criminal law

Published on January 2017 | Categories: Documents | Downloads: 69 | Comments: 0 | Views: 470
of 256
Download PDF   Embed   Report

Comments

Content


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 uxiviisiry 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-s::-ïo:oo-+
isnx-:+ ,•-c-s::-::s::-s
© Larry Laudan 2006
2006
Information on this title: www.cambridg e.org /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-s::-::s::-c
isnx-:c c-s::-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
www.cambridge.org
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.
deathpenalty.org.
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

Sponsor Documents

Or use your account on DocShare.tips

Hide

Forgot your password?

Or register your new account on DocShare.tips

Hide

Lost your password? Please enter your email address. You will receive a link to create a new password.

Back to log-in

Close