requires that witnesses, whether for the pros- ecution or the defense, can claim no unconditional right to privacy in deciding which questions they will answer and which they will ignore. No court in the land would allow a witness to deflect a pertinent question about her private life by responding that the question invades her privacy. If it is germane to some point of contention in the trial, she must answer it, whether it falls in her “private enclave” or not. The fact that her answer may embarrass or degrade a witness – that it may cost her her job, shipwreck her marriage, or prompt a Mafia contract on her life – is never acceptable legal grounds for refusing to answer. As the great scholar of the law, John Wigmore, once put it, “For more than three centuries it has now been recognized as a fundamental maxim that the public . . . has a right to every man’s evidence.” 11 The Supreme Court itself 10 David Dolinko, Is There a Rationale for the Privilege against Self-Incrimination? UCLA L. Rev. 1063, at 1109 (1986). 11 8 John Wigmore, Treatise on the Anglo-American System of Evidence in Trials at Common Law (3rd ed., 1940), 2192. P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 222 flawed rules of evidence and procedure has said repeatedly and explicitly that witnesses have no privacy rights. Here is how they put the point in U.S. v. Calandra: Ordinarily, of course, a witness has no right of privacy. . . . Absent some recognized privilege of confidentiality, every man owes his testimony. He may invoke his Fifth Amendment privilege against compulsory self-incrimination, but he may not decline to answer on the grounds that his responses might prove embarrassing or result in an unwelcome disclosure of his personal affairs. 12 Inexplicably, Goldberg – and the rest of the 1964 court, since he was writing for them all – would give a criminal defendant vast privacy rights enjoyed by no other participant in a criminal trial. Whatever the motive for doing so, it cannot be that the criminal law generally respects the privacy of those who know something about a crime. It does not. Given the weakness of the privacy arguments in favor of Silent Defendant, we might be inclined to wonder why it continues as a fundamental part of American criminal law. The usual response is to say that it is a part of the Bill of Rights, specifically the Fifth Amendment. For many constitutional lawyers, that is enough to end the discussion. We simply have to live with it, they will say. (Fortunately, this book is not bound to defer to constitutional requirements since our concern is not with what the founding documents say but rather with the truth-promoting credentials of the criminal justice system.) That said, however, I will discuss briefly the question of whether the Fifth Amendment, as originally understood, had anything to do with Silent Defendant, as currently understood. Fortunately, several eminent legal historians have already covered this terrain in the last two decades. 13 What their work makes very clear is that the framers of the Bill of Rights had no intention of putting in place a system that would both enable the defendant to avoid telling his story and then demand that the jury ignore his silence. The language of the Fifth Amendment seems pretty clear on this point. The totality of what it says about Silent Defendant is this: No person . . . shall be compelled in any criminal case to be a witness against himself. . . . To understand what the framers were getting at, we must remind ourselves that, when the Bill of Rights was adopted, it was unusual to have defendants rep- resented by counsel. Defendants typically conducted their own defense, cross- examining prosecution witnesses, explaining their version of events, responding to questions from judge, jury, and prosecutor, and generally pleading their case 12 U.S. v. Calandra, 414 U.S. 338 (1974). 13 Best known among these studies is Leonard Levy, Origins of the Fifth Amendment (1968). Likewise pivotal are the investigations by Akhil Amar and Renee Lettow, Fifth Amendment First Principles, 93 Mich. L. Rev. 857 (1995); John Langbein, The Historical Origins of the Privilege against Self-Incrimination at Common Law, 92 Mich. L. Rev. (1994). P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 Dubious Motives for Flawed Rules 223 as best they could. Had defendants chosen to remain silent, they would have had no defense at all. Juries would almost certainly have convicted in such a situation, unless the prosecution’s case was unpersuasive in its own terms. Defendants, apart fromconducting their defense, could also testify if they chose, although, unlike prosecution witnesses, defendants could not give testimony under oath. Probably to avoid putting guilty defendants in the position where they would have to commit perjury (and open themselves up to further prose- cution), the traditional rule in the common law was that defendants should say what they had to say, but not under oath. So, one aim of the Fifth Amendment was to ensure that defendants could not be obliged to give sworn testimony. That, however, was very different from giving no testimony at all. A more important motive behind the Fifth Amendment, and why it speaks of compulsion or coercion, was to ensure that the practice of forcing suspects to confess bytorture andintimidation(still commonat the time inEurope) foundno place in the American criminal trial. From the beginning, the Fifth Amendment has been a bulwark guarding against forced confessions. As Albert Alschuler, a prominent historian of the privilege against self-incrimination, writes: The privilege in its inception was not intended to afford criminal defendants a right to refuse to respond to incriminating questions. Its purposes were far more limited. . . . its goal was simply to prohibit improper methods of interrogation. 14 This is why Justice Benjamin Cardozo, in the epigraph to Chapter 5, argued that, provided protections against torture were in place, the right to silence could be jettisoned without undermining the values to which the framers of the Constitution were committed. Justice Oliver Wendell Holmes had the same reading of the Fifth Amendment: “The prohibition of compelling a man in a criminal court to be a witness against himself is a prohibition of the use of physical or moral compulsion to extort communications from him.” 15 In short, the Fifth Amendment was never intended to give the defendant an unlimited right to silence, so we ought not suppose that this “right” must simply be taken as a given. Even if we suppose that the defendant has the right (or at least the power) to say nothing either after his arrest or at his trial, need we likewise be committed to the view that this right would cease to be a right unless steps were taken to ensure that jurors never drew an adverse inference from a defendant’s silence? The idea seems to be that a right is a right only if there are no negative costs possibly associated with its exercise. With all due respect, I fail to see the connection. There are, after all, many circumstances in which the exercise of a right may carry negative repercussions for the person exercising it. Consider, 14 Albert Alschuler, A Peculiar Privilege in Historical Perspective in R. Helmholz et al., The Privilege against Self-Incrimination, 184–5 (1997). 15 Holt v. U.S., 218 U.S. 245, at 252 (1910). P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 224 flawed rules of evidence and procedure most importantly, the counterpart to the right to silence: the right of a defendant to speak in his own defense. If he exercises this right, it may well be that his own testimony convicts him. His taking the stand likewise enables the prosecution to introduce (where it otherwise could not) evidence of his previous criminal record, of his general character, and of coerced confessions. 16 Everyone agrees that the defendant’s exercise of his right to speak can potentially carry severe liabilities for his chances of acquittal. (That, unfortunately, is the principal reason that so many defendants choose silence.) Yet none of that is taken to imply that the defendant doesn’t really have the right to speak in his defense. Why, then, should the defendant’s exercise of the right to silence be guaranteed to be cost-free when the exercise of his right to speak in his own defense (which courts should be trying to encourage) is so costly? Policing the Police There is a commonly held view (especially among judges) that one important function of the courts is to ensure that the activities of the police comply with the law. On this analysis, the courts have a responsibility not only to preside over cases in which police officers are charged with breaking the law (which is fair enough) but also to punish the police, even when they are not on trial, by refusing to admit the fruits of their labors if they appear to have bent the law to their own purposes. Indeed, much of evidence law consists of rules of exclusion aimed specifically at sending a message to the police. Consider, once again, the case of the Miranda rights. The Supreme Court has conceded more than once that the Miranda “rights” are not genuine rights found in the Constitution. Instead, according to the Court, they are “prophylac- tic” measures, designed to discourage the police fromengaging in unacceptable practices of interrogation. Precisely the same analysis applies to the exclusion of a confession freely offered by someone who was arrested without probable cause. The proffered confession, however well corroborated and however vol- untary, will be excluded, not because the defendant enjoys a right to have it excluded (the courts concede that he does not), but because its exclusion may, in the view of the courts, persuade the police that they should not go around arresting citizens without probable cause. As the Supreme Court explained this deterrence doctrine in Michigan v. Tucker: The deterrent purpose of the exclusionary rule necessarily assumes that the police have engaged in willful, or at the very least negligent, conduct which has deprived the defen- dant of some right. By refusing to admit evidence gained as a result of such conduct, 16 The Supreme Court, in Oregon v. Hass, held that an illegally obtained confession can be presented to the jury in order to discredit a defendant’s testimony (420 U.S. 714, at 722 [1975]). P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 Dubious Motives for Flawed Rules 225 the courts hope to instill in those particular investigating officers, or in their future counterparts, a greater degree of care toward the rights of an accused. 17 In Miranda and many similar cases, we see the evidence rules being used, not to promote accuracy in verdicts, but as devices for fostering desirable police conduct. This immediately prompts two questions: Is the exclusion of relevant but improperly seized evidence an effective way to correct errant police behav- ior? And, if it is, is this sufficient to counterbalance the high epistemic costs exacted by the exclusion of incriminating evidence? We scarcely even need to bother to pose the second question, since the answer to the first is probably negative. Even the most fervent advocates of the “policing the police” policy concede that the exclusion of evidence is, at best, a mild and indirect deterrent to unacceptable police investigative practices. After all, it is not the police themselves but prosecutors who feel the brunt of evidentiary exclusion rules. Police investigators tend to regard a case as solved when they are persuaded that they have found the guilty party. While the police may be dismayed when a jury – with access to less evidence than the police have seen – acquits someone whom they regard as guilty, a police officer’s career is not usually going to be significantly impacted because the courts sometimes exclude evidence that she has collected. While exclusion can often deal a body blow to a prosecutor, it is, at best, an indirect slap on the wrist to the cop on the beat. Courts have sometimes conceded as much. This is howthe Supreme Court in 1954 viewed the wisdom of such practices as are now under discussion: It must be rememberedthat petitioner is not invokingthe Constitutiontoprevent or punish a violation of his federal right . . . to recover reparations for the violation. He is invoking it only to set aside his own conviction of crime. That the rule of exclusion and reversal results in the escape of guilty persons is more capable of demonstration than that it deters invasions of right by the police. The case is made, so far as the police are concerned, when they announce that they have arrested their man. Rejection of the evidence does nothing to punish the wrong-doing official, while it may, and likely will, release the wrong- doing defendant. It deprives society of its remedy against one lawbreaker because he has been pursued by another. It protects one against whom incriminating evidence is discovered, but does nothing to protect innocent persons who are the victims of illegal but fruitless searches. The disciplinary or educational effect of the court’s releasing the defendant for police misbehavior is so indirect as to be no more than a mild deterrent at best. 18 In short, in deciding whether to exclude evidence to discipline the police or to include the evidence to seal the conviction of a guilty defendant, we are weighing a known and serious cost against an uncertain and probably modest gain. That should be an easy call. 17 Michigan v. Tucker, 417 U.S. 433, at 447 (U.S. 1974). 18 Irvine v. California, 347 U.S. 128, at 136–7 (U.S. 1954). (Emphasis added.) P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 226 flawed rules of evidence and procedure Safeguarding the Moral Integrity of the Courts We now turn to the most common, but perhaps most disingenuous, justification of all for the warped character of so many exclusionary rules: the claim that the admission of evidence that was obtained illegally or in a way that violated the rights of the defendant would have the court stooping to the level of the criminal himself. Judicial integrity – said to be essential for fostering a respect for the law in broader society – demands that the courts not become tainted by admitting any evidence that smacks of the illegal or the deviously obtained. Justices Louis Brandeis and Oliver Wendell Holmes, in the 1920s, were among the first to articulate this new standard of purity for the judiciary. In their famous dissent in Olmstead v. U.S., they added their enormous influence to the idea. Brandeis, for his part, said: In a government of laws, existence of the government will be imperiled if it fails to observe the law scrupulously. Our Government is the potent, the omnipresent teacher. For good or for ill, it teaches the whole people by its example. Crime is contagious. If the Government becomes a lawbreaker, it breeds contempt for law; it invites every man to become a law unto himself; it invites anarchy. 19 Justice Holmes added, for good measure: We must consider the two objects of desire, both of which we cannot have, and make up our minds which to choose. It is desirable that criminals should be detected, and to that end that all available evidence should be used. It also is desirable that the Government should not itself foster and pay for other crimes, when they are the means by which the evidence is to be obtained. . . . We have to choose, and for my part I think it a less evil that some criminals should escape than that the Government should play an ignoble part. . . . If the existing code does not permit district attorneys to have a hand in such dirty business it does not permit the judge to allow such iniquities to succeed. 20 Thirty-two years later, Chief Justice Earl Warren was sounding the same theme. Excluding illegally acquired evidence, he insisted, “serves another vital function – ‘the imperative of judicial integrity’.” 21 In this ruling from 1960, he was content to hold that illegally seized evidence should be excluded to keep the court’s hands clean. By 1968, the chief justice had honed his rhetoric from the prosaic to the sublime. Illegally seized evidence had to be excluded, he said, because to include it would have the “necessary effect of legitimizing the conduct which produced the evidence, while an application of the exclusionary rule withholds the constitutional imprimatur.” For good measure, he added that “Courts which sit under our Constitution cannot and will not be made party to lawless invasions of the constitutional rights of citizens.” 22 19 Olmstead v. U.S., 277 U.S. 438, at 485 (1928). 20 Ibid., at 470. 21 Elkins v. U.S., 364 U.S. 206, at 222 (1960). 22 Terry v. Ohio, 392 U.S. 1, at 13 (U.S. 1968). P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 Dubious Motives for Flawed Rules 227 If I seem unduly cynical about Justice Warren’s treatment of this question, it is because it is impossible to square his discourse of “preserving judicial integrity” with scores of practices that are routinely admitted by the courts, that have been endorsed countless times by the Supreme Court itself, and that show a remarkable indifference to tainting by illegality. Confronted with Holmes’s dilemma between a) excluding relevant evidence, if illegally obtained, and b) admitting such evidence and the government fostering illegal acts by accepting tainted evidence, the legal system does not uniformly prefer a to b. That is, the courts routinely make themselves “party to lawless invasions of the constitu- tional rights of its citizens.” Consider just a fewof the circumstances in which courts showlittle reticence about using illegally seized evidence: r Evidence seized illegally by the police fromalmost anyone except the defen- dant himself will be unproblematically admitted into evidence. 23 r Evidence seized illegally from the defendant himself by someone other than the police will be freely admitted. r Evidence seized illegally from the defendant by the police may be admitted if the defendant testifies in his own defense. r If the defendant is illegally arrested by the police, he can still be made to stand trial, even though he is in police custody only by virtue of an illegal act. r Courts voice few if any qualms about the use of “tainted,” illegally seized evidence in grand jury proceedings nor at the sentencing hear- ing for the defendant (even though decisions in sentencing hearings can impose costs on the defendant as devastating as those associated with a conviction). r At a preliminary hearing before a judge – which may issue in a war- rant for arrest or search – the suspect is not allowed to object to the introduction of evidence “on the ground that it was acquired by unlawful means.” 24 r If a defendant, on bail awaiting trial, skips bail, then the court authorizes the bondsman to take almost any steps that are necessary to return the defendant for trial, regardless of whether those measures break the law or violate the rights of the defendant. 23 The Supreme Court wrote in U.S. v. Payner: The supervisory power of the federal courts does not authorize a court to suppress otherwise admissible evidence on the ground that it was seized unlawfully froma third party not before the court. Under the Fourth Amendment, the interest in deterring illegal searches does not justify the exclusion of tainted evidence at the instance of a party who was not the victim of the challenged practices (447 U.S. 727 [U.S. 1980]). 24 Federal Rules of Criminal Procedure, Rule 5.1(a). P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 228 flawed rules of evidence and procedure This last case is a particularly intriguing one, since it is the courts them- selves that are authorizing in advance the illegal behavior in question. Here is what the still reigning judicial opinion says about the powers of bail bondsmen: Whenever they choose to do so, they may seize [the defendant] and deliver him up in their discharge; and if that cannot be done at once, they may imprison him until it can be done. . . . They may pursue him into another State; may arrest him on the Sabbath; and . . . may break and enter his home for that purpose. . . . The bail [bondsmen] have their principal on a string, and may pull the string whenever they please, and render him in their discharge. 25 Although penned more than a century ago by a U.S. district court judge, this opinion continues to describe the awesome, lawbreaking, and rights-flouting authority that the courts routinely give to bondsmen. Evidently, these policies do not implicate the courts in the associated wrongdoing, since courts will happily accept for trial an illegally seized person who jumped bail. If these practices do not trouble the courts, I cannot conceive why the admission of highly relevant but illegally seized evidence should do so. To be clear, I am not criticizing the courts for having these policies of admit- ting various sorts of evidence, without inquiring too meticulously into howthey were obtained. Any robust theory of evidence would do the same thing. What I find troubling is the unevenness with which this policy is applied. It is as if, whenever the courts find it convenient or otherwise suiting the justices’ fancies, they mount their moral high horses and exclude relevant evidence, invoking the “transcendent” need to protect the image of the courts and to avoid the taint of illegality. In other, common enough, situations, they turn a decidedly blind eye to questions about the provenance of the evidence, now draping themselves in the importance of admitting relevant information. There is nothing visibly principled about this opportunism. There might be, if, for instance, the courts were willing to overlook minor infractions of the law or violations of defendant rights while scrupulously enforcing the major ones. This could be seen as a proportionate response to the level of police wrongdoing. But that is not what divides the cases we have been examining. On the contrary, the most objectionable formof police illegality is arguably that of seizing not a person’s property, but his very person without just cause. The deprivation of one’s liberty is surely more troubling than having one’s house entered or one’s phone conversations tapped without a warrant. Nonetheless, courts will routinely throw out evidence from illegally seized entries or from recordings made without probable cause, while they will virtually never let a defendant, who was illegally apprehended without probable cause, escape trial if there is substantial evidence against him. 25 83 U.S. (16 Wall.) 366, at 371–2 (1872). P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 Dubious Motives for Flawed Rules 229 In many jurisdictions, the courts then add insult to injury by insisting that if John Q. Citizen, arrested without probable cause, is subsequently tried and convicted of a crime, he cannot turn around and sue the police for false arrest. The system denies him the one traditional remedy provided by the common law to protect his rights. The courts that endorse and enforce this policy are the same ones who argue that the exclusion of illegally seized but relevant evidence is essential to preserve the clean hands of the justice system. Examples of this sort clearly do not resolve the problem posed by the two horns of the dilemma that Justice Holmes described a few pages earlier. They simply accentuate the difficulty of the choices faced by the judicial system. I suspect that, confronted with some of these examples, few of us would pursue the purist line that the judicial system must never be seen to be a party to illegality. If (for example) the police falsely arrest Jones for loitering only to discover, on getting him to the police station, that he is a wanted serial killer, fewwould hold that Jones should be released fromcustody, even though, but for the false arrest, he would not now be in the hands of the police and the courts. In other cases (use by police of torture and intimidation to secure a confession), our intuitions incline us the other way. So, I think we must reject Holmes’s way of posing the dilemma. He writes as if we must choose between steadfastly pursuing epistemic values or rigidly maintaining the integrity of the courts. I would prefer a solution that acknowl- edged that a significant degree of compromise was called for, rather than a pol- icy of consistently promoting one value over the other. Consider the following examples: Concerning the defendant’s silence, we could recognize that as a right but leave juries free to make what they will of that silence (given the kind of warnings to jurors about overinterpretation discussed in Chapter 6). Concerning confessions, we could stipulate that all interrogations of arrested subjects must be videotaped and preserved. No police station confession would be introduced into evidence unless it was so recorded. Judges can then drop the charade of the tests of voluntariness and corroboration. The jurors can watch the confessions themselves and judge whether they were made under unacceptable conditions of coercion or intimidation. The right not to be retried after a first trial could be retained, on the under- standing that the first trial has not ended until the relevant appeal fromthe losing party has been heard (whether that is the defendant or the prosecutor). As for the relationship privileges, we might retain the attorney-client privi- lege because it may (and I stress the modal verb) work to promote the interests of truth finding. All the other privileged relationships should be dropped in an epistemically ideal world, but I would not oppose their retention too vigorously, provided jurors could be informed (and left free to draw whatever conclusions seem appropriate) whenever the defendant has exercised such a privilege. P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 230 flawed rules of evidence and procedure As for illegally seized evidence, the balancing act becomes more difficult. The current compromise of excluding such evidence until and unless the defen- dant testifies – and then releasing it in abundance – is madness, not least because such a policy massively discourages testimony from the one person most likely to have knowledge of the crime. One alternative would be to retain the exclusion of such evidence per se, while dropping the poison fruit doctrine, as has already been done in the case of Miranda rights. I continue to believe that – given the other avenues for rectifying this problem (especially through civil actions) – relevant evidence seized from the defendant should be admitted, regardless of how it was obtained. The Supreme Court, in a unanimous opinion from 1976, seemed to concur: The costs of applying the exclusionary rule even at trial . . . are well known: the focus of the trial, and the attention of the participants therein, are diverted from the ultimate question of guilt or innocence that should be the central concern in a criminal proceeding. Moreover, the physical evidence sought to be excluded is typically reliable and often the most probative information bearing on the guilt or innocence of the defendant. 26 That notwithstanding, a reluctance to do anything about the problem persists. Courts could temper their worries about taint in this case by stressing that the victims of illegal searches have a variety of other channels through which they can seek redress and satisfaction, provided they can prove that their rights were violated. None of these fixes would satisfy the zealots on either side of this debate. Perhaps that is not such a bad thing. You can draw your own conclusions about which trade-offs are appropriate. What is indisputable is that far fewer false verdicts would issue from these compromises than from the current regime, while traditional (as opposed to certain recent and court-invented) rights of defendants would remain largely intact. Conclusion Half a century ago, the Supreme Court attempted to grapple with the charge that the law of evidence was a mess: We concur in the general opinion of courts, textwriters and the profession that much of th[e] law is archaic, paradoxical and full of compromises and compensations by which 26 Stone v. Powell, 428 U.S. 465, at 489–91 (U.S. 1976). They added: The disparity in particular cases between the error committed by the police officer and the wind- fall afforded a guilty defendant by application of the rule is contrary to the idea of proportionality that is essential to the concept of justice. Thus, although the rule is thought to deter unlawful police activity in part through the nurturing of respect for Fourth Amendment values, if applied indiscriminately it may well have the opposite effect of generating disrespect for the law and administration of justice (at 490–1). P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 Dubious Motives for Flawed Rules 231 an irrational advantage on one side is offset by a poorly reasoned counterprivilege to the other. But somehow it has proved a workable even if clumsy system when moderated by the discretional controls in the hands of a wise and strong trial court. To pull one misshapen stone out of the grotesque structure is more likely simply to upset its present balance between adverse interests than establish a rational edifice. 27 This is nothing but delusionally wishful thinking parading as sage advice. Ours is supposed to be a government of laws, not of men. Yet, the Supreme Court tells us to rest assured: The profoundly compromised law of evidence and procedures is just fine, because it is “in the hands of wise and strong” trial judges. But how many of those are there? Wisdom seems to me to be in rather short supply in every walk of life. One ought no more to expect it from judges than from anyone else. Indeed, if judges were genuinely and uniformly wise, we would have little need for evidence law. We have rules of evidence and procedure precisely because we don’t want to trust the outcome of a trial to the intuitive hunches of the old and the wise. In this passage, the Supreme Court is pleading to leave the system as is. We are asked to believe that it has evolved in such a marvelous way that the seemingly dysfunctional and contradictory parts all work together to yield a viable system of which we may all be proud. Any change would upset its delicate equilibrium. I beg to differ. The Federal Rules of Evidence currently consist of sixty-seven rules, many with numerous subclauses. The Federal Rules of Criminal Procedure involve some sixty additional rules. Most of the former articulate exceptions to the rule of relevance and are thus truth thwarting. Many of the latter impose procedures that make it harder than it should be for juries to reach valid verdicts. Beyond this, there is a huge body of appellate jurisprudence, imposing precedents that frequently give an even less truth-conducive twist to the rules of evidence and procedure than the explicit statement of those rules already implies. It thus comes as no surprise that prospective criminal attorneys in lawschools – instead of studying the nuances of relevance and how to evaluate it – learn in exquisite detail how easy it can be to get relevant evidence excluded. Repeatedly, the Supreme Court has insisted, as it did in 1979, “that the function of legal process is to minimize the risk of erroneous decisions.” 28 The analysis in this book raises serious doubts about whether the courts have made a good faith effort to put that commendable philosophy into practice. Everywhere we look – whether to the standard of proof, the rules of evidence, the rules of procedure, or the asymmetry of appellate review – we see a system doing precious little to reduce the likelihood of erroneous decisions. Throughout this book I have focused on the justice system as a system of inquiry. I have tried to identify some features of the system that make it less likely that truthwill emerge or that error will be avoided, andhave recommended 27 Michelson v. United States, 335 U.S. 469, at 485–6 (1948). 28 Addington v. Texas, 441 U.S. 418, at 425 (1979). P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 232 flawed rules of evidence and procedure changes in the rules to make them more truth sensitive. The depressing fact is that this review of selected rules of evidence and procedure has left virtually unmentioned many other rules and practices that are as epistemically deviant as those discussed here. The law’s treatment of hearsay and of eyewitness testimony and its handling of expert witnesses are but three obvious examples that could be added to the already long litany of obstacles to truth seeking. Worse, it is not the trial alone and the preliminary hearings leading up to it – which have been my foci here – that demand epistemic analysis. During the pre-arrest stage of a criminal investigation, there are fascinating questions about probable cause and probable suspicion that I have barely touched upon here. Likewise, sentencing hearings and grand jury proceedings have their own sets of rules and their own standards of proof. All these topics could profit from serious epistemological scrutiny. I alleged in the first chapter that the American criminal justice systemis not a systemthat anyone principally concerned with finding out the truth about crimes would have devised. The criminal SoP is ill thought out, barely explained to jurors (if at all), and not integrated with those other basic doctrines (the BoD, the ratio of errors, the PI, and the burden of proof) to which it is intimately related. Many of the rules of evidence and of procedure pose quite unnecessary obstacles to finding out the facts about a crime. I trust that the arguments in this book bear out the claim that the rules of evidence, like the proverbial Topsy, just grew up to be the way they are, often lacking either rhyme or reason. Despite strenuous efforts over the last century to codify the rules, they remain a patchwork of truth-enhancing and truth-thwarting compromises and half-measures. More importantly, I hope that the conceptual machineryI have beenexploring here will encourage others to participate in the thought experiment of critically thinking about what a justice systemwould look like if it were really committed to the idea that the aim of a trial is to find out the truth about a crime. At the moment, we still have only the most rudimentary conceptual machinery for talking about the law in an epistemically subtle way. The message of this book – that the rules of evidence and procedure and the various principles for distributing error need to be drastically rethought – is scarcely a newone. Two centuries ago, Jeremy Benthamwas trumpeting it to all who would listen. In the early-twentieth century, a host of scholars in the lawof evidence, most prominently Charles McCormick, called for a vast overhaul of the rules that bar jurors’ access to relevant evidence. 29 The gloomy reality that 29 McCormick, optimistically but not very presciently, opined: “So we have said that the hard rules of exclusion will soften into standards of discretion to exclude. But evolution will not halt there. Manifestly, the next stage is to abandon the systemof exclusion” (Charles T. McCormick, Tomorrow’s Lawof Evidence, 24 A.B.A. J. 507, at 580–1 [1938]). Sixty-plus years on, we are still encumbered with most of the exclusionary rules McCormick was rail- ing against, and a nontrivial bevy of new ones. McCormick’s older contemporary, Charles Chamberlayne, likewise predicted – commendably but erroneously – that the exclusionary P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 Dubious Motives for Flawed Rules 233 these earlier calls to action produced very limited results must give us pause about the amenability of the legal system to change, however outrageous the current conditions. Powerful forces, most obviously the bar itself, are arrayed against any change that might make the system simpler, more transparent, and less subject to cynical manipulation by the actors within it. When even the Supreme Court can say, with a straight face, that, while many individual rules look cockeyed or otherwise ill thought out, taken together they produce a “workable” system of criminal justice, it is hard to see where change might emerge. The proper riposte to such a silly argument is that trial by ordeal was considered workable for several centuries as was the system of judicial torture that replaced it. Neither, unfortunately, was very good at finding out the truth. Nor is ours. However limited the prospects for practical reformmay be, it remains impor- tant to do the theoretical spadework that is necessary simply to grasp how well or badly the current system is working. That activity, the epistemology of the law, is inexplicably still a nascent subject. It deserves to be more than that, for any purported system of inquiry that does not bother to discuss candidly the legitimacy of its claims to truth and rationality is no system of inquiry at all. This book has not aimed to end that conversation with a definitive story. I would be delighted if it simply rekindles it. rules would soon atrophy since “any rule which excludes probative or constituent facts actually necessary to proof of proponent’s case is scientifically wrong” (The Modern Law of Evidence and Its Purpose, 42 American L. Rev. 757, at 765 [1908]). P1: JZZ 0521861667c09 CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 18:15 234 P1: JZZ 0521861667ind CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 19:15 Index adversarial system, 24–5, 167, 173, 203 affirmative defenses, 35, 86, 110–14, 125, 127, 143 Allen, Ronald, xii, 54–5, 70, 73, 82–3, 89, 107, 147 Alschuler, Albert, 223 Amar, Akhil, 154, 163–4, 194, 211, 222 appellate asymmetry, 3, 17, 96, 131, 138–9, 194, 196–207, 209–11, 231 Ayling, Corey, 128–9, 134 bail hearing, 95 benefit of the doubt, xi, xv, 29–30, 59, 64–6, 76, 89, 91, 109, 117, 123–7, 131–2, 136, 138–9, 141, 232 Bentham, Jeremy, 89, 147, 151, 168–9, 171, 232 Berger, Maria, 159, 161 Black, Hugo, 34, 131, 186, 204 Blackmun, Harry, 35, 45 Blackstone, William, 63, 66, 74–5, 88, 119, 130, 196 Brandeis, Louis, 189, 226 Brennan, William, 45, 90, 132–3 capital punishment, 55, 57–61, 197 Cardozo, Benjamin, 147, 162, 223 Cassell, Paul, 126, 182 certainty moral, 33, 35, 38–40, 51, 54, 80 Chambers, Harry, 11 civil law, 8, 44, 48, 51, 55–6, 64, 66–7, 97, 109–10, 130–1, 161, 190–1, 219, 230 Cohen, John, 42, 47–8 common law, xii, 1, 4, 6, 18, 32–4, 57, 81–2, 90, 120, 134, 151, 155–6, 173–4, 183, 188–9, 196, 223, 229 confessions, 8, 15, 19–20, 65, 81, 125–6, 128–37, 151, 166–7, 171–84, 189, 193, 209, 214–15, 217–18, 223–4, 229 corroboration of, 172–3, 178–82, 184, 214, 229 Miranda rule, 126, 148, 157, 173–8, 180–3, 192, 209, 216, 224–5, 230 voluntariness test for, 65, 125, 132–3, 135, 137, 172–8, 180–1, 184, 209, 224, 229 Damaˇ ska, Mirjan, 217–18 Dekay, Michael, 66, 68, 73, 119 Dripps, Donald, 71 due process, 31, 64, 91–2, 126, 156 Dworkin, Gerald, 9 epistemology, xi, 2–3, 7, 9, 30, 33, 58, 60, 64, 118, 121, 157, 171, 179, 233 legal, 2, 89, 110, 117, 122, 214 235 P1: JZZ 0521861667ind CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 19:15 236 index error, xiii, 1–5, 7–10, 12–14, 16–17, 20–1, 29–30, 41, 46, 48, 58–60, 68–9, 72–5, 87, 89, 92, 99, 113–14, 117, 119, 123–4, 127–34, 136, 142, 144, 155, 160, 164, 190, 194–206, 209–11, 214, 230–2 distribution of, 2–3, 29–30, 56, 64, 66–8, 73–7, 89, 105, 117–18, 123–4, 127–38, 142, 144–5, 177, 211, 214, 232 false acquittals, xv, 1, 11–12, 14–15, 20, 25, 29, 67–76, 88, 108–9, 112, 117–19, 123–4, 126, 128, 130–2, 134, 138–40, 144, 150–2, 160, 177, 190, 196, 198, 204–6, 209, 211 false convictions, xv, 1, 11–12, 20, 25, 29, 55, 58–9, 67–77, 85–8, 108–9, 111–12, 114, 117–19, 123–6, 128–32, 134, 138–40, 144–5, 150, 152, 160, 163, 177, 195–7, 204, 207, 209–11, 214 false verdicts, 1, 4, 8, 10–11, 13, 15, 126, 130, 137, 140, 205, 210, 230 invalid verdicts, 13–14, 71, 142, 195, 204, 210 reduction of, 1–3, 117, 119, 123–4, 130, 136, 204 relative costs of, xv, 8, 13, 22, 25, 33–4, 43, 51, 55–6, 61, 66, 68–76, 87–9, 105, 111–13, 117, 123, 126–8, 130, 132–3, 136, 138, 140, 144, 157, 164, 183, 202, 209, 216 evidence circumstantial, 54, 82–4, 163 rules of, 3, 5, 7, 14–15, 73, 76, 108, 117–20, 122–4, 127–32, 136, 138, 141, 145, 147, 155, 177, 183, 191, 194–5, 207–10, 213–14, 216, 231–2 eyewitness testimony, 17–19, 23, 33, 41, 43, 53–4, 79, 82, 84, 97, 120, 122, 129, 131, 134, 137, 140, 143, 147, 151–2, 155, 157–9, 162–6, 169, 172–3, 187–8, 199–200, 206, 216–17, 219, 222–4, 230, 232 false arrest, 97, 173, 229 Feeney, Floyd, 70 Ferrer, Jordi, xii Frankfurter, Felix, 93, 155 Friedland, Martin, 202 Friedman, Richard, 100, 129, 203 Friendly, Henry, 152, 163 Galilei, Galileo, 5–6 Ginsberg, Ruth Bader, 35 Gold, Victor, 22 Goldberg, Arthur, 132, 152–4, 160, 220–2 Goldwasser, Katherine, 129–30 grand jury, 15–17, 86, 95–7, 219, 232 Grano, J., 154 guilt material, xv, 10–13, 67, 69–70, 72–3, 75, 83, 85, 96, 100–4, 106–8, 114, 118–20, 135, 145, 150–1, 211 probatory, xv, 12, 96, 102, 105 Hand, Learned, 27, 78 Hart, H. L. A., 9 Huff, Ronald, 71 In re Winship, 34–5, 45, 51, 56, 64, 73, 80, 91 innocence material, xv, 10–12, 66, 68, 73, 75, 85–6, 96–108, 113–14, 118–20, 132, 146, 151 presumption of, xi, xv, 11–12, 29–30, 43, 56, 89–98, 100–1, 103–7, 109, 111–12, 117, 125, 180, 207, 232 probatory, xv, 12, 96, 98–9, 103–7, 110 inquisitorial system, 152, 156 jury system, 7, 24, 34–5, 53, 214–15, 218–19 Kamisar, Yale, 70, 192 Kelsen, Hans, 9 Kramer, Geoffrey, 49 Langbein, John, 222 Laudan, Larry, 57, 81 Laufer, William, 91, 98 Leipold, Andrew, 70 P1: JZZ 0521861667ind CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 19:15 Index 237 Leiter, Brian, xi, xiii Leo, Richard, 126, 179 Levy, Leonard, 222 Lillquist, Erik, xii, 60, 66 Maimonides, Moses, 63, 144 Marshall, Thurgood, 90, 132 May, W., vii Mayo, Deborah, xiii McCormick, Charles, 117, 121, 165, 169, 190, 213, 232 Newman, Jon, 29 Peirce, Charles Sanders, 207–9 Posner, Richard, 15, 105, 158, 169, 192 Powell, Lewis, 1, 4, 7, 204, 230 principle of indifference, 75, 97, 106, 123–4, 128, 135, 144, 169, 206, 227 privacy, right to, 154, 162, 219–22 probability, 19–20, 33, 44–7, 56, 65–6, 72–3, 75, 77–9, 81, 84–5, 100–1, 104–7, 128, 140, 182 probable cause, 86, 95–6, 133, 137, 173, 178, 182, 185–6, 190, 192, 219–20, 224, 228–9, 232 probative value, 5, 13, 20–2, 24–6, 49, 120–2, 135, 157, 160–1, 177 procedure, rules of, 141, 143, 207, 231 proof beyond a reasonable doubt, xi–xii, xv, 4, 10, 15, 29–57, 60–2, 64–6, 71, 77–9, 81, 85–7, 90–2, 95–7, 99, 101–2, 104, 107–14, 129, 132–5, 141–2, 150, 154, 161, 195, 198, 201, 219 burden of, xi, xv, 29–30, 33, 43, 45, 48, 51, 89–92, 94, 99, 109–12, 114, 117, 162, 232 clear and convincing evidence, xv, 56, 64, 86, 112, 133 preponderance of the evidence, xv, 44, 56, 64–5, 78, 86, 110–12, 132, 135, 145, 178 standard of, xii, xv, 4, 10, 13, 15, 17, 29–32, 34–5, 44–5, 47, 50–1, 54–7, 60–1, 64–8, 70, 72–83, 85–9, 91–3, 97, 104–5, 107–9, 112–14, 117–19, 121, 123–8, 144–5, 154, 195, 214, 231–2 objective, 81 subjective, 77, 107 questions of fact, 216 questions of law, 206, 216 Rawls, John, 9 Rehnquist, William, 93, 192 relevance, xi, 4–5, 10, 13, 15–26, 44, 66, 73, 75, 84, 86, 97, 102, 104, 109–10, 117, 120–6, 128–9, 134–41, 143, 145, 149, 151–2, 157, 159, 161, 163–9, 173, 175–6, 178–85, 187–91, 196, 198, 201, 205, 210, 214–21, 225, 227–32 reliability, xi, 8, 17–18, 31, 50, 77, 82, 86, 120–3, 129, 136, 138, 141, 144–5, 147, 149, 158–9, 171, 176, 179–81, 184, 190–1, 213, 215, 217, 230 rights, 1–4, 6–9, 14, 18, 31–2, 34–5, 44, 47, 51, 55, 59, 64, 66–7, 69, 71, 78, 90–1, 93–6, 118–19, 123, 128, 132–4, 137–9, 143, 148–58, 160–2, 164, 170, 173, 175–8, 183, 185–6, 189–90, 194, 196–201, 203–4, 213–15, 217–230 rules discovery, 91, 124, 143 exclusionary, 10, 15–23, 25–6, 86, 91, 120–2, 125, 129–30, 132–40, 147, 149, 151–3, 156–61, 164–70, 172, 174–7, 179, 181–6, 188–91, 199, 206, 210, 214–17, 220, 222–4, 226–7, 229–30, 232 procedural, 141, 194, 218 Saltzburg, Stephen, 131, 135 Seeberger, R., 180 Seidman, Daniel, 130 Shapiro, Scott, 194 Shaviro, Daniel, 46 P1: JZZ 0521861667ind CUNY392B/LAUDAN 0 521 85930 1 April 19, 2006 19:15 238 index Shaw, Lemuel, 33–5, 38 Sheppard, Steve, 98 Silent Defendant, ix, 147, 150–3, 155, 159, 222 Silent Witness, ix, 147, 162–3, 171, 221 Souter, David, 45 Starkie, Thomas, 46, 80 Stein, Alex, 129–30, 175 Stephan, Paul, 112, 114 Stith, Kate, 202, 205–6 Sundby, Scott, 98 Taruffo, Michele, xii Thaler, Jeff, 98 thought experiment, 5–7, 120, 123, 147, 191, 213, 232 Tribe, Laurence, 47, 98 Underwood, Barbara, 78, 114 Uviller, Richard, 154 van Kessel, Gordon, 157, 171 verdicts false, 3, 6, 8, 13, 131, 177 true, 3, 6, 8, 13, 32, 88, 123, 125, 131, 144, 177 valid, 12–14, 31, 62, 71, 73, 88, 130, 138, 141, 162–3, 190, 195, 220, 231 Voltaire, 63 Warren, Earl, 209, 226–7 White, Byron, 20, 91–2, 115, 177 Wigmore, John Henry, xii, 164, 221 Zeisel, Hans, 70, 139