The Language of Statutes: Laws and Their Interpretation 9780226767987

Citation preview

The Language of Statutes

the chicago series in law and society Edited by John M. Conley and Lynn Mather Also in the series: specializing the courts  by Lawrence

Baum

asian legal revivals: lawyercompradors and colonial strategies in the reshaping of asian states by

speaking of crime: the language of criminal justice by Lawrence M. Solan

and Peter M. Tiersma

Yves Dezalay and Bryant G. Garth

human rights and gender violence: translating international law into social justice by Sally Engle Merry

invitation to law and society: an introduction to the study of real law

just words, second edition: law, language, and power by John M. Conley

by Kitty Calavita

and William M. O’Barr

belonging in an adopted world by

distorting the law: politics, media, and the litigation crisis by William

Barbara Yngvesson

making rights real: activists, bureaucrats, and the creation of the legalistic state by Charles R. Epp lawyers on the right: professionalizing the conservative coalition by Ann Southworth arguing with tradition: the language of law in hopi tribal court by Justin B. Richland

Haltom and Michael McCann

justice in the balkans: prosecuting war crimes in the hague tribunal by

John Hagan

Additional series titles follow index

The Language of Statutes Laws and Their Interpretation

l aw r e n c e m . s o l a n

the university of chicago press    chicago and london

lawrence m. solan is the Don Forchelli Professor of Law and director of the Center for the Study of Law, Language, and Cognition at Brooklyn Law School. He is the author or coauthor of three books, including The Language of Judges and Speaking of Crime: The Language of Criminal Justice. The University of Chicago Press, Chicago 60637 The University of Chicago Press, Ltd., London © 2010 by The University of Chicago All rights reserved. Published 2010 Printed in the United States of America 19  18  17  16  15  14  13  12  11  10   1  2  3  4  5 While the bulk of the work reported in this volume is new, some of it is adapted from earlier publications, which include The New Textualists’ New Text, 38 Loyola L.A. L. Rev. 2027 (2005); Private Language, Public Laws: The Central Role of Legislative Intent in Statutory Interpretation, 93 Geo. L.J. 427 (2004) (reprinted with permission of the publisher, Georgetown Law Journal © 2004); Jurors as Statutory Interpreters, 78 Chi.-Kent L. Rev. 1281 (2003); Statutory Inflation and Institutional Choice, 44 Wm. & Mary L. Rev. 2209 (2003); and Law, Language, and Lenity, 40 Wm. & Mary L. Rev. 57 (1998). isbn-13: 978-0-226-76796-3 (cloth) isbn-10: 0-226-76796-5 (cloth) Library of Congress Cataloging-in-Publication Data Solan, Lawrence, 1952–   The Language of statutes: laws and their interpretation  /  Lawrence M. Solan.    p. cm.   Includes bibliographical references and index.   isbn-13: 978-0-226-76796-3 (cloth : alk. paper)   isbn-10: 0-226-76796-5 (cloth : alk. paper)  1. Law—Interpretation and construction.  I. Title.   k290.s65 2010   348'.02—dc22 2010005711 a The paper used in this publication meets the minimum requirements of the American National Standard for Information Sciences—Permanence of Paper for Printed Library Materials, ansi z39.48–1992.

for my parents, harold and shirley solan

Contents Acknowledgments  ix chapter 1. Laws and Judges  1 chapter 2. Why We Need to Interpret Statutes  16 chapter 3. Definitions, Ordinary Meaning, and Respect for the Legislature  50 chapter 4. The Intent of the Legislature  82 chapter 5. Stability, Dynamism, and Other Values  120 chapter 6. Who Should Interpret Statutes?  160 chapter 7. Jurors as Statutory Interpreters  196 chapter 8. Legislatures, Judges, and Statutory Interpretation  223 appendix.

Two Mail Fraud Jury Instructions  231

Notes  237 List of Cases  283 Index  285

Acknowledgments

T

his book has been a long time in the making, and as a result, I owe a debt of thanks to many people whose ideas and suggestions have made their way into this publication. Many colleagues at Brooklyn Law School will recognize their input. Among them are Neil Cohen, Susan Herman, Claire Kelly, and Jason Mazzone. Throughout my career at Brooklyn, I have received generous support. I am grateful to the institution, and especially to Joan Wexler, the dean of Brooklyn Law School. I am also grateful to many research assistants who contributed valuable work, in particular Lauren Davis and Kim Finneran. Some of this work was completed while I was a visiting professor at Yale Law School. Bill Eskridge was especially helpful to me, both there and since my visit, including providing comments on the manuscript. An important insight by Jed Rubenfeld also plays a prominent role in the book. Steven Shepard’s work as a research assistant was invaluable. Alan Hurst will recognize his contribution with ease. I also conducted a great deal of the research for this book while a visiting scholar in Princeton’s Department of Psychology. Much of the psychological research discussed in this book was developed after many fruitful conversations with John Darley and Sam Glucksberg. Other colleagues have shared their views and thoughts about the ideas in this book. No doubt I will leave out some names and regret it. But that is not a good reason not to thank people with whom I have worked over the years and who have reacted to my work, whether in conversation or by reading earlier drafts. In this regard, I would especially like to thank Tony Sebok, Peter Tiersma, Spencer Waller, and Deb Widiss. Michael Sinclair twice provided suggestions that have made their way into this book, and I



acknowledgments

am very grateful to him for his thoughtful and careful reading of an earlier version. I have also benefited from presenting portions of the book at various universities. Among them are the law schools of Case Western Reserve University, Yale University, the University of Minnesota, the University of Southern California, University Pompeu Fabra, and Cardiff University. I have also presented portions of this book to the language and law programs at Cardiff University and Aston University, both in the UK, and as the Freeman Lecture at the University of Massachusetts. All of those experiences provided me with insightful comments, many of which have made their way into the book.

chapter one

Laws and Judges

T

his book is about the relationship between lawmakers and judges. More specifically, it is about how judges judge disputes about laws. In cases involving the common law, judges determine how well the facts of a dispute fit into the earlier body of decisions and attempt to reach the best outcome in the new case, which then becomes part of the tradition for the next judge to consider. Depending on which court in the judicial hierarchy makes the decision, the decision not only is advisory but also has official legal status as binding precedent under the doctrine of stare decisis. But now, much of the law is made not by judges but by legislatures. Laws governing crime and punishment, trademarks, patents, copyright, securities, corporations, taxation, environmental regulation, antitrust, the sale of goods, and insurance are all enacted through legislation. When a legislature passes a law in a particular domain, common-law judges must give the statute priority over their own values and defer to the legislative judgment. At times, the legislature delegates to agencies the authority to write rules to implement the statute. These too have the force of law. For the most part, courts are also obliged to subordinate their own judgment to those of the rule makers. Common-law judges are also charged with interpreting statutes, which they have done for centuries.1 Some believe that judges, so accustomed to having the last word in common-law cases, have not been willing to adjust



chapter 1

to the role of taking a back seat to the legislature. Instead, judges attempt to legislate beyond their authority by imposing their own glosses and values on statutes that should simply be applied as the legislature wrote them. The most prominent such critic is Supreme Court Justice Antonin Scalia. In his book A Matter of Interpretation, he remarks: “But though I have no quarrel with the common law and its process, I do question whether the attitude of the common-law judge—the mind-set that asks, ‘What is the most desirable resolution of this case, and how can any impediments to the achievement of that result be evaded?’—is appropriate for most of the work that I do, and much of the work that state judges do. We live in an age of legislation, and most new law is statutory law.”2 Justice Scalia is by no means alone in his concern. For example, Adrian Vermeule suggests that judges should eschew virtually all interpretive principles. They should instead, he claims, apply laws by their plain language when possible and defer to administrative agencies (including prosecutors) when there is some uncertainty about a statute’s meaning.3 Because judges bent on finding the intent of the legislature have no reliable methodologies that will ensure that they have accomplished their goal successfully, the argument goes, the system would be better off if judges applied rules mechanically and let the experts take over when decisions must be made.4 Others, in sharp contrast, believe that the common-law tradition provides a special opportunity for judges to continue to do justice, even though so much of the law is statutory. Guido Calabresi takes this position in his book A Common Law for the Age of Statutes, whose title states his thesis. Calabresi asks: “What, then, is the common law function to be exercised by courts today? It is no more and no less than the critical task of deciding when a retentionist or revisionist bias is appropriately applied to an existing common law or statutory rule. It is the judgmental function . . . of deciding when a rule has become sufficiently out of phase with the whole legal framework so that, whatever its age, it can only stand if a current majoritarian or legislative body reaffirms it.”5 This position argues that because legislatures move slowly— or not at all—it is only by some reasonable sharing of power between the two branches of government that any reasonable legal system can sustain itself without irreparably compromising the goal of doing justice. William Eskridge, in his important book Dynamic Statutory Interpretation, also focuses on the need for courts to recognize doctrinal development notwithstanding laws enacted pursu-

laws and judges



ant to a constitutionally mandated process, which, at least in principle, is intended to give the laws primacy over the value systems of individual judges.6 Whether one believes that judges should wield more power or less power in the age of statutes, there can be no doubt that judges actually do continue to sound like common-law judges even when they are interpreting laws. Although they purport to defer to legislative judgment, they indeed make their own judgments about which pieces of the legislative history and other social facts surrounding the enactment of a law tell us what the legislature really had in mind; they adhere to the common-law principles of precedent, so that an earlier interpretation of a statute— even a demonstrably bad interpretation— continues to have binding effect on future cases; they create all kinds of “canons of construction,” ranging from assumptions about the resolution of grammatical ambiguity to the rule of lenity, which calls for ambiguities in criminal statutes to be resolved in favor of the defendant; and they impose on legislators “plain statement rules,” through which they warn lawmakers to draft disfavored provisions in an especially “clear” manner if they expect courts to enforce them. Even when the legislature reacts to a court decision by changing the law to override the judges, courts often continue to refer to their own precedents, construing the new law narrowly, as if the legislature were an inconvenience whose effect on the decision making of judges should be minimized to the extent possible.7 When the language of the law leaves uncertainty, and it predictably does leave uncertainty, then discretion is unavoidable, whether we like it or not. For those concerned with there being a crisp rule of law conveyed in language that we can understand and comply with, this is unfortunate, since these interpretive gaps permit argument on both sides of an issue that is consistent with some legitimate understanding of the language, on the one hand, and with at least some set of values sufficient to justify the interpretation, on the other. No wonder the personal values of the individual judge seep in to the statutory analysis. Judges exercise even more judicial power when the language of a law appears sufficiently at odds with a reasonable application in a particular situation that it seems only fair to construe the law as not applying. As Judge Richard Posner observes, no one would arrest a prosecutor for possession of child pornography when her only reason for having it is to use it as evidence against a pornographer.8 Judges and scholars disagree about how



chapter 1

far such judicial power may be taken. Some would limit it to the erasure of absurd results, while others would go a little further in an effort to enforce the statute’s intent even when the literal application is not so bad as to be absurd. Ostensible errors in the language used by the legislature opens the door to judicial interpretation. Despite all this judicial activity in the realm of statutory law, as the argument of this book unfolds I hope to show that laws work fairly well. Most of the time most people understand their obligations well enough, and most of the time the law’s application is clear. Hard cases arise because of a gap between our ability to write crisp yet appropriately flexible laws and the design of our cognitive and linguistic faculties. That gap is small enough to permit some to fall prey to the illusion that it is even smaller than it actually is. But it does exist, and as we will see, beginning in chapter 2, most of the time it manifests itself in linguistically predictable circumstances. Yet it would be a mistake to make too much of this gap. If one focuses only on the highly contested cases decided by the U.S. Supreme Court, one might sensibly infer that legal interpretation is a mess, full of arbitrary decisions. But if one focuses on all the times that a law applies without even generating a dispute, and on the times that it is clear enough to resolve a dispute without much difficulty, then the hard cases become just that: difficult situations embedded in a system of order that appears to be working fairly well. I argue in this book that this is what happens, and thus I am neither terribly worried about the fact that judges must actually make decisions some of the time or about their seemingly having too little power to do so. As long as judges constrain themselves by staying within the range of reasonable interpretations that the language of the statute affords or, in unusual situations, by articulating a good reason for not doing so (such as a legislative error or an obviously anomalous result), legislators really will be given primacy over judges in forming and applying statutory law. And judges typically do so. Judges of all political stripes repeatedly articulate legislative primacy as an overarching value in the decisionmaking process. They often disagree about what evidence of the legislative will is legitimate and useful, and they weigh differently other values, such as fair notice, coherence, adherence to constitutional norms, and stability. All this disagreement, however, comes within a fairly narrow range of discourse in which it is a relatively straightforward matter to identify the many simple cases and to apply laws without controversy. As for the hard cases, only a naïve apologist could ignore the fact that judges’ per-

laws and judges



sonal values contribute to their decisions. It can be no accident that the five conservative Supreme Court justices, in recent cases, construed an ambiguous labor statute in favor of the employer9 and a regulatory statute in favor of the tobacco industry and against regulation,10 whereas the four more liberal justices did just the opposite. In both of these cases, the dueling justices evoked the intent of the legislature to justify their opposing positions. Judges are acutely aware of the ramifications of their decisions11 and cannot help but steer the legal system in a direction they believe to be the best course when more than one outcome is licensed by a statute whose application is not sufficiently clear in a particular case. How much one is troubled by this observation should depend upon the extent to which judges are constrained and the relative frequency of straightforward cases to ones in which political judgment is permitted to masquerade as close legal analysis. I maintain that because language works quite well as a vehicle for conveying rules, and because judges indeed take seriously the concept of legislative primacy as a value, the residue of unrestrained political judgment is well within tolerable limits. Because my argument depends upon the salience of legislative primacy as a value, much of this book is devoted to developing that concept and defending its legitimacy. For now, though, let us turn to some easy cases.

Laws Work . . . Most of the Time In our everyday lives, we can generally tell in advance what we are required to do, what we are permitted to do, and what benefits we have from the obligations of others. Consider the person who commutes to work from New Jersey to New York by train. A host of laws regulate his commute. Passengers are not permitted to get on or off a moving train,12 they must pay the fare,13 and they are not allowed to walk on the tracks.14 Passengers must also obey the criminal law and refrain from violence toward the train crew, and they must heed the signs that say it is prohibited to pull the emergency stop cord without a good reason. We barely notice such rules. They conform to our own norms of conduct, so for the most part we would act in accordance with them even if they were not articulated. This may not be so for everyone, however. Some people might be thrillseekers who would get pleasure out of walking on the railroad tracks with a train close by. Others might be antisocial enough to enjoy pulling the



chapter 1

emergency stop cord for no legitimate reason. The rules are written to proscribe conduct that violates ordinary social norms but in which some people would engage if they were not told that they cannot do so. Compliance with these norms would be expected even if there were no laws making it illegal to violate them. Their clear articulation serves to legitimize legal sanctions against violators. Laws do not always work that well, however. Let us say that the commute also includes a ride on the New York City subway. It has long been illegal to ride between cars on the subway. Since 2005, it has also been illegal to move between cars, even when the train is stopped. The Transit Authority Web page has the following listed under its Rules of Conduct: “It is a violation to . . . [m]ove between end doors of a subway car whether or not train is in motion, except in an emergency or when directed by police officer or conductor.”15 Arrests for riding or moving between cars jumped from about 700 in 2005 to 3,600 the next year, and to more than 17,000 in 2007, reflecting a crackdown and the fact that most riders did not know about the new rule and were caught off guard when they behaved as they always had.16 Riders made aware of the new rule complained. There is a long tradition in New York of moving between cars to evade loud passengers, unwanted entertainment, broken air conditioners, and foul odors. New Yorkers are also often in a hurry and move between cars so they are closer to the appropriate exit in their destination station. Yet the rule has a purpose: to impede those who wish to engage in crimes on the subway system. A newspaper article quotes a Transit Bureau police chief: “We’ve tried to really hammer down on moving between cars. A lot of bad guys move between cars, really, prowling for victims.”17 The rule also addresses safety concerns. Deaths and injuries have occurred by virtue of passengers riding between cars, sometimes resulting in lawsuits against the Transit Authority. Small signs are posted on the end doors of subway cars. The signs have a cartoon-like picture of a person riding between cars with a diagonal red stripe through him to indicate that his conduct is prohibited. Until the new rule was put in place, they bore the words, “Riding Between Cars Prohibited.”18 Subsequently, the Transit Authority replaced these signs with ones that state the rule verbatim, but even several years after the rule was changed, some of the cars continued to have the sign pictured on the next page Moreover, the Web page entitled Subway Safety has only the following

laws and judges



[To view this image, refer to the print version of this title.]

figure 1.1.  Photo by Shannon Stapleton/ The New York Times/Redux. Reproduced with permission.

warning: “When you’re inside a moving train, never ride between cars or lean against doors. When you are standing, always hold on.”19 It does not mention the new rule. How should the law apply to a passenger who, without knowledge of the rule, walks from one car to another to avoid heat, noise, or smells while the train is stopped at a station? On the one hand, the rule is clear, and the passenger has violated the rule. It generally should not matter that he does not know the rule, since ignorance of the law is no excuse for disobedience unless the legislature has made such knowledge a prerequisite for conviction.20 On the other hand, the sign might give a misimpression as to what is really prohibited, and the Transit Authority’s Web page contains conflicting information, depending on whether one reads the safety tips or the rules themselves. The best result would be, I believe, for the police to explain the law and to let the passenger go. He meant no harm and will not move between cars again. Thus, the police themselves act as the first line of statutory interpreters. If they conclude that the purpose of the law is not being served by an arrest, they can simply not make the arrest and avoid the need for a decision about the legality of their actions. When police, prosecutors, or, for that matter, plaintiffs’ lawyers refrain from bringing legal cases that raise



chapter 1

difficult questions about the proper application of a law, they permit the system to treat laws that are unclear at the margins as though they were clear. We will see in chapter 6 that, for example, federal prosecutors bring very few criminal copyright cases, and the ones they bring are for clear, egregious violations of the Copyright Act. The result is that whatever uncertainty in meaning the copyright laws contain becomes irrelevant in the criminal context, since prosecutors do not raise such issues. They would rather devote the resources needed to try to expand judicial interpretation of the copyright laws elsewhere. Let us say, however, that the passenger is arrested and charged with moving between cars. He decides to fight the case because he regards himself as a law-abiding citizen and he is angry about the entire incident.21 The problem is that the government has enacted a law but has communicated its substance to the public in a way that misleads them into thinking that the law is narrower than it really was written to be. As one group of scholars has put it, the information that legislators wish to convey in a statute must be compressed into a signal and later expanded by the recipient into an understanding of that information. In this case, the signal containing the compressed information could not be expanded by the recipient to reveal the information that the legislators wished to convey.22 This sometimes happens in actual cases. For example, the Internal Revenue Service assists taxpayers in meeting their obligations, but having received misinformation from an IRS agent is no excuse for not paying one’s taxes, both in full and on time.23 Similarly, in chapter 5 I discuss United States v. Locke, a case in which the government had misled a miner about the deadline for renewing his claim to federal land.24 The Supreme Court held that a deadline is a deadline no matter who said what but still found a clever way to do justice and get Mr. Locke back his mine. In a more recent case, the Court was far less generous with a convicted felon attempting to appeal a trial court’s denial of his petition for habeas corpus relief. In Bowles v. Russell, decided in 2007, a prisoner convicted of murder had applied to the federal courts for habeas corpus relief, claiming that his constitutional rights had been violated at his trial, which had occurred in the state courts of Ohio.25 His application was denied, and he failed to appeal. Sometime later, he applied to reopen the period under which he could appeal the denial of his application under a statute that permits the courts, under certain circumstances, to reopen the time to appeal for a period of fourteen days.26 A judge granted his application but, for some unknown reason, gave him seventeen days to do so. Bowles

laws and judges



did appeal within that seventeen-day period, ordered by the judge, but he did not do so until the sixteenth day, outside the fourteen-day statutory period. The question facing the Supreme Court was whether the judge’s order—seemingly made in error— could trump the statute, whose deadline was clear and absolute. The case resulted in a five-to-four decision against Bowles. Writing for the majority, Justice Clarence Thomas held that the statute granting judges the right to extend the period to appeal by fourteen days is jurisdictional in nature. That means that the judge who made the error cannot confer jurisdiction upon the appellate court to hear the appeal of the denial of the habeas corpus application, regardless of whether this creates an unfair result to this particular individual. In dissent, Justice David Souter made the following remark: “The District Court told petitioner Keith Bowles that his notice of appeal was due on February 27, 2004. He filed a notice of appeal on February 26, only to be told that he was too late because his deadline had actually been February 24. It is intolerable for the judicial system to treat people this way, and there is not even a technical justification for condoning this bait and switch. I respectfully dissent.” 27 Obviously, not everyone agrees that such rulings are “intolerable” for a judicial system, including five of the justices. Yet Justice Souter asks the right question. Given a law that appears to be quite specific, are there values that might override fidelity to the language of a statute when the law’s substance was miscommunicated? Our subway story, while in some sense a novelty, illustrates the same core problem in the interpretation of statutes raised in Locke and Bowles: the legislature intends to convey a message imposing an obligation on members of society, but somehow that message does not come through. What is unusual about these stories is that one branch of government enacted a clear statute, but another later muddied the waters by issuing a conflicting statement. In the ordinary case, the statute itself fails to communicate adequately just what is expected. Most often this happens because the statute uses words that were chosen with a particular set of scenarios in mind, and a disputed event comes within the scope of a word’s meaning but does not fit the scenarios that led to the statute’s enactment; or something happens that fits the scenario but does not come within the scope of the word’s meaning; or it comes within the word’s meaning in one sense but not another. These problems most often result not from sloppy drafting or bad judging but from a partial cognitive mismatch between our ability to regulate

10

chapter 1

ourselves with precision through detailed laws whose language we take seriously and the architecture of our linguistic and conceptual faculties. We flexibly absorb new situations into categories that we have already formed. In the realm of law, however, those new situations might not have been intended to come within the law. By the same token, a new event might seem not to fit a legal category when, from a perspective of justice, it should. These observations are by no means new. Aristotle observed the problem in the Nicomachean Ethics: [E]very law is laid down in general terms, while there are matters about which it is impossible to speak correctly in general terms. Where it is necessary to speak in general terms but impossible to do so correctly, the legislator lays down that which holds good for the majority of cases, being quite aware that it does not hold good for all. The law, indeed, is none the less correctly laid down because of this defect; for the defect lies not in the law, nor in the lawgiver, but in the nature of the subject matter, being necessarily involved in the very conditions of human action.28

Aristotle recognized that because of this “defect” in “the nature of the subject matter” decision makers might have to make adjustments in applying the law to avoid anomalous results, a recognition that seemed obvious enough to him. He explained: “When, therefore, the law lays down a general rule, but a particular case occurs which is an exception to this rule, it is right, where the legislator fails and is in error through speaking without qualification, to make good this deficiency, just as the lawgiver himself would do if he were present, and as he would have provided in the law itself if the case had occurred to him.”29 Judges cannot help but contribute to the meaning of legislation. Aristotle had no problem with the concept, which is a matter of considerable discomfort to contemporary thinkers concerned about judges usurping the legislative role. Also controversial, this time among psychologists and philosophers, is the extent to which the problems lie in the nature of the thing, as Aristotle saw it, or in the nature of the mind. Both must be in play.30 We do not say that a window is a type of ball at least in part because windows are not balls. Thus, the nature of things plays a role in how we conceptualize the world. But when we have trouble deciding whether having a gun in the trunk of the car should count as “carrying” the gun to the site of a drug crime, at least part of the decision appears to lie, not in some independent

laws and judges

11

essence of carrying that can be discovered through, say, scientific research, but rather in some aspects of how we fit situations into the categories that we have formed.31 That psychology, and the role it plays in statutory interpretation, are the subject of much of chapters 3 and 4. Given that the core problem in the interpretation of laws has been recognized since antiquity, it is no wonder that the scholarly literature and law school casebooks are full of discussion of old cases. In the realm of statutory interpretation, the classic one is Church of the Holy Trinity v. United States, an 1892 case that continues to receive significant attention in the literature and that is discussed in chapter 3 of this book.32 The question was whether the church had violated a law prohibiting the importation of people performing a service or labor of any kind when it paid to bring a minister to New York from England. The case raises the same problem as whether the defendant in the previous paragraph was carrying a firearm and, for that matter, the same problem that Aristotle observed. When such things happen, judges must decide whether to enforce the law’s plain meaning (when the language is unequivocal) or its ordinary meaning (when there is more than one interpretation, but one of them seems more typical than the others) or to enforce it according to some other set of values, such as the purpose of the statute, evidence of what the enacting legislature had in mind, coherence with other laws, consistency with earlier decisions of other courts, the need for evolution so that the law remains responsive, and so on. I develop those arguments and the debates over which should be given priority in chapters 3 through 5. In thinking about them, it is important to keep in mind the examples contained in the beginning of the subway story. These are the rules that we understand without controversy, such as not pulling the emergency cord unless there is an emergency. The clear rules, too, can generate hard cases when, for instance, a passenger concludes that some sort of physical discomfort constitutes an emergency, but the train crew does not, or it becomes necessary to walk across the tracks to avert harm. Most of the time, however, the rules work so well that the possibility of concocting unusual situations in which they do not provide unequivocal answers goes unnoticed. That is generally true of laws that codify social norms. It is less true of laws that attempt to regulate behavior in ways that are counterintuitive or in ways to which people would rather not conform. Such situations sometimes create a game of cat and mouse, where the legislature attempts to set standards and the regulated attempt to comply with the letter of the law but to thwart its intent by engaging

12

chapter 1

in conduct that is largely equivalent to what is not allowed but is different enough in form to come outside the law. We see this in such areas as tax shelters33 and the manufacture of weapons that circumvent the statutory definitions of assault weapons,34 to take two of many examples. The availability of competing values is also why our subway story pre­ sents a difficult problem. It would be perfectly reasonable to hold the commuter to the new law as written and enforce the summons for moving between cars. The law was duly enacted and is clear, and it might cause a disproportionate administrative burden if the 17,000 people charged with violating this rule in 2007 were given grounds to challenge their violations. As for the sign, it does not say that people are permitted to move between cars, only that they are not permitted to ride between cars. This, one can argue, is not enough to remove the general requirement that people become familiar with the laws or face the consequences. On the other hand, the sign is seriously misleading in a way that courts routinely acknowledge. The principle expressio unius est exclusio alterius recognizes that when a law proscribes one thing and not another, we typically infer that the missing event was intentionally omitted. Of course, the sign is not itself a law, but we nonetheless tend to draw the inference. I would certainly think that it is all right to move between cars when the train is stopped if I saw that sign on the door of a subway car. Fair notice is an important value in the criminal law, and this surely is not fair notice. If I were a judge listening to a challenge by a subway passenger, I would rule in the passenger’s favor but suggest first that the Transit Authority seriously consider dropping all charges to avoid the precedential effect of my ruling. Perhaps many readers would rule differently. I have little doubt that my own political tendencies are at work here. Those who place a higher value on enforcing the legislative will than on fair notice might feel differently. Without question, however, both solutions come within the range of legitimate judgments that a court could make.

This Book This book, then, will not present a method of statutory interpretation that claims to come up with the right answer in each hard case. By their nature, hard cases are ones in which ordinary legal principles—most often construing the language of the statute— do not force a single interpretation.

laws and judges

13

At that point, decisions are made by weighing various values that the legal system deems legitimate, the most significant of which is legislative primacy. How any individual decision maker assigns the weights necessarily depends upon the values implicated by the case, their relative importance in legal analysis, and the personal values of the decision maker herself. It is a mistake to pretend that this interpretive space does not exist, just as it is a mistake to make so much of the indeterminacy as to believe that statutes never have a single legitimate application in a particular situation. Rather, I will explore the problems—largely stemming from our psychology—that lead to recurrent difficulties in statutory interpretation and will argue that a rather expansive array of tools is appropriate to use in resolving disputes that arise. Just as importantly, I will discuss the dogs that do not bark—the aspects of lawmaking that appear to present very little problem and that, therefore, make it possible for us to govern ourselves more or less successfully in a complex legal system based on finely articulated rules. Finally, I will discuss the question of who should interpret statutes and whether certain drafting strategies can reduce the burden on judges by better defining the range of discretion that the legislature wishes to delegate to others. The basic argument of the book is that laws generally work well; when they fail to provide us with sufficient information to know our rights and obligations, it is usually (but by no means always) because of uncertainties in how well the concepts contained in a statute’s words match the events that are in dispute. That is, most problems of statutory interpretation, including most of the famous cases, are about problems of conceptualization. Chapter 2 presents this argument through a case study: the federal bribery statute. Most of the terms of the bribery statute have been litigated to the hilt, but little else in the language of a long, syntactically complex statute has caused the legal system any problems. I describe some of the psychology that underlies this distribution of difficulty. As we will see over and over again, many of the problems of statutory interpretation have at their roots a gap between our capacity to make laws that are crisp enough to apply and our capacity to make laws that are flexible enough to deal with new situations. At the same time, the organization of our language faculty makes some aspects of lawmaking second nature. Throughout this book, I will take issue with some of the claims that modern textualists make about the ease with which statutory language produces clear results. Yet such claims, even if they are wrong when it comes to hard cases, are

14

chapter 1

sufficiently grounded in everyday experience to resonate to the point of attracting supporters. In chapter 2, we will see what makes such theories attractive and identify their limits. Chapters 3 and 4 are the heart of the book. Chapter 3 portrays many of the important statutory interpretation cases as instances of grappling with borderline instances of word meaning. I argue that there is actually little difference between textualists and contextualists on the large issue: the principal goal of the statutory interpreter is to be loyal to the legislator. Just about all judges and scholars believe that the best first step in achieving this goal is to pay close attention to the language of the law. Just about all judges and scholars believe that some combination of extrinsic evidence, consequentialist reasoning, and substantive values must play a role when the language of the statute, as applied to the case at hand, does not yield a clear answer. Differences arise with respect to what kind of extrinsic evidence is considered legitimate evidence. Even there, just about all judges and scholars make reference to the circumstances facing the legislature when the law was enacted. The biggest difference at this time is the propriety of reference to legislative history. I support the use of legislative history in statutory interpretation but, as a general matter, find there to be far more agreement than disagreement about which tools are appropriate for judges to use. Chapter 4 delves more deeply into these issues by exploring the concept of legislative intent. Many have argued that legislative primacy must be achieved without reference to the intent of the legislature (whether by studying legislative history or otherwise) because the concept of legislative intent makes no sense. Intent is not enacted—laws are enacted. Moreover, how can more than five hundred people with all kinds of unannounced personal agendas be said to have a single intent? In chapter 4, I defend the notion of intent against claims that it is not democratic and against claims that it is incoherent. We routinely talk about the intent of groups. Using literature from psychology and philosophy, I show how doing so makes perfect sense in the context of the legislature. Chapter 5 returns to a theme discussed earlier in this chapter: what values other than legislative primacy contribute to the interpretation of statutes? Those explored include the competing values of stability and responsiveness to changing circumstances, coherence, constitutional values, fair notice, support for law enforcement, and the personal values of judges. The last two are outside the scope of what is typically considered to be legitimate. Yet it is difficult to ignore their role and important to

laws and judges

15

determine the extent of their contribution to the legal decision-making process. Chapters 6 and 7 move to the question of who should interpret statutes. Chapter 6 deals with suggestions that other branches of government replace judges as statutory interpreters. Some have proposed that judges have too much of a role in statutory interpretation and that various agencies would do a better job. Others propose that the legislature keep more control over the interpretation of laws by specifying how they wish the laws to be construed in the event that a dispute occurs. In fact, some states have enacted laws requiring courts to ignore evidence of legislative intent other than the statute’s language upon a finding that the statutory language is clear. I argue that these laws are not an unconstitutional usurpation of the judicial function, despite the fact that it is typically judges who have the power to interpret statutes. I also argue that it is a bad idea. Because the problems that lead to difficult statutory cases are conceptual in nature, a legislature might inhibit judges from arriving at what it would regard as reasonable interpretations of the laws if it takes away important interpretive tools. As the law is currently applied, judges appear to find sufficient ambiguity to retain their discretion in hard cases. Chapter 7 addresses the role that juries have in the interpretation of statutes. Traditionally, the interpretation of laws is a matter for the judge, and the application of the law as interpreted by the judge is left to the jury. But it does not always happen that way. Jury instructions often leave to the jury all kinds of interpretive decisions, which I discuss and evaluate. The book concludes with chapter 8, where I summarize some of the major points and ask what might make the system better. In particular, I focus on what each branch of government might do to improve the entire regime of statutory interpretation. The tone of this chapter, however, is more positive than it is didactic. Statutory interpretation in the United States does a reasonably good job of effectuating the will of the legislature that enacted the laws and of promoting other values that most would agree are appropriate. By focusing only on the most difficult and controversial cases — the ones that lead to political divides — the legal and academic communities sometimes have lost sight of the fact that laws work quite well in everyday life.

chapter two

Why We Need to Interpret Statutes

I

n a different world, there would be no study of statutory interpretation. There would merely be statutory application. It would be clear what laws require of us ex ante, and it would be clear whether we met our obligations ex post. Justice Scalia expresses this ideal: “A text should not be construed strictly, and it should not be construed leniently; it should be construed reasonably to contain all that it fairly means.”1 His approach to statutory interpretation, dubbed the “new textualism,” attempts to implement as much of that ideal as possible in actual practice.2 Of course, laws are not always so crisp that we can easily tell what our obligations are, or whether we have met them. Eighty years ago, Benjamin Cardozo lamented that “life is too complex” to make attainable “a code at once so flexible and so minute, as to supply in advance for every conceivable situation the just and fitting rule.”3 Yet Scalia’s approach to the role of laws has its appeal. Its most convincing support comes not from the published opinions of appellate courts, which constitute a small percentage of those disputes resolved by the judiciary, which in turn reflect a small percentage of disputes initially brought to court, which themselves fail to reflect all the disputes resolved outside court “in the shadow of the law.” Rather, the best evidence that the law works predictably enough most of the time is that our rights and obligations are so often sufficiently clear that disputes do not occur at all. When lawsuits are initiated, reality often

why we need to interpret statutes

17

sets in before the trial, leading to settlement as the parties learn more about the strengths and weaknesses of their case.4 As George Priest and Benjamin Klein argue in their important study of civil suits, “The Selection of Disputes for Litigation”: “In litigation, as in gambling, agreement over the outcome leads parties to drop out. There are no betting lines or handicaps for legal disputes. Plaintiffs win all or lose all. Where either the plaintiff or defendant has a ‘powerful’ case, settlement is more likely because the parties are less likely to disagree about the outcome. Settlement negotiations will most often fail, however, where the dispute is most problematic whatever the applicable decision standard.”5 Because only the least predictable cases will be brought to verdict, the argument continues, one would expect the rate of plaintiff success to be about 50 percent, however pro-plaintiff or pro-defendant the substantive law appears. And sure enough, that is what happens, according to an impressive set of empirical data that the authors gather. Others have since proposed competing, more nuanced models of litigation outcomes, taking into account the type of case, among other factors. Peter Siegelman and Joel Waldfogel found that plaintiffs tend to win more often in some kinds of cases, such as contract and intellectual-property cases, and to lose more often in others, such as prisoner’s rights and torts cases.6 Nonetheless, they, too, found that only a small percentage of cases filed are adjudicated to a final outcome, whether on motion or after a trial. The rest all settle. The dynamics of the criminal-justice system are sufficiently different from those of civil litigation that we would not expect defendants to prevail in half of all criminal trials. With life or liberty at stake, there may be reasons for taking a case to trial even with little chance of winning. Conversely, there may be reasons for pleading guilty to a lesser crime even when the defendant believes himself to be innocent, when the risk of a conviction and long prison sentence looms, and the plea includes a relatively light punishment.7 Even taking into account any incentives to try one’s luck before a jury, remarkably few cases brought to the courts are actually resolved by them. In 2003, for example, only 4.3 percent of the criminal cases brought by federal prosecutors went to trial.8 And only a small percentage of those cases resulted in appeals that were decided on the merits and explained in published opinions describing the reasoning of the appellate court. In 2007, the percentage of federal criminal cases going to trial remained

18

chapter 2

4.2 percent, despite Supreme Court decisions in the intervening years that diminished the significance of the federal Sentencing Guidelines, perhaps making trial a more attractive option for some defendants.9 The fact that so few cases make their way through the system suggests that in a significant percentage of cases there is really very little to say about the application of the law to the facts, although the facts themselves may be seriously disputed. When a young adult is arrested for mugging a woman at an ATM machine, he may argue that the police got the wrong guy, but he is not likely to argue that the assault laws do not apply. Whatever may account for the uncertainties in the application of laws to situations that arise in the real world, the explanation should also recognize that we often agree on the application of statutes to actual events. The fact that we agree so often about a law’s application rightly gives us confidence in our ability to live under a rule of law that defines our rights and obligations. Statutes are generally written as classical definitions, which in turn are familiar to us as rules that tell us the conditions that are necessary and sufficient for us to use a word appropriately. For example, the American Heritage Dictionary of the English Language defines the verb “lie” as follows: 1. To present false information with the intention of deceiving. 2. To convey a false image or impression: Appearances often lie.

To meet the requirements of the first definition, it is necessary (1) to present false information, (2) with the intention of deceiving. A person who has met both of those criteria has (3) lied. To the extent that these conditions accurately define the term, each is individually necessary, and together they are sufficient. Now let us turn to the federal perjury statute, which says: Whoever— (1) having taken an oath before a competent tribunal, officer, or person, in any case in which a law of the United States authorizes an oath to be administered, that he will testify, declare, depose, or certify truly, or that any written testimony, declaration, deposition, or certificate by him subscribed, is true, willfully and contrary to such oath states or subscribes any material matter which he does not believe to be true; is guilty of perjury.10

why we need to interpret statutes

19

Thus, if a person 1. has taken an oath to testify truthfully under the circumstances described, and 2. has willfully stated or subscribed any material matter which that person does not believe to be true, then 3. that person is guilty of perjury.

We analyze this statute in terms of conditions that are individually necessary and together sufficient to establish perjury. In this way, laws operate as definitions. Here again, the law works quite smoothly in paradigm cases. If a witness in a criminal case, after taking the oath, swears that the defendant was in Omaha at the time of the crime, knowing that the defendant was actually in Boulder, where the crime occurred, then that witness has committed perjury, assuming that the information was material, as the statute requires.11 To the extent that we can analyze alleged criminal activity to determine whether it includes all the definitional elements of a statute, the rule of law really is a law of rules, to use Justice Scalia’s expression.12 Yet we fight over the meanings of statutes, and we fight about how to fight over the meanings of statutes. Most problems arise because, while the laws themselves are structured as classical definitions that tell us all and only the circumstances in which they apply, the meanings that we assign to the words that make up the laws are not fully characterized by classical definitions. Rather, although some aspects of word meaning can be described by definitional features that are necessary or sufficient, words are also understood in terms of such features as typicality. An important part of word meaning consists of judgments we make about the fit between what we experience in the world and the concepts that words convey based on our prior experience. These are not describable in terms of necessary and sufficient conditions. Definitions tell us too much. They describe the universe of all possible meanings. In the real world, words take on a more particular subset of possible meanings, determined by the context in which the word is used. We form mental models of our concepts, which contain all kinds of information.13 In the next chapter, we will see how definitional features and prototypical features of legally relevant concepts compete for attention from judges. At this point, let us limit ourselves to where the controversies lie in the interpretation of statutes. If there was ever a group of people bound and determined to twist the meanings of words to mean something other than what they mean, that

20

chapter 2

group would be those accused of committing crimes (and the lawyers acting on their behalf). Indeed, we can think of the set of all persons accused of violating a particular criminal statute as subjects of a vast experiment that would be unethical if it were not already part of real life. Below, I will focus on a single statute—the bribery statute—to see just which disputes occur and which ones do not.

An Unethical Experiment Conducted with Real-Life Subjects The history of science is replete with ugly moments.14 Although linguistics has played a very minor role in this history, perhaps the very first unethical experiment ever proposed was an experiment designed to determine how we use and acquire language. Would a child raised without language be recognizably human? Though a fascinating question, any experiment that would test it was ethically “forbidden,” even in a Europe that knew serfdom and slavery. Instead, early linguists looked to rare, naturally occurring phenomena—boys and girls separated from their parents by disaster and raised by wolves or goats. Though possibly apocryphal, these tales of lost children—and the tantalizing possibility they offered to early scientists of conducting the first Forbidden Experiment without themselves committing evil— captured the imaginations of generations of natural philosophers.15 Title Eighteen of the United States Code, in which are codified the federal felonies, gives us the chance to run a similar experiment to investigate how we understand statutory language. That experiment might go something like this: You have been accused of violating a law by doing x. The law says the following: [quote the law]. Unfortunately, a jury has convicted you of violating the law, and you are now serving a five-year prison sentence. Your only hope for getting out is to convince a court that what you did is not clearly prohibited by the law, and therefore not illegal. If you can come up with some reasonable uncertainty in the law’s interpretation, the judge will let you go. To help you take this task seriously, we will keep you here in our lab until you come up with a plausible argument for why the law does not apply to your conduct. Good luck.

My proposed experiment is cruel. We do not, however, need to conduct it. The criminal-justice system has already done it for us. People claim

why we need to interpret statutes

21

to have been unfairly charged with crimes because, even if they have engaged in the conduct that is alleged, the conduct falls outside the statute. A rule — the rule of lenity (discussed later in this chapter)—holds that ambiguities should be resolved in favor of the accused, both because a defendant is entitled to fair notice and because only the legislature has the power to define what constitutes a crime. Therefore, if we want to find out whether a particular set of linguistic issues predominate in disputes over the interpretation of statutes, we can design a study with the following structure: 1. Choose a statute over which there have been a significant number of disputes resulting in published opinions. 2. Categorize the issues in the disputes linguistically (e.g., word meaning, syntactic ambiguity, reference, amount of evidence required, contradiction). 3. Determine the extent to which each category is represented in the disputes over the application of the statute.

I have conducted such a study using the federal bribery statute, section 201 of the federal Criminal Code.16 I chose that law because it is old enough and complex enough to invite all kinds of disputes, and long enough to be linguistically complicated. The law has existed more or less in its current form since 1962.17 Under the statute, it is illegal to give a bribe to a public official, and it is illegal for a public official to solicit or to accept a bribe. Both crimes are defined in section (b) of the statute: (b) Whoever— (1) directly or indirectly, corruptly gives, offers or promises anything of value to any public official or person who has been selected to be a public official, or offers or promises any public official or any person who has been selected to be a public official to give anything of value to any other person or entity, with intent— (A) to influence any official act; or (B) to influence such public official or person who has been selected to be a public official to commit or aid in committing, or collude in, or allow, any fraud, or make opportunity for the commission of any fraud, on the United States; or (C) to induce such public official or such person who has been selected to be a public official to do or omit to do any act in violation of the lawful duty of such official or person;

22

chapter 2 (2) being a public official or person selected to be a public official, directly or indirectly, corruptly demands, seeks, receives, accepts, or agrees to receive or accept anything of value personally or for any other person or entity, in return for:

(A) being influenced in his performance of any official act;

(B) being influenced to commit or aid in committing, or to collude in, or

allow, any fraud, or make opportunity for the commission of any fraud, on the United States; or

(C) being induced to do or omit to do any act in violation of the official

duty of such official or person;

*  *  * shall be fined under this title or not more than three times the monetary equivalent of the thing of value, whichever is greater, or imprisoned for not more than fifteen years, or both, and may be disqualified from holding any office of honor, trust, or profit under the United States.18

Other provisions of the statute make it a crime to bribe witnesses.19

What People Fight About When disputes arise over the application of this statute, where do the problems lie? They lie in the task of drawing inferences about the defendant’s state of mind; they lie, occasionally, in the relationship between the bribery statute and other rules of law; and they lie, most frequently, in the meanings of the key words. There have been no disputes alleging ambiguity resulting from aspects of meaning driven by the law’s syntax, although one of the definitional sections of the statute, as we will see, has such an ambiguity that has led to litigation. Nor has there been a single dispute alleging ambiguity of reference, which has both rulelike and pragmatic aspects, as we will see below. People do not have trouble figuring out who’s who, even in this long, complex law. Of course, discovering what goes wrong only identifies the problem. Once a dispute arises, the linguistic and cognitive foundations that led to the problem take a back seat to legal and political considerations that suggest how to solve it. We will get to that in later chapters. For now, let us look at what triggers the disputes over the application of the bribery statute.

why we need to interpret statutes

23

The Meanings of Words To some extent, it may be surprising that people argue so much about the boundaries of the words in the bribery statute, since it contains its own internal glossary precisely to reduce disputes over word meaning by defining some of the key terms: “public official,” “person who has been selected to be a public official,” and “official act.”20 However, definitions can solve the problem of meaning only in part. The definitions are themselves written in words, and the application of the definitional words is not always clear in situations that end up in court. When people are not fighting over the words in the statute proper, they fight over the words in the definitions of the words in the statute proper. In addition, the definitions consist of complex sentences, which unfortunately contain their own syntactic ambiguities that must be resolved from context. When the context does not make it obvious how this should occur, the courts must step in. Let us look at some of the problems that courts have been asked to resolve with respect to words that are defined and words that are not. “public official.”  Despite the definition of “public official” in the glossary, courts have been asked to decide whether various people and entities should count as a “public official” under the bribery statute. The statute defines “public official” as follows: (1) the term “public official” means Member of Congress, Delegate, or Resident Commissioner, either before or after such official has qualified, or an officer or employee or person acting for or on behalf of the United States, or any department, agency or branch of Government thereof, including the District of Columbia, in any official function, under or by authority of any such department, agency, or branch of Government, or a juror.21

Now consider the following stories: A corporation was hired by a city to administer block grants awarded to the city by the federal Department of Housing and Urban Development. Part of the job was to recommend contractors to complete projects for which the grant was awarded. The executive director of the corporation received kickbacks from potential contractors. He claimed not to be a public official. The Supreme Court held that he was a public official for purposes of the bribery statute,22 resolving a conflict among lower courts in similar cases that had been decided earlier.23

24

chapter 2

A baggage porter working for a railroad owned by the United States took money to deliver a trunk checked as baggage, which was outside the scope of his permitted duties. The court held that he was not a public official.24 He may have moonlighted illegally by performing services not permitted under the terms of his employment, but he did not perform them in his role as public employee. These cases illustrate the kinds of borderline situations that dominate statutory litigation. They also show how definitions can sometimes be of little or no help in deciding whether a person comes within the purview of a statute. “anything of value.”  People also fight about what “anything of value” means.25 Among the most celebrated bribery scandals in recent American history was Abscam, a 1979 sting operation that led to nineteen convictions, including five members of the House of Representatives and one senator.26 Government agents and informants tempted these legislators by offering them money in exchange for introducing private immigration bills that would permit individuals posing as sheiks from the Middle East to remain in the United States. The cases are remembered in part for their having tested the limits of the entrapment defense. Government agents played such a large role in the operation that it is not clear whether the targets of the investigation would have accepted bribes absent the elaborate sting operation.27 Harrison A. Williams, then a U.S. senator from New Jersey, claimed that the stock he received in exchange for his votes was objectively worthless and therefore not within the scope of the term “anything of value.”28 The trial court disagreed and instructed the jury that the words “anything of value” could include even objectively worthless objects, so long as the jury found that the senator believed that the stock had value. The court of appeals affirmed the conviction. Similarly, a federal prosecutor accused of bribery tried—but failed— to convince the federal courts that the loans he received from a criminal defendant were not “things of value” because he later repaid those loans in full.29 Thus, courts tend to be very expansive in their interpretation of that term. “in return for . . . being influenced.”  Among the people accepting money in the Abscam operation was Congressman Michael Myers of

why we need to interpret statutes

25

Pennsylvania.30 Part of Myers’s defense was that he never intended to follow through with his promise to introduce the immigration legislation and therefore had not accepted the money “in return for being influenced in any official act,” as the statute requires. If “in return for” is understood to describe the public official’s intent, then Myers had a legitimate point. But if “in return for” is understood to describe the message that the public official conveyed to the people making the payment, then Myers’s defense was without merit. The U.S. Court of Appeals for the Second Circuit decided that the latter understanding better reflected the intent of Congress, based on the legislative history of the 1962 amendments to the statute. This understanding is also more consistent with the way the law treats promises generally. When a person communicates that he accepts money in exchange for promising to perform a service, his private, unexpressed intention not to live up to his commitment does not negate the fact that he has, nonetheless, made the commitment. This is a core concept of contract law. Ian Ayres and Gregory Klass describe these statements as “insincere promises.”31 Because insincere promises are promises nonetheless, it is fair to say that Congressman Myers accepted money in return for being influenced, whether or not he was sincere.32 “whoever.”  Believe it or not, even the statute’s first word has been a matter of controversy. The question is whether it should include the federal government when the government pays informants to gather information and then to testify about what they have found. The provision of the statute that refers to paying witnesses to testify reads: Whoever . . . directly or indirectly, gives, offers or promises anything of value to any person, for or because of the testimony under oath or affirmation given or to be given by such person as a witness upon a trial, hearing, or other proceeding, before any court . . . authorized by the laws of the United States to hear evidence or take testimony . . . shall be fined under this title or imprisoned for not more than two years, or both.33

The words of the statute, if interpreted literally, would certainly appear to apply to a government official who agrees to pay an informant. Yet in case after case the courts have held that the statute is not intended to apply to the U.S. government’s payment to witnesses in sting operations.34 Statutes are not ordinarily construed to interfere with governmental prerogatives.

26

chapter 2

“corruptly.”  People also fight about what it means to offer or accept a bribe corruptly. The issue often arises in the context of disputes over jury instructions. The statute does not define the word, so the courts have stepped in. Because the scenarios involving people’s states of mind are wide-ranging and instructive, we will look at a few of these cases in detail. Consider the following stories: Darrell Tomblin and some confederates promised Glen Mauldin, the administrative assistant to Senator Jacob “Chic” Hecht of Nevada, a $50,000 campaign contribution, a 10 percent stake in their new bank venture, and a trip to Texas, all in exchange for the senator’s help in bypassing regular channels needed for approval of their takeover of a savings and loan association.35 Because their acquisition of that bank had been illegal, they would not have gained approval without inappropriate intervention. The problem facing the court was that it is not illegal to give campaign contributions, even contributions made with the hope of being treated well as a result of the payment. Only contributions made in exchange for a promise to perform a particular official act are bribes. Tomblin testified that his intent was to put himself in good favor by making the campaign contribution and paying for Mauldin’s trip to Texas but that the terms of any quid pro quo were so vague that they could not have amounted to the corrupt state of mind that the statute requires. The court instructed the jury on the meaning of “corruptly” but was not very clear on how specific the quid pro quo must be: An act is “corruptly” done if it is done intentionally with an unlawful purpose. Therefore, in evaluating whether the Government has proved beyond a reasonable doubt that the Defendant bribed or intended to bribe a public official, you must focus specifically on the Defendant’s mental state, regardless of the mental state of the public official. Merely because the public official accepts a thing of value with corrupt intent does not mean that the Defendant made the offer with corrupt intent. Indeed, the public official may accept an offer as a bribe, while the Defendant may not possess the same corrupt intent in offering something of value. Again, your focus must be on the Defendant’s state of mind.36

Tomblin was convicted of bribery and conspiracy to commit bribes. Tomblin’s point was not without merit. While nothing the trial court said was incorrect, this instruction did not really focus the jury on the fact that the bribe must be for the purpose of influencing an identifiable official act. Instead, the trial court merely directed the jury to consider Tomblin’s

why we need to interpret statutes

27

“unlawful purpose,” which could easily be interpreted to mean Tomblin’s more general purpose to take over a savings and loan association unlawfully rather than the specific unlawful purpose of paying for the senator’s help. The appellate court recognized the force of Tomblin’s argument but rejected it nonetheless. Tomblin had also been indicted for aiding and abetting Mauldin in Mauldin’s efforts to take bribes. The instruction to the jury on that count stated that for Tomblin to have done so, Mauldin must have “demanded, sought or received the item of value corruptly in return for being influenced in the performance of an official act.”37 This, the appellate court stated, was good enough to let the jury know of the quid pro quo requirement with respect to Tomblin’s conviction for making a bribe. Perhaps the jury did have a good sense of just what state of mind Tomblin must have had to be guilty, but if it did, it surely was not because the court made the requirements of the statute at all clear. Now consider another case. David Barash, a lawyer and accountant, was indicted for bribing internal revenue agents to provide favorable audits to his clients. The agents themselves were corrupt and were cooperating with the government in the case against Barash in order to receive more favorable treatment when it came time for them to be sentenced for the crimes that they had committed. Barash’s defense was, in part, that the agents had been shaking him down and that he gave them money only because he felt that he had no choice if he wanted his clients to be treated fairly.38 He claimed that he had given the money only under duress, which, if accepted as a legitimate defense and believed, could exonerate him. But the court ruled out duress, since that defense required some threat of physical harm. Barash was convicted. Later, the appellate court agreed with Barash that “if a government officer threatens serious economic loss unless paid for giving a citizen his due, the latter is entitled to have the jury consider this, not as a complete defense like duress but as bearing on the specific intent required for the commission of bribery.”39 That intent, as the court points out, includes a “corrupt intent to influence.”40 None of this is in the bribery statute. Rather, it reflects the courts’ use of their best judgment in applying the statute when new, unanticipated situations arise in the interpretation of “corruptly” in the bribery statute. Or consider Senator Daniel Brewster’s appeal of his conviction for accepting illegal gratuities.41 Brewster had taken a number of contributions from entities that made it clear that their goal was to influence his vote on

28

chapter 2

legislation to raise postal rates. These individuals were convicted of bribery. As for Brewster, he claimed that although he understood their motives in making these contributions to his campaign fund and other funds, he did not accept the money in return for being influenced. Section 201 not only makes it illegal to accept a bribe but also makes it illegal to accept a gratuity. While a bribe requires that the payment be demanded “corruptly . . . in return for . . . being influenced in the performance of any official act,”42 the gratuity section does not use the word “corruptly” and requires that the thing of value be accepted “because of” an official act.43 Senator Brewster was charged with both crimes but convicted only of accepting a gratuity. On appeal, he claimed that the charge to the jury did not adequately define the various states of mind corresponding to accepting a bribe, accepting a gratuity, and accepting a legal contribution, even from a contributor who obviously wanted something in return. The court agreed and reversed the conviction. As the court noted, “No politician who knows the identity and business interests of his campaign contributors is ever completely devoid of knowledge as to the inspiration behind the donation.”44 As for the right way to tackle the problem, the court found the original instruction “indigestible” and offered no “complete recipe or formula” adequate to stating these distinctions clearly in a coherent set of instructions.45 It somewhat pragmatically cautioned prosecutors that the safer course might be to charge defendants with one crime or the other but not both. When one crime requires that an act be done corruptly, and another that the same act be done “because of” an official act but not corruptly, it is so hard to separate the two in words understandable to the average person that the best solution is to avoid the situation entirely. Reading the Minds of Others: Problems of Proof Even when we know what a statutory word means, it is not always easy to tell whether the facts before us fit the definition, because we cannot be sure that we have drawn the right inferences from the evidence. Nowhere is this truer than when a statute requires proof of a defendant’s state of mind. We are all dedicated mind readers. In everyday life, we routinely ascribe intentions to others. When we say such things as “I can’t believe you’re so angry,” “He’s dishonest,” or “She’s a generous person,” we project states of mind on others. As we will see in chapter 4, this plays an important role

why we need to interpret statutes

29

in how we learn new concepts,46 which in turn explains in part why it is so natural for judges to ask what the legislature intended. The issue also arises, not surprisingly, when crimes include a state-of-mind element and jurors must decide whether the defendant acted accordingly. Generally, problems of proof are not problems of statutory interpretation. Whether the teenaged defendant actually was the one who mugged someone at the ATM has little to do with how the boundaries of the law’s language should be understood. Yet when courts are asked to decide whether the evidence proves that the defendant acted with a particular state of mind, they ultimately must determine whether the circumstances of the case fall within the statute’s state-of-mind requirement. And deciding whether a set of facts comes within a statute is just what courts do when they engage in statutory interpretation. Returning to the bribery statute, people rarely say, “I hereby give you a bribe in exchange for your promise, as a government official, to provide me with documentation to which I am not legally entitled.” Rather, they say something less direct, and inferences must be drawn to determine whether their acts of speech violate the statute.47 These inferences, pragmatic in nature, are not always clear-cut. Most involve circumstances suggesting that the defendant gave or took a bribe “corruptly,” as the statute requires. Some are close calls, and the decisions of the courts are as much about the circumstances that the law is willing to describe as indicating corruption as they are about a matter of proof. For example, a major scandal of the 1980s involved a New York–based defense contractor, Wedtech Corporation. Mario Biaggi, then a congressman from the Bronx, was also a partner in the law firm that represented Wedtech.48 Over time, the line between being a lawyer for a contractor and being a member of Congress seems to have blurred, and Biaggi began advocating in Congress for Wedtech, in exchange for payments of various kinds. Among them was the transfer of 2.5 percent of Wedtech stock into an account in the name of Biaggi’s son, Richard, at the completion of Wedtech’s initial public offering.49 Mario Biaggi’s conviction for bribery and other crimes was affirmed on appeal. But his son’s conviction was reversed. At the trial, the government had demonstrated adequately that the congressman knew that Richard was receiving the Wedtech stock as a nominee for his father and that the congressman knew why the stock had been transferred. However, no proof had been offered at trial to demonstrate that Richard himself knew that his receipt of the stock was in exchange for his father’s being

30

chapter 2

corruptly influenced in his legislative duties. As for what is required for such an inference, the court said: “[I]t is possible that the Congressman told his son the true reason. But knowledge, though inferable from circumstances, must be based on evidence, not speculation.”50 Since all evidence of another person’s state of mind must be indirect, it is not clear where, in the court’s estimation, speculation from circumstances ends and evidence begins. Judges must acquire a “feel” for this, and it is not surprising that they do not always share the same intuitions. Thus, Peter Neglia, a Small Business Administration official, did not fare as well. He argued that at most, in accepting the job offer from Wed­ tech, he had violated the gratuity statute. There was no evidence, he claimed, that he had done anything “in return for” the promise of future employment after the offer had been made. The court recognized this as a close call but affirmed the conviction based largely on a taped statement to Neglia from a Wedtech officer in which the officer said, “We always take care of our friends when they are in government and we never forget them after they leave government.”51 Moreover, the court noted, some of Neglia’s corrupt acts occurred after the job promise. Comparing the fates of Richard Biaggi and Peter Neglia, one must conclude that there is no hard science involved in drawing inferences about a person’s intent. Drawing inferences from the circumstances, we do the best we can. These are not the only cases that raise this issue.52 Some are straightforward, others more difficult. They all involve the same question: when do the facts of a case permit an inference of the defendant’s corrupt state of mind to be drawn beyond a reasonable doubt? The question is an important one in that it demonstrates that a rule of law that concerns itself with people’s thoughts requires us to be able to draw conclusions about their mental processes.

The Dog That Didn’t Bark: The Semantics of a Complicated Law If all we knew about the bribery statute is that it has generated so much litigation and has absorbed so much judicial effort resolving disputes, there would be reason for pessimism about the ability of the legal system to provide a set of rules that are clear enough to put us on notice of our obligations. But a closer look at the cases that do arise demonstrates that some potential problems occur only occasionally and others not at all. Like Sherlock Holmes, we can draw conclusions not only from the sounds

why we need to interpret statutes

31

of the dogs that did bark but also from the other dogs’ silence. Of all the things that might go wrong in the interpretation of a complex law, very few do. Most notably, even though the bribery statute is written as a very complex and very long sentence, there are virtually no disputes over meaning that derive from the statute’s syntactic structure. Each of the two crimes—giving a bribe to a public official and taking a bribe as a public official—is defined in a single sentence. Each sentence begins with the word “whoever,” which is the subject of the sentence. The sentences then continue with either (1) or (2), which are actually relative clauses that modify the subject, “whoever.” Each sentence ends with the predicate of the sentence, set off at the end, which begins, “shall be fined. . . .” When the words are added up, section 201(b)(1), which criminalizes the giving of bribes, consists of a single sentence of 199 words; section 201(b)(2), which criminalizes the taking of bribes, consists of a single sentence of 156 words. To what extent has this extremely complex set of clauses and phrases embedded in other clauses and phrases caused problems for statutory interpreters? It has not done so. Let us explore some of what remains uncontested in the bribery statute by looking more closely at a truncated version. In abbreviated form, the statute reads: Whoever corruptly gives anything of value to a public official, or promises a public official to give anything of value to any other person, with intent to influence any official act, shall be fined.

Without even being aware of it, when we read the bribery statute, we “know” the following, among other things: 1. The statute is a sentence. 2. The subject of the sentence is “whoever, directly or indirectly, corruptly gives, offers, or promises anything of value to any public official, with intent to influence any official act,” and the predicate is “shall be fined.” 3. The subject of the sentence contains a relative clause. “Whoever” is the subject of that clause, and “directly or indirectly, corruptly gives, offers, or promises anything of value to any public official, with intent to influence any official act” is the predicate of that clause. 4. The verbs “give” and “offer” involve three arguments: the subject of the sentence, a direct object, and an indirect object. The subject is “whoever,” the direct object is “anything of value,” and the indirect object is “any public official.”

32

chapter 2

5. “Corruptly” modifies the predicate “gives, offers, or promises anything of value.” 6. “With intent to influence any official act” modifies “whoever.” 7. The logic of the word “or” requires that “with intent to influence any official act” apply to both disjuncts.

The fact that none of this is litigated tells us a lot about what makes it possible to govern ourselves under a rule of law comprising detailed and complex descriptions of our rights and obligations. Whatever else goes wrong, our ability to understand those aspects of language that tell us about the relationships among the players appears to function very well. This is not to say that the statute lacks ambiguity. For example, only corruptly giving a gift to a public official or a person chosen to be a public official is bribery. But what about promising a public official to give something of value to someone else? It is not clear from the language itself whether such a promise must be corrupt, although it is fair to assume that Congress intended to impose that state of mind when it enacted the statute. The ambiguity arises because one cannot tell by reading this sentence whether the structure of the sentence is as (1) or as (2): 1. Whoever corruptly [(gives anything of value to a public official) or (promises a public official)] . . . shall be fined. 2. Whoever [(corruptly gives anything of value to a public official) or (promises a public official)] . . . shall be fined.

Does “corruptly” modify both “giving” and “promising” or only the former? As (3) illustrates, the statute is ambiguous in just the way sentences of that structure generally allow both interpretations: 3. He stealthily placed the envelope in the drawer and left the room.

Did he leave the room stealthily? The sentence is ambiguous. If we really care, we must turn to context and draw inferences about what the speaker intended to communicate. As for the statute, it only makes sense to assume that Congress intended “corruptly” to modify both verbs, but not because the statutory language forces us to draw that conclusion. By the same token, the statute does not unambiguously tell us who must have the intent to influence a public official. Is it the defendant (the “whoever”), or is it the person who has been promised a thing of value? Context tells us it must be the defendant. But the language itself does not

why we need to interpret statutes

33

compel us to understand it that way. There are other ambiguities very much like these distributed throughout this long and rather unruly law. Yet people are such good contextualizers that most ambiguities are resolved quickly and go unnoticed.53 Context certainly does not resolve every ambiguity, but it does so most of the time. Were we to look at all two hundred words of the bribery statute, uncover all the syntactic ambiguities, the potential garden paths that we could take with all the various conjunctive and disjunctive phrases, and everything else that could go wrong, we would have to conclude that very little does go wrong. Sometimes syntactic ambiguity in a statute does matter. Earlier, we looked at cases in which the statutory definition of “public official” failed to resolve borderline cases. Here, we will see that it is syntactically ambiguous as well. The definition is repeated below: (a) For the purpose of this section— (1) the term “public official” means Member of Congress . . . , or an officer or employee or person acting for or on behalf of the United States, or any department, agency or branch of Government thereof, including the District of Columbia, in any official function, under or by authority of any such department, agency, or branch of Government, or a juror.54

The problem is that one cannot tell whether “in any official function” applies (1) to the entire set of potential defendants, (2) to “an officer or employee or person acting for or on behalf of the United States,” or, less broadly, (3) to a “person acting for or on behalf of the United States.” The provision is subject to the following three analyses: 1. [(member of Congress . . . ) or (officer or employee or person acting for or on behalf of the United States) . . . in any official function] 2. [(member of Congress . . . ) or (officer or employee or person acting for or on behalf of the United States . . . in any official function)] 3. [(member of Congress . . . ) or (officer or employee) or (person acting for or on behalf of the United States . . . in any official function)

Context tells us that the first interpretation is not likely. Members of Congress should be considered public officials whether or not they are acting in an official function. Not every interaction with a senator may be a bribe, but the senator is surely a public official.

34

chapter 2

How have courts decided between the other two options? Consider the case of Gjergj Gjieli, who tried to pay an agent of the federal Bureau of Alcohol, Tobacco, and Firearms (ATF) $100,000 to arrange to spring a friend from state prison in Michigan so the prisoner could return to Albania.55 He admitted that he had attempted to pay an officer of the United States but claimed that this did not violate the statute because the ATF officer could not have been acting in an official function since he had no authority over the state prison system in Michigan. Gjieli was convicted. In affirming that decision, the U.S. Court of Appeals for the Sixth Circuit argued first that if “any official function” were applied only to officers and employees of the United States, then there really would be no reason to have included that group in the statute at all. If an ATF agent is an officer of the United States performing an official function, then surely he is a “person” performing an official function. It is a fair assumption that Congress did not intend “officer” and “employee” to be merely superfluous words in the statute. The only way to avoid that being the case is to construe “in any official function” as modifying only the last set of individuals, in keeping with interpretation 3. The court also referred to the legislative history, which makes specific mention of “other persons carrying on activities for or on behalf of the Government.”56 In fact, the Supreme Court has observed that early versions of the current statute would have omitted the phrase “person acting for or on behalf of the United States.”57 However, an analysis by the Department of Justice persuaded Congress to retain the phrase in the 1962 revisions to the law, so that the statute would apply to persons who were “acting on behalf of the federal government, but were not direct federal employees.”58 Moreover, the record suggests that the 1962 Congress approved of earlier court holdings that read the bribery statute expansively to include the acts of those not directly employed by the federal government, and disapproved more narrow court opinions.59 Finally, this reading is consistent with the purpose of the statute, which is to ensure that federal employees do not use their position in an unlawful manner. In an interesting dissent, Judge Lively looked at the legislative history differently. Relying on a Senate report, he argued that the purpose of the 1962 revisions to the bribery statute was to add some provisions not relevant to this dispute and to consolidate the rest of bribery law into a single, comprehensive provision. He read a number of earlier cases as consistent with his position.

why we need to interpret statutes

35

On balance, the majority probably had the better of the argument, although other courts have noted that arguments are available in support of both sides of the dispute.60 Most noteworthy is that while the problem with the statute is linguistic in nature, the solution is only partly so. What this suggests, then, is that certain legal problems arise because our linguistic capacity leaves us with ambiguous laws, but the solutions depend, not on linguistics or psychology, but rather on legal and political concerns such as what the legislature intended and how to construe the statute to promote coherence. Cases involving linguistic ambiguity are far less common than those involving conceptual vagueness, but they are not absent altogether. Stepping outside the bribery statute, consider Circuit City Stores v. Adams, which the Supreme Court decided in 2001.61 The Federal Arbitration Act requires courts to enforce arbitration clauses in “[a] written provision in any maritime transaction or a contract evidencing a transaction involving commerce.”62 But an exception says that the law does not apply “to contracts of employment of seamen, railroad employees, or any other class of workers engaged in foreign or interstate commerce.”63 The law was enacted in 1923, and at the time, transportation workers, such as those listed in the exception, were the only class of workers deemed to be engaged in foreign or interstate commerce. Currently, interstate commerce is understood much more expansively, and the class of workers engaged in interstate commerce includes a person like Saint Clair Adams, who was a sales counselor in California for Circuit City Stores, a national chain of electronics stores. When Adams tried to sue Circuit City for employment discrimination, Circuit City invoked both the arbitration clause in Adams’s employment contract and the Federal Arbitration Act, arguing that the only remedy was to go to arbitration. Adams, in turn, argued that the exception to the act for employees engaged in interstate commerce applied. The case made its way through the courts, and ultimately the Supreme Court ruled in favor of Circuit City in a five-to-four decision. The statute is ambiguous. It can be understood to apply only to transportation workers, as the majority held. The expression “or any other class of workers engaged in foreign or interstate commerce” can be construed as shorthand for whatever such workers were members of that class at the time the statute was enacted, just as “the starting lineup of the New York Mets” can be understood to refer to the nine individuals who made up the starting lineup at the time of the utterance.64

36

chapter 2

Alternatively, as the four dissenters would have it, the expression can be understood as a sort of variable expression, referring to whatever groups of people happen to meet the description at any given time. (The “starting lineup” example is ambiguous in the same way.) Under this reading, what is relevant is that Adams, at the time he brought his lawsuit, was an employee engaged in interstate commerce. In determining which of the two readings should prevail, both sides had reasonable arguments. The majority was able to make use of the fact that the examples of jobs mentioned in the statute were all in the area of transportation, legitimizing an inference that Congress intended the statute to be limited to that group. A canon of construction, ejusdem generis, calls for general terms to be construed to include categories similar in nature to those enumerated in the statute. The majority also argued that the principal statute includes contracts “evidencing” interstate commerce, while the exception applies to those who are “engaged in” interstate commerce. Since the former expression seems to be somewhat broader than the latter, the majority believed that a narrow construal of the exception was appropriate. The dissent had on its side the fact that the variable reading, by which anyone with a job engaged in interstate commerce comes within the statute’s language, is the more natural reading: reading the statute today, Adams’s job is most comfortably regarded as one engaged in interstate commerce. Moreover, when Congress wrote the statute decades ago, it appeared to have as its goal the exemption of employment contracts to the full extent that such contracts would otherwise be covered. With both sides making points about how to use context to resolve the ambiguity in the statute, it appears that politics won out. Those in the majority were the more conservative justices, and those in dissent the more liberal ones. Statutory cases do not always resolve themselves this way, but cases in which there are two perfectly legitimate interpretations lend themselves to this kind of division. Justice Stephen Breyer, in his book Active Liberty, characterizes this case as a conflict between textualism on the part of the majority and concern about purpose on the part of the dissenters.65 My analysis is slightly different. I agree that the dissenters’ position was more loyal to the purpose for which Congress appeared to have enacted the exception to the arbitration requirement. However, the strength of their argument, I believe, is bolstered by the fact that the statute is susceptible to a legitimate reading that is more consistent with what Con-

why we need to interpret statutes

37

gress was apparently attempting to accomplish. Both sides concerned themselves with the statute’s language, but the dissenters had the better of the argument about how to resolve the ambiguity in a manner more loyal to the enacting legislature. The majority position, in contrast, was more consistent with the recent history of Supreme Court decisions favoring arbitration, which has continued in the years following the Circuit City decision.66 Thus, ambiguity in statutory language provides the opportunity to argue the relative importance of other values (such as coherence and loyalty to legislative purpose), while at the same time promoting positions consistent with each justice’s own political orientation. This is not to say that political loyalty is always predictive of how a judge will resolve ambiguity in statutory cases. We will see in chapter 3, for example, a number of cases in which Justice Scalia agreed with the dissenters, arguing that lenity should be applied when the language of a criminal statute is ambiguous. One such case is Muscarello v. United States,67 in which Justice Scalia not only signed on to Justice Ruth Bader Ginsburg’s dissent, but Justice Breyer, a moderate on the court, wrote the majority opinion in favor of the government’s position. Nonetheless, Adams illustrates both that linguistic indeterminacy other than nuances in word meaning can create difficult statutory disputes and that when such disputes occur, it is neither unusual nor surprising to find the personal values of individual judges playing a role in how that indeterminacy is to be resolved. Thus, the battles fought over the meanings of words are not the only ones fought in the realm of statutory interpretation. Yet the distribution of cases under the bribery statute is too skewed to ignore. It is a fair inference that most statutory problems are conceptual: they result from laws written as all-or-nothing propositions for people whose judgments are graded. Philosophers distinguish this problem — the problem of vagueness—from the problem of ambiguity.68 Let us look a little more closely at what underlies this distinction.

Lawmaking and the Modular Mind As one reads stories about the consequences of stroke and traumatic head injury, whether in the technical or popular literature, it is impossible to escape the conclusion that different parts of our brain appear to perform different functions.69 Otherwise, each injury would impair brain function

38

chapter 2

in some general way across the board rather than knocking out or compromising a particular function. Philosophers, linguists, and psychologists often describe this compartmentalization by speaking of the mind as a series of modules. Often, these modules, at least in part, are associated with particular areas of the brain, although research also shows that the brain is very plastic in its ability to compensate for the loss of a function because of an injury to one region by developing that function in another region. All this remains a matter of some consensus, much controversy, and the subject of fruitful research. Nonetheless, at least in some meaningful way, it continues to make sense to speak of the mind as modular, with complex interfaces and executive functions combining the outputs of various modules to enable us to experience a complex world with subtle nuance.70 Psychologist Steven Pinker has characterized some of the modules that make up our language capacity by distinguishing between words and rules.71 The distinction, which in many respects tracks the work and thinking of linguistic theory over the past half century, can be useful in characterizing the disputes that arise over the application of laws to situations in the real world. Some aspects of our language faculty are rulelike, much the same way that the perjury statute is rulelike. Syntax is among them. Others, like our use of language to express concepts, are what Pinker calls wordlike in nature. They cannot be characterized solely in terms of rules that operate definitionally by virtue of a set of necessary and sufficient conditions. Rather, as we will see in more detail in chapter 3, they reflect the mental models we form based at least in part on how we experience the world in addition to definitional conditions. Pinker’s taxonomy is itself not without controversy. Some suggest that words and rules are not discrete categories but rather operate on a continuum.72 Others think that very little of language is rulelike at all.73 Still others think that it is more rulelike than Pinker concedes.74 Nonetheless, Pinker’s illustrations of each type of phenomenon provide a useful way to discuss the various modules of our linguistic capacity. Rulelike properties apply whenever the linguistic environment that triggers them is present. Most relevant for our purposes are those aspects of meaning that derive from the structure of language. As we just saw, these aspects of meaning are usually clear, and when they are unclear, they are “ambiguous” in that there are a discrete number of possible meanings. Typically, the possible meanings are remote from each other, making it relatively easy to determine the intended meaning through context. Word-

why we need to interpret statutes

39

like aspects of meaning are also usually clear. But when they are unclear, they are “vague” in that there are innumerable possible meanings, and these are not necessarily as easily resolved by context. That is because they involve borderline cases and require subtle judgments of line drawing to determine whether one interpretation or another fits a situation best. The meaning that derives from syntax is so ubiquitous and works so well that it might seem trivial. But it is not. Consider the following examples: “The senator was easy to bribe.” “The senator was eager to bribe.”

We know, without any conscious analysis, that the senator was the recipient of the bribe in the first example and was the potential source of the bribe in the second. Although context suggests that the senator was more likely the bribee than the bribor, context does not change our minds about what the sentences say. Or consider the following: “The guard was taken from the courtroom to the prison.”

We know here that although guards typically are responsible for taking others to prison, in this sentence, which is in the passive voice, the guard, although the grammatical subject, is the logical object of the sentence. The logical subject is left unexpressed but could be expressed as the object of the preposition “by,” for example, “by the police.” We infer that it is the guard in this case who is in trouble. And we do this automatically and unself-consciously. For that matter, we have no trouble recognizing that the guard is being taken from the courtroom to the prison, and not from the prison to the courtroom. Real-world context is irrelevant to these aspects of meaning. Anyone saying or hearing this sentence will understand the relationships among the various phrases and words strictly by virtue of the grammatical roles that they play in the sentence. If someone used our example sentence to tell us that the guard took the defendant from the courthouse to the prison, or from the prison to the courthouse, we would say that the speaker has not spoken accurately. This is not to say that syntax and the aspects of meaning that derive from it are without ambiguity. A classic example is Chomsky’s “Flying

40

chapter 2

planes can be dangerous”75 and the variation “Visiting relatives can be annoying.” These sentences are ambiguous in just the way described above: we can organize them into precisely two alternative structures, which result in two distinct meanings, both of which become obvious to us on even short reflection. We all experience examples of syntactic ambiguity in everyday life, sometimes entertaining ones. At the gym I once used are signs that say, “Please do not leave your locker or personal items unattended to prevent theft.” Common sense tells us to ignore the interpretation in which the proprietors are encouraging people not to prevent theft. Syntactic ambiguity sometimes creates problems for the interpreters of statutes. Consider Liparota v. United States.76 The defendant had purchased food stamps from an undercover agent for less than face value, which is against the law. The relevant statute said, “whoever knowingly . . . acquires . . . [food stamps] in any manner not authorized by this chapter or the regulations . . . has committed a felony.”77 In Liparota, the defendant claimed not to know that his purchase was illegal. Although ignorance of the law does not excuse criminal activity, this statute is syntactically ambiguous, and under one of the readings, knowledge of the law is an element of the crime. The Supreme Court held that this ambiguity must be resolved in the defendant’s favor. But this kind of ambiguity is not often the subject of statutory disputes. Instead, battles over statutory meaning are nearly always about the “wordlike” aspects of the statute’s language rather than the syntactic, rulelike aspects of the statute. Some of the classic cases that we will discuss in later chapters involve the following questions. Does a minister’s work count as “labor” for purposes of a statute that bars the payment of transportation for any individual performing labor of any kind?78 Should an airplane in 1931 have been considered a vehicle for purposes of a statute outlawing the transportation of stolen vehicles across state lines?79 Has a person “used a firearm” when he has attempted to trade a machine gun for cocaine?80 If we approach these questions in a rulelike manner, we would perhaps consult a dictio­ nary and determine whether preaching satisfies the conditions set forth in the definition of “labor.” But if we approach them in a wordlike manner, we would instead ask, for example, how close a fit the concept of “preaching” makes with the concept of “labor.” The problem with this statute is not that it is ambiguous in the way that “Flying planes can be dangerous” is syntactically ambiguous. The problem is that the concept of “labor”— its word meaning—becomes vague at the margins, making it hard to tell whether we would consider it fair to equate preaching and labor.

why we need to interpret statutes

41

Word meaning is not chaotic though. We form our concepts the same way, so that within a community there is relative consensus about the application of words to situations in clear cases. As the philosopher Jerry Fodor has put it, we must have a cognitive architecture that causes us to “lock” on to the same kind of concept “from experience with good examples.”81 Nonetheless, word meanings are additionally shaped by experiences that disturb this consensus in ways that may not be present in rulelike phenomena.82 It now seems well established that significant aspects of word meaning, especially what are called “prototype effects” after Eleanor Rosch’s pioneering work of the 1970s, are based in large part on experience and lead to uncertain and conflicting judgments as we try to apply words to unusual situations.83 Psychologists who study conceptualization and linguists who study lexical semantics continue to investigate how we conceptualize and categorize. We take a closer look at this research in chapter 3. For now, it is enough to observe that we all employ similar psychological processes, even if those processes yield somewhat different judgments at the margins.

Words, Rules, and Lenity Judges do not write about the difference between rulelike problems and wordlike problems in the resolution of statutory disputes.84 Nonetheless, they do not treat these problems the same way. Some of the differences reveal themselves in the application of the rule of lenity: the rule that requires ambiguities in criminal statutes to be resolved in favor of the defendant. The question of lenity arose early in American judicial history. Chief Justice John Marshall’s analysis in United States v. Wiltberger, decided in 1820, contains a highly perceptive understanding of the language issues and is still instructive today.85 The issue in that case was whether the federal courts had jurisdiction over a prosecution for manslaughter when the homicide had occurred on an American merchant marine vessel located on the Tigris River in China. Section 12 of the 1790 act in dispute criminalized manslaughter “upon the high seas” for purposes of federal jurisdiction. The question was whether this should be construed as including rivers in foreign countries. The government’s principal argument was that section 8 of the act gave the courts jurisdiction over prosecutions for robbery committed “upon the high seas, or in any river, haven, basin or bay.”86

42

chapter 2

It only made sense, the argument continued, to conclude that Congress used the expression “high seas” in section 12 as shorthand for the collection of bodies of water described in section 8. The Court had to decide whether it should expand the language of a criminal statute to include a class of defendants whom a reasonable legislature in all likelihood would have wished to include. Marshall answered this question negatively, thus establishing a linguistic wall. Criminal statutes will not be interpreted more broadly than their language reasonably permits, even if the legislature may have intended to criminalize additional conduct. Marshall did not limit himself solely to that issue, however. Rather, he used Wiltberger as a vehicle to discuss an entire array of interpretive problems. First, Marshall stated with elegance the rule of lenity and its motivations: The rule that penal laws are to be construed strictly, is perhaps not much less old than construction itself. It is founded on the tenderness of the law for the rights of individuals; and on the plain principle that the power of punishment is vested in the legislative, not in the judicial department. It is the legislature, not the Court, which is to define a crime, and ordain its punishment.

But Marshall then continued: It is said, that notwithstanding this rule, the intention of the law maker must govern in the construction of penal, as well as other statutes. This is true. But this is not a new independent rule which subverts the old. It is a modification of the ancient maxim, and amounts to this, that though penal laws are to be construed strictly, they are not to be construed so strictly as to defeat the obvious intention of the legislature. The maxim is not to be so applied as to narrow the words of the statute to the exclusion of cases which those words, in their ordinary acceptation, or in that sense in which the legislature has obviously used them, would comprehend.87

Marshall’s warning that an unnaturally narrow reading of a statute should be avoided is dicta. In Wiltberger, the language of the statute was clear. The caveat nonetheless demonstrates a high level of sophistication. Marshall recognized that statutory terms become vague at the margins. Lord Blackstone had written about judges applying lenity to a statute that criminalized “stealing sheep or other cattle.” The court determined that the general words “or other cattle” were “much too loose to create a capi-

why we need to interpret statutes

43

tal offence, and the act was held to extend to nothing but mere sheep.”88 Marshall wanted to make sure that his opinions would not lead once again to judicial findings that cows are not cattle. To discern the intent of the legislature, we look at the words “in their ordinary acceptation, or in that sense in which the legislature has obviously used them.” That is, we draw the inference that the legislature has used statutory words in their prototypical sense. Inquiry into such matters left Marshall with only a single choice. After all, rivers are not the high seas. As far as Marshall was concerned, then, the statute really presented no interpretive problem. It was simply a narrowly drafted law. An 1838 opinion by Justice Story shows exactly the kind of interpretive problem we saw in our exploration of the bribery statute and why it is difficult to find a satisfactory solution. United States v. Winn involved a statute making it unlawful for “any master or other officer” of an American ship to unjustifiably beat, wound, or imprison “any one or more of the crew of such ship.”89 Story used Marshall’s exact approach in refusing to apply the rule of lenity. The defendant, master of a ship, had been convicted of unjustifiably imprisoning the ship’s chief officer. The issue was whether the chief officer should be considered among the “crew” of the ship. Story acknowledged the rule of lenity but added: “I agree to that rule in its true and sober sense; and that is, that penal statutes are not to be enlarged by implication, or extended to cases not obviously within their words and purport. But where the words are general, and include various classes of persons, I know of no authority, which would justify the court in restricting them to one class, or in giving them the narrowest interpretation, where the mischief to be redressed by the statute is equally applicable to all of them.”90 He then proceeded to determine that the word “crew,” in its “general and popular sense,” meant the ship’s company. To reach this conclusion, Story referred to various dictionaries and to other provisions in the United States Code that used the word “crew,” much the way the Supreme Court has determined the appropriate meanings of words in recent years. Story found that statutes generally used the word “crew” to refer to the entire company on a ship. Statutes that used the word more narrowly usually drew a distinction by referring separately to the officers and the crew. Turning to the purpose of the statute, Story then asked rhetorically: “Why should we resort to the narrowest possible sense of the words, instead of the general sense, if there is the same mischief in each case to be suppressed, and the same public policy in the protection of the

44

chapter 2

commercial interests of the country?”91 Combining analyses of the ordinary meaning of the word in dispute, its use in the United States Code, and the purpose of the statute, Story concluded that the rule of lenity should not apply and let the conviction stand. I focus on these early judicial decisions because they are perceptive and because very little has changed over the past two centuries. Both Marshall and Story knew that lenity could convert statutory interpretation into a language game every time the scope of a word is placed in issue. They were unwilling to infer a meaning broader than the words that Congress used would reasonably permit, even at the cost of accepting an undesirably narrow interpretation of a statute. However, when Congress has spoken in words that arguably, but not necessarily, encompass the conduct before a court, simply choosing the narrowest interpretation will not do. Contemporary cases repeat this sentiment, frequently reminding us that, in applying the rule of lenity, courts may give a criminal statute a broader interpretation when there exist “nontextual factors that make clear the legislative intent where it is within the fair meaning of the statutory language.”92 In fact, the Supreme Court often rejects the rule of lenity when the linguistic issue is the breadth of word meaning. For example, Fischer v. United States, a case decided in 2000, involved the interpretation of yet another bribery statute, this one governing those who receive benefits of more than $10,000 under a federal program.93 The defendant was the president of a company that conducted billing audits for medical providers who received payments under the federal Medicare program. One of those providers gave Fischer’s company a loan, and Fischer in turn paid an official of that provider a $10,000 kickback for helping him obtain the funds. He was eventually indicted for misusing the funds and for bribery. There was no doubt that he had given a bribe, but the statute applies only when “(b) . . . the organization, government, or agency receives, in any one year period, benefits in excess of $10,000 under a Federal program involving a grant, contract, subsidy, loan, guarantee, insurance, or other form of Federal assistance.”94 Fischer argued that the ordinary meaning of “benefit” should not include a hospital that receives Medicare in exchange for providing services. Rather, it should apply only to the direct recipients of those services. In reasoning that resembles that of Justice Story, the Supreme Court disagreed, refusing to give the word its narrowest meaning: “That one beneficiary of an assistance program can be identified does not foreclose the existence of others, however. In this respect petitioner’s construction would give incomplete meaning to the term ‘benefits.’ Medi-

why we need to interpret statutes

45

care operates with a purpose and design above and beyond point-of-sale patient care, and it follows that the benefits of the program extend in a broader manner as well.”95 Or consider Bryan v. United States, a 1998 case that involved the meaning of the state-of-mind adverb “willfully.”96 Federal law requires that those who sell firearms first obtain a license and subjects to criminal sanctions any person who “willfully violates . . . any provision of this chapter.”97 Related criminal conduct requires only that the defendant performed the acts “knowingly.” Bryan indeed sold firearms without a license and did so knowing it to be unlawful, based on his sneaky conduct. But he had no idea that his behavior violated any particular law. In other cases, the courts have interpreted “willfully” to require some level of knowledge of the law that is being broken. Recognizing that “[t]he word ‘willfully’ is sometimes said to be ‘a word of many meanings’ whose construction is often dependent on the context in which it appears,” the Court decided that general knowledge of unlawfulness was enough to meet the statutory requirement.98 Justice Scalia dissented, arguing that this level of linguistic uncertainty should result in application of the rule of lenity. Only in a few cases has the Court ruled in favor of a defendant when the linguistic issue was a borderline case of word meaning. In United States v. Aguilar, the Court, in an opinion by Chief Justice William Rehnquist, interpreted the obstruction-of-justice statute narrowly to hold that a judge making false statements to FBI agents in the course of a criminal investigation was not “endeavor[ing] to influence, obstruct, or impede . . . the due administration of justice” required for conviction under the statute.99 The Court held that the false statements must have a nexus with judicial proceedings, such as a grand jury. Virtually all of the statute’s other provisions made reference to such proceedings. The majority held that, absent proof that the judge knew that his false statements would influence grand jury investigations and the like, he had not obstructed justice. Rehnquist never mentioned the rule of lenity, but he did mention its rationales of “deference to the prerogatives of Congress” and “fair warning . . . to the world in language that the common world will understand.”100 It was Justice Scalia, in dissent, who argued correctly that the majority really was applying lenity through the back door. To Scalia, the statute was so clear that even a liberally applied rule of lenity was not necessary to decide the case.101 Similarly, consider the prosecution of Dewey Jones, who threw a Molotov cocktail through his cousin’s window and was prosecuted under a federal statute that makes it a crime to damage or destroy, “by means

46

chapter 2

of fire or an explosive, any . . . property used in interstate or foreign commerce or in any activity affecting interstate or foreign commerce.”102 The case made it to the Supreme Court, where the issue was whether torching a private residence fit within the statute. All nine justices agreed that it did not, although an argument could be made that even a private home is “used” in activities affecting interstate commerce. Consistent with both lenity and the principle that statutes should be interpreted to avoid close constitutional questions (in this case a potential violation of the commerce clause),103 the Court held that commercial activity is required for the statute to apply. Looking at the ordinary meaning of the statute, the Court noted: “It surely is not the common perception that a private, owner-occupied residence is ‘used’ in the ‘activity’ of receiving natural gas, a mortgage, or an insurance policy.”104 Perhaps the Court would have made the same decision without invoking the rule of lenity, relying instead on the rule that words are to be afforded their ordinary meaning, a principle addressed in detail in chapter 3. At best, then, courts are ambivalent about applying lenity when a statute is vague, that is, when the difficulty in applying the law to a set of facts is wordlike. The application of lenity to these cases inescapably requires that difficult line-drawing decisions be made, and it is easy enough to thwart the legislative will by imposing too narrow an interpretation. The Court, in contrast, has been more generous in applying lenity to cases in which the statute is ambiguous, that is, when the problem is rulelike. Liparota (the food stamp case discussed above) is a good illustration. Consider also United States v. Granderson, a case decided by the Supreme Court in 1994.105 Granderson, a letter carrier, was convicted of destruction of mail and sentenced to five years’ probation. The maximum prison sentence under the federal Sentencing Guidelines would have been six months. While on probation, he was caught with illegal drugs. A statute states: “the court shall revoke the sentence of probation and sentence the defendant to not less than one-third of the original sentence.”106 What is “one-third of the original sentence”? Taken out of context, the most natural reading is that probation is revoked, and the individual is resentenced to a much shorter period of probation. It can also mean, as the government argued, that the individual must serve a prison sentence of at least one-third of the time that he would have been on probation had he not possessed drugs during his probation period (one-third of five years in this case). Alternatively, it can mean, as Granderson argued, that the individual must serve a sentence of at least one-third of the sentence that he

why we need to interpret statutes

47

would have received had he been sentenced to prison in the first place and not to probation (one-third of six months). The Court accepted Granderson’s interpretation. The majority of justices agreed with the government that it would be absurd to interpret the statute to require Granderson’s sentence to be reduced to an even shorter period of probation.107 But the Court also rejected the government’s request for a longer sentence, in part using the rule of lenity to resolve the ambiguity between the remaining two interpretations. In 1994, the year Granderson was decided, Congress amended the statute to require a court to resentence the defendant under the Sentencing Guidelines.108 Similarly, in Ratzlaf v. United States, also decided in 1994, the Supreme Court used lenity to resolve an ambiguity.109 A regulation requires that bank deposits of $10,000 or more in cash be reported to the government. A statute makes it illegal to structure transactions to avoid the reporting requirement, and another statute imposes criminal sanctions on a person “willfully violating this subchapter.”110 Ratzlaf had won money gambling and intentionally deposited just under $10,000 in a number of banks to avoid the reporting requirements. He was charged with and convicted of willfully structuring the transactions to evade reporting. The conviction was affirmed by a U.S. court of appeals,111 but the Supreme Court reversed. The question before the Court was whether Ratzlaf had “willfully evaded violating the subchapter” that prohibits structuring. He argued that he did not know that it was a crime to avoid reporting the deposits, so he did not act willfully with respect to “violating this subchapter.” In fact, the language is ambiguous in a way that is well studied by linguists and philosophers. In the transparent reading of the statute, one can simply substitute the reporting requirements for the words “this subchapter,” in which case Ratzlaf clearly acted willfully. In the opaque reading, the words “this subchapter” are opaque to the statute’s content. This is the same linguistic ambiguity that we saw earlier in Circuit City Stores v. Adams, the case involving the obligation to arbitrate employment disputes.112 Thus, on one reading of the statute, unless Ratzlaf knew that he was violating this particular law, he should not be convicted. The Court resolved this twoway ambiguity in favor of the defendant. In both Granderson and Ratzlaf, the Court applied the rule of lenity. In Granderson, the Court interpreted favorably for the defendant the expression “original sentence,” and in Ratzlaf, it narrowly defined a “willful violation” of the cash transaction structuring requirements. Such decisions

48

chapter 2

are far less threatening to the rule of law. For one thing, close cases involving ambiguity occur far less frequently than do close cases involving vagueness. For another, these rulelike ambiguities leave open, not a range of options that require refined judgment, but rather two or three clear choices. When the Court elects one of them, it by definition eliminates the others forever, resolving the interpretive issue once and for all unless the legislature revises the statute.

Where This Leaves Us This review is by no means encyclopedic. For example, disputes often arise about the relationship between a statute and other aspects of the legal system, such as the Constitution, the relationship between one law and another, sentencing schemes, defenses, and so on. In fact, constitutional questions have arisen—and been resolved—about how far the bribery statute can delve into the legislative activities of members of Congress.113 Some interesting generalizations have already presented themselves, however. Most significantly, we can now begin to see why the statutory system works as well as it does, and to see what goes wrong. Statutory drafters typically use words that best capture the recurring scenarios that triggered the statute’s enactment. Most of the time, for example, it is obvious enough who is a public official and who is not a public official. A police officer is; my dry cleaner is not. I know all this and have no trouble deciding who should receive a gift to ensure good service in the future and who should not. By the same token, we generally understand the meaning that derives from a statute’s syntactic structure automatically and with very little difficulty, resolving ambiguities from context so quickly that we never even notice them. This begins to explain why statutes work as well as they do: they are written to address recurring situations, using language that captures those situations well enough in the paradigmatic cases. The biggest problem with statutes is that they are written in words whose application becomes uncertain in situations that stray from the typical case. This is a well-known phenomenon in the psychology of language, to which we return in chapter 3. Questions of state of mind also play a significant role, both in disputes over the meaning of the statutory language and in evidentiary questions about when we can be confident that we know what another person was thinking. Both of these issues will resurface in chapter 4, which deals with the issue of legislative intent.

why we need to interpret statutes

49

It is impossible to write a statute whose words will not be subject to debate at the margins. Nor is it possible to solve the problem by defining the terms in the statute, because the words of the definitions will themselves be subject to the same problems. Nonetheless, over time, as recurrent problems are litigated, and the outcomes of the cases become precedents for future disputes, the interpretation of the words tends to stabilize. Although bribery continues to occur, the legal framework for it is now fairly well established. The current version of the bribery statute was enacted in 1962, many cases were decided in the next two decades, and since then there has been relative calm.114 I will examine the trend toward stability in the history of a statute’s interpretation in chapter 5. As for rulelike problems, they are both less frequent and more easily resolved. Thus, they generally present less of a threat to rule-of-law values. A syntactic ambiguity typically leaves only two or three reasonable interpretations of a statute. Once a court decides which of them wins out, the problem goes away. The same holds true for constitutional challenges. The law does not itself acknowledge the difference between wordlike problems and rulelike problems. But it recognizes them tacitly in the way it treats various situations.

chapter three

Definitions, Ordinary Meaning, and Respect for the Legislature

A

s we have seen, most disputes over the meanings of statutes are about     the fit between events in the world and the words in the statute.1 We focused on a single statute in chapter 2, but the generalization holds across a wide sample of statutory cases: The defendant did x. The statute says it is a crime to do y. Should x be considered a member of the category y for purposes of interpreting the statute? Sometimes an event appears to fit within the language of the statute, but it seems absurd to think that the legislature intended to make the disputed activity a crime. At other times, the defendant has done something similar to, and just as bad as, the activities included in the statutory language, but it seems like a stretch to say that the words include the event in question. In still other cases, it is simply hard to know what to do. Questions about how to resolve disputes in statutory cases have generated a polarized debate in both the courts and legal academic literature between those who regard themselves as textualists and those who advocate for courts to consider a broader range of information.2 In simplest terms, textualists claim that it is illegitimate for courts engaged in statutory interpretation to rely upon the intent of the legislature, and especially upon legislative history as evidence of such intent.3 Their various opponents, in contrast, maintain both the legitimacy and the usefulness of these tools.4 At times, evidence of the legislature’s intent appears to contradict

definitions, ordinary meaning, and respect for the legislature

51

the statute’s plain meaning. In these cases, advocates of each position pose different arguments concerning how a dispute should be decided. Gone largely unnoticed in the battles between these camps during the past quarter century is the fact that both sides in the debate agree upon almost everything when it comes to statutory interpretation. Most of those whom textualists call “intentionalists” are really not that at all. Rather, they take a pragmatic, eclectic approach to the interpretation of statutes, relying upon whatever information appears to provide an interpretation that is loyal to the language of the statute and the intent of its drafters and is coherent with the code in general.5 Chief Justice John Marshall’s statement that “[w]here the mind labors to discover the design of the legislature, it seizes every thing from which aid can be derived” characterizes this approach.6 Jonathan Molot criticizes textualists for exaggerating the extent to which those who seek to further the purpose of the statute are willing to stray from any fair reading of a law to infer intent, and I agree with this criticism.7 What the two sides share is a commitment to legislative primacy as the core value in statutory interpretation. Once a statute is enacted, the common-law court’s authority is curtailed, and its obligation is to effect the decision that the legislative body has made. This is not the only value that judges take into account, however. In chapter 5, we will look at values that compete with legislative primacy, including fair notice, constitutional values, stability, flexibility, law enforcement, and others. However, the primacy of the legislature is a shared value and is the principal consideration that courts articulate in statutory cases. Most of the debate in the interpretation of statutes, then, is not what a court should try to do but rather how a court should accomplish what everyone agrees should be accomplished. For example, Supreme Court justices of all political stripes routinely begin statutory interpretation by analyzing the language of the statute. Justices with political orientations as diverse as Chief Justice Roberts’s8 and the late Chief Justice Rehnquist’s on the right9 and Justice Ginsburg’s on the left10 say so using virtually the same words. Often enough, the inquiry need go no further. Yet at times, information that may seem helpful in isolation conflicts with other such information. Whatever their philosophical orientation, courts faced with this situation must exercise discretion. Ordinarily, however, they are not in conflict, and the language of the statute is applied without controversy.

52

chapter 3

By the same token, textualists do not eschew all contexts in the interpretation of statutes, as a cartoon-like description of the approach may suggest. To the contrary, proponents of both approaches find no difficulty looking at the earlier interpretive decisions of courts, background assumptions shared by the relevant community, constitutional considerations, questions of coherence with related statutes, and a host of other considerations. Only some context disturbs textualists, in particular, legislative history adduced as evidence of legislative intent, which they regard as illegitimate.11 The two camps, then, espouse similar but competing approaches to statutory interpretation that differ largely in the willingness of one but not the other to use a particular species of evidence. Thus, I agree with Caleb Nelson’s assessment that the differences between textualists and their opponents are “less categorical than either textualists or their critics generally acknowledge, and they are correspondingly harder to describe.”12 The goal of the textualist program is to reach results in disputed cases that are sensitive to a statute’s purpose, and thus respectful of the primacy of the legislature, without resorting to extratextual materials that create both evidentiary and conceptual difficulties. By adopting an enriched approach to language as an initial matter, a view that considers context an important element of word meaning itself, textualist practice can internalize a great deal of contextual information while at the same time maintaining procedures less likely to lead courts into a decision-making process that conflicts with basic values such as separation of powers. As John Manning explains: Even the strictest modern textualists properly emphasize that language is a social construct. They ask how a reasonable person, conversant with the relevant social and linguistic conventions, would read the text in context. This approach recognizes that the literal or dictionary definitions of words will often fail to account for settled nuances or background conventions that qualify the literal meaning of language and, in particular, of legal language. Accepting this more modern understanding of textual interpretation, I believe, offers a firmer and more legitimate basis for cutting off many problems of absurdity at the threshold.13

This approach to statutory interpretation, reflected in many opinions by Justice Scalia, incorporates a context-sensitive perspective on word meaning that helps blunt the bite of reducing the universe of evidence permitted

definitions, ordinary meaning, and respect for the legislature

53

in the interpretive process, thereby increasing the power of the methodology.14 The key intellectual decision is to focus on the ordinary meaning of statutory words rather than on dictionary definitions that describe, using necessary and sufficient conditions, the outer limits of a word’s proper usage without regard to the circumstances in which the word is used in a particular instance. I refer to these two approaches as the ordinarymeaning and definitional-meaning approaches to statutory language. Both are based entirely on the language of the statute but in many cases lead to opposite conclusions. The two correspond respectively to the wordlike and rulelike aspects of our linguistic capacity discussed in chapter 2. What makes ordinary meaning ordinary? Experience. In the circumstances in which a word is used, the ordinary meaning is the one that stands out, since that is the meaning people generally intend to convey when they use that word in circumstances like those. Some dictionaries actually capture aspects of ordinary meaning by using such modifiers as “usu[ally],” “typ[ically],” and “often.” Textualist judges (and academics) who rely on ordinary meaning have therefore allowed context into their analysis through the back door without ever looking at anything outside the text. The shift from focusing on dictionaries to ascertaining ordinary usage should be seen as an additional step toward reconciling textualist methodology with the goal of providing an interpretation that helps to carry out a statute’s purpose. If so, it is worth asking how well it achieves this goal. I argue here that in many circumstances it does so very well. Nonetheless, there are several recurring situations in which the textualist effort falls short, even on its own terms. First, it is not a simple matter to reject the notion of definitional meaning in favor of ordinary meaning. In everyday life we think both ways, and judges— even textualist judges— do so as well. Moreover, as we saw in chapter 2, laws are written as definitions. Construing the words with which a law is written by reference to the conditions under which they are properly used furthers the notion that laws can be applied analytically, element by element. Thus, despite the intellectual attractiveness of focusing on ordinary meaning, judges continue to rely upon definitions based on necessary and sufficient conditions to a great extent in their opinions. The ordinary-meaning approach is most attractive to judges when it is independently clear that the legislature really did have the ordinary meaning in mind when it enacted the statute. But that, of course, is exactly the inquiry that textualists argue courts should not make.

54

chapter 3

Second, it is not always easy to decide what makes ordinary meaning ordinary. Just as the use of legislative history is riddled with evidentiary problems that result from the temptation to pick snippets from a long record to support a particular position, corresponding problems arise when attempting to find the ordinary meaning of a statutory term. Thus, while the linguistic orientation of the new textualists is likely to provide results acceptable to a broad spectrum of legal analysts in a broad range of circumstances, there will remain a residue of disagreement about which interpretation is the ordinary one. When this happens, it is inevitable that statutory interpreters will turn to the intent of the legislature. Legislative intent is the subject of chapter 4. Here, we explore some of the advantages and disadvantages of an approach to text that builds context into the meanings of words.

Two Approaches to the Meaning of Statutory Words Let us begin with Church of the Holy Trinity v. United States, the famous case decided by the Supreme Court of the United States in 1892.15 A statute made it a crime “in any manner whatsoever, to prepay the transportation . . . of [an] alien . . . to perform labor or service of any kind in the United States.”16 A church was convicted of violating this statute, having prepaid the transportation of its rector from England. The Supreme Court reversed the conviction. The opinion expresses the conflict that courts experience when they confront problems of statutory meaning. Writing for a unanimous Court, Justice David Brewer first noted that “the act of the [church] is within the letter of this section.” On the next page, he continued: “It is a familiar rule, that a thing may be within the letter of the statute and yet not within the statute, because not within its spirit, nor within the intention of its makers.” He concluded: “No one reading such a title would suppose that Congress had in its mind any purpose of staying the coming into this country of ministers of the gospel, or, indeed, of any class whose toil is that of the brain.”17 Brewer understood at an intuitive level that the minister’s activities were within the definitional meaning of the statute. They fit within the outer boundaries of the concept, “labor or service of any kind.” However, they were not within the ordinary meaning of the words. People using the word “labor” in the context of this statute are most likely to have in mind physical labor. Brewer chose the ordinary-meaning approach over

definitions, ordinary meaning, and respect for the legislature

55

the definitional-meaning approach when the two conflicted. What is so special about Church of the Holy Trinity is that the same judge expressed competing views of word meaning on adjacent pages of a single opinion. How did Brewer know the ordinary meaning of “labor”? He relied on his intuition as a native speaker of English based on his experience in contemporary society and assumed that others shared his view. Scholars often discuss this case for the fact that the Court used legislative history to undermine the statute’s plain meaning.18 It is also possible to understand this case, however, as an example of a Court struggling between these two different notions of word meaning and choosing the ordinarymeaning approach over the definitional-meaning approach, much in the style of contemporary textualists.19 The tension between these two approaches to the meanings of statutes — definitional meaning versus ordinary meaning — appears in a number of classic legal decisions on statutory interpretation.20 Consider McBoyle v. United States, the 1931 Supreme Court case that inspired the “no vehicles in the park” debate between Hart and Fuller.21 Justice Oliver Wendell Holmes, writing for a unanimous Court, refused to extend the National Motor Vehicle Theft Act to include theft of an airplane. The act defined “motor vehicle” to “include an automobile, automobile truck, automobile wagon, motorcycle, or any other self-propelled vehicle not designed for running on rails.” Holmes looked at legislative history (no mention of airplanes), looked at related statutes (different treatment of airplanes), and applied the principle of ejusdem generis (statutory list appears to include only land vehicles). He concluded that airplanes should not be considered vehicles for purposes of the statute. Consistent with the ordinary-meaning approach, Holmes stated: “But in everyday speech ‘vehicle’ calls up the picture of a thing moving on land.”22 The notion of adjudicating by calling up a mental picture might seem nonlegalistic, but some three decades later, Wittgenstein articulated the same approach to word meaning: “What really comes before our mind when we understand a word?—Isn’t it something like a picture? Can’t it be a picture?”23 Holmes knew very well that the statute referred to the theft of vehicles and that airplanes are vehicles. Nonetheless, he rejected the definitional approach to word meaning in favor of analyzing what Congress seemed to have regarded as the ordinary understanding of what counted as a vehicle at the time. In more recent cases, the Court has occasionally recognized the issue and addressed it head on. In Dolan v. United States Postal Service, decided

56

chapter 3

in 2006, the issue before the Court was whether Ms. Dolan, who tripped over some mail negligently left on her porch by her letter carrier, could sue the Postal Service.24 The Federal Torts Claims Act waives sovereign immunity, thus allowing actions against the government. The version of the statute in place when Ms. Dolan brought her suit permitted actions “for claims against the United States . . . for injury or loss of property, or personal injury or death caused by the negligent or wrongful act or omission of any employee of the Government while acting within the scope of his office or employment, under circumstances where the United States, if a private person, would be liable to the claimant in accordance with the law of the place where the act or omission occurred.”25 Surely Ms. Dolan’s claim fits within that definition. There are exceptions, however, and one of them involves negligence by the post office: “The provisions of this chapter and section 1346(b) of this title shall not apply to . . . any claim arising out of the loss, miscarriage, or negligent transmission of letters or postal matter.”26 It is easy enough to conceptualize Ms. Dolan’s accident as coming within the negligent transmission exception. In fact, the lower courts had done just that. But it is also possible to understand the exception as preventing people from suing the post office for damaged mail, late deliveries, or deliveries to the wrong address—not for physical injury caused by ordinary negligence. The Supreme Court agreed with this approach and based its decision largely on a rejection of dictionary definitions in favor of ordinary understanding. After conceding that the dictionary definition of “transmission” can well include the current case, Justice Anthony Kennedy’s majority opinion went on: The definition of words in isolation, however, is not necessarily controlling in statutory construction. A word in a statute may or may not extend to the outer limits of its definitional possibilities. Interpretation of a word or phrase depends upon reading the whole statutory text, considering the purpose and context of the statute, and consulting any precedents or authorities that inform the analysis. Here, we conclude both context and precedent require a narrower reading, so that “negligent transmission” does not go beyond negligence causing mail to be lost or to arrive late, in damaged condition, or at the wrong address.27

Like Justice Brewer 114 years earlier, the Court chose, not context over language, but a contextualized understanding of word meaning over a

definitions, ordinary meaning, and respect for the legislature

57

purely definitional one that extends to a word’s outermost conceptual boundaries. Among the most interesting examples of this struggle are those in which the members of the Court divide between the two approaches. We see this in the 1993 Supreme Court case Smith v. United States.28 John Smith drove from Tennessee to Florida, where he planned to buy cocaine from a contact he had there. Unfortunately for Smith, his contact had become a government informant, who arranged a meeting with undercover police officers. At the meeting, Smith offered to trade his unloaded machine gun, which he had in a case, for the drugs. The undercover agents pretended to agree and said they would return soon with the cocaine. Smith, however, became uncomfortable with the situation and drove off in his van. The police caught him, and the government prosecuted him for attempting to procure cocaine using a firearm. A federal statute required an enhanced sentence for anyone who “uses or carries a firearm” “during and in relation to any crime of violence or drug trafficking crime.”29 The court found that Smith’s attempted barter transaction came within that definition. The statute imposed a minimum of five years in prison for those using an ordinary firearm and a minimum of thirty years for those using a machine gun.30 The court sentenced Smith to the thirty-year statutory minimum. The case made its way to the Supreme Court, which affirmed the conviction by a vote of six to three.31 Writing for the majority, Justice Sandra Day O’Connor first recited the rule of law that “[w]hen a word is not defined by statute, we normally construe it in accord with its ordinary or natural meaning.”32 But O’Connor’s inquiry was not really about the ordinary meaning. Rather, she looked up the word “use” in a number of dictionaries: “Webster’s defines ‘to use’ as ‘[t]o convert to one’s service’ or ‘to employ.’ ”33 Black’s Law Dictionary contains a similar definition: “[t]o make use of; to convert to one’s service; to employ; to avail oneself of; to utilize; to carry out a purpose or action by means of.”34 Earlier Supreme Court decisions had adopted similar definitions.35 Justice Scalia dissented.36 Relying upon his judgments about ordinary meaning that come from everyday use, Scalia echoed Brewer’s perspective from Church of the Holy Trinity, decided one hundred years earlier. He wrote: In the search for statutory meaning, we give nontechnical words and phrases their ordinary meaning. To use an instrumentality ordinarily means to use it

58

chapter 3 for its intended purpose. When someone asks, “Do you use a cane?,” he is not inquiring whether you have your grandfather’s silver-handled walking stick on display in the hall; he wants to know whether you walk with a cane. Similarly, to speak of “using a firearm” is to speak of using it for its distinctive purpose, i.e., as a weapon. To be sure, “one can use a firearm in a number of ways,” including as an article of exchange, just as one can “use” a cane as a hall decoration—but that is not the ordinary meaning of “using” the one or the other. The Court does not appear to grasp the distinction between how a word can be used and how it ordinarily is used. It would, indeed, be “both reasonable and normal to say that petitioner ‘used’ his MAC-10 in his drug trafficking offense by trading it for cocaine.” It would also be reasonable and normal to say that he “used” it to scratch his head. When one wishes to describe the action of employing the instrument of a firearm for such unusual purposes, “use” is assuredly a verb one could select. But that says nothing about whether the ordinary meaning of the phrase “uses a firearm” embraces such extraordinary employments. It is unquestionably not reasonable and normal, I think, to say simply “do not use firearms” when one means to prohibit selling or scratching with them.37

Scalia also referred to the dictionary, but he did so only to point out how many definitions of “use” it contains to further emphasize his point that “use” gets most of its meaning from context.38 O’Connor replied to Scalia’s remarks and disrespectful tone: There is a significant flaw to this argument. It is one thing to say that the ordinary meaning of “uses a firearm” includes using a firearm as a weapon, since that is the intended purpose of a firearm and the example of “use” that most immediately comes to mind. But it is quite another to conclude that, as a result, the phrase also excludes any other use. Certainly that conclusion does not follow from the phrase “uses . . . a firearm” itself. As the dictionary definitions and experience make clear, one can use a firearm in a number of ways. That one example of “use” is the first to come to mind when the phrase “uses . . . a firearm” is uttered does not preclude us from recognizing that there are other “uses” that qualify as well. In this case, it is both reasonable and normal to say that petitioner “used” his MAC-10 in his drug trafficking offense by trading it for cocaine; the dissent does not contend otherwise.39

The exchange contains important substantive issues, along with some entertaining ironies. Foremost among the latter is the fact that Scalia, in his

definitions, ordinary meaning, and respect for the legislature

59

1997 book on legal interpretation, ridiculed Brewer’s opinion in Church of the Holy Trinity, especially its reference to the “spirit of the law.”40 At the same time, Scalia used his dissenting opinion in Smith as an illustration of how statutory interpretation should be conducted.41 Scalia never recognized that Brewer’s approach was identical to his own. Far more important are the substantive issues. Although O’Connor cast her analysis as one of ordinary meaning, her opinion is more consistent with a definitional approach to legal interpretation as embodied in Brewer’s “letter of the law” analysis. This approach to interpretation often manifests itself in the “plain-meaning rule.” A typical statement appears in Caminetti v. United States, a 1917 Supreme Court decision: “It is elementary that the meaning of a statute must, in the first instance, be sought in the language in which the act is framed, and if that is plain, and if the law is within the constitutional authority of the law-making body which passed it, the sole function of the courts is to enforce it according to its terms.”42 Courts continue to employ this rule today, and they routinely refer to the “plain language” of a statute or other document as an argument supporting one interpretation over another.43 The Smith Court never discusses the tension between definitional and ordinary meaning as a phenomenon in its own right. But in this and other cases discussed below, the tension lies just under the surface, embedded in a larger controversy over the extent to which statutory language can be seen as self-contained, or autonomous, as Peter Tiersma puts it.44 The more self-contained it is, the less the role for judges, who are sometimes seen as an antidemocratic force bent on making their own policy notwithstanding the will of the people reflected in the laws their elected representatives enact. Scalia’s “new textualism” discourages judges from referring to some contextual information, especially information about the statute’s legislative history leading up to its enactment.45 That is why he remains so bothered by Brewer’s reference to “the spirit” of a law more than a century ago. Or consider Small v. United States, decided by the Supreme Court in 2005. A law makes it “unlawful for any person . . . who has been convicted in any court of, a crime punishable by imprisonment for a term exceeding one year . . . to . . . possess . . . any firearm.”46 Gary Small had served a prison term in Japan after having been convicted there of weapons charges. The issue in Small was whether the statute bars those who have been convicted in foreign courts from owning firearms or whether “in any court” should

60

chapter 3

be understood as referring to federal and state domestic courts. Relying on the presumption that statutes are written with domestic considerations in mind, and concerned about forcing courts to determine on a case-bycase basis whether any particular violation of foreign law corresponds well enough with American legal principles to trigger the federal law, the Court ruled in Small’s favor, rejecting a rulelike approach to statutory interpretation in favor of an approach that draws inferences from ordinary usage. Justice Breyer wrote for the majority: “The[se considerations] simply convince us that we should apply an ordinary assumption about the reach of domestically oriented statutes here—an assumption that helps us determine Congress’ intent where Congress likely did not consider the matter and where other indicia of intent are in approximate balance. . . . We consequently assume a congressional intent that the phrase ‘convicted in any court’ applies domestically, not extraterritorially. But, at the same time, we stand ready to revise this assumption should statutory language, context, history, or purpose show the contrary.”47 As for the expression “any court” in the statute, the majority observed that “any” must be understood in context. If I say that I am happy to go to “any film,” I am likely to mean any film among those that are playing locally, as the Court observed. Let us say that after discussing four or five films we might see, I told you that I would go to “any film.” If you took me to a film not among those we discussed and that you knew would be distasteful to me, I would be perfectly justified in saying that I did not mean “any” that broadly, and that you wrongly took advantage of my words to force me into something that you knew I did not want to do. The Court bolstered this point by citing earlier opinions in which “any” was construed less broadly than its outer limits would permit. The majority opinion provoked a vigorous dissent by Justice Thomas, joined by Justices Scalia and Kennedy. Basing his opinion on the statute’s “plain terms,”48 Thomas relied on other cases in which “any” had been interpreted broadly, some of which quoted the definition of “any” from Webster’s Third New International Dictionary. Various dictionaries were also used to define “court,” on the theory that the meaning of “any court” can be determined by concatenating the dictionary definitions of “any” and “court.” Of course, it is impossible to interpret laws out of context. As John Manning accurately notes, textualists have no quarrel with this fact and permit a limited range of tools to be used in statutory interpretation.49 These tools include reference to dictionaries, to the use of the same words

definitions, ordinary meaning, and respect for the legislature

61

elsewhere in the statute, to the use of the same words in other statutes, to court decisions, and to a set of canons of construction, including a rule that assumes that, in enacting statutes, the legislature had in mind the ordinary meanings of the words they used.50 Yet because ordinary meaning takes into account the ways in which people are most likely to use a word, the concept of ordinary meaning is probabilistic. For the only thing that is ordinary about ordinary meaning is the increased likelihood that the meaning the author intended to convey will be understood as such in the context in which the word was used. In fact, judges who employ the ordinary-meaning approach justify its use in just that way. Consider the following statement by Justice Scalia: “The question, at bottom, is one of statutory intent, and we accordingly ‘begin with the language employed by Congress and the assumption that the ordinary meaning of that language accurately expresses the legislative purpose.’”51 To Scalia, then, the ordinary meaning of statutory language provides a method for drawing an inference about the legislature’s intent without engaging in extratextual investigation.52 He elaborates on the methodological advantages of his approach in a dissent: “I thought we had adopted a regular method for interpreting the meaning of language in a statute: first, find the ordinary meaning of the language in its textual context; and second, using established canons of construction, ask whether there is any clear indication that some permissible meaning other than the ordinary one applies. If not—and especially if a good reason for the ordinary meaning appears plain—we apply that ordinary meaning.”53 The definitional approach, in contrast, asks only whether the disputed events fit cleanly within the outer boundaries of the disputed word’s meaning. Context is not important except, presumably, to distinguish between distinct senses of a word. It takes context to recognize even the obvious: banking law is the law governing certain financial institutions and is not the law governing certain airplane maneuvers. Even remaining within a single sense of a word, if one relies upon definitional meaning alone, one is likely to confront absurd results when the words incorporate a situation that the legislature clearly did not intend to make illegal. A classic example is United States v. Kirby, an 1868 case cited by the Supreme Court in Church of the Holy Trinity.54 A statute made it illegal to “knowingly and willfully obstruct or retard the passage of the mail, or of any driver or carrier, or of any horse or carriage carrying the same.”55 Kirby was a local sheriff in Kentucky who chased a man named Farris, who worked as a letter carrier and was wanted for murder. Kirby

62

chapter 3

arrested Farris while Farris was on his delivery route and was later prosecuted for obstructing delivery of the mail. A unanimous Court reversed his conviction because it made no sense to think that Congress would have wanted it otherwise.56 Thus, the consequences of choosing between the definitional-meaning and ordinary-meaning approaches to statutory interpretation depend largely upon what other evidence a court is willing to consider. The less extratextual material a court takes into account, the bigger the danger of injustice posed by the definitional approach, and the more the need for the ordinary-meaning approach. Even the staunchest textualists seem to recognize this. Making things more difficult for the legal system, neither courts nor legal scholars typically distinguish between these two concepts analytically. Often, for example, courts use the expression “plain meaning” to express either or both of these approaches to word meaning. Although some scholars recognize and rely upon the distinction,57 it is not unusual to read that the words of a statute should be given their “plain and ordinary meaning,” assuming that these two terms mean the same thing. For example, the Supreme Court has spoken of the “plain and ordinary meaning” of the word “under” in the expression “an adjudication under sec­tion 554.”58 Other courts use the expression routinely.59 Often enough, the definitional and ordinary meanings will coincide, and contextual information will serve to reinforce the interpretation. That is because the statutes themselves were written with their most typical applications in mind, and typical cases are per se both within the statute’s definitional and ordinary meanings. But at least in some instances, as we have seen, the difference between the two interpretive approaches matters. To see why this might happen, let us look at the psychology underlying plain and ordinary meaning.

Definitional and Ordinary Meaning: A Psycholinguistic Account Traditionally, linguists and philosophers regarded the meaning of a word as the set of conditions that must obtain for a statement using that word to be true. To take a classic example, a bachelor is an unmarried adult male. These conditions are each necessary and, taken together, are sufficient to define bachelorhood. This approach to meaning, sometimes called the classical approach because it is based on Aristotelian logic, is consistent

definitions, ordinary meaning, and respect for the legislature

63

with both the definitional-meaning approach to statutory interpretation and, as we saw in chapter 2, with everyday notions of the rule of law.60 A law consists of various elements, each necessary, together sufficient. Only if you disobey them all do you violate the law. For the past quarter century, psychologists, linguists, and philosophers have questioned the classical approach from a variety of perspectives. First, it is very challenging, if not impossible, to define words with conditions that are both necessary and sufficient. Philosopher Jerry Fodor has been making this point for twenty-five years.61 To take a classic example from the legal literature, try to define the word “vehicle” in the statute “no vehicles in the park,” and you will find it very difficult.62 More basically, try to define “book” or “pen” or anything else in your immediate reach so that your definition includes all instances of the concept and not much else; you will find the task both daunting and time-consuming. In short, we are not good at defining things, which is one reason that judges run to the dictionary so often. The linguist Charles Fillmore showed how some problems with definitions apply even to seemingly easy examples, like “bachelor.”63 All bachelors are unmarried adult males, but we are uncomfortable calling some people bachelors, such as the pope, Tarzan, and homosexual men. From this, one may conclude that even “bachelor” is a category better described by ordinary usage than by definitions. Legal writers making similar points often refer to Wittgenstein’s family resemblance categories64 to capture the fact that some categories are better understood in terms of a network of relationships that are typical but are neither necessary nor sufficient to define membership.65 Second, in the 1970s, Berkeley psychologist Eleanor Rosch began to question the psychological reality of the classical model.66 She observed that classical definitions often understate our understanding of concepts. Not only do we know whether a concept obtains, but we know how well it obtains. Rosch conducted experiments in which she asked subjects to rate how good an exemplar a concept is of a category. For instance, she found that people generally agreed that chairs and tables were good examples of furniture but that lamps were questionable examples. Yet this knowledge cannot be characterized by reference to defining conditions, whether necessary or sufficient. Rosch argued that prototypes, rather than definitions, better capture what people really know about categorization. Third, concepts become fuzzy at the margins. We can elongate a chair into a love seat and a love seat into a sofa. People might find it hard to

64

chapter 3

know what to call in-between sizes and are likely to disagree with one another. While examples like these may appear to threaten the use of definitions as a psychologically plausible theory of word meaning, other research shows that it is too early for the funeral to begin. Returning to the definition of “bachelor,” the classic definition is not all bad, even according to critics. No one thinks that anyone other than an unmarried adult male can be called a “bachelor.” So, at worst, the definition contains conditions that are each necessary but not together sufficient. Other attacks on definitions, such as Fodor’s, recognize this fact.67 Not only do definitional features seem to play at least some role in how we conceptualize and categorize, but the extent of the role of prototypes in the psychology of word meaning has been the subject of some debate. Rosch demonstrated that people are willing, when asked, to grade natural categories. For example, a robin is a good example of a bird, an ostrich a poor one.68 Armstrong, Gleitman, and Gleitman demonstrated, however, that although prototype effects indeed occur with categories such as “bird,” when later asked, subjects do not think that a prototypical bird is any more a bird than a nonprototypical one, and they disagree with the proposition that “bird” is a graded category at all.69 Moreover, complex concepts are compositional in a way that stereotypes are not. The prototypical pet fish is neither the prototypical pet nor the prototypical fish, as scholars have observed.70 Thus, although prototypes are part of our psychology, they do not seem to provide a full explanation of how we form concepts. Most psychologists now believe that we think both ways. We think in terms of prototypes in some circumstances and in terms of rules in others. Medin, Wattenmaker, and Hampson found that people prefer to rely on defining features when they do not have much information about the surrounding circumstances. They also found, however, that when people have greater information about context, they use family resemblance models based on prototypes.71 Psychologist Steven Sloman suggests that people employ both rule-based and associative systems in reasoning and that conflicts between the two occur frequently in everyday life. As an example, he suggests the dilemma of deciding whether to wear a seat belt for a car ride of a very short distance,72 say moving one’s car from one parking space to an adjacent one. An individual can rely upon experience-based intuitions about danger and not don the seat belt or can apply a rule: always wear a seat belt. Many other psychologists have reached similar conclusions and

definitions, ordinary meaning, and respect for the legislature

65

have suggested that people also include in their concepts such things as partial causal theories.73 Even dictionary writers recognize that our concepts consist of both definitional and prototypical features. Consider this definition of “book” in Webster’s Third: “a collection of written, printed, or blank sheets fastened together along one edge and usu. trimmed at the other edges to form a single series of uniform leaves; specif: a collection of folded sheets bearing printing or writing that have been cut, sewn, and usu. bound between covers into a volume.” Consisting of a collection of sheets is a necessary condition for a thing to be a book, but other aspects of the definition speak in terms of prototypes by means of signals such as “usu.” and “specif.” Put in terms of the dichotomy used in chapter 2, some aspects of word meaning are “wordlike,” while other aspects of word meaning are “rulelike.” What all this means for legal interpretation is that the choice between definitional and ordinary meaning is only natural. It appears that people reason about concepts from both the top down, consistent with the classical model, and from the bottom up, consistent with prototype analysis. Consider the concept of lying, touched upon in chapter 2. The Clinton scandals provoked a societal debate over the nature of truth and lying. Clinton’s critics argued that a false statement intentionally made is a lie and that lying under oath is perjury.74 His supporters retorted that some lies are worse than others.75 We routinely tolerate white lies, exaggerations, and other untruths. In fact, the perjury statute, which refers to “material” false statements, recognizes this fact.76 No doubt Clinton lied. However, his doing so to avoid disclosure of sexual misconduct that was sure to hurt his family did not, at least to the majority of people at the time, seem relevant to his performance as president and was not the worst kind of lie.77 A poll conducted in 1998 showed that most Americans agreed that Clinton had committed perjury both before the grand jury (71 percent vs. 23 percent) and in the Paula Jones case (64 percent vs. 27 percent) but did not believe that either was an impeachable offense (40 percent thought that his testimony to the grand jury was an impeachable offense, 37 percent thought that his testimony in the Paula Jones case was an impeachable offense).78 People do distinguish between paradigmatic lies and other sorts of false statements in everyday life. The two approaches to legal construction — definitional meaning and ordinary meaning— capture both the fact that people reach different conclusions and the ambivalence within each individual.79 Judges routinely use

66

chapter 3

both methods, although the definitional-meaning approach is the more common. This is not a surprising observation given that the definitional approach appears, at least superficially, to be the more “lawlike” of the two. As we saw in the previous chapter, laws are themselves structured as definitions. If, to win a conviction, the state must prove that a defendant has violated all elements of a crime, then it is natural to analyze similarly the words that make up the elements. In this way, the state must prove all the elements of all the elements. Prototype, or ordinary-meaning analysis, is to some extent methodologically inconsistent with this standard approach to applying laws in that it reduces the uniformity with which the law applies.

Challenges to Statutory Interpreters Although the ordinary meaning approach reduces the likelihood of judicial decisions at odds with what the legislature would have wanted, problems remain. First, it is not always as easy as it may seem to abandon the definitional approach, and it is not always clearly legitimate to do so if the goal is to establish a reliable proxy for actual legislative intent. Second, once one chooses the ordinary-meaning approach, it is not always clear what the ordinary meaning is. This section will also look at some promising approaches to interpreting legal texts that solve some, but not all, of these problems. Choosing between Definitional and Ordinary Meaning Given the way our minds work, we should not expect it to be a simple matter to eschew the definitional-meaning approach to word meaning in favor of the ordinary-meaning approach as a matter of doctrinal imperative. They are both firmly embedded in the way we think, and a definitional approach to meaning feels consistent with our basic notions of how laws work. This exposes two problems. The first is that the legislature may not always have intentionally limited the scope of a statute to the ordinary occurrences of the events that come within statutory language. For example, Adrian Vermeule argues that Congress really did intend a broader understanding of “service or labor of any kind” in the statute disputed in Church of the Holy Trinity, even if applying the statute would lead to banning the transportation of preachers.80

definitions, ordinary meaning, and respect for the legislature

67

To take a more recent case that has been the source of controversy, consider Chisom v. Roemer, a 1991 case interpreting section 2 of the Voting Rights Act.81 The act prohibits states from affording protected classes of people “less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice.”82 In Chisom, the question was whether the act applies to judicial elections as well as to legislative elections. At issue, then, was the scope of the word “representative” in the statute. In a six-to-three decision holding that judicial elections are within the bounds of the Voting Rights Act, the majority noted that the statutory language came from an earlier Supreme Court opinion, White v. Regester.83 That opinion, however, used the word “legislators” where the statute in Chisom uses the word “representatives.”84 The Court inferred that the change in language is best explained by a desire to expand the scope of the statute to elections other than legislative elections, a goal consistent with the overall purpose of the Voting Rights Act. In his dissent, Justice Scalia scolded the Court for straying from the ordinary meaning of “representative,” which he determined from looking at the dictionary.85 He noted that “[t]here is little doubt that the ordinary meaning of ‘representatives’ does not include judges.”86 Without question, Scalia was right about that much. When we think of elections of representatives, we ordinarily think of legislators, not judges. He further stated: The Court, petitioners, and petitioners’ amici have labored mightily to establish that there is a meaning of “representatives” that would include judges . . . and no doubt there is. But our job is not to scavenge the world of English usage to discover whether there is any possible meaning of “representatives” which suits our preconception that the statute includes judges; our job is to determine whether the ordinary meaning includes them, and if it does not, to ask whether there is any solid indication in the text or structure of the statute that something other than ordinary meaning was intended.87

The issue then becomes whether courts must limit their interpretation of statutory words to prototypical instances, even in the face of evidence that the legislature had a more expansive meaning in mind. To the extent that Scalia justifies his position on the premise that the ordinary meaning of a statute serves as an adequate proxy for the intention of the legislature, the argument is questionable. Legislators are not such consistent probabilistic reasoners that we can always assume that any

68

chapter 3

instance of a statutory word that strays from the prototype is necessarily outside a statute’s scope. They, too, use words with both prototypical and defining features in mind. While the ordinary-meaning approach provides a useful rule of thumb as to how a word was most likely used, it can do no more than that. A study by Miranda McGowan of Scalia’s dissenting opinions finds that he is not consistent in his use of the ordinary-meaning approach to statutory interpretation.88 If Scalia is correct that ordinary meaning is useful because it is likely to approximate intent much of the time, then it is entirely appropriate that he not apply the approach mechanically when to do so may sometimes serve to undermine the principle of legislative primacy. Legislators at times surely think in broad, definitional terms, and courts recognize that fact.89 Just as the rule of lenity tells courts to construe criminal statutes narrowly, a corresponding canon tells courts to construe remedial statutes broadly to effectuate the legislators’ remedial goals. Courts do not do so consistently. For that matter, there is no particular consensus of what makes a statute “remedial” for purposes of the rule. Nonetheless, courts recognize this principle and adduce it frequently when they believe that the legislature did indeed have in mind the outer boundaries of a statutory word’s definition.90 In fact, consistent with the Supreme Court’s broad reading in Chisom, a U.S. court of appeals has declared the Voting Rights Act such a statute and has interpreted it to include private recall petitions on the theory that “as a remedial statute, the Voting Rights Act is to be broadly construed so as to achieve the Act’s objectives.”91 Or consider United States v. E. I. Dupont De Nemours & Co., decided by the U.S. Court of Appeals for the Third Circuit in 2005.92 The issue there was whether “the Comprehensive Environmental Response, Compensation and Liability Act of 1980 (CERCLA), authorizes the United States to recover costs incurred in the course of supervising a hazardous waste cleanup conducted by . . . private parties” that CERCLA holds responsible for the cleanup.93 The statute holds responsible parties liable for “all costs of removal or remedial action incurred by the United States government or a State or an Indian tribe not inconsistent with the national contingency plan” and “any other necessary costs of response incurred by any other person consistent with the national contingency plan.”94 Although the court had earlier held that a clear statement by Congress is needed before the costs of government supervision of private cleanup may be recovered, it reconsidered its position after other courts disagreed and in light of the fact that “CERCLA is a broad remedial statute.”95 While

definitions, ordinary meaning, and respect for the legislature

69

cost recovery statutes may ordinarily exclude the costs of government supervision unless the statute specifically includes them, the court employed a broader reading of “costs” that includes cost recovery beyond the prototype. Thus, the ordinary-meaning approach to statutory interpretation may act as a reasonable initial hypothesis for determining the intent of the legislature, but it is no more than a rule of thumb whose application is inappropriate in a wide range of situations. Moreover, as many commentators have noted, legislators are not the only ones who cannot help thinking in definitional terms. Judges cannot help themselves either. Since Justice Scalia’s appointment in 1986, references to dictionary definitions have not declined in favor of ordinarymeaning analysis.96 To the contrary, they have increased, with Scalia himself being the largest contributor.97 Recall Justice O’Connor’s majority opinion in Smith, the “use a firearm” case. First she professed a commitment to seeking the ordinary meaning of the word “use,” then she went right to the dictionary to confirm its meaning in the broadest, definitional sense. In fact, she bolstered her argument by referring to a number of dictionaries. O’Connor is by no means alone in this respect. For example, MCI Telecommunications Corp. v. AT&T required the Court to construe the word “modify” in the Federal Telecommunications Act, which authorizes the Federal Communications Commission (FCC) to “modify” the requirement that carriers file schedules of charges.98 If construed broadly, it would permit the FCC to grant to small carriers total exemptions from onerous tariff publication requirements. If construed narrowly, such exemptions would be beyond the authority vested in the agency by the statute. If ambiguous, the agency’s interpretation would prevail under familiar principles of administrative law.99 Various dictionaries, including Webster’s Third, the dictionary that Thomas relied upon in dissent in Small, provide definitions that support a broad view, which might include the elimination of certain obligations entirely, and other dictionaries, including The Random House Dictio­nary of the English Language, define “modify” in terms of small changes, which appears to be more in keeping with a narrower interpretation.100 The case became a battle over which dictionary should be considered the more authoritative. Scalia’s majority opinion rejected what it called “the peculiar Webster’s Third definition”101 and, for that matter, the interpretation of the FCC, and ruled that, as a matter of law, modification is limited to “modest” change. The opinion does not consider that a change may be

70

chapter 3

modest in that it applies to only a small percentage of the overall tariffs. We will return to such battles below. For now, though, my point is that judges, including textualist judges, are susceptible to construing words in terms of classical definitions: conditions that must obtain for the word to be used appropriately. Cases like these make it difficult to rely too heavily on pronouncements that textualism’s methodology uses an enlightened vision of language. They also suggest a more complex interaction between definitional and prototypical features.102 In Small, for example, the majority relied upon the central tendency of courts to interpret laws to apply within the territory—an argument from ordinary meaning. Once it made that decision, however, it used a definitional feature (territoriality) to determine which meanings are ordinary. One can look at many of the cases discussed in this chapter as trading between plain and ordinary meaning in just this way. In Smith, Scalia’s dissent is motivated by the fact that ordinarily we use a firearm as a weapon — not merely as a thing of value — an argument from ordinary meaning. Once we accept ordinary usage as legally relevant, it becomes possible to articulate a narrower rule: for purposes of the statute, we assume that one has used a firearm only when it has been used for its ordinary purpose. This rule is definitional in nature, although it is a narrow definition dependent upon prototypical meaning. Given that we routinely apprehend concepts using both definitional and prototypical features, this dynamic feels natural enough to us. Ordinary Meaning Is Hard to Find A second, evidentiary question arises when judges rely upon the ordinary meaning of statutory terms. When a court decides to base its decision on ordinary meaning, how does it decide what the ordinary meaning is? The answer, somewhat to the embarrassment of the American legal system, is that courts find ordinary meaning anywhere they look, and judges are not restrained in deciding where they are willing to look. To see this dynamic in play, let us return to the statute at issue in Smith v. United States, which made it a crime to use or carry a firearm during and in relation to a drug-trafficking crime.103 Smith is actually the first of four cases in which the Supreme Court construed the statute. The second, Bailey v. United States, was decided in 1995, two years after Smith.104 Bailey also involved a person convicted of using a firearm during a drug-

definitions, ordinary meaning, and respect for the legislature

71

trafficking crime. In that case the gun was in the trunk of the car, and the drugs were in the front of the car with Bailey. The government argued that this constituted using a gun because Bailey was taking advantage of the weapon to embolden himself during any drug transaction. A similar case, heard with Bailey, involved a drug dealer keeping a weapon in a trunk in her bedroom while she engaged in drug transactions in the front of her apartment. The Supreme Court decided nine to zero that Bailey was not “using” a firearm. The Court first reaffirmed its earlier decision in Smith and made mention of the various dictionaries. But most of the opinion, written by O’Connor, who also wrote the majority opinion in Smith, concerned the need to interpret statutes in context. She explained that, in the context of the statute, “use” implies some kind of active use that goes beyond mere possession.105 In fact, the Court attributed this sense to the dictionaries as well. Interestingly, the strongest argument was a linguistic one. The Court gave the following example: “I use a gun to protect my house, but I’ve never had to use it.” The Court surmised that in enacting the statute, the legislature contemplated the second occurrence of “use” in that sentence: active use of some kind. If the argument sounds more linguistically sophisticated than we should expect from a judge not trained in linguistics—it is. As the Court was deciding Bailey, Clark Cunningham, a law professor, and Charles Fillmore, a linguist, published an article using the precise example contained in the Court’s opinion.106 They made the linguistic argument that the court relied upon, although the opinion did not mention the article. So one answer to the question of where courts find ordinary meaning is that they use their judgments as native speakers, enhanced occasionally by serious linguistic analysis performed by linguists when such analysis can be helpful.107 Bailey’s reliance on the analysis of linguists, however, is unusual. More typical is the Supreme Court’s analysis in the third “use a firearm” case, Watson v. United States, decided twelve years after Bailey.108 Watson was the flip side of Smith. The defendant this time tried to trade the controlled drug OxyContin for a weapon. The issue was whether he used the firearm in the transaction. Once again, a unanimous court said no, and once again it relied on the ordinary use of the statutory language: “With no statutory definition or definitive clue, the meaning of the verb ‘uses’ has to turn on the language as we normally speak it. . . . The Government may say that a person ‘uses’ a firearm simply by receiving it in a barter transaction, but no one else would. A boy who trades an apple to get a granola bar is sensibly

72

chapter 3

said to use the apple, but one would never guess which way this commerce actually flowed from hearing that the boy used the granola.”109 Much more difficult was the determination of ordinary meaning in the Supreme Court’s “carry a firearm” case, Muscarello v. United States, decided in 1998.110 The facts are much like those in Bailey: drugs in the front of the car, a weapon in the back. But this time, probably because of the Bailey decision three years earlier, the government charged Muscarello, not with using a firearm, but with carrying one, which the statute also covers.111 In defense, Muscarello argued that the ordinary meaning of “carry a firearm” is to carry the gun on one’s person. The government argued that carrying a gun in a vehicle is the more ordinary sense of the word. The Court had to decide which sense of “carry” should be considered the ordinary one for purposes of interpreting the statute. Ultimately, the Court ruled in favor of the government by a five-to-four vote. In the majority opinion, written by Breyer, the Court first established that the issue before it was the ordinary meaning of the disputed expression: We begin with the statute’s language. The parties vigorously contest the ordinary English meaning of the phrase “carries a firearm.” Because they essentially agree that Congress intended the phrase to convey its ordinary, and not some special legal, meaning, and because they argue the linguistic point at length, we too have looked into the matter in more than usual depth. Although the word “carry” has many different meanings, only two are relevant here. When one uses the word in the first, or primary, meaning, one can, as a matter of ordinary English, “carry firearms” in a wagon, car, truck, or other vehicle that one accompanies. When one uses the word in a different, rather special, way, to mean, for example, “bearing” or (in slang) “packing” (as in “packing a gun”), the matter is less clear. But, for reasons we shall set out below, we believe Congress intended to use the word in its primary sense and not in this latter, special way.112

Note that the Court framed the issue by answering its own question without the benefit of supporting evidence. What the Court called “the first, or primary, meaning . . . as a matter of ordinary English” is precisely the meaning that the Court would have to find to be ordinary in order to reach the result it did. Next, the Court set out to answer its question more fully. Not surprisingly, it first turned to the dictionaries and quoted several definitions that support its position. The Court acknowledged that some dictionaries, including Black’s Law Dictionary, appear to support the defendant’s posi-

definitions, ordinary meaning, and respect for the legislature

73

tion.113 Black’s defines “carry arms or weapons” as follows: “[t]o wear, bear, or carry them upon the person or in the clothing or in a pocket, for the purpose of use, or for the purpose of being armed and ready for offensive or defensive action in case of a conflict with another person.” 114 But the Court rejected these as providing what it called “special definitions.”115 It opted instead for the “primary definitions” contained in the dictionaries cited by the majority. To bolster its choice of dictionary, the Court turned to the etymology of the word “carry” and found that it shares its Latin origin with the word “car”— good news for the government, assuming (without presenting a reason) that word origins should make a legal difference.116 For anyone questioning the relevance of this history, the Court brought out the Bible, with two quotations: “His servants carried him in a chariot to Jerusalem”; and “They will carry their riches upon the shoulders of young asses.”117 Apparently, we should infer that if this use of the word “carry” was good enough for the Bible, it is good enough for the Supreme Court. It was irrelevant that the Book of Kings and the Book of Isaiah were originally written in Hebrew and that the ordinary way of carrying a person is in a vehicle, while the ordinary way of carrying a small object may be otherwise. Not to offend the non-Jews and non-Christians among us, the Court also found some examples from Daniel Defoe’s Robinson Crusoe and from Melville’s Moby Dick. Perhaps somewhat more to the point, the Court observed that it previously used the word “carry” in its opinions to refer to the transportation of drugs in a vehicle.118 The Court also did its own lexicography, using a computerized database of newspaper and magazine articles that contained sentences using “carry,” “vehicle,” and “weapon.” It found “that many, perhaps more than one-third, are sentences used to convey the meaning at issue here, i.e., the carrying of guns in a car.”119 In dissent, Justice Ginsburg, writing for herself and three others, cited her own dictionaries, her own passages from the Bible, and her own literary allusions.120 Each supported the defendant’s position that the ordinary meaning of “carry a firearm” means to carry the weapon on one’s person. She further pointed out that many English translations of the Bible do not use the word “carry” in the passages quoted by the majority.121 In response to the majority’s survey of newspaper and magazine articles, she wrote: “The Court’s computer search of newspapers is revealing in this light. Carrying guns in a car showed up as the meaning ‘perhaps more than one-third’ of the time. One is left to wonder what meaning showed up

74

chapter 3

some two-thirds of the time.”122 The dissent also considered the legislative history and the use of “carry” in other portions of the statute. Ultimately, it concluded that the language of the statute is hopelessly ambiguous, and neither meaning is clearly more obvious than the other. Thus, the rule of lenity should be applied, as it had been in Bailey. I dwell on this case largely because it illustrates how bankrupt courts are when they must actually decide just what makes ordinary meaning ordinary. The argument resembles a food fight in a school for children with disciplinary problems more than a serious argument among distinguished jurists. One problem may be that the majority is not entirely sincere when claiming that it actually sought the ordinary meaning. In response to the dissent’s marching out dictionaries like Black’s Law Dictionary, the majority said: “These special definitions, however, do not purport to limit the ‘carrying of arms’ to the circumstances they describe. No one doubts that one who bears arms on his person ‘carries a weapon.’ But to say that is not to deny that one may also ‘carry a weapon’ tied to the saddle of a horse or placed in a bag in a car.”123 But this statement has nothing to do with discovering the ordinary meaning. Rather, it says only that the majority’s interpretation falls within the range of the statute’s possible meanings. The statement is more consistent with the definitional-meaning approach, which leads to broader, less nuanced interpretations. At least in part, the dispute over ordinary meaning may be only an apparent battle. The real fight is over the rule of law: is it acontextual and available from reading the statute and applying definitions of its words to their outer limits, or do we need to introduce context, whether from the ordinary-meaning rule or from extratextual material? Where Should Courts Look? Where courts should look for word meaning depends on what they are looking for. If one really believes the best way to execute the legislative will is by assuming that the legislature used statutory words in their most ordinary sense, simple introspection is generally an adequate way to discover that sense. After all, linguists rely heavily on their own knowledge of their native languages. They have enjoyed great success exploring their own judgments about grammaticalness, felicity, and preferences of one structure or meaning over another. With the help of research assistants, I examined, decade by decade, cases in which the U.S. Supreme Court relied upon the ordinary-meaning rule in statutory interpretation through

definitions, ordinary meaning, and respect for the legislature

75

the turn of this century. Although our study no doubt missed some, the cases we did examine form a reasonable corpus for evaluation. We looked at 122 cases, the first one occurring in 1817.124 Forty-seven (39 percent) were decided after 1980, demonstrating the important role that textual analysis has played in recent times. During most of American judicial history, the predominant methodology for discovering ordinary meaning has been introspection. Without fanfare, judges simply rely upon their own sense of how common words are typically used. Although controversies sometimes emerge, as in Muscarello, for the most part what judges say rings true. This did not change significantly until the 1980s, when Scalia was appointed to the Supreme Court. Although the 1890s, the decade the Court decided Church of the Holy Trinity, was also a time when the Court experimented with different methodologies, in the past twenty-five years dictionaries, precedent, and the use of similar language in the same and other statutes have gained prominence, with introspection declining in popularity. This suggests that Scalia’s textualism, so influential in American jurisprudence, is a departure from the legal tradition, a point to which I return later. First, let us look briefly at the various techniques that courts employ. Sometimes courts look to their own earlier decisions to find the ordinary meaning of statutory terms. The notion is that if the courts have already interpreted the word in the statute, or in other related statutes, then the earlier decision is determinative of later cases. For example, in Bryan v. United States (discussed in chapter 2) the Supreme Court, as part of an effort to construe the word “willfully” in a statute requiring a license to sell certain firearms, reviewed earlier cases in which the Court had construed the word.125 And when the legislature enacts a law adopting language already construed in a prior court decision, it is presumed that the legislature intended the words in the statute to be understood as the court had earlier construed them.126 Similarly, when the same word is used multiple times in a statute, and the meaning is clear in some instances, courts draw an inference that the legislature intended the word to have that interpretation in the other instances. This assumption may not always be well founded, but it is sensible enough as a rule of thumb.127 Moreover, legislators, knowing that a court will impose this interpretation, are in a position to draft statutes that take these issues into account. Somewhat more problematic are inferences drawn from how statutes remote from the one in dispute use the same word. The Supreme Court has done this occasionally throughout history, but it has become a popular

76

chapter 3

method of interpretation only in the past two decades. William Buzbee has criticized this method.128 If meaning depends in part on context, then why would a mental-health statute provide the appropriate context for interpreting a word in, say, a statute dealing with the regulation of the mining industry? One function of the ordinary-meaning approach is to use prototypical experiences as a proxy for intent. Searching for ordinary meaning in contexts remote from the one in dispute seems doomed to cause misinterpretations, as Buzbee convincingly demonstrates. Without question, though, the biggest change in the search for word meaning is the almost obsessive attention courts now pay to dictionaries, using them as authority for ordinary meaning. Until the late twentieth century, Supreme Court justices only infrequently used the dictionary as a source of ordinary meaning. For example, in the almost two hundred years prior to Scalia’s appointment to the Supreme Court in 1986, the Court referred to “ordinary meaning” in close proximity to “dictionary” only six times. From the time of his appointment through 2008, the Court did so an additional twenty-one times.129 In recent years, Thomas appears to be taking the lead, frequently dissenting in civil cases in which he believes that the majority has interpreted statutory words wrongly. He instead bases his analysis on ordinary meaning as found in the dictionary.130 The problem with using dictionaries to determine the ordinary meaning of a word is that the principal purpose of a dictionary is to determine the outer boundaries of appropriate usage for each sense of a word. But most of the time, the issue before a court is not whether the legislature intended one distinct sense of a word as opposed to another (e.g., “riverbank” vs. “savings bank”). Rather, the question is whether the facts of the case are sufficiently close to the circumstances in which it feels comfortable to use the statutory word that it is fair to conclude that the legislature most likely would have expected the statute to apply. Other than providing an articulate expression of the general meaning to assist in that inquiry, dictionary definitions most often do little to aid in that inquiry. Once judges begin to fight over which dictionary to consult, the use of dictionaries to determine ordinary meaning is virtually futile. As we have seen, lexicographers sometimes insert information about typicality into their definitions, but this practice surely is not systematic. Amid all this squabbling, however, are some instances of creative analysis. Among them is the use of linguistic analysis, which the Court adopted in Bailey, having borrowed the example “I use a gun to protect my house, but I’ve never had to use it” from an article by Cunningham

definitions, ordinary meaning, and respect for the legislature

77

and Fillmore.131 As Cunningham and Fillmore point out, it is no accident that the words “to protect my house” appear with the nonprototypical occurrence of “use.”132 Generally, unmodified forms reflect the prototypical, or unmarked, case in language. Compare their example to “I use a gun, but I’ve never had to use it to protect my house.” This implies that the speaker has experience shooting a gun but has not done so to ward off intruders. Now let us return to Muscarello and the ordinary meaning of “carry.”133 Is carrying a gun in a glove compartment or trunk of a car within the “ordinary” meaning of “carry”? The question answers itself. The very fact that one needs to add “in a glove compartment” or “in a trunk” to avoid the natural implication that the gun is being carried on the individual’s person suggests that these are not the prototypical uses of the word. To see why, let us return to the biblical examples that the majority opinion used to support its position: “His servants carried him in a chariot to Jerusalem,” and “They will carry their riches upon the shoulders of young asses.”134 Putting aside that both of these are poor translations of the Hebrew version,135 in each example a prepositional phrase describing how the carrying is to occur modifies the word “carry.” In the first, it is “in a chariot.” In the second, it is “upon the shoulders of young asses.” Compare these to the following: “His servants carried him to Jerusalem,” and “They carried their riches.” The default assumption is that they physically carried the corpse or the riches. Further support of this analysis comes from sentences like the following: “The chariot carried him to Jerusalem.” Here, the verb is also unmodified and the carrying is direct, suggesting further that the most ordinary use of “carry” implies direct carrying. There are, of course, other ways to carry, but to express them additional information is typically required. Otherwise, the inference will be that the most natural way to carry was the meaning that the speaker intended to convey. Recall that the statute speaks of “us[ing] or carr[ying] a firearm” “during and in relation to . . . [a] drug trafficking crime” but says nothing about the manner of carrying the firearm.136 The absence of any mention of a means of transportation suggests one of two possibilities: either Congress intended that the ordinary meaning apply, in which case Muscarello should not have been convicted, or Congress intended the statute to apply to more than ordinary instances of carrying a firearm. This method of analysis, which we might call linguistically motivated introspection, is not new to judges. For example, the dissent in Muscarello makes the following argument: the word “transport” better conveys what the majority says is the ordinary meaning of “carry.”137 In fact, Congress

78

chapter 3

used the word “transport” in other sections of the same statute.138 Therefore, to give effect to the legislature’s choice of words, “carry” should receive the narrower interpretation. Had the legislature wanted the courts to apply the statute broadly, the argument goes, it would have used “transport” instead of “carry” as it had done elsewhere in the law. Interpretive Communities: Technical versus Ordinary Language Finally, a technologically more sophisticated method of determining ordinary meaning suggests itself in Breyer’s reliance on databases of periodicals to discover the ordinary meaning of “carry.” While the dissent rightly criticizes the conclusion that ordinary meaning can be inferred from approximately one-third of examples,139 the use of this method is both creative and promising. Over the past quarter century, a great deal of work has been published in the growing field of “corpus linguistics,” both for basic research140 and in legal settings.141 Like Breyer, researchers in this area amass data from actual language usage, although the principal goal of the basic research is to allow inferences about the ways in which various linguistic phenomena are organized in our minds. This is not the place to explore the use of corpus linguistics thoroughly. Yet even brief consideration raises questions that remain hidden in current statutory analysis. For example, there are now extensive linguistic corpora of English in both the United Kingdom and the United States that include both written and spoken language. These corpora can be further sorted by the source of the language, whether literary texts, spoken language from law enforcement officers, newspaper and magazine articles, and so on. When the legal system decides to rely on the ordinary meaning of a word, it must also determine which interpretive community’s understanding it wishes to adopt.142 This choice is made tacitly in legal analysis but becomes overt when the analysis involves linguistic corpora, because the software displays the provenance of the usage on a screen in front of the researcher. Nix v. Hedden, in which the Supreme Court had to decide whether a tomato is a fruit or a vegetable, illustrates this point.143 Lexicographers and botanists call it a fruit. American people generally call it a vegetable, even those who know that it is botanically a fruit. At the time, higher tariffs were imposed on imports of vegetables than on imports of fruit. The government maintained that imported tomatoes were subject to the higher

definitions, ordinary meaning, and respect for the legislature

79

duty. The importer said otherwise; that they should be considered fruit. Looking at the dictionary as potential evidence of ordinary meaning for the first time, the Court nonetheless chose the ordinary meaning above the technical meaning.144 Somewhat ironically, this means that the ordinary people lost. Tomatoes would cost more. Commenting on another case in which the dispute was whether beans should count as seeds or vegetables, the Court observed: “Both are seeds in the language of botany or natural history, but not in commerce nor in common parlance.”145 This suggests that the distinction between ordinary meaning and technical meaning is dependent upon the audience and its relation to the speaker. Access to computers now makes it relatively simple to see how words are used in commerce and in common parlance. This allows judges to become their own lexicographers. If they perform that task seriously, they stand to learn more about how words are ordinarily used than by today’s method of fighting over which dictionary is the most authoritative. But if they do their own research, they still must decide whether to apply the plain or ordinary meaning and, if the ordinary meaning, whose ordinary meaning.

Conclusion The new textualist methodology relies heavily on a vision of language that itself contains an enriched vision of context. In particular, it has replaced the definitional approach to word meaning, with the probabilistic approach gleaned from the word’s ordinary usage. The result is that it is possible to rely on language judgments alone for a great deal of contextsensitive analysis of statutes. I believe that this approach has a great deal to say for itself. For one thing, the principles upon which it relies are both sophisticated and traditional. Through most of American history, courts have looked for the ordinary meaning of the words in a statute to determine what the legislature intended the scope of the statute to be.146 For another, most of textualist methodology is highly consistent with rule-of-law values. Yet matters are not so simple. For one thing, it is not entirely clear that legislatures always have the prototypical meanings of statutory language in mind. In some instances, they may use concepts more broadly. The ordinary-meaning rule serves only as a proxy for the intent of those

80

chapter 3

who wrote the statute. There is no reason to overestimate the success of using this proxy, especially when evidence to the contrary presents itself in particular cases. Second, judges, including textualist judges, are unable to stick with their program. Because we typically use both definitional and prototypical information in understanding the meanings of words, it is simply impossible to eschew a dictionary-like analysis, even if one’s political vision says that one should do so. Finally, it is not a simple matter to determine what constitutes the ordinary meaning of a word in a close case. When such debates occur, the courts are in complete disarray, with unbecoming fights over the status of one dictionary over another, one literary allusion over another, or, still worse, one biblical reference over another. Ultimately, the choice is between insisting upon a standard set of methodologies, sensible enough most of the time but sure to result in errors, even on its own terms, and living with a more relaxed set of evidentiary standards, less able to constrain judicial discretion but better able to head off results that are likely at odds with what an enacting legislature intended its law to accomplish. How one weighs in on this debate is largely a matter of political preference, but at least to some extent, it is an empirical matter. For that reason, I agree strongly with those scholars who have called for more empirical research into the real likelihood of mischief when judges resort to legislative history or perhaps other species of evidence that textualists reject.147 However one decides such matters for oneself, it is a healthy start to acknowledge that we will not develop methods or standards that will permit us to dispense with judicial discretion and still have a system that does justice. Language, whether ordinary or plain, works well—but not that well. It is also important to recognize that while the problems that trigger difficult questions of statutory interpretation are often psychological and linguistic, decisions about how statutory interpretation should proceed are legal and political decisions, as a number of writers have pointed out in the recent literature.148 For example, the fact that a minister seems to be performing labor in some sense of the word but not in the ordinary sense is a psychological fact. But application of the rule of lenity in a case like Church of the Holy Trinity is a legal decision. What makes the cases discussed in this chapter especially interesting is that the legal system has decided to adopt linguistic principles as part of its repertoire of legal rules. It thus becomes relevant for legal theorists, even those who have little independent interest in such things as linguistics and the psychology of

definitions, ordinary meaning, and respect for the legislature

81

language, to ask about the nature of legal constructs such as definitional meaning and ordinary meaning, which the courts use routinely. In the next chapter, we continue this exploration and find that these concepts are intricately tied in with the notion of intent. Since reference to legislative intent has been controversial in recent years, it is important to see how central it is to those concepts that appear to be used routinely.

chapter four

The Intent of the Legislature

T

he previous chapter explored how disagreement can arise even when a statutory interpreter relies upon nothing but the language of the statute. Yet all those engaged in these debates claim to be acting in the service of fidelity to the legislature. Statutory interpreters care so much about the language of the legislature because they regard it as their duty to defer to the body that made the law. This chapter delves more deeply into what it means to be faithful to the legislature.1 The greatest controversy over statutory interpretation during the past two decades concerns the use of legislative history as evidence of the in­ tent of the legislature. Opponents say that relying upon the historical rec­ ord of the lawmaking process is undemocratic (committee reports are not enacted),2 unreliable (history is often conflicting and may even be planted to influence judges in the future),3 and incoherent (the record does not represent the views of all members of the legislature, so it cannot be evi­ dence of legislative intent).4 More basically, the concept of legislative intent has itself been subject to attack.5 Whose intent? What difference does it make what people in­ tended as long as we know what they said? Oliver Wendell Holmes’s 1899 statement that “we do not inquire what the legislature meant; we ask only what the statute means,” continues to be widely quoted today.6 In this chapter I defend the use of legislative intent in statutory inter­ pretation and explain its ubiquity. It is not incoherent: we routinely at­

the intent of the legislature

83

tribute intent to a group of people based on the intent of a subset of that group, provided that there is agreement in advance about what role the subgroup will play. The legislature is a prototypical example of the kind of group to which this process applies most naturally. Moreover, resort to legislative intent need be no less democratic than the legislative process itself. To the extent that a small group of legislators determines the details of a law with the acquiescence of the legislative body, it does no harm to democratic principles for a court to recognize the process for what it is. On the other hand, judicial reliance on portions of the legislative record that do not reflect delegation of planning to a subgroup, such as inciden­ tal remarks of individual legislators, cannot be justified on these grounds. Thus, evidentiary arguments that courts sometimes misuse legislative his­ tory may have merit and should be dealt with on a case-by-case basis. In contrast, arguments against any reference to legislative intent do not have merit and should be rejected.

Legislation or Legislative Intent? The question of legislative intent has gained recent prominence in large part because it has been a major item on the agenda of Justice Antonin Scalia since his 1986 appointment to the Supreme Court. However, al­ though criticism of legislative intent is generally associated with politically conservative judges and academics, critiques appear throughout the politi­ cal spectrum. For example, William Eskridge, who is generally regarded as a progressive theorist on questions of statutory interpretation, expresses concern that resort to legislative intent can render the interpretation of statutes inappropriately unresponsive to changes in the interpretive envi­ ronment over time.7 And the debate is by no means new. About seventy-five years ago, in a biting article published in the Harvard Law Review, legal philosopher Max Radin wrote: A legislature certainly has no intention whatever in connection with words which some two or three men drafted, which a considerable number rejected, and in regard to which many of the approving majority might have had, and often demonstrably did have, different ideas and beliefs. That the intention of the legislature is undiscoverable in any real sense is al­ most an immediate inference from a statement of the proposition. The chances

chapter 4

84

that of several hundred men each will have exactly the same determinate situa­ tions in mind as possible reductions of a given determinable, are infinitesimally small. The chance is still smaller that a given determinate, the litigated issue, will not only be within the minds of all these men but will be certain to be se­ lected by all of them as the present limit to which the determinable should be narrowed.8

Radin argued from the perspective of the legal realist movement, which espoused that judicial rhetoric often masks the real reasons behind judges’ decisions, which are frequently political. Today, the same critiques are made by judges and scholars throughout the political spectrum, but more often from the right. Yet despite the large body of law and commentary addressing these questions, Radin’s observations about the problems with attributing a single intent to an entire legislature have still not been fully answered by proponents of legislative intent. Defenders of legislative intent often argue that when the language of a statute is not clear, a court charged with interpreting the statute should resort to extrinsic evidence to see how best to effect the statute’s goals. 9 That intent can often be inferred, at least in part, from the circumstances surrounding the statute’s enactment. Among those circumstances are the legislative proceedings in which the statute was written, during which its substance and form were debated. In fact, not taking this sort of informa­ tion into account increases the likelihood of a court’s accepting an inter­ pretation that is at odds with the intentions of the enacting legislature. Justice (then Judge) Stephen Breyer makes these and other arguments in his important article defending the use of legislative history by judges.10 Breyer catalogues situations in which the language of a law simply does not provide an adequate basis for interpretation. Sometimes, the legisla­ ture has made an obvious mistake. In other cases, the statute is ambiguous. In still others, the language of the statute seems applicable to a particular situation, but it would be absurd to apply the statute as written. In all these situations, there is more or less consensus that a court must look beyond the words of a statute to determine how to apply the law.11 Concern for legislative primacy argues that inferring and then enforcing the intent of the enacting legislature is appropriate when statutory language alone does not suffice. The counterargument goes something like this: much of the time the legislature probably had no intent with respect to the law’s application in the particular situation.12 That is what makes the situation a hard case. In

the intent of the legislature

85

any event, the legislature is not a person, like the governor or the presi­ dent. The legislature consists of a group of people with diverse goals and intentions. It cannot itself have “an intention.”13 In all likelihood, most legislators who voted for the bill did so because the party leadership told them to do so as a matter of party discipline: “Vote for bill number 802, and then we’ll see if we can move some of your pet projects forward before the legislative session ends.” Or perhaps: “The president wants this one badly. If you don’t support it, I can’t promise that the party leadership will help you to win the next primary election. We may prefer a candidate who will cooperate better than you have.”14 When a judge looks at legislative his­ tory to determine the intentions of the enacting legislature, the argument continues, what she is really looking at is a set of documents prepared by staff people, speeches made by individual members, or other informa­ tion that cannot possibly be considered representative of the thinking of “the legislature” as a whole. As Judge Patricia Wald famously quoted her late colleague Judge Harold Leventhal, citing legislative history is “akin to looking over a crowd and picking out your friends.”15 At worst, legislative history will be corruptly unreliable, perhaps planted by those who wish to influence courts without complying with the onerous process of legislative enactment.16 Often, it will be self-contradictory, allowing first lawyers and then judges to pick and choose among statements that support a party’s position.17 In any event, poring through thousands of pages of material will run up the costs of litigation with the risk of doing more harm than good.18 Justice Scalia summarizes the danger of using legislative history in a number of his writings. In A Matter of Interpretation, he writes: “My view that the objective indication of the words, rather than the intent of the leg­ islature, is what constitutes the law leads me, of course, to the conclusion that legislative history should not be used as an authoritative indication of a statute’s meaning.”19 In other words, the concept of legislative intent is irrelevant to legal interpretation, so the use of legislative history to deter­ mine legislative intent must be irrelevant too. He continues: As I have said, I object to the use of legislative history on principle, since I reject intent of the legislature as the proper criterion of the law. What is most exasper­ ating about the use of legislative history, however, is that it does not even make sense for those who accept legislative intent as the criterion. It is much more likely to produce a false or contrived legislative intent than a genuine one. The first and most obvious reason for this is that, with respect to 99.99 percent of the

chapter 4

86

issues of construction reaching the courts, there is no legislative intent, so that any clues provided by the legislative history are bound to be false.20

He finishes his argument by noting that committee reports and the like probably do not influence the votes of many in the legislature. Scalia’s philosophy comes out strongly in his opinions as well, often in concurrences in which he agrees with the Court’s result but objects to its looking at congressional proceedings in reaching that result. Consider Conroy v. Aniskoff: state laws that permit municipalities to sell a house to collect the tax owed allow for a period of redemption during which the owner can rescue the situation by paying the back taxes.21 A federal stat­ ute, the Servicemembers Civil Relief Act, was enacted in the context of World War II’s huge military conscription and suspends the running of this redemption period while the owner is in the military.22 Sensible enough. But what happens when the homeowner is a career military person? Should he get twenty-five years of protection from paying his local taxes? The statute, read literally, suggests that he should. It draws no distinction between the draftee fighting in the north of France and the deadbeat who happens to be in the army. In deciding to interpret the statute broadly to protect all members of the military notwithstanding the context of its enactment, the Supreme Court looked at the legislative history of the relevant statute and related statutes. It found no serious contradictions between the historical record and the plain language, and some historical evidence in further support of the literal reading. For example, the relevant provision was reenacted after World War II ended, during a time of peace. The Court thus inter­ preted the statute according to its plain language, even though this meant enforcing the statute in circumstances broader than the policy behind its enactment would have demanded. In his concurrence, Justice Scalia accused the majority of selecting snippets from the history and misunderstanding the role of the judiciary. He remarked: The greatest defect of legislative history is its illegitimacy. We are governed by laws, not by the intentions of legislators. As the Court said in 1844: “The law as it passed is the will of the majority of both houses, and the only mode in which that will is spoken is in the act itself.” . . . [emphasis added]. But not the least of the defects of legislative history is its indeterminacy. If one were to search for

the intent of the legislature

87

an interpretive technique that, on the whole, was more likely to confuse than to clarify, one could hardly find a more promising candidate than legislative history.23

Scalia pointed to examples from the historical record that, if accumulated, could create the impression that Congress intended the opposite of what the plain-language reading suggests. The defenders of legislative history respond, a little more weakly, to this and related arguments: Of course, we should be honest in presenting the historical record, including the possibility that the record is internally inconsistent and does not teach us much in some instances. But laws are written in language and language can be understood only in context. The thinking of those who supported and proposed the law in the first place may not reflect the will of every legislator, but it certainly can make some contribution to statutory interpretation if used wisely. At the very least, it can help us to determine whether the difficulty in applying the statute re­ sults from an unfortunate choice of statutory language chosen to effectu­ ate a legislative goal that becomes clear once one investigates the matter. And it can be used to confirm that decisions made on other grounds are not likely to fly in the face of what the statute was intended to accomplish. In response to one of Justice Scalia’s more aggressive concurring opinions, Justice Byron White replied in a footnote: As for the propriety of using legislative history at all, common sense suggests that inquiry benefits from reviewing additional information rather than ignor­ ing it. As Chief Justice Marshall put it, “where the mind labours to discover the design of the legislature, it seizes every thing from which aid can be de­ rived.” Legislative history materials are not generally so misleading that jurists should never employ them in a good-faith effort to discern legislative intent. Our precedents demonstrate that the Court’s practice of utilizing legislative his­ tory reaches well into its past. We suspect that the practice will likewise reach well into the future.24

The passage that White quotes dates back to an 1805 opinion written by Chief Justice John Marshall.25 When used cautiously, the argument goes, legislative history can lead to a more thoughtful analysis of a complex legal problem. Why should courts not embrace every opportunity to make a more reasoned decision? Assume, as I will argue later, that understanding

88

chapter 4

language is very much a matter of striving to understand the intent of the speaker, just as speaking is an effort to facilitate our hearer’s efforts to understand our message. If this is true, then statutory language can be no different, and our laws can be no more than efforts at communication based on the intention of the drafters. Moreover, as Justice Breyer points out, we routinely attribute beliefs and intentions to groups of people as if they were a single indi­ vidual:26 “What’s IBM planning for next year?” “Should the philosophy department hire a Descartes scholar?” Doing so when the group is a legis­ lature is just a special case of an ordinary psychological process. As opponents of legislative intent have long recognized, the larger stakes in the debate over legislative history and legislative intent are about our conception of language and how it relates to our ability to govern our­ selves by enacting laws. Judge Frank Easterbrook expressed the problem in an opinion interpreting the federal Bankruptcy Code: “Statutes are law, not evidence of law.”27 As Easterbrook understood, once we start talking about intent, then language, including statutory language, can be noth­ ing other than evidence of that intent. It might be good evidence, and we might want to privilege it in some way as being more reliable than other evidence (such as legislative history), but it is still just evidence. This is a serious problem for a jurisprudence based upon the assumption that government defines our obligations through a democratically legitimate process, requiring such procedures as bicameralism and presentment as set forth in Article I of the U.S. Constitution. It is a lot messier if we have to start thinking about the intent of the legislature because the enacted words are not good enough.28 Whether the quest for intent is for the good is a separate issue. The legal system often has to deal with situations in which our psychological propensities are at odds with the system’s goals. For example, the rules of evidence recognize that jurors are likely to make too much of a criminal defendant’s prior criminal record. Therefore, except for limited purposes, such as to impeach a defendant’s credibility on the witness stand, the rules prohibit the government from introducing evidence of prior convictions.29 It may well be that judges are prone to make more of legislative intent and legislative history than their probative value justifies, and that they should therefore be barred from doing so. I take up that possibility in chapter 6. For now, let us focus on just how natural it is to talk about groups as if they were individuals.

the intent of the legislature

89

Understanding Groups as Individuals To see how we regard groups as units in everyday life, consider the follow­ ing story. Mario and Adela Rossi are both engineers. They met in graduate school in Italy, married, and immigrated to the United States seven years ago when they were both offered jobs with large companies. You and I are friends of the Rossis, and we both admire them for the exciting vacations that they take. We also both know that Adela is the one who makes all the plans. Mario just tags along, knowing that Adela is certain to have put together something great, whether a trip to an exotic part of the world or a visit to a country inn within an hour of their home. You say to me: “Where are Mario and Adela planning to go this year?” I respond: “I just talked with Adela yesterday. They’re planning a trip to a wonderful small resort in Virginia. She told me all about it.” In fact, as you and I both know, Mario does not have a clue about this year’s vacation, except for the barest outline. It would be very odd to say that he is planning anything —Adela’s the one doing the planning. Yet our conversation about the Rossis’ plans seems perfectly natural. We can think of Adela making plans, we can think of Mario making plans, and we can think of Mario and Adela making plans as a unit even if it is not the case that they are both, individually, making plans. In this case, it is fair to say that Adela and the plural subject, the Rossis, made plans, but not that Mario made plans. The philosopher Margaret Gilbert uses examples very similar to this to describe what she calls “plural subject theory.”30 Gilbert makes an im­ portant point about situations like that of Adela and Mario. While Mario may not have personally made any plans, he has committed himself, as a member of the couple, to the plans that Adela made on their behalf. So if Mario later makes separate plans with a friend for the same time and says to Adela, “I think I’ll skip Virginia this year and go fishing with Fred instead,” it would be more than an insult. Mario would be backing out of a commitment he made when he agreed to accept Adela’s plans as the couple’s plans. Even though only half of the couple has made plans, those plans count as plans for the couple as an entity, because they have agreed in advance that they should count as such. Put somewhat differently, the couple really did make plans when Adela made them. Gilbert writes of the ramifications of her work for social theory. But the ramifications for statutory interpretation are also significant. Mario and

90

chapter 4

Adela tell us that we can and do speak of the intent of a unit consisting of more than one person, whether or not all the people in that unit share that intent. We can similarly speak of the intent of the legislature without com­ mitting ourselves to the proposition that every legislator shares the intent that we attribute to the entity. Consider another example. Your local government is planning to put a large sculpture in a park near your home. You have seen pictures of the sculpture, and you like it. But many of your neighbors are against it, including people who are unofficially considered the leaders of the neigh­ borhood. You have no idea what percentage of the neighborhood supports the installation and really have not thought very much about it. When I ask you about the sculpture, you say: “I like it, but the neighborhood’s against it.” This would be a perfectly natural thing to say. But it is also a very complex thing to say. Your answer attributes aesthetic responses to the neighborhood as though it were a single person. In addition, your answer remains reasonable even if you are unable to say exactly who should count as being in the neighborhood and how you draw the line. Some neighbor­ hoods have natural boundaries, like a river, and others have formally de­ fined boundaries, like SoHo in New York City.31 But many neighborhoods have no such definitions as to their boundaries. Yet we can speak of the at­ titudes of “the neighborhood” without regard to these indeterminacies. Moreover, when you say that the neighborhood is against the sculpture, you do so without any formal criteria for determining the neighborhood’s position on the sculpture. You have not taken a poll. Why would you? You thus have no proof that a majority of the people in the neighbor­ hood oppose the sculpture, although you assume that to be the case. All you know for sure is that people who are usually thought to speak for the neighborhood as a whole (whatever that means, given that you do not know where the neighborhood begins and ends) are speaking out against the sculpture. And somehow, that seems good enough to make your state­ ment, “the neighborhood’s against it,” seem natural and reasonable. If you later find out that only a few people agree with the informally recognized leaders in this case, you may decide that your initial conclusion about the neighborhood’s negative reaction was premature and incorrect. But there is no doubt that taking a position at all is a coherent thing to do. Similarly, as the philosopher Michael Bratman points out, a group can jointly intend something without having worked out the details of the plan.32 The day before the trip, Mario and Adela can intend to go to Virginia even if they have not agreed on a departure time, where to stop

the intent of the legislature

91

on the way, and many other such details. For that matter, they can intend to go to Virginia even if they have delegated the details of the trip to a travel agent, who will then let them know where they have various dinner reservations and where they will be stopping on the way.33 But in order for each to say “we” intend to go, each by then has to intend to go, each has to know that the other intends to go, and whatever subplans exist must mesh reasonably well. Bratman further points out that people can jointly intend without shar­ ing a common purpose for their intent. Let us say that Mario and Adela participated equally in planning their trip to Virginia and are in complete agreement about where they will go. However, Mario’s excitement about the trip comes from his love of the mountains, and Adela’s comes from her love of the wonderful small town with its antique stores. According to Bratman, these differences do not interfere with our saying that they jointly intend to take a vacation in Virginia, as long as the details of the trip mesh well enough.34 My intuitions accord with his. Now let us return to the first Mario and Adela, with Adela planning the trip and Mario tagging along. What would happen if they lost the direc­ tions to the inn en route and could not remember the name of the small town? Perhaps while one of them was looking at the piece of paper, it blew out the car window. They stop at a gas station and Adela asks the attendant whether he knows of a small town within a certain distance that is known for having a number of interesting antique stores. The attendant knows of the town and directs the couple there, and they find the inn. The Rossis’ inquiry is motivated by the history of the planning pro­ cess. The sheet of paper that blew out the window said nothing of antique stores. It had a set of driving directions and the address and phone number of the inn. Significantly, Mario had no complaint that the question that Adela asked was about her planning history. There would be no point in getting upset about that, because Mario had no planning history. It would do no good to ask about going to the mountains or other such vague state­ ments. If Mario and Adela want to carry out their joint intent, the missing details will have to come from looking at the intentions of whoever was involved in developing those details. That is how the details got there in the first place. Metaphors that legal theorists have used in describing how courts en­ gage in statutory interpretation fit this story well. Richard Posner writes of the field general whose communication with headquarters is inter­ rupted before an important strategic decision must be made.35 Alexander

92

chapter 4

Aleinikoff employs a nautical metaphor: “Congress builds a ship and charts its initial course, but the ship’s ports-of-call, safe harbors and ul­ timate destination may be a product of the ship’s captain, the weather, and other factors not identified at the time the ship sets sail. This model understands a statute as an on-going process (a voyage) in which both the shipbuilder and subsequent navigators play a role. The dimensions and structure of the craft determine where it is capable of going, but the current course is set primarily by the crew on board.”36 Ronald Dworkin writes about creating the next chapter in a chain novel, based on every­ thing that happened earlier but with the goal of advancing the story.37 All these theorists support using historical context to make the next deci­ sion. When crucial information is missing from the statute itself, context becomes a valued source for drawing inferences about the legislature’s intentions, which each theorist, despite otherwise divergent perspectives, regards as central to statutory interpretation. Let us now look a little more closely at the circumstances in which we are likely to treat a group of individuals as a single entity and attribute to that entity the ability to engage in ordinary human psychological pro­ cesses. Not every collection of people is perceived as a group with a single mind and a unified set of beliefs. If you see a bunch of people standing together at a bus stop, you are not likely to think they hold a collective belief about anything of importance. Even if they share beliefs, such as displeasure at the sorry state of their public transportation system, they do so as individuals with experiences and values in common, not as members of a coherent group. In contrast, we perceive Mario and Adela as a unit by virtue of their being married and by virtue of commitments they have made with respect to what counts as their collective intent. We understand the neighborhood as a group because of physical proximity, a common interest, and a tacit understanding about whose statements represent the views of the neighborhood. By the same token, the legislature is a group by virtue of a host of legal and social institutions, voting practices, and understandings about how its members’ purpose is represented during the legislative process. During the past decade, a number of psychologists, expanding on ear­ lier research, have been exploring when we are most likely to regard a group of individuals as an entity and what it is we attribute to that entity. Some fifty years ago, Donald Campbell, in an important article, referred to the properties of a group that make it more likely to be conceptual­ ized as an individual unit as “group entitativity.”38 These include similar­

the intent of the legislature

93

ity, proximity, formation of a symmetrical pattern, and “common fate.” More recently, Robert Abelson and his colleagues mercifully shortened the term to “entitivity,” which is what it is now called in the psychological literature. They summarize the implications of entitivity: “The greater the exposure to these factors, the higher the perceived entitivity. It will seem like a kind of person, in the sense that it will have the same sorts of fea­ tures that a person has: a unitary identity and personality, and in the real world, a web of past, present, and future relationships with other groups. It will have a set of goals and characteristic behaviors in the service of those goals.”39 Interestingly, one of the most important factors that con­ tributes to judgments about this “unitary identity” is whether the group is perceived as being “highly organized with a specific purpose or intention that drives [its] behaviors.”40 For our purposes, the most relevant characteristic of individuals that we attribute to groups is the property of acting intentionally, as if the group were a person with thoughts and goals of its own. Psychologists have begun to study how this happens. Matthew O’Laughlin and Bertram Malle ob­ serve that “people view each other as agents capable of intentional action, and intentionality is in turn conceptualized by reference to the agent’s beliefs, desires, and intentions.”41 Given this, there are two ways of ex­ plaining a person’s behavior. One is simply to attribute an intention to the person. If a person storms out of a meeting in a fit of anger, we are likely to say that he became angry at what someone said and left. But sometimes we look not at the person’s own beliefs and intentions but rather at what O’Laughlin and Malle refer to as the “causal history of reason.” We also may say, “He often behaves this way when there’s too much pressure on him at his firm. If he loses this case, the company will fire him.” One ex­ planation attributes an intentional state of mind to the actor; the other, a set of circumstances in which it may be reasonable for a person to act the way he or she did. O’Laughlin and Malle found that when given a choice of explanations, people more often choose the one that attributes a reason to the person. That is, we explain the behavior of other people in terms of their intent. The surrounding circumstances act as a backup when we do not know about a person’s motives or when a person appears to have acted pecu­ liarly. Thus, when subjects were presented with a fictitious conversation about a young woman named Nina in which it came out that Nina had been using drugs, the subjects were more likely to account for the drug use by explaining Nina’s intent than by describing historical circumstances.

94

chapter 4

How does this distinction apply to groups? The more we perceive a group as deliberating and acting jointly, the more likely we are to attrib­ ute the actions of a group to the joint intent that we have attributed to the group as a unified agent. When O’Laughlin and Malle substituted, for example, “high school seniors” for “Nina,” the number of explana­ tions that involved circumstances rather than intent increased to about half. On the other hand, when a group appears to deliberate and act as a unit, people prefer intent-based reasons over historical background—as they do with individuals. Thus, “the department faculty” is more likely to produce explanations of intent than is “department chairpersons in the United States,” although both are groups. The authors explain: “The per­ ception of jointly acting groups as unified agents, and the explanation of their actions with a preponderance of reasons, may derive from the nature of coordination in group activity. When acting together, such groups must make their intentions and reasons explicit in order to ensure participa­ tion and coordinated action by group members. When explaining such coordinated action, social perceivers will then use reason explanations to capture the deliberate and reasoned nature of the group action.”42 Legisla­ tures would seem to be a prototypical example of this kind of group. That explains why speaking of legislative intent comes so easily.

The Legislature as an Entity with Intent At this point, we can easily enough transpose our stories about groups from examples about small talk to a discussion about the enactment and subsequent interpretation of a law. Assume that a small group of legisla­ tors who are members of the party in power believe that it would be a good idea to pass a law increasing (or decreasing) the amount of certain chemicals that can be poured down the drain without having to use expen­ sive procedures for disposing of hazardous waste. Most members of the party are sympathetic to the idea, but only those working on the project really understand the details. The language of the bill is worked out in the relevant committees. Compromises are reached with moderates from the opposing party, but in the end the bill passes along party lines.43 It is a complicated, technical law. The committee reports explain some of the fine points, but most of the law is opaque to the legislators who vote ei­ ther for or against the bill, largely based on instructions from their party’s leadership.44

the intent of the legislature

95

Some time after the law is enacted, a dispute arises as to whether it should be applied to the activity of a particular company. The language of the statute does not answer the question adequately. A lawsuit is brought (perhaps a prosecution for failure to comply with the law), and it is left to a court to determine whether or not the law was violated. The judge reads the committee report and concludes that the law’s sponsors did not intend for it to apply to situations such as the one now in litigation. In her opin­ ion, the judge writes: “The committee report suggests that the legislature did not intend the statute to reach the situation that is the subject matter of this case.”45 From what we have seen thus far, the judge’s remark is both natural and coherent, even though almost no one in the legislature had any intent at all with respect to this issue—just about every legislator either voted for the bill without understanding it or voted against the bill. Just as I could say that the Rossis had planned to go to Virginia, and you could say that the neighborhood opposed the sculpture, the judge could say that the leg­ islature did not intend the law to apply in this sort of situation and that the legislature would accept the notion that its intent is reflected in the intent of those who framed and negotiated the final terms of the law. Thus, it re­ ally does seem to be the case that we glean legislative intent from both the words of a statute and the circumstances surrounding its enactment. One need not rely on hypothetical situations to find this kind of ar­ gumentation. In Bank One Chicago, N.A. v. Midwest Bank & Trust Co., Justice Stevens, in his concurring opinion, made the following remarks in support of reliance on legislative history as evidence of the intent of Congress: Legislators, like other busy people, often depend on the judgment of trusted colleagues when discharging their official responsibilities. If a statute . . . has bi­ partisan support and has been carefully considered by committees familiar with the subject matter, Representatives and Senators may appropriately rely on the views of the committee members in casting their votes. In such circumstances, since most Members are content to endorse the views of the responsible com­ mittees, the intent of those involved in the drafting process is properly regarded as the intent of the entire Congress.46

Charles Tiefer calls this justification for reference to legislative intent the “busy Congress” model and cites numerous cases in which the court employs it, largely in opinions written by Justices Stevens and Breyer.47

96

chapter 4

Thus, as Breyer points out, those who oppose citation to the intent of the legislature engage in a category mistake: they confuse the fact that legisla­ tors have different reasons for supporting a bill with their having different perspectives on how the law should be carried out. In reality, legislators do not generally think much about small details, understanding that these have been delegated to a small group of members.48 Moreover, unlike the Rossis’ reliance on Adela to work out the de­ tails of their vacations or the neighborhood’s informal social hierarchy, the legislature’s reliance on committees to work out the details of legislation is a formal part of the process and always has been. Article I, section 5, of the Constitution states, “Each House may determine the Rules of its Proceedings.” Throughout this country’s history, John Roberts points out, both houses of Congress have operated through committees formed pur­ suant to this power.49 Thus, not only does the legislature form its plans through the work of a small number of its members, but it is structured to do things just that way. This should not be much of a surprise. It is difficult to see how a large body with such a broad mandate could function without a committee system. Even a much smaller departmental faculty at a university cannot function without committees. Given that the government conducts its leg­ islative business in this manner, and does so openly, it serves no demo­ cratic function to require the courts to pretend otherwise by ignoring this aspect of the legislative process. Despite the primacy of the committees, however, attribution of intent is not logically bound to any particular set of events within the legislative process. What matters, as Gilbert suggests, is that a particular statement be adopted by the group as representing its intent. While committees will often be at the center of the inquiry, this will not always be the case. Some­ times, for example, the administration may propose legislation through members of Congress. When that happens, the relevant committees may adopt statements from the executive branch as reflecting the bill’s pur­ pose.50 In other instances, the bill’s journey through committees, floor debate, and conference is complicated, with particular moments in the process being crucial to passage of the bill.51 Significantly, it is not necessary that all members of the legislature have the same reasons for supporting a bill, as long as there is general recog­ nition that those who ushered the bill through the process did so with particular subplans that deserve to be honored. Just as Adela’s planning is most useful when the couple loses the directions to the inn, the historical

the intent of the legislature

97

record of a committee or agency that developed the details of a statute is typically useful evidence of that subgroup’s, and thus the entire group’s, in­ tent. If a bill was introduced to increase the amount of pollutants that can be disposed of without a permit, the legislator who votes for it solely be­ cause there is a chemical plant in her district will typically recognize that implementation of the law’s details will and should make reference to the work of the bill’s planners. Empirical research supports this perspective. A study by Eskridge shows that Congress is far more likely to enact legisla­ tion overriding a court decision based on a statute’s plain language than it is to enact legislation overriding a court decision based on a statute’s legislative history or its purpose and policy.52 But stray remarks from individual legislators most likely do not reflect even the intent of the subgroup and are most often not probative of much of anything, as critics of using legislative history in statutory interpretation point out. Reed Dickerson writes: “Among the least reliable kinds of leg­ islative history are floor debates. Not only are they laden with sales talk, but their frequent reference to what a provision means is an unconscious effort to finesse the courts in performing their constitutional function of having the last word on what the statute means. Besides, it would be rare for the authors of a statute to take such references into account.”53 Al­ though I have argued that legislative intent properly plays a role in statu­ tory interpretation, the arguments I have made require that the evidence of intent be tied to what members of the legislature would legitimately regard as constituting the formation of the statute’s details. Thus, except in unusual circumstances, I agree with the limitations that Dickerson and others have imposed on discussion of legislative intent. In contrast, the expressed intentions of the subplanners indeed provide relevant context, whether the subplanners are members of a congressional committee or administrative agency or some other persons or subgroup. It may be true that many who voted for a bill did so on instructions from the party leadership or because the bill contained some benefit for people in their district or for some other reason having nothing to do with what the bill’s authors and planners had in mind. Nonetheless, the bill’s planners gave it content. When disputes arise, it would be odd for a member who voted for the bill without knowing what was in it to complain that the court was looking at the details of the planning process, just as it would be odd for Mario to complain that Adela considered her own planning when she asked for directions. Indeed, there is no such general outcry by legislators about courts’ regular reference to the intent of the legislature.

98

chapter 4

Just as experimental subjects rely upon both intentional and histori­ cal circumstantial explanations to account for group action, so do judges, even when they eschew legislative history. For example, in Smith v. United States, discussed in chapter 3, a majority of six justices held that the de­ fendant’s attempt to trade a machine gun for some illegal drugs should be considered “[using] a firearm during and in relation to a drug trafficking crime.”54 In perfect textualist fashion, the Court nowhere relied on com­ mittee reports or other legislative history to support its decision. But it did rely upon a report of the American Enterprise Institute, a conservative think tank, to describe the context in which Congress made the decision to enact the statute.55 When judges attribute intent to the legislature as though it were a single individual with a mind of its own, they are simply doing what we all do when we talk about deliberative groups.

Looking into the Minds of Others For many years, developmental psychologists have been studying how children acquire words at such an incredible rate and seemingly with so little effort.56 Children build theories of words from a combination of their experience and the way their minds organize the information gleaned from that experience. Adults do the same thing.57 As we saw in chapter 3, our theories of words consist of definitional features, salient features, background information, and sometimes even recognition that we have gaps in our knowledge. For example, I know that table salt is sodium chloride. I know that the chemical symbol for this molecule is NaCl. But that is about all I know about the chemistry of table salt. Yet I believe that there is much more to the concept of table salt and that there are people who know about such things.58 Just as courts make reference to plain and ordinary meaning, they sometimes make reference to the expertise of others in construing technical terms of which the judges themselves may have incomplete knowledge.59 Moreover, our understand­ ing of words is replete with background information that reduces the cost of learning new concepts. To say that burglary is a crime makes sense only if we already have a theory of social organization into which such a state­ ment fits. How do we develop these theories of meaning in our minds? An im­ portant part of what makes it so easy to acquire new words and concepts is our innate ability to draw conclusions about what is in the minds of others.

the intent of the legislature

99

Psychologists call this the “theory of mind” approach to the acquisition of words.60 The claim that children come to understand the world through the minds of others may at first seem radical, but it really does accord with our experiences with young children. Consider this summary by Gopnik, Meltzoff, and Kuhl, who have written an accessible book that discusses this issue: When babies are around a year old, then, they seem to discover that their ini­tial emotional rapport with other people extends to a set of joint attitudes to­ward the world. We see the same objects, do the same things with those objects, even feel the same way about those objects. This insight adds a whole new dimension to the babies’ understanding of other minds. But it also adds a whole new di­ mension to babies’ understanding of the world. One-year-old babies know that they will see something by looking where other people point; they know what they should do to something by watching what other people do; they know how they should feel about something by seeing how other people feel.61

To illustrate, in one experiment, an adult looked with a happy expres­ sion into one box and with an expression of disgust into another. She then pushed the boxes to the baby participating in the study. The infants were more likely to open the box that made the adult happy than they were to open the other one.62 Since the child did not know in advance what was in the boxes, she could have based her decision only on an assessment of the experimenter’s reaction. That is, she had to form a theory of what the experimenter was thinking in order to decide what to do. Developmental psychologist Paul Bloom takes a similar position: “Children use their naive psychology or theory of mind to figure out what people are referring to when they use words. Word learning is a species of intentional inference or, as Simon Baron-Cohen has put it, mind read­ ing.”63 Of course, as Bloom acknowledges, we do not only concern our­ selves with the minds of others when we learn new words and concepts. When a mother points to a caterpillar and says, “That’s a caterpillar,” the caterpillar itself plays an essential role in the acquisition of the concept. Nonetheless, inferring what others have in mind contributes to the acqui­ sition of the meanings of words. Assessing the minds of others is not limited to children. It is how we interact in social contexts generally. As Steven Pinker puts it: “A mind unequipped to discern other people’s beliefs and intentions, even if it can

100

chapter 4

learn in other ways, is incapable of the kind of learning that perpetuates culture. People with autism suffer from an impairment of this kind. They can grasp physical representations like maps and diagrams but cannot grasp mental representations — that is, they cannot read other people’s minds.”64 In fact, recent work has begun to trace the progression from young child to adult in the ability to derive conclusions about peoples’ states of mind from their actions.65 In short, we are accustomed to under­ standing new concepts by reference to the intent of the person who intro­ duced the concept to us. We form a theory of the person’s mind.

Can Statutory Interpreters Avoid Looking into the Minds of Others? It is not easy to stop thinking in terms of the minds of others. We start doing it as babies, and we still do it every day to understand our world. Judges are no different in this respect. Regardless of the academic contro­ versy, judges, including Supreme Court justices, often talk of the “intent of the legislature” or what the legislature “meant.” Consider the 2003 Su­ preme Court case Roell v. Winthrow. The case involved a dispute between a group of prisoners and the medical staff at the prison. The prisoners challenged an unfavorable ruling by a magistrate judge on the grounds that the magistrate judge did not have jurisdiction because not all medi­ cal staff defendants had formally waived their right to have the case tried before a district court judge. After looking at the language of the relevant statute, Justice Souter, writing for a majority of five, stated: These textual clues are complemented by a good pragmatic reason to think that Congress intended to permit implied consent. In giving magistrate judges case-dispositive civil authority, Congress hoped to relieve the district courts’ “mounting queue of civil cases” and thereby “improve access to the courts for all groups.” S.Rep. No. 96–74, p. 4 (1979); see H.R.Rep. No. 96–287, p. 2 (1979) (The Act’s main object was to create “a supplementary judicial power designed to meet the ebb and flow of the demands made on the Federal judiciary”). At the same time, though, Congress meant to preserve a litigant’s right to insist on trial before an Article III district judge insulated from interference with his obligation to ignore everything but the merits of a case.66

The highlighted words all attribute intent.

the intent of the legislature

101

table 4.1.  Use of Intentional Language by Federal and State Courts From 1998 to 2007 “Legislature” or “Congress” within six words of:

Federal courts

State courts

“intent,” “intend,” etc. “means,” “meant” “believe,” “believes”

30,000+ 2,621 646

30,000+ 2,249 519

Note: The searches for the respective cells in the table were as follows: Search 1: ((Congress or legislature) w/6 inten!) and date bef 2008 and date aft 1997 Search 2: (Congress mean! or legislature mean!) and date bef 2008 and date aft 1997 Search 3: (Congress believe! or legislature believe!) and date bef 2008 and date aft 1997

Cases like this are standard fare. Although it would be difficult to quan­ tify how frequently courts discuss the legislature’s state of mind be­cause there are so many ways to do so, table 4.1 gives the results of some word searches conducted on LexisNexis in both the federal and state court libraries. Most noteworthy, state and federal courts each used the words “legis­ lature” or “Congress” within six words of “intend” words (“intend,” “in­ tent,” “intention,” and so forth) more than 3,000 times per year, or at least 30,000 times for the decade.67 That means courts used this language at least 60,000 times in the ten-year period of the search. These numbers are so large that my research assistant and I conducted the same search for the previous ten-year period (1988–97) to make sure the results reported in the table are not aberrational. They are not. There were more than 60,000 occurrences in the earlier period as well. That is how judges think and how they express their thoughts. Other words attributing to the legislature a single state of mind are also used often, but not as much as words of intent, as the table indicates. One way of investigating the extent to which intent is embedded in our thinking is to see whether people who have rejected legislative intent as relevant to legal analysis continue to make reference to it anyway. Con­ sider again the famous statement by Oliver Wendell Holmes: “We do not inquire what the legislature meant; we ask only what the statute means.”68 Maybe so. But here are a few excerpts from Holmes’s opinions, with inten­ tional language in boldface: When it is considered that at the time the Act allowing the drawback was passed the tax was collected wholly by stamps, it seems evident that Congress meant to carry the policy of the Constitution against taxing exports beyond its strict requirement and to let the event decide about the tax.69

102

chapter 4

Considering that only the principal of mortgages was taxed when the law was passed and that in those days no one thought of an income tax; that any contract of exemption must be shown to have been indisputably within the intention of the Legislature; that it is difficult to believe that the Legislature meant to barter away all its powers to meet future exigencies for the mere payment of a mort­ gage recording tax. . . .70 We see no sufficient ground for supposing that Congress meant to open the questions that the other construction would raise.71

There are many more like these. I am not calling Holmes a hypocrite. Holmes’s approach to statutory in­ terpretation is more complex than simple reliance on an acontextual read­ ing of the statutory text.72 Yet the frequent mention of legislative intent in his opinions does seem inconsistent with his famous pronouncement about statutory interpretation. Holmes wrote about intent because we think in terms of intent when language leaves us unsure. Because Holmes was not committed to textualist methodology, he did not make a post hoc effort to censor his remarks.

Where Intent Matters Most Intent and the Textualist Canons What about the argumentation of Justice Scalia, who is a committed tex­ tualist? He does not slip up often, but it happens to him as well, especially in his dissenting opinions. Here are a few examples, with the intentional language again in boldface: When, Chevron said, Congress leaves an ambiguity in a statute that is to be ad­ ministered by an executive agency, it is presumed that Congress meant to give the agency discretion, within the limits of reasonable interpretation, as to how the ambiguity is to be resolved.73 Such a system of justice seems to me so arbitrary that it is difficult to believe Congress intended it. Had Congress meant to cast its carjacking net so broadly, it could have achieved that result—and eliminated the arbitrariness—by defin­ ing the crime as “carjacking under threat of death or serious bodily injury.”

the intent of the legislature

103

Given the language here, I find it much more plausible that Congress meant to reach—as it said—the carjacker who intended to kill.74

In fairness, Scalia may have been responding to the intentionalist argu­ ments of others in some of these examples and therefore structured his arguments accordingly. Still, intentionality has even made its way into statutory analysis that appears prototypically textual. Consider this seem­ ingly standard analysis by Scalia in a case involving the obligations of individuals who received more Supplemental Social Security benefits than they were entitled to receive: “If Congress had in mind only shortfalls or excesses in individual monthly payments, rather than in the overall pay­ ment balance, it would have been more natural to refer to ‘the correct amount of any payment,’ and to require adjustment ‘with respect to any payment . . . of less [or more] than the correct amount.’ This termi­nology is used elsewhere in [the statute].”75 Scalia has made a routine textual ar­ gument— Congress would have used other words to convey the meaning that the losing party would ascribe to the disputed statute—but he has done so in intentionalist terms. We infer from the language, Scalia suggests, that Congress did not have the losing party’s repayment scheme “in mind.” And in doing so, we necessarily attribute intentionality to Congress as an entity. The same holds true for other text-oriented interpretive devices. Take, for example, the “ordinary-meaning” rule discussed in chapter 3. What is the rationale for this rule? It is based on the assumption that legisla­ tive drafters are most likely to use words that way. If a court adopts that assumption, it will be more likely to make a decision that is loyal to the legislature’s intention. Again, consider Scalia’s explanation of the rule: “The question, at bottom, is one of statutory intent, and we accordingly ‘begin with the language employed by Congress and the assumption that the ordinary meaning of that language accurately expresses the legisla­ tive purpose.’ ”76 Scalia argues that we are likely to honor the legislature’s intention if we construe words in their ordinary sense because that is how the legislature probably intended them to be understood. It is very difficult to jettison this perspective in statutory interpretation other than in cases in which there is no legitimate controversy.77 Even arguments of statutory coherence, which, as discussed in chap­ ter 5, imply values other than enforcing the will of the legislature, are often stated as good surrogates for intent, including by Scalia:

104

chapter 4

The meaning of terms on the statute books ought to be determined, not on the basis of which meaning can be shown to have been understood by a larger hand­ ful of the Members of Congress; but rather on the basis of which meaning is (1) most in accord with context and ordinary usage, and thus most likely to have been understood by the whole Congress which voted on the words of the statute (not to mention the citizens subject to it), and (2) most compatible with the surrounding body of law into which the provision must be integrated—a compatibility which, by a benign fiction, we assume Congress always has in mind. I would not permit any of the historical and legislative material discussed by the Court, or all of it combined, to lead me to a result different from the one that these factors suggest.78

One reason for a court’s interpreting laws to foster a coherent code is the assumption that the legislature would have wanted it to do so. Other can­ ons, such as the rule against construing statutory language as surplusage, are equally oriented toward the probable intent of the legislature and are just as acceptable to textualist judges, as Caleb Nelson points out.79 Despite his frequently acrimonious tone in deriding the kind of inten­ tionalist rhetoric he actually uses on occasion, I do not believe Scalia is being hypocritical either. He has taken an intellectual position that cer­ tain kinds of analysis are not appropriate in statutory interpretation; the problem is that it is almost impossible to avoid thinking in the intention­ alist terms that he would outlaw. Indeed, he cannot always do it either. Text-based canons of construction, then, are by no means an alternative to considering legislative intent, although they are commonly represented as such.80 On the contrary, even according to their defenders, they are best seen as default rules for assessing the likely intent. Legislative Errors and Absurd Results Intent frequently plays a role in statutory interpretation when a court decides that the legislature has made a mistake. Consider United States v. Granderson, discussed in chapter 3.81 Granderson, a letter carrier, was convicted of destruction of mail and sentenced to five years’ probation; the maximum prison sentence under the federal Sentencing Guidelines would have been six months. While on probation, he tested positive for co­ caine, in violation of his probation conditions. The district court sentenced him pursuant under the relevant statute, which states: “If the defendant violates a condition of probation . . . the court shall revoke the sentence

the intent of the legislature

105

of probation and sentence the defendant to not less than one-third of the original sentence.”82 What is “one-third of the original sentence”? The Court accepted Granderson’s interpretation: the individual must serve a sentence of at least one-third the sentence that he would have received had he been sen­ tenced to prison in the first place rather than to probation—that is, onethird of six months. The majority of justices agreed with the government that it would be absurd to interpret the statute to require Granderson’s sentence to be reduced to an even shorter period of probation, which is the most natural reading of the words taken literally. The Court, though, rejected the government’s position that Granderson should receive a prison sentence at least one-third as long as his original probation period, in part using the rule of lenity to resolve the ambiguity between the re­ maining two interpretations. In so doing, the majority looked at the legis­ lative history: Congress drafted this provision hastily, and it appeared that the drafters may have had in mind an earlier sentencing scheme in which probation was granted by imposing a sentence and then revoking the sen­ tence in favor of a probationary period of good conduct. The intermediate position was truest to general principles of interpretation and to what the legislature was likely trying to accomplish. John Manning argues that it is possible to maintain a version of the absurd-results doctrine without resort to the intent of the legislature.83 Instead of concerning themselves with intent, Manning argues, courts could focus instead on statutory language as understood in its social con­ text. In Granderson, that would amount to saying that in the context of this statute, members of the relevant linguistic subcommunity typically would not use the expression “one-third of the original sentence” to mean one-third of the original period of probation. Thus, without referring to the intention of any particular legislator or group of legislators, it is possible to reject this interpretation as absurd within textualist methodol­ ogy. The argument is akin to reliance on ordinary meaning in lieu of an investigation into the intentions of the legislature. Manning is surely correct that a more nuanced, context-sensitive un­ derstanding of language can reduce resort to extratextual material. The more information contained in the language itself, the less the need to look elsewhere. The problem with this analysis, however, is its denial that intent underlies it. Why would the relevant linguistic subcommunity not understand “one-third of the original sentence” as reducing the sentences of probation violators? After all, because there was only one sentence—

106

chapter 4

probation — the most natural or ordinary interpretation of the phrase would indeed lead to such a reduction. The principal reason to reject this interpretation is that it makes no sense to conclude that those in Con­ gress responsible for drafting this language would have wanted to reduce the sentences of probation violators. That is, the social context matters only because it provides significant clues as to what the author intended to communicate within the range of possible interpretations. In this case, a linguistically less natural interpretation of the phrase “one-third of the original sentence” is appropriate precisely because Congress is unlikely to have intended the most natural one. Just as the ordinary-meaning canon is repeatedly justified by textualists as good evidence of legislative intent, the absurd-results rule acts as a safety valve when resort to ordinary mean­ ing fails. It may be possible, as Professor Manning further argues,84 to bypass this entire discussion and rely on values other than intent to justify the rejection of absurd results. In Granderson, for example, one may reject as unconstitutionally irrational an interpretation of the statute that calls for higher sentences for those who abide by their conditions of probation than for those who violate them, regardless of what the legislature may have intended. This strikes me as a legitimate alternative analysis. Yet problems of intent continue to arise. For example, United States v. X-Citement Video, Inc. dealt with a stat­ ute in which Congress had not drafted an appropriate state-of-mind re­ quirement making it illegal to deal in child pornography.85 The majority of justices assumed that Congress intended to write a constitutional statute, and they implied the appropriate state-of-mind requirements to rescue the law from constitutional infirmity. Chief Justice Rehnquist’s majority opinion is replete with inferences about legislative intent.86 Only because of that presumed intent did the majority consider itself justified in amend­ ing the statute judicially. Justice Scalia, in dissent, regarded that position as too generous to Congress, which may well have intended the law to apply according to its plain, unconstitutional language.87 He would have declared the statute unconstitutional. How this debate would play out in a case like Granderson is not at all clear. Once constitutional values are invoked to reject the “absurd” reading of the statute, a choice would arise between declaring the statute unconstitutional based on its ordinary meaning or finding a constitutional reading of the statute at least plausibly consistent with the statutory lan­ guage. Without resort to intent, it is difficult to see how a court could do

the intent of the legislature

107

anything other than declare the statute in X-Citement unconstitutional, a choice that eight of the justices did not make. Similarly, in Green v. Bock Laundry Machine Co., Inc., a literal reading of Rule 609 of the Federal Rules of Evidence would have been absurd.88 In that case, Paul Green, a prisoner, was working at a car wash on a work release program. He was involved in a serious accident when a dryer at the car wash literally tore off his arm. He sued the manufacturer of the machine that injured him. In the course of the trial, lawyers for the dryer manufacturer wanted to bring to the jury’s attention the fact that he was a convicted felon. Green objected. The rule at that time permitted the fact of a criminal conviction to reach the jury “only if the crime (1) was punishable by death or imprisonment in excess of one year under the law under which the witness was convicted, and the court determines that the probative value of admitting this evidence outweighs its prejudicial effect to the defendant, or (2) involved dishonesty or false statement, regardless of the punishment.”89 The testimony would obviously be prejudicial, since Green’s status had nothing to do with whether the machine was designed defectively. But, as written, the rule would require only a balancing of the conviction’s “prejudicial effect to the defendant.” Since Bock Laundry Machine Company was the defendant, and the evidence was not the least bit prejudicial to it, the plain language of the rule would appear to make the evidence admissible. None of the justices on the Supreme Court interpreted the rule as writ­ ten. The Court was not about to declare an important rule of evidence unconstitutional for its arbitrariness, creating an immeasurable problem for the judiciary until the problem could be solved legislatively. The only issue was whether the balancing was relevant only to criminal cases (i.e., “defendant” means “criminal defendant”), or whether it should be ap­ plied to all cases (i.e., “defendant” should be read “party”). Research­ ing the record for clues of congressional intent, the Court opted for the former, and the evidence was deemed properly admitted (and Green lost his case). Justice Scalia concurred, objecting to the majority’s reliance on historical investigation.90 At the root of this discussion is the observation that it is intent that makes a mistake a mistake. We all have experiences like “Did I say Tues­ day? I meant Thursday.” Alternatively, we sometimes believe that we were mistaken when, in retrospect, we failed to think through the consequences of our statements in unforeseen circumstances. The two types of mistake are quite different, and textualists draw a distinction between them.91 In

108

chapter 4

the first instance, the person says, “I didn’t mean to say what I said.” In the second, the person says, “I meant to say what I said, but now I see that it was a very bad idea.” Sometimes, in the statutory context, it is not easy to tell which type of error occurred, and judges may differ in both how they characterize the statute and how they vote. Consider, for example, United States v. Locke, discussed in chapter 1, which the Supreme Court decided in 1985.92 Sec­ tion 314 of the Federal Land Policy and Management Act of 1976 states that a notice of intention to hold one’s unpatented mining claim must be filed with the Bureau of Land Management (BLM) “prior to December 31 of each year.”93 Locke filed his notice on December 31, 1980. He therefore suffered the statute’s sanction for a late filing: a conclusive presumption that the mine has been abandoned. Ironically, the BLM had made the same mistake in its 1978 questionand-answer brochure, in which it advised miners that the annual filing must be made on or before December 31. Moreover, at Locke’s request, Locke’s daughter had called the local BLM office and was told that the fil­ ing had to be made “on or before December 31, 1980.”94 Locke complied with this advice. Locke and his family had been operating the mine since the 1960s. Refusal to grant a one-day reprieve seemed to mean that they were to lose their family business. Writing for the majority, Justice Thurgood Marshall held that a dead­ line is a deadline: “Going behind the plain language of a statute in search of a possibly contrary congressional intent is ‘a step to be taken cautiously’ even under the best of cir­ cumstances.” When even after taking this step nothing in the legislative history remotely suggests a congressional intent contrary to Congress’ chosen words, and neither appellees nor the dissenters have pointed to anything that so sug­ gests, any further steps take the courts out of the realm of interpretation and place them in the domain of legislation. The phrase “prior to” may be clumsy, but its meaning is clear. Under these circumstances, we are obligated to apply the “prior to December 31” language by its terms.95

Absent clear evidence that the language of the statute did not reflect the intent of the legislature, no error would be inferred, so there was noth­ ing to correct. Although some, including Justice Stevens, who dissented in this case, might find Marshall’s threshold too severe because the leg­ islature’s intent seems obvious enough from the circumstances, adhering

the intent of the legislature

109

to the Constitution’s procedures for enacting laws is surely a value that judges take seriously. In fact, the outcome of this case was not as bad as it looked for the Locke family. Locke has been widely criticized for its “wooden” adher­ ence to language that “creates a trap for valuable property rights.”96 But a footnote in the majority opinion and Justice O’Connor’s concurrence hint that the decision may have been much fairer to the Locke family than initially appears. Stevens argued in dissent that the phone call of Locke’s daughter to the BLM should play a crucial role in the case’s outcome.97 In footnote 7, Marshall, responding to this argument, suggested that it might have been appropriate for the district court to have applied the doctrine of equitable estoppel against the government for its poor advice about the deadline.98 Since the district court had not considered this estoppel claim, the Court remanded the case for further consideration. O’Connor’s concurrence was based entirely on the possibility of equitable estoppel.99 Interestingly, Locke was never actually decided on remand. Having read footnote 7, the government decided to abandon the case, and the Lockes got back their mine.100 One may come to different conclusions about which, if either, of these types of error should be corrected judicially. In chapter 5, we will see that respect for the legislative process routinely trumps concern for the broad purpose of the statute. Notwithstanding Locke, courts are most likely to correct legislative errors when they independently conclude that the lan­ guage of the statute is inconsistent with what the legislature would have intended in the particular case at hand. Indeterminacy in Meaning A third set of cases raises the issue of intent somewhat differently: those in which the language of a statute is subject to more than one interpreta­ tion. If a court pays attention to the ordinary meaning of a statutory term, it will reach one conclusion about how the statute should apply in the case before it. If, on the other hand, it asks whether the legislature would have wanted the statute to apply in a broader range of circumstances, it must reach the opposite conclusion. This is exactly what happened in an important 1991 Supreme Court case, Chisom v. Roemer, discussed in chapter 3 in connection with the ordinary-meaning rule.101 Section 2 of the Voting Rights Act prohibits states from affording protected classes of people “less opportunity than

110

chapter 4

other members of the electorate to participate in the political process and to elect representatives of their choice.”102 The question in Chisom was whether the act applied to judicial elections as well as to legislative elec­ tions. Based on an analysis of legislative intent, the Court held six to three that the act did apply to a judicial election in Louisiana. In a sharp dissent, Scalia argued: “There is little doubt that the ordinary meaning of ‘representatives’ does not include judges.”103 Scalia was right about the statute’s ordinary meaning. But Scalia’s dissent was just as intent oriented as the majority opinion. He explained: “[O]ur job is to determine whether the ordinary meaning includes them, and if it does not, to ask whether there is any solid indication in the text or structure of the statute that something other than ordinary meaning was intended.”104 Thus, Scalia’s rationale for using ordinary meaning is that it serves as a surrogate for legislative intent. In this case, competition among the can­ ons—resort to ordinary language against resort to extrinsic evidence in the face of ambiguity—is merely evidentiary in nature. Both sides purport to base their decision on the intent of the legislature; they simply disagree about how to find that intent.

Private Language, Public Laws The ubiquitous presence of intent talk in all these contexts suggests that there may actually be less difference between intentionalists and textual­ ists than the rhetoric would cause us to believe. If so, we need to explain why Holmes’s warning—that we care about what the statute means, and not what the legislature meant— continues to resonate today. Much of the debate over this issue, I suggest, stems from the high stakes involved in the dispute: how we are to govern ourselves under a rule of law ar­ ticulated pursuant to constitutional procedures. Seen in this broader light, disagreement over the role of legislative intent serves as a surrogate for a larger disagreement over whether government by laws is at least in part an illusion. If statutes have meaning on their own without inquiry into the intent of their makers, then decision makers need only read the statute to de­ termine when it should apply. But if determining intent is a necessary element of interpretation, then statutory language can be no more than evidence of that intent. As soon as we begin to speak of “legislative intent” or “what the legislature meant” or “had in mind” or any other such thing,

the intent of the legislature

111

we automatically reduce the statute’s language to “evidence of law,” to use Easterbrook’s words. The need to look beyond a statute’s language to understand the law is a serious concession. It means that constitutional procedures for enacting laws are not enough to determine the rights and obligations of the citizenry in a significant set of circumstances. The textu­ alists recognized this inference long ago and have been trying to eliminate intent as part of the interpretive enterprise ever since. Political concerns, however, do not alter the nature of the human lan­ guage faculty. Those who remember the Watergate scandal will recall Senator Sam Ervin’s presence as chair of the Senate select committee in­ vestigating President Nixon’s activities. At one point in the proceedings Ervin remarked that the president’s duties did not include criminal activ­ ity. When asked how he knew that, he replied: “Because the English lan­ guage is my native tongue, and I understand it.”105 What does this mean? When I say that I know the English language as “my native tongue,” I mean that I have in my mind the vocabulary, rules, and principles that entitle me to call myself a speaker of English.106 If I also identify you as a native speaker of English, I assume that your mind is similarly endowed and configured. When I say something to you, I expect that you will un­ derstand it more or less as I would if you said it to me. Similarly, when you say something to me, I expect that you intended to express more or less what I understood you to have said. But that is as far as I can go. When we communicate, all we have are intentions and some confidence that our language faculties are more or less the same.107 Noam Chomsky refers to this concept of linguistic knowledge as “I(nternalized)-language.”108 Put in these terms, Senator Ervin’s remark is best interpreted: “Anyone who can legitimately claim to have English as his native tongue would understand this statement the same way that I do.”109 Our language faculty consists of many components: vocabulary, a set of sounds, rules governing the sound system, syntactic rules and principles, conventions of inference. With respect to at least some of these, I can be very sure that you and I will be in accord. If I say, to take a simple, standard example, “The cat is on the mat,” I have every reason to believe that you will understand the relationship between the cat and the mat the same way I do. You will not, for example, think that the mat is on the cat. While there may be small differences at the margin, it appears that knowledge of syntactic relations is quite uniform among speakers of a given language.110 There may be words that you know that I do not, or that you and I appear to use in slightly different ways. For example, is the cat on a bathmat, a

112

chapter 4

placemat, or a doormat? For that matter, what you call a mat may for me be a small rug. Yet most of the time we use words the same way. We may pronounce some words differently, but we seem able to make adjustments for this. In all these cases, our experience is individual and personal. Nonethe­ less, I do have the intuition that when I say I understand English, there is something external to my mind that constitutes English. This, however, is the result of an illusion. To the extent that we can say with complete confidence that you and I will understand an English utterance the same way, it is as though there really is something called English “out there,” apart from our intentions and the similar configurations of our minds. There is no need to think about intended meaning, possible differences between us, or the context in which a statement is made if there can be no controversy over the statement’s meaning even after taking those things into account. Unless I make a mistake, I certainly do not mean that the mat is on the cat when I say, “The cat is on the mat.” By the same token, as we saw in chapter 2, when a bribery statute makes it illegal for any person to “corruptly . . . give anything of value to a public official,”111 we do not mistakenly think it means that a government official cannot give a Christ­ mas present to her dry cleaner. The grammatical relations unquestionably inform us about the direction of the banned payments. The point becomes clearer if we compare the following two questions, which roughly track the distinction that Holmes made between the intent of the speaker and the meaning of the language: 1. What did you mean? 2. What does that mean?

Philosophers sometimes distinguish between these two senses of mean­ ing: some call the first “intentional” and the second “extensional”;112 oth­ ers distinguish between “speaker’s meaning” and “word meaning.”113 The distinction seems to disappear on close analysis. It is rather a matter of whether we focus on the individual intent of the speaker or on the intent of the speaker as a member of a group with shared knowledge. If you use an English word that I do not know, I can ask either of these questions appropriately. If I ask question 1, I am, sensibly enough, asking for your intent, given that you are trying to convey some thought to me; if question 2, I am assuming that everyone with command of that word would have a similar purpose in uttering it, and I am asking what that pur­

the intent of the legislature

113

pose is. They both amount to the same thing. In contrast, if you are speak­ ing French to a group of people and say something that goes beyond my knowledge of that language, I am more likely to use question 2 than 1 be­ cause I assume that French speakers by and large would understand your utterance the same way, by virtue of sharing an I(nternalized)-language, but that my state of knowledge is deficient. The problem is the difference I perceive between myself and the community of people who can legiti­ mately say that they speak French. If, however, you say something vague, ambiguous, or otherwise unclear in a language in which I consider myself a competent speaker, question 1 is more appropriate. In fact, question 2 sounds rather nasty in many contexts: it more or less accuses you of not having a good enough command of the rules and vocabulary of English to use it as a native speaker. In other contexts, question 2 serves a didactic purpose, as when a teacher is prodding a student to explore more fully the ramifications of an idea.114 Crucially, if you say something that I understand perfectly well, then neither question 1 nor 2 is appropriate, and for exactly the same reason. I believe, as a speaker of English, that I understand what you have said and that I understand it for the same reason that everyone else who speaks English can say that they understand it. This situation is exactly what courts have in mind when they invoke the plain-language rule. A commonly quoted example appears in Caminetti v. United States, a 1917 Supreme Court decision: “It is elementary that the meaning of a statute must, in the first instance, be sought in the language in which the act is framed, and if that is plain, and if the law is within the constitutional au­ thority of the law-making body which passed it, the sole function of the courts is to enforce it according to its terms.”115 When the Court says that the language is “plain,” it is saying that no one would disagree about what someone using that language intended to say. Let us return to the bribery statute. Assume consensus about the re­ lationships among the parties mentioned in the statute. This agreement does not mean that discerning communicative intent is irrelevant to our understanding of the statute’s meaning. It means only that we do not have to worry about any distinction between language and intent, because na­ tive speakers of English would have only a single intent if they used the language sincerely and without error. In fact, courts at times actually state the plain-language rule in terms of the “best evidence” of legislative intent, although the Supreme Court has not done so. For example, the U.S. Court of Appeals for the Third Circuit characterized the role of plain language in

114

chapter 4

statutory interpretation as follows, quoting a leading treatise: “As always, the most authoritative indicators of what Congress intended are the words that it chose in drafting the statute. See 2A N. Singer, Sutherland Statutory Construction § 46.03 (‘What a legislature says in the text of a statute is con­ sidered the best evidence of the legislative intent or will.’). Here, we think that Congress’s use of the word “purchaser” by itself connotes an intent to include only volitional transactions.”116 The Third Circuit regards the language of the statute, not as the law in Easterbrook’s sense, but as “[an indicator] of what Congress intended.” This is so even when the language is plain. It is easy enough to find similar statements elsewhere, in the opin­ ions of both federal117 and state118 courts. Thus, when faced with a dispute about whether a law applies in a par­ ticular situation, our first impulse is to examine what the law says. After all, laws are intended to be both autonomous and authoritative. If a law is clear enough, so also is our obligation to obey it in particular circum­ stances. Our second impulse is to ask what the legislature meant by what it said. These two impulses are so closely related that they are really part of the same task. For this reason, Scalia’s concurrence in Conroy v. Aniskoff, asserting that “we should not pretend to care about legislative intent (as opposed to the meaning of the law), lest we impose upon the practicing bar and their clients obligations that we do not ourselves take seriously,” misapprehends what we do when we interpret the language of others.119 We do indeed care about the meaning of the law. We care so much that we look to the intent of the legislature to see what that meaning is, just as we look to the intent of others to understand most everything else in social contexts. As Scalia regularly reminds us, the language that the legislature used is often the only evidence needed to determine what the legislature had to say. But that does not mean that intent is irrelevant, and it does not mean that the language of the statute is always sufficient to allow us to make a final decision about the statute’s applicability. Most of the time, whether or not statutory interpreters speak of legis­ lative intent makes little substantive difference. The language of the law as applied to the situation at hand leaves little room for controversy. The stakes go up, however, in those cases in which different interpretive ap­ proaches may lead to different results. Once we encounter even the small­ est amount of controversy or uncertainty, we routinely begin thinking in terms of intent. This explains why courts talk of intent so often. Whether we have re­ course to the ordinary-meaning canon, search out the legislature’s use of

the intent of the legislature

115

similar language elsewhere in the code, determine that another expres­ sion would have conveyed a particular meaning better, or, for that matter, examine legislative history, we are searching for intent. We do so because that is the only question to ask if we care about what someone was trying to say and the words are not enough to answer the question. For these reasons, I do not believe that any judge or commentator can consistently maintain that courts should dispense altogether with discussion of legisla­ tive intent. The concept is just too deeply embedded in the way we see the world, and it is not a serious threat to rule-of-law values, as I hope to have demonstrated earlier.

The Legislative History Debate Whether legislative history is good evidence of legislative intent is another matter. Scalia’s disapproval of courts’ relying on legislative history has two parts. The first is that legislative history can be used only to provide evidence of legislative intent, and legislative intent is not a legitimate subject of inquiry for statutory interpreters. We have been dealing with that problem in this chapter and return to it in chapter 6. The second prong, addressed here, is that even if it is legitimate to speak of legislative intent, legislative history is a notoriously bad way to investigate intent. This challenge has not yet been adequately answered. Much of the debate about legislative history focuses on the question of how good a window into legislative intent it really is— once we accept that legislative intent is a legitimate inquiry in the first place. Opponents, focusing largely on cases in which courts appear to pick and choose snip­ pets of the Congressional Record to make a point, argue that it is not reli­ able at all. For the most part, I do not believe this issue to be very important, despite all the attention it has received in both the case law and the schol­ arly literature. How often does a dissenting or concurring opinion, or a well-researched law review article, show that judges actually misuse legis­ lative history in a way that seriously threatens a legal system based on ac­ ceptable legal values? It happens, but not very often. For example, Scalia’s remarks in his concurring opinion in Conroy v. Aniskoff, discussed above, suggest that the majority did a sloppy job in selecting snippets of legisla­ tive history.120 Scalia’s point strikes me as valid. Because the statute was clear on its face, and the historical research shoddy, perhaps the research

116

chapter 4

should not have been reported as part of the basis of the decision. Yet the history was used only to confirm the decision based on the statute’s language. Remember, Scalia concurred in the judgment, as he often does when the majority reviews a statute’s legislative history. On the other side of the equation, to make the strongest case for legis­ lative history it is necessary to determine just how often it is demonstrably useful and, when it is, how often it does enough good to justify the cost of digging it up. To the best of my knowledge, no one has investigated these questions systematically, although some academics have called for research into these kinds of issues.121 I agree with these scholars. That is the only way to separate the facts from ideologically based biases asserted with no serious empirical foundation. In just this context, an exciting debate has developed over whether legislative history was misused by the Supreme Court in Church of the Holy Trinity v. United States.122 Chapter 3 discussed this case as an ex­ ample of the tension between definitional meaning and ordinary mean­ ing in statutory interpretation. It is also probably the first Supreme Court case in which legislative history played a major role in trumping one inter­ pretation in favor of another. Recall that the statute in question made it illegal “in any manner whatsoever, to prepay the transportation . . . of [an] alien . . . to perform labor or service of any kind in the United States.”123 The Court held that the statute did not apply to the church, based in large measure on the ordinary meaning of “labor.” The Court also relied on a committee report that suggested that Con­ gress had intended the statute to apply to manual labor but did not have time to make the language more precise before recessing for the term. Adrian Vermeule, who reviewed the entire legislative history surrounding the enactment of the statute at issue, argues that the Court got it wrong. Congress knew perfectly well what it was doing when it did not include members of the clergy among a list of statutory exceptions.124 Vermeule argues further that the work it takes to learn the full history—and the likelihood of error from picking and choosing—makes legislative history an unreliable source and, for that matter, a very expensive way of learning very little, an argument that has resonated with English courts in modern times.125 In response, Carol Chomsky has written an article defending the Court’s use of legislative history.126 She suggests that when the enactment history and social context are taken together, the Court’s decision is vindicated. She further claims that this sort of investigation is necessary if the inter­

the intent of the legislature

117

preter wishes to base a decision on a full understanding of the context in which the statute was enacted. Controversy over the use of legislative history is most stark when there is no serious allegation that the history was strategically culled to create a false impression, and the historical analysis suggests that the legislature did not want the statute to be applied as the most natural reading of its language would seem to demand. The issue arises most often when the history is at odds with the ordinary meaning of a statute’s words but not with every possible reasonable interpretation of the statutory language, as happened in Chisom v. Roemer, discussed above.127 The Court there chose to look at extrinsic evidence of legislative intent, which ultimately trumped the most natural reading of the statute. The analysis presented here suggests that as a default position courts should indeed take legislative intent into account and then make a decision as to whether evidence of legislative intent that contradicts the ordinary understanding of the statute should be given priority over the language it­ self. If even plain language is by and large only evidence of communicative intent, albeit privileged evidence, then this is really the only sensible way to proceed. In this way, the tension between rule-of-law values and the in­ tent of the legislature dissolves. After all, intent is a rule-of-law value. As for legislative history as evidence of legislative intent, evidentiary issues will always remain: When is it the best evidence? When is it supportive but equivocal? How much weight a court should give to each type of evidence will depend on the facts of the case and the political orientation of the judge. That cannot be helped. In most cases, the language of the statute does lead to the conclusion that only one interpretation is possible. Consider United States v. Jefferson, decided by the U.S. Court of Appeals for the Seventh Circuit.128 A stat­ ute made it illegal to “sell or otherwise dispose of” a firearm to a known felon.129 The defendant was convicted of violating this statute after buying two guns and giving one to his brother, a convicted felon, for temporary safekeeping. The question was whether this transfer should count as “dis­ posing of” the gun under the statute. The language supports, but does not require, a finding that the defendant violated the statute. In affirming the conviction, the Seventh Circuit held: “Although we must keep in mind that ambiguity in criminal statutes should generally be resolved in favor of lenity, this maxim must of course yield to the paramount consideration— to follow congressional intent. Here, our best evidence of congressional intent is the legislative history cited above, which indicates that Congress

118

chapter 4

wanted to broaden the reach of the gun control statute to cover a wider range of firearms transfers.”130 The court in Jefferson probably overstated its position. Assume that the statute, instead of saying “dispose of,” said “transfers permanently” or some such thing. Assume further that the legislative history clearly indi­ cated that Congress intended the statute to reach far beyond the statutory words, to cases like this one. I know of no judge who would permit the intent of the legislature to criminalize conduct clearly outside the scope of any reasonable interpretation of the statute. In chapter 2, I called this the linguistic wall. But this does not mean that intent is irrelevant to statutory interpretation or that legislative history cannot be credible evidence of in­ tent. It means only that a host of substantive considerations play a role in determining the weight that each type of evidence should be given in each case. Obviously, there is likely to be disagreement over the weightings. When legal values conflict, choices always have to be made. Sometimes, these choices will be result oriented, as Radin suggested eighty years ago131 and as cases like Chisom strongly imply. Furthermore, some judges will ordinarily give more weight to one value than to the other, as a matter of political values and cognitive style. But, as I hope to have shown, it is very hard indeed to find a judge who can consistently claim to eschew intent as a value. All we can do when we interpret language is hope that we have absorbed whatever a speaker or writer has intended to convey. Even those who try to avoid intent talk do not always succeed and often use intent as the rationale for rules that they claim to be alternative approaches to interpretation. This does not mean, however, that the language of a statute does not count for much. Nor does it create an open season for those who would like to see courts create policy without regard to what the legislature wrote. For one thing, an overriding goal of the legislative process is to write a law of general application. It is no accident that each federal statute is given a “Public Law” number. For this reason and because, as textualists point out, laws must be enacted pursuant to a set of rigid constitutional proce­ dures, courts (and others) construing statutes should clearly give dispro­ portionate weight to the statute’s language. And they do. Nor does this mean that intent should automatically trump other val­ ues, such as fair notice, when these values conflict with one another. It may well be, for example, that the rule of lenity should be applied in some criminal cases regardless of whether the enacting legislature would have wanted a statute to be applied more broadly. But this does not make intent

the intent of the legislature

119

irrelevant to interpretation either. Rather, it makes intent one of a num­ ber of competing values, which may be given different weights in different cases. In chapter 5, we turn to some of the values that compete not only with intent but more generally with legislative primacy in the interpreta­ tion of statutes.

chapter five

Stability, Dynamism, and Other Values

T

hus far, my argument has focused on fidelity to the legislature. There is passionate debate about how to be faithful, but no question about whether to be faithful. In this chapter, I consider other values that contribute to decisions about statutory interpretation. These include fair notice; stability for its own sake; the contradictory value of responsiveness to changes in the legal and broader cultures; the purpose for the legislation; respect for the legislative process; coherence; adherence to constitutional norms; effective law enforcement; and political values. We will see that all these values compete with fidelity to the will of the legislature. In those relatively unusual cases in which fidelity and other values conflict, fidelity often—but not always—wins. Some values that come into play in statutory interpretation are not, at least in a narrow sense, strictly legal in nature. For example, in some criminal cases, it is difficult to avoid the conclusion that a court has decided to err on the side of assisting the law enforcement community when the language of the statute and other extrinsic information allow some latitude in interpretation. More generally, it simply does not appear to be an accident that when a statute is linguistically susceptible to more than one interpretation, conservative jurists most often resolve the indeterminacy in favor of a conservative outcome, and liberal jurists most often resolve the indeterminacy in favor of a liberal outcome. Many have made this observation, from the legal realists to Richard Posner in a recent book.1

stability, dynamism, and other values

121

Whether we find this fact disturbing should depend on the extent to which it determines outcomes. As we will see, the level of indeterminacy that justifies purely political outcomes does not occur in the typical case and is far more prevalent in higher courts. The presence of competing values in statutory interpretation should not be surprising to those who study constitutional interpretation. As Richard Primus observes, most constitutional scholars acknowledge that the original intent of the framers should have at least some influence on interpretation, with some minority of scholars thinking that intent is everything and another minority finding original intent irrelevant as a matter of principle.2 How other values—such as adherence to earlier judicial precedent and the desire for coherence in the structure of government— contribute depends in part on how the decision makers view the relationship between the framers and the people now being governed and how changes in society over time might affect constitutional questions. Primus likens the array of available constitutional arguments to a tool kit, with certain tools more appropriate than others in a given situation. Likewise, in the realm of statutory interpretation, the intent of the legislature might not be important to today’s decision maker when the statute was written in the nineteenth century and barely survives scrutiny under current constitutional doctrine. Similarly, decisions that would tend to render a particular body of law incoherent or that will thwart reasonable law enforcement efforts without promoting countervailing values such as fair notice might be unattractive to some decision makers even if they did serve to enforce the legislative will. I begin by looking at fair notice. I then consider the dynamics of stability and change and turn later to other values.

Fair Notice and the Marshall Test While fidelity to the legislature remains a principal value in statutory interpretation, it is not inviolate. Judges who disagree about such issues as the use of legislative history or the difference between plain and ordinary meaning all acknowledge that their ultimate goal is to effectuate the will of the legislature. Yet these same judges become uncomfortable effecting the legislative will when the law is not clear enough to have put the citizenry on notice. The Supreme Court has related the constitutional doctrine that a statute may be void for vagueness to the rule of lenity in criminal cases,

122

chapter 5

fitting both of these doctrines under the broader conceptualization of fair warning.3 To see just how strong a value fair notice is, let us return to Chief Justice John Marshall’s decision in United States v. Wiltberger, a case discussed in chapter 2. In that case, Marshall held that a law that made manslaughter committed on the high seas a federal crime did not apply to manslaughter committed on a vessel docked on a river in China. Provisions of the statute governing other crimes also referred to rivers, basins, and other bodies of water, but these were mysteriously left out of the manslaughter section. Marshall held that the rule of lenity should apply, even though Congress likely intended to write a statute broader in scope than the words conveyed: The probability is, that the legislature designed to punish all persons amenable to their laws, who should, in any place, aid and assist, procure, command, counsel, or advise, any person or persons to commit any murder or piracy punishable under the act. And such would have been the operation of the sentence had the words, “upon the land or the seas” been omitted. But the legislature has chosen to describe the place where the accessorial offence is to be committed, and has not referred to a description contained in any other part of the act. The words are, “upon the land or the seas.” The Court cannot reject this description.4

To Marshall, then, the value of providing fair notice to criminal defendants was important enough to override the intent of the legislature, and he was willing to say so bluntly. We can generalize from Marshall’s holding in Wiltberger by creating a test that will provide evidence of whether a court seriously values other considerations above the intent of the legislature. Let us call it the Marshall Test and define it as follows. For any given value proposed to replace legislative intent in the interpretation of a statute, that value supersedes legislative intent if a decision maker would accept the following proposition: “There are values more important than intent at stake here, so I am willing to undermine what I know to be the intent of the legislature in order to promote those competing values.”

Marshall put fair notice into this category. Justice Holmes kept it there in 1931, with his statement in McBoyle v. United States, the case discussed

stability, dynamism, and other values

123

in chapter 3 in which an airplane was not considered to be within the statute making it a federal crime to transport a stolen vehicle across state lines. Holmes remarked: “Although it is not likely that a criminal will carefully consider the text of the law before he murders or steals, it is reasonable that a fair warning should be given to the world in language that the common world will understand, of what the law intends to do if a certain line is passed. To make the warning fair, so far as possible the line should be clear.”5 Neither Holmes’s sentiment nor his statement of it, for that matter, are old hat. In 1995, the Supreme Court held that a judge had not obstructed justice when he lied to the FBI in a criminal investigation since the judge had no knowledge that his answers would be used in a grand jury.6 The Court did not mention the rule of lenity but quoted the Holmes passage verbatim. As far as I know, no other values stand up to the Marshall Test routinely, although many values become part of the mix of considerations that courts use in deciding cases, as William Eskridge has shown.7 When lenity can apply to avoid the lack of notice, courts at least sometimes answer the Marshall Test inquiries affirmatively. If the language of a statute does not criminalize particular conduct, courts are not likely to expand its scope, even if the enacting legislature mistakenly wrote the statute more narrowly than it intended. In contrast, other values are often used to bolster a result that reliance on legislative intent would have reached in the first place. These cases tell us very little about the role that such values play on their own. In other cases, legislative intent is adduced to support resort to other values, sometimes disingenuously, as in the claims in some of the baseball cases that Congress had intended the antitrust laws to exclude baseball.8 These interactions among various seemingly legitimate values are what make statutory interpretation complicated and are what cause judges to be accused of being activists.

Stability If one looks only at contemporary appellate opinions, it is easy enough to draw the conclusion that statutory interpretation often appears to be chaotic and arbitrary. However, it is a mistake to look only at the universe of cases over which there is maximum controversy and about which the law is maximally undetermined. If outcomes have precedential effect and

124

chapter 5

become part of the meaning of the statute being interpreted, the law should become more stable over time even if pockets of uncertainty and conflict remain. In this section, I will argue that it does just that. In chapter 2, we saw that a small set of recurrent problems— controversy over word meaning and difficulty determining a defendant’s state of mind—account for a large proportion of all published judicial decisions interpreting the federal bribery statute.9 Steven Shepard, who worked as my research assistant while a law student at Yale, has reviewed justice statistics to determine the number of annual prosecutions, the number of appeals, and the number of published appeals in cases brought under that statute.10 Kimberly Finneran, while a student at Brooklyn Law School, conducted additional analysis of the data. The total number of appeals reflects instances in which defense counsel concluded that there was a point worth making to an appellate court. The total number of published opinions reflects instances in which an appellate court, consistent with the court’s rules governing publication, determined that the case resolved a legal issue of sufficient importance to merit publication.11 Table 5.1 displays some of the results. The first thing to notice is that the number of prosecutions has increased, almost linearly—from 126 to 677— during the thirty-five-year period studied. In contrast, the total number of published appeals has decreased over time, but not in a smooth curve. The history of appeals under this statute generally appears to be a downward slope with periods of punctuated activity. Prosecutions with respect to a particular scandal (such as Abscam or Wedtech, both discussed in chapter 2) or in which prosecutors attempt to introduce a new theory (such as counting contractors performing government functions as “government officials”) generate increased activity for a period of time. But once those cases are decided, the issues they raised remain more or less resolved. Eventually, fewer convictions, whether as the result of a guilty plea or a trial, raise important legal issues. This is not to deny the presence of other factors within the legal system that might influence the number of appeals taken. For example, the federal Sentencing Guidelines took effect in 1987.12 Because the guidelines placed significant emphasis on the acceptance of responsibility as a factor in reducing sentences,13 fewer defendants in federal cases have chosen to go to trial, reducing the grounds for appeal throughout the system. Nonetheless, the Sentencing Guidelines cannot explain the downward trend in the rate of appeal in the four years preceding their adoption, nor the continuing downward trend after they had been firmly established. Moreover,

stability, dynamism, and other values

125

table 5.1.  Bribery Prosecutions and Appeals in Selected Years Year 1971 1975 1979 1983 1987 1991 1995 1999 2003 2005

Prosecutions

Published appeals

Total appeals

126 134 198 195 342 387 403 486 650 677

11 5 1 5 1 2 2 1 4 2

31 51 19 80 32 28 50 37 25 19

while the guidelines increased the rate of plea bargaining and corresponding reduction in the number of trials, the trend had been in that direction for some time.14 Thus, the adoption of the guidelines does not appear to explain the decrease in the rate of appeals taken by defendants in federal bribery cases. Because of the increase in the number of prosecutions over time, the absolute number of appeals and published appellate decisions is not as meaningful as the number of appeals and published opinions as a percentage of total prosecutions. Figure 5.1 presents this information and shows that both the total appeals and the total published appellate decisions have decreased as a percentage of total prosecutions.15 To the extent that this study of the bribery statute is indicative, it suggests that the combination of statutory language and judicial decision making leads to stability over time. This is not to say that new theories will not arise from time to time or that the resolution of recurrent situations by appellate courts is dispositive of all new, unforeseen circumstances that may arise later and generate additional disagreement about a statute’s application. Nonetheless, the story of the bribery statute is a positive one for those concerned about the development of certainty and stability. Not all would agree with this assessment, however. An interesting study by Stefanie Lindquist and Frank Cross draws somewhat contrary conclusions.16 Their study examines the relationship between adherence to precedent in statutory cases, on the one hand, and reliance on ideology, on the other. The statute they studied is 42 U.S.C. § 1983, which gives a cause of action to those whose federal rights have been violated “under color of state law.” Lindquist and Cross set out to determine whether Ronald Dworkin’s famous chain-novel metaphor describes the reality when large

126

chapter 5

figure 5.1.  Total appeals as a percentage of total bribery prosecutions and published appeals as a percentage of total bribery prosecutions

numbers of cases are examined. Dworkin likens statutory interpretation to the writing of a chain novel. For Dworkin, interpreting a law resembles what happens when “a group of novelists writes a novel seriatim; each novelist in the chain interprets the chapters he has been given in order to write a new chapter, which is then added to what the next novelist receives, and so on.” The writers “aim jointly to create, so far as they can, a single unified novel that is the best it can be.”17 Implied in this analysis is an important role for precedent, for earlier decisions are part of the history that the author of the latest chapter will consider. Focusing on cases decided by U.S. courts of appeals between 1961 and 1990, Lindquist and Cross found that ideology plays a significant role in cases of first impression. Conservative judges make conservative decisions; liberal judges, liberal decisions. This should not be surprising: when there is no precedent to take into account, judges’ political orientation plays a more prominent role. What is especially interesting in Lindquist and Cross’s study is what happens over time. In the period immediately following interpretive deci-

stability, dynamism, and other values

127

sions, courts of appeals generally respect precedent. But over time, adherence to earlier precedent dissipates and ideology begins to play a larger role again. They explain their results: “As more time passes and more precedents are decided, however, the proliferation of available prior decisions in turn expands judges’ discretion to decide cases in accordance with their attitudes simply because they have more precedents from which to choose. The influence of precedent could thus be conceptualized as quadratic or curvilinear over time.”18 But this dynamic does not describe the cases decided under the bribery statute examined earlier in this chapter. Perhaps one difference is that the bribery statute, although it involves the prosecution of political figures, conjures up less ideology, for the most part. Once the Supreme Court has decided, for example, that a contractor performing a governmental function should count as a “public official” under the statute, lower-court judges will have little motivation to come up with unnaturally narrow understandings of “contractor.” Not enough is at stake. In contrast, the proper scope of the civil rights statute that Lindquist and Cross studied has been the subject of a great deal of controversy over the years, some of which has played out in linguistic games.19 It should not be surprising that politics plays a greater role when a statute is both broadly worded and controversial. Part of the difference between my results and theirs may also reflect an important issue of methodology. Lindquist and Cross show, somewhat ironically, that extensive precedent sometimes frees judges to rely upon ideology. Once enough appellate decisions seemingly go in both directions, it becomes easier for a judge performing his job earnestly to choose some precedents as persuasive and to distinguish others as less relevant to reach a result without thwarting traditional rule-of-law values. Missing from their study, however, is information about how frequently these decisions occur as a function of the number of cases brought in the lower courts and the total number of appeals taken, whether published or not. It may be the case, for example, that the percentage of section 1983 cases resulting in published appeals is quite low, but that those cases that do make it that far through the system are susceptible to multiple analyses based on current precedent and are therefore ripe for ideological considerations. If this is so, it would serve to confirm the importance of precedent in statutory decisions. To determine the relationship between prosecutions and appeals over time, Steven Shepard (who assisted me in my study of the bribery statute)

128

chapter 5

conducted his own study of four federal statutes: a statute making it a crime to engage in a “continuing criminal enterprise”; the Hobbs Act, which makes it a crime to “obstruct, delay, or affect commerce” by means of “robbery or extortion”; the federal credit card fraud statute; and the federal money-laundering statute.20 Shepard conducted analyses similar to the one presented above in which the number of published appeals was analyzed in terms of the number of prosecutions for each year. Shepard’s study differed from the bribery study reported here in two respects. First, the bribery statute was enacted in 1962, but data rich enough to perform these analyses exist only beginning in 1971. The statutes that Shepard studied, in contrast, were not enacted in their current form until after 1971, so he was able to track trends from the beginning of the statutes’ histories. Second, Shepard looked at the relationship between published appeals and total convictions rather than total prosecutions. This difference should not be of great significance. Shepard’s results for each of the four laws match the results concerning the bribery statute reported earlier in this chapter. Prosecutions and convictions increase over time, but the ratio of published appeals to convictions decreases over time. The federal money-laundering statute, for example, which was enacted in 1986, saw a steady rise in convictions over its then twenty-year existence.21 However, the percentage of convictions resulting in published appellate opinions reached a peak of about 45 percent in 1992 and decreased steadily to about 10 percent in 2005. As the most recurrent issues are resolved, there is simply less for appellate courts to contribute. There remains ample opportunity for thoughtful empirical work on how statutory language and judicial decision making interact to bring certainty to the law. For example, one might expect less stability with respect to broadly worded statutes, such as the antitrust laws and the civil rights laws. Nonetheless, preliminary results certainly suggest that the legislature and the courts, working overtime as a team, tend to decrease uncertainty in the law.

Stability as a Value It is hard to imagine that a court today would not consider an airplane to be a vehicle for purposes of a statute outlawing the interstate movement of stolen vehicles. Yet that is exactly what the Supreme Court did in 1931

stability, dynamism, and other values

129

in McBoyle v. United States, as we saw.22 A court making the same ruling today would surely subject itself to criticism for elevating formalism over common sense. There may be reasons, however, to value consistency by adhering to decisions made long ago even if they do not seem to be the best ones in today’s world and, for that matter, even if in retrospect they do not today seem to have been good decisions in the first place. Law students learn this perspective on statutory interpretation by reading Flood v. Kuhn, a 1972 Supreme Court case that reaffirmed the Court’s 1922 decision that the antitrust laws do not apply to professional baseball because baseball is not a business engaged in interstate commerce.23 The original decision was probably ridiculous then and certainly is unanchored to reality now. Even if you know nothing about sports, you can confirm the current reality with a quick visit to Major League Baseball’s Web page.24 Nonetheless, courts stick to their earlier interpretations of statutes for the sake of consistency as a value in itself and to protect reliance interests. For example, in 1953, the Supreme Court reaffirmed its earlier antitrust decision on just these grounds: “Congress has had the ruling under consideration but has not seen fit to bring such business under these laws by legislation having prospective effect. The business has thus been left for thirty years to develop, on the understanding that it was not subject to existing antitrust legislation. The present cases ask us to overrule the prior decision and, with retrospective effect, hold the legislation applicable. We think that if there are evils in this field which now warrant application to it of the antitrust laws it should be by legislation.”25 In the meantime, there have been numerous unsuccessful efforts over the years to amend the antitrust laws legislatively to override the 1922 decision and its progeny, but to no avail, despite the fact that these laws have been held to apply to other professional sports, including football and basketball. By the time Flood v. Kuhn made its way to the Supreme Court, the only legitimate reason for continuing the fiction that baseball is not a business that affects interstate commerce was the ironic fact that so much interstate commerce had developed in reliance on the holding. Although the majority opinion clung to earlier holdings that Congress never intended the antitrust laws to apply to professional baseball, Chief Justice Warren Burger jettisoned that insupportable claim in his concurrence: [L]ike mr. justice douglas, I have grave reservations as to the correctness of Toolson v. New York Yankees, Inc.; as he notes in his dissent, he joined that holding but has “lived to regret it.” The error, if such it be, is one on which the

130

chapter 5

affairs of a great many people have rested for a long time. Courts are not the forum in which this tangled web ought to be unsnarled. I agree with mr. justice douglas that congressional inaction is not a solid base, but the least undesirable course now is to let the matter rest with Congress; it is time the Congress acted to solve this problem.26

Whether one agrees with the holding in this case, it surely shows that values other than fidelity to the legislature have the potential to drive the interpretation of statutes. One of these values is stability. Common-law courts apply the principle of stare decisis to their decisions about statutory interpretation just as they apply it to their common-law and constitutional decisions. As we have just seen, however, taking judicial precedent as binding in subsequent statutory cases has the effect of separating the interpretation of the statute from the legislative process, thus aggrandizing the role of the courts. And all judges, textualist or contextualist, appear content to leave this power in place. The result is inevitably a sharing of legislative power between the two branches: the legislature writes laws that, at best, define rights and obligations in relatively clear cases, and the courts construe the statute when the statute is not clear, thus creating binding glosses on the statute. The legislature then gets the last word: it can override judicial decisions when they go too far afield. But resource limitations restrain the legislature from overriding often, leaving judicial decisions in place as part of the meaning of statutory law. Over time, reliance concerns often trump second thoughts about either legislative primacy or policy, leaving judge-made decisions in place for indefinite periods of time, both good decisions and bad.

The Dynamics of Change over Time At the same time that consistency and stability drive statutory interpretation toward respect for precedent, flexibility and responsiveness to changing times tug in the opposite direction. William Eskridge’s book Dynamic Statutory Interpretation provides strong arguments for the legitimacy of changes in interpretation over time, even if the new interpretations are inconsistent with the intent of the enacting legislature.27 Eskridge argues that even when linguistic change has not occurred, values reflected in judicial practice over time may take priority over the en-

stability, dynamism, and other values

131

acting legislature’s intent.28 Eskridge’s principal examples involve the civil rights laws, and especially cases concerning affirmative action. Although the Congress of 1964 that enacted the original federal Civil Rights Act could not be said, as a body, to favor affirmative action, the statute’s early enforcement history, Eskridge argues, made it appropriate to accept affirmative action as necessary to enforce the civil rights laws without creating perverse incentives that would actually thwart the integration of nonwhite workers into the workforce. Eskridge is not alone. As we saw in chapter 1, Guido Calabresi has written an important book suggesting that courts be given the right to update old statutes to avoid obsolescence.29 And Dworkin’s chain-novel metaphor discussed earlier as an engine of stability necessarily involves change as well.30 The very fact that there is a next chapter in the chain novel means that developments in the law should be expected. The novel must balance fidelity to what happened in earlier chapters against new developments in plot and character, which are understood in the context of the prior developments. Although the interpretation of statutes clearly does change over time in some instances, I do not consider dynamism, whether in Eskridge’s or Dworkin’s sense, to be a “theory” of statutory interpretation. Rather, I believe dynamism occurs when certain values are important enough to the statutory interpreter that they trump legislative primacy. The most dramatic instances are the ones about which Eskridge writes: times change, and old understandings become untenable, even if it is crystal clear that the enacting legislature would not have wanted such a result. The situation becomes a bit easier when the language of the statute, as currently understood, would allow a result contrary to one contemplated by the enacting legislature. In those cases, fair notice to the parties is less of an issue, and apparent conflicts between intent and language of the kind discussed in chapter 4 do not arise. But these are not the only ways in which statutory interpretation evolves over time. Statutes change in meaning when courts look to the language of their own earlier interpretive opinions rather than to the language of the statute itself. Peter Tiersma writes about this phenomenon, which he calls “the textualization of precedent.”31 In addition, many statutes are written with dynamic interpretation in mind. Assault weapon bans, fraud statutes, and many others actually contemplate that courts or agencies will respond to innovations in bad conduct, updating the law within bounds

132

chapter 5

that the statute spells out. Law enforcement concerns can lead to evolving interpretations. And delegation to administrative agencies of authority to regulate and enforce a statute’s goals leads both to flexibility and to various dynamics that depend upon the institutional setting in which the statute is enforced. When Times Change but Laws Do Not Capable of trumping the intent of the legislature are arguments that a statute’s meaning has changed over time, or that its enforcement over time has resulted in a legal landscape worthy of preservation even if that means interpreting a statute in a way that makes sense in context but would not have been acceptable to the enacting legislature. Below I describe a case that illustrates the kinds of change over time that Eskridge describes. In Gay & Lesbian Advocates & Defenders v. Attorney General, the Supreme Judicial Court of Massachusetts was asked by advocacy groups to declare that state’s sodomy statute unconstitutional.32 The court declined to do so, holding that the statute should be interpreted to ban only acts that are either public or nonconsensual. None of the parties in the case before the court claimed to engage in such conduct. The decision was in harmony with an earlier decision of the same court, Commonwealth v. Balthazar, that refused to impose criminal liability under a statute banning “an unnatural and lascivious act with another person” to private, consensual behavior.33 What counts as an unnatural and lascivious act can surely change over time, making it at least possible to argue that the legislature left the statute with room to develop as times change. But there is little likelihood that the enacting legislature in 1887 would have endorsed any decision to legalize sodomy, even if both consensual and private. The decision’s justification must lie in aspects of today’s culture in which enforcement is sought. In other words, cultural change over the course of a century has trumped intent, leading to a prototype of Eskridge’s dynamic model. By the same token, the argument that a nineteenth-century legislature had in mind that lewd and lascivious conduct is a flexible concept subject to change as social values develop has no basis in fact either. It is a legal fiction licensed by the gratuitous linguistic fact that, in retrospect, values really have changed over the past century. The value of legislative primacy is so important in statutory interpretation that courts are sometimes willing to attribute to the legislature a flexibility that it cannot possibly have had.

stability, dynamism, and other values

133

Laws Written for a Changing World Sometimes, however, that flexibility is no fiction. Laws are written intentionally to permit subsequent interpreters to deal with changing circumstances. Consider the federal mail fraud statute. It criminalizes mailing “any matter or thing” or causing “any matter or thing” to be mailed for the purpose of “executing” a “scheme or artifice to defraud.”34 The very purpose of this statute is to remain responsive to new, creative ways of committing fraud. It is the prime example of what Dan Kahan calls a federal common law of crimes.35 Such laws constitute legislative delegations to the courts, whose responsibility will be to determine what conduct should be criminalized, within broadly drawn guidelines. Let us look at the leading mail fraud case, Schmuck v. United States. Wayne T. Schmuck had been involved in a fraudulent scheme: he purchased automobiles, set back the odometers, and resold them to car dealers, who would in turn sell the cars. However, only after Schmuck had reaped the benefit of his fraudulent scheme did a mailing occur. As Justice Harry Blackmun explained: “To complete the resale of each automobile, the dealer who purchased it from Schmuck would submit a title-application form to the Wisconsin Department of Transportation on behalf of his retail customer. The receipt of a Wisconsin title was a prerequisite for completing the resale; without it, the dealer could not transfer title to the customer and the customer could not obtain Wisconsin tags. The submission of the titleapplication form supplied the mailing element of each of the alleged mail frauds.”36 The question was whether this scheme met the statute’s requirement that the mailing be in furtherance of a fraudulent scheme. Schmuck argued that the fraud was complete by the time the mailing occurred. The Supreme Court disagreed and upheld the conviction, reasoning that “[a] rational jury could have concluded that the success of Schmuck’s venture depended upon his continued harmonious relations with, and good reputation among retail dealers, which in turn required the smooth flow of cars from the dealers to their Wisconsin customers.”37 Justice Scalia dissented: “[I]t is mail fraud, not mail and fraud, that incurs liability. This federal statute is not violated by a fraudulent scheme in which, at some point, a mailing happens to occur—nor even by one in which a mailing predictably and necessarily occurs. The mailing must be in furtherance of the fraud.”38 Schmuck is an instance of expansive interpretation of a criminal statute. Not only was lenity not a consideration, but the mail fraud statute was interpreted as broadly as “a rational jury” would take it. Courts do

134

chapter 5

not ordinarily leave such decisions in the jury’s hands, a fact to which we will return in chapter 7. Yet the Supreme Court’s interpretation recognizes that at least some statutes are designed to be construed flexibly, as long as there is fair notice. Surely, Mr. Schmuck could not argue that he thought his behavior was legal, a fact that apparently was enough to permit a majority of justices to stretch the language and, for that matter, the traditional division of labor between judge and jury. Lawmakers do not have an easy time writing laws that both meet dueprocess concerns and are sufficiently flexible to block circumvention. Moreover, it is not simple to draw the line between legitimate avoidance of an undesired obligation and evasion of a duty. Tax shelters, for example, reflect a battle between legislators and skilled tax lawyers over opportunities to circumvent the tax code by creating transactions that do not appear to be taxable.39 Legislation banning assault weapons shares this dynamic.40 If the statute is not precise enough, it will be stricken as unconstitutionally vague. But if it is too precise, it will leave opportunities to engineer new weapons that perform just as the banned weapons do but that do not come within the statute’s bounds. Since its repeal in 2004, there has been no federal ban on assault weapons. But many states have such bans, and their different approaches illustrate the difficulty in legislating against efforts to engineer around a statute. Consider, for example, California’s statute, which contains an extensive list of banned weapons (including weapons with “minor differences” from those specifically listed) and a separate list of definitional features that further characterize illegal assault weapons.41 Among them are the following: (1) A semiautomatic, centerfire rifle that has the capacity to accept a detachable magazine and any one of the following: (A) A pistol grip that protrudes conspicuously beneath the action of the weapon. (B) A thumbhole stock. (C) A folding or telescoping stock. (D) A grenade launcher or flare launcher. (E) A flash suppressor. (F) A forward pistol grip.42

There are many more.

stability, dynamism, and other values

135

An earlier version of the statute—repealed by the California legislature in 2006—permitted the attorney general of the state to update the list of banned weapons as new ones became known to law enforcement officials. This, in turn, allowed a brief period of safe harbor between the discovery of the weapon and the completion of the administrative procedure banning the new weapon.43 Since the repeal, the state must rely on the list of banned weapons and the definitions alone. Quite clearly, the task of keeping this statute dynamic enough to continually ban all and only assault weapons is a difficult one. Change through Precedent in “Common-Law Statutes” Lawyers often have the experience of recognizing, prior to an appellate argument, that there are enough cases supporting both positions to permit the judges to rule in either direction. As this happens, statutory reasoning moves toward becoming common-law reasoning, at least with respect to cases that stray from prototypical situations. The legal question no longer arises solely from the language of the statute. Rather, the question arises from the language of the statute and a great deal of analogical reasoning, in which the most heated inquiry is whether the case before the court is more like the ones that had earlier been decided in one way than in the other. Now let us look at an extreme version of this dynamic that results not from the proliferation of precedent but rather from precedent on an important statutory question becoming so central that future decisions turn toward the interpretation of the earlier precedent rather than the statute itself. This happens most vividly when precedent becomes “textualized,” to use Tiersma’s term.44 The textualization of precedent is more a phenomenon of common law and constitutional argument than of statutory interpretation. Nonetheless, it also occurs frequently in the interpretation of what Judge Posner has called “common-law statutes,” laws that either codify a developing area of common law or are written in such broad terms as to invite further development by the courts.45 Below I describe two examples from statutes governing intellectual property. the nonobviousness requirement in patent law.  Consider the following difficult question of patent law: when is an invention or improvement to an invention creative enough to deserve patent protection? In an important nineteenth-century case, the Supreme Court held, quite sensibly,

136

chapter 5

that the substitution of a porcelain doorknob for the standard wood doorknob is not good enough.46 After almost a century of struggling with this problem, in 1941 the Court held: “[T]he new device . . . must reveal the flash of creative genius.”47 This expression became textualized, not as a matter of statutory interpretation, but under the common law of patents, which then controlled. It is not difficult to find courts of appeals denying patent protection because a new invention fails to reveal the flash of creative genius.48 Possibly in response to this court-imposed standard, which made patent protection quite difficult to obtain, Congress enacted the Patent Act of 1952, which contains the following requirement: “A patent may not be obtained . . . if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art.”49 This is referred to as the “nonobviousness” requirement. The Court first interpreted this language fifteen years after its passage, in Graham v. John Deere Co.50 There, the Court considered several patents filed by William Graham and his small Texas company, Graham Plow. Graham’s patents were all variations of a “spring clamp” that would, when attached to the shank of a chisel plow, permit the shank to better withstand its impact with rocks buried in the soil. In 1950, Graham obtained his first patent on a spring clamp. In 1953, he obtained a second patent on a slightly refined version of the clamp. When the John Deere Company began manufacturing a spring clamp very similar to Graham’s second, 1953 patent, Graham sued for an injunction. Deere defended by claiming that Graham’s 1953 patent was invalid under section 103, because the slight modifications on Graham’s 1950 design would have been “obvious” to any “person having ordinary skill in the art.”51 In holding that Graham’s 1953 patent was invalid because its modification of Graham’s 1950 patent was “obvious,” the Court noted that section 103 “lends itself to several basic factual inquiries.” The words of the statute call for courts to determine: (1) the “scope and content of the prior art,” that is, the existing patents and other knowledge available at the time of the patent; (2) “differences between the prior art and the claims [of the patent] at issue”; and (3) the “level of ordinary skill in the pertinent area.”52 These three factual questions are all set forth in the statute itself. The Court went on, however, to note several “secondary considerations” that might “have relevancy” as “indicia of . . . nonobviousness,” namely:

stability, dynamism, and other values

137

(a) “commercial success” of the patented invention; (b) “long felt but unsolved needs” that were met by the invention; and (c) “failure of others” to meet those needs through a similar invention. Neither (a), (b), nor (c) appears in the text of section 103; for support for these “secondary considerations,” the Court cited only a 1964 student law review note.53 The Federal Circuit, however, has come to treat these secondary considerations as a required fourth element in section 103 analysis.54 For example, in Greenwood v. Hattori Seiko Co., the Federal Circuit stated: “[C]ertain factual predicates are required before the legal conclusion of obviousness or nonobviousness can be reached. The underlying factual determinations to be made are (1) the scope and content of the prior art, (2) the differences between the claimed invention and the prior art, (3) the level of ordinary skill in the art, and (4) objective evidence of nonobviousness, such as commercial success, long-felt but unsolved need, failure of others, copying, and unexpected results.”55 There is now a lively debate in the academic literature on patent law about which of these secondary considerations is most important. 56 And courts continue to quote this language as part of the law. A Lexis search reveals that federal courts have used the expression “long-felt but unsolved need” more than seven hundred times.57 Their importance, however, derives from the textualization of a court decision interpreting a statute—not directly from the language of the statute itself. may décor constitute a false advertisement?  Section 43(a) of the 1946 Lanham Act creates a civil cause of action for anyone harmed by another’s false advertising: Any person who, on or in connection with any goods or services, or any container for goods, uses in commerce any word, term, name, symbol, or device, or any combination thereof, or any false designation of origin, false or misleading description of fact, or false or misleading representation of fact, which . . . [i]s likely to cause confusion, or to cause mistake, or to deceive as to the affiliation, connection, or association of such person with another person, or as to the origin, sponsorship, or approval of his or her goods, services or commercial activities by another person . . . shall be liable in a civil action.58

Though nothing in section 43(a) expressly refers to trademarks (which are covered elsewhere in the Lanham Act), courts and commentators have come to refer to section 43(a) as providing “federal common-law”

138

chapter 5

protection for trademarks that have not yet been officially registered with the Patent and Trademark Office.59 Consider an important Supreme Court case interpreting section 43(a), Two Pesos, Inc. v. Taco Cabana, Inc.60 The parties to the case were rival chains of Mexican restaurants in Texas. The older chain, Taco Cabana, opened its doors in 1978 with a distinctive décor: various colors, artifacts, and painted designs that, taken together, created a “festive eating atmosphere.”61 The newer chain, Two Pesos, allegedly copied Taco Cabana’s décor. Incensed at the loss of business to the upstart Two Pesos chain, Taco Cabana sued under section 43(a). The words of section 43(a) would lead a normal reader to raise four factual questions: 1. Was Two Pesos’s interior a “word, term, name, symbol, or device, or any combination thereof”? 2. Was the interior “use[d] in commerce”? 3. Was the interior used “on or in connection with” Two Pesos’s “goods or services, or any container for goods”? 4. Was Two Pesos’s interior “likely to cause confusion, or to cause mistake, or to deceive . . . as to the origin, sponsorship, or approval of” Taco Cabana’s “goods, services or commercial activities”?

Now consider the five questions that the trial court actually presented for the jury to answer: 1. Did Taco Cabana have a “trade dress”? (The jury answered, “Yes.”) 2. Was Taco Cabana’s trade dress, taken as a whole, nonfunctional? (“Yes,” again.) 3. Was Taco Cabana’s trade dress inherently distinctive? (“Yes,” the jury said.) 4. Had Taco Cabana’s trade dress acquired a secondary meaning in the Texas market? (“No,” the jury found.) 5. Finally, was there a “likelihood that customers might associate or confuse a Taco Cabana restaurant with a Two Pesos restaurant”?62 (“Yes.”)

Where do all these new words come from? The term “trade dress” is shorthand for “the total image of the business”63 —in this case, the restaurant’s décor. But the terms “nonfunctional,” “inherently distinctive” and “secondary meaning” come not from the statute itself but from the textualization of past court precedents in the field of trademark law. Only the

stability, dynamism, and other values

139

trial court’s final question was drawn from the actual text of section 43(a). When Taco Cabana took its loss on question 4 to the Supreme Court, Justice White framed the question using that vocabulary: “The issue in this case is whether the trade dress of a restaurant may be protected under § 43(a) . . . based on a finding of inherent distinctiveness, without proof that the trade dress has secondary meaning.”64 Let us now examine how this question came to supplant the questions posed by the actual text of section 43(a). This transportation took place in two steps. The first step was to decide that “the general principles qualifying a [trade]mark for registration under § 2 of the Lanham Act are for the most part applicable in determining whether an unregistered mark is entitled to protection under § 43(a).”65 In other words, the Court decided that section 43(a) would protect Taco Cabana’s trade dress only if that dress was sufficiently distinctive that it could have been registered as a trademark despite the limits on registration set forth in section 2 of the Lanham Act.66 This first step was not motivated by the plain text of section 43(a), because the words “mark” and “trademark” are nowhere to be found in section 43(a). The first step thus shifted the question from an inquiry into the match between the facts of the case and the language of the statute to whether section 2 would have barred that registration if Taco Cabana had attempted to register its décor as a trademark. Section 2(e) forbids registration of marks that are either “merely descriptive” or “functional.” Section 2(f), however, carves out an exception: even “merely descriptive” marks can still be registered if they have become “distinctive of the applicant’s goods in commerce.” Even “distinctive” marks, however, may still be canceled if they become “generic” or are “abandoned” by their owners.67 The Court’s second step was to follow the hugely influential 1976 opinion of Judge Henry Friendly in Abercrombie & Fitch Co. v. Hunting World, Inc., in which he distilled the standards set out in section 2 and the holdings of cases decided under prior trademark statutes into five categories, complete with helpful examples to make the standards seem concrete. In increasing order of distinctiveness, Judge Friendly explained that marks could be (1) generic (e.g., the word “ivory” used to describe a product made from elephants’ tusks); (2) descriptive (e.g., the term “deep bowl” used to describe a spoon); (3) suggestive (“Tide” laundry detergent is a modern-day example); (4) arbitrary (“Camel” cigarettes is a modern example); or (5) fanciful (“Kodak” cameras).68

140

chapter 5

Though merely descriptive marks cannot normally be registered, they become registrable when buyers recognize the mark’s “secondary meaning”—that is, when buyers recognize that the product comes from a single source. This occurs, for example, when consumers recognize a bag labeled “Tender Vittles” as containing a unique cat food. Returning to the text of the statute, a “secondary meaning” saves a merely descriptive mark from cancellation under section 2(e) by causing it to become “distinctive” under section 2(f).69 Having taken these two steps away from the actual text of section 43(a), the Court had opened the door to just such a confusing jury verdict as the one rendered in favor of Taco Cabana: while the jury found the décor to be “inherently distinctive,” the jury also found that the décor did not have a “secondary meaning” in the minds of Texas customers. Two Pesos argued to the Court that this verdict was incoherent: Taco Cabana’s décor could only be “inherently distinctive” if it had taken on a “secondary meaning.” Lack of “secondary meaning” logically entailed a lack of inherent distinctiveness. The Court disagreed and upheld the jury’s verdict. If trade dress is “inherently distinctive,” the Court held, there is no need to show that it also has “secondary meaning” in the minds of consumers. The courts continue to rely heavily on the language of Judge Friendly’s taxonomy, now deeply entrenched in trademark law, and continue to develop a common law of trademarks by further parsing their own language rather than the legislature’s.70 The result of this textualization of precedent is that the actual language of section 43(a) offers very little guidance to the owner of an unregistered trade dress. Statutory Inflation Let us now consider another way in which laws change over time.71 Many laws, such as the Food, Drug, and Cosmetics Act, are remedial in nature. They promote welfare by setting standards of conduct for industry, the health professions, and so on. A number of these statutes, in addition to permitting various civil remedies, have provisions that make it a crime to violate the law willfully. In fact, such laws are commonplace and include the antitrust laws, the Copyright Act, various environmental laws, and the securities laws.72 In general, remedial laws are interpreted broadly, and criminal laws are interpreted narrowly according to the rule of lenity. A dilemma arises when a single statute is both remedial and criminal. One possibility is that the statute will be interpreted narrowly in criminal and

stability, dynamism, and other values

141

civil cases alike on the theory that the rule of lenity is needed even in civil cases to ensure the narrow interpretation of the statute when criminal cases are later brought. This has happened a number of times over the past two decades, although it is not the norm.73 Courts do not routinely interpret statutes narrowly in civil cases to avoid the broad interpretation of criminal statutes in later cases. Rather, they often interpret these statutes expansively in civil cases according to the perceived intent of the legislature and later incorporate these broad rulings into criminal prosecutions. I call this dynamic statutory inflation. Statutory inflation is most likely to occur when an administrative agency is charged with the civil enforcement of a statute, and a bureau of the Department of Justice is empowered to bring criminal cases. In this institutional setting, government agencies have the greatest incentive to try to convince judges to construe statutes expansively. Sometimes they succeed. Perhaps the most dramatic example of statutory inflation lies in the area of securities law. Rule 10b-5, the principal antifraud regulation promulgated under the securities laws, says nothing about insider trading.74 The rule addresses deceit and fraud generally. Nonetheless, the enforcement division of the Securities and Exchange Commission (SEC) began charging corporate directors who traded on inside information with Rule 10b-5 violations in civil enforcement proceedings.75 In 1968, the Second Circuit affirmed an SEC insider-trading decision in Securities and Exchange Commission v. Texas Gulf Sulphur Co. In so doing, it held that “the securities laws should be interpreted as an expansion of the common law both to effectuate the broad remedial design of Congress . . . and to insure uniformity of enforcement.”76 Subsequently, civil plaintiffs, using the implied private right of action under Rule 10b-5, began bringing successful actions against those who disseminated and traded on inside information. In Shapiro v. Merrill Lynch, Inc., the Second Circuit held that its reasoning in Texas Gulf Sulphur also applied in private lawsuits. For support, it quoted the Supreme Court’s admonition that the securities laws be construed broadly to effectuate their remedial purpose. It also expanded insider-trading doctrine further by not requiring that the plaintiffs prove that the actual shares they purchased on the open market were the same shares that the defendants sold. Rather, the defendants had breached a duty to “all persons who during the same period purchased Douglas stock in the open market without knowledge of the material inside information which was in the possession of defendants.”77

142

chapter 5

It was not until 1980 that the Supreme Court approved prosecution of insider trading as a criminal violation of Rule 10b-5. In Chiarella v. United States, the Court reversed the conviction of an employee of a financial printer who bought stock based on information he had read in the course of printing announcements of corporate takeover bids that had not yet been made public.78 Though the Court held that the employee owed no duty, it embraced the position that insider trading could constitute a violation of Rule 10b-5 and that prosecution for such violators was appropriate. The Court relied on earlier administrative decisions and circuit court decisions in civil cases, including Texas Gulf Sulphur. It did not mention the fact that Chiarella was a criminal case, whereas all the earlier ones were civil, and it certainly did not consider the rule of lenity.79 Thus, criminal application of Rule 10b-5 in the context of insider trading grew out of the broad interpretation of the rule in civil cases, in part as the result of aggressive administrative enforcement actions brought earlier by the SEC.80 As noted, the inflation of criminal liability without further legislative action typically arises from institutional settings in which an agency has an incentive to drive expansive readings in civil litigation, which can then feed subsequent criminal prosecutions.

Other Values in Statutory Interpretation The Purpose of the Legislation Chapter 4 addressed the question of intent. Sometimes, intent is contrasted with purpose. The practical difference between the two is not always clear, because the legislature’s intent is to write a law that fulfills the legislative purpose. To the extent they differ, “purpose” typically refers to more abstract goals that the legislature was trying to accomplish when it enacted a law. “Intent,” when contrasted with “purpose,” is generally used to describe a more specific state of mind toward a situation at hand. Consider once again Church of the Holy Trinity v. United States, discussed in chapter 3. A law banning the transportation of labor of any kind had been enacted shortly after the Irish potato famine. The law was almost certainly intended to stifle an influx of cheap labor that would compete with American workers for jobs and depress wages by increasing supply. The Supreme Court recognized this purpose and construed the law narrowly, in keeping with the legislative goals: “It was never suggested that we had in this country a surplus of brain toilers, and, least of all, that the market

stability, dynamism, and other values

143

for the services of Christian ministers was depressed by foreign competition.”81 It could not have conceivably advanced the purpose of this statute to have held that the law banned a church from paying to bring its next minister to New York from England, and the Court agreed unanimously that it would not so hold. In civil-law countries, purposive interpretation is considered the norm,82 although courts are no more consistent elsewhere than they are in this country in approaching statutory cases.83 Here, it is also a constant refrain. A combined search of state and federal court cases for the use of “purpose” within five words of “Congress or legislature” yielded more than three thousand hits for the first five years of the millennium.84 For that matter, Supreme Court justices talk about congressional purpose routinely. For example, in Begay v. United States, decided in 2008, the Court construed a provision of the Armed Career Criminal Act, which calls for a minimum sentence of fifteen years in prison for armed offenders who had earlier been convicted of three or more violent felonies.85 The term “violent felony” is defined in relevant part as a felony that is “burglary, arson, or extortion, involves use of explosives, or otherwise involves conduct that presents a serious potential risk of physical injury to another.”86 Begay, who was convicted of illegally possessing a firearm, had earlier been convicted a dozen times for driving under the influence of alcohol in New Mexico. The question was whether driving under the influence should count as a “violent felony” under the statutory definition since it presents a serious potential risk of physical injury, as the statutory definition requires. In a six-to-three ruling, the Court held that it should not and invoked the purpose of the statute to justify its holding. Criticizing the three dissenting justices for ignoring the four named offenses, the majority used the principle of ejusdem generis (without so naming it) and limited the statute to the sorts of crimes that were listed in order to “effectuate Congress’ purpose to punish only a particular subset of offender, namely career criminals.”87 Judge Posner espouses this purposive approach to statutory interpretation as in keeping with legal pragmatism.88 The goal of pragmatism is for decision makers to focus on the consequences of their decisions rather than on formalistic considerations, such as minute nuances in language. As Posner notes, purposivism in statutory interpretation is very inconsistent with the kind of rigid textualism discussed in chapters 3 and 4. To take one of Posner’s examples, read literally, a law criminalizing the possession of child pornography applies to prosecutors and court personnel who have

144

chapter 5

seized such material from a wrongdoer and are using it to prosecute. But no one would ever even consider holding such people criminally liable, because the possession of pornography by the law enforcement community has nothing to do with the functions and purpose of the law. Earlier, I argued that even the most devoted textualist must be an intentionalist in practice, because the only serious justification for being a textualist in the first place is the claim (often, but not always, correct) that the words the legislature uses are the best evidence of what it intended. This is not the case when it comes to a statute’s broader purpose, however, as Jonathan Siegel points out forcefully in a recent article.89 Judge Easterbrook explained: “We interpret texts. The invocation of disembodied purposes, reasons cut loose from language, is a sure way to frustrate rather than implement these texts.”90 Thus, although chapter 4 is replete with examples of Justice Scalia making reference to legislative intent, he does not do the same when it comes to construing statutes in light of the larger legislative goals. His concurrence in Begay, for example, did not endorse the majority’s purposive analysis. Rather, it expressed his ambivalence in calling drunk driving a dangerous activity per se. It is rare to find Scalia referring to the legislature’s purpose in his majority opinions, except to criticize those who would refer to it in rendering a decision. For example, in MCI Telecommunications Corp. v. AT&T, discussed in chapter 3, Scalia explained, in rejecting the FCC’s interpretation of its authority to relax the rules requiring publication of tariffs: “But we (and the FCC) are bound, not only by the ultimate purposes Congress has selected, but by the means it has deemed appropriate, and prescribed, for the pursuit of those purposes.”91 It is not unusual, on the other hand, for dissenters, in response to a majority opinion written by Scalia, to evoke the purpose for which a statute was written in response to what seems to them an unjustifiably narrow interpretation of the law. That is what happened in Great-West Life Assurance & Annuity Co. v. Knudson.92 The case was one of many in the Supreme Court that involved interpreting provisions of the Employee Retirement Income Security Act (ERISA), the statutory scheme that regulates employee benefit plans. ERISA limits private rights of action but permits civil enforcement “by a participant, beneficiary, or fiduciary (A) to enjoin any act or practice which violates any provision of this subchapter or the terms of the plan, or (B) to obtain other appropriate equitable relief (i) to redress such violations or (ii) to enforce any provisions of this subchapter or the terms of the plan.”93

stability, dynamism, and other values

145

In Knudson, the plan provided for the payment of medical expenses in the event that an employee was injured in an accident but also permitted the plan to recover from the employee amounts paid by the plan that the employee recovered from a third party. Knudson’s wife was injured in an accident and sued the individual who caused the accident. The plan attempted to recover the amount that she was to receive for medical expenses that the plan had paid for on the theory that such recovery constituted “appropriate equitable relief” to prevent violation of the plan. The plan lost in both the district court and the court of appeals because a lawsuit to recover money damages is not considered equitable relief. Rather, it is considered legal relief. Five Supreme Court justices agreed and affirmed the lower court. Relying heavily on the historic distinction between law and equity, Justice Scalia rejected even considering the fact that a ruling against the plan would thwart the purposes of the statute. Quoting from an earlier case, he wrote, “vague notions of a statute’s ‘basic purpose’ are . . . inadequate to overcome the words of its text regarding the specific issue under consideration.”94 Justice Ginsburg’s dissent, in sharp contrast, relied heavily on the law’s purposes, which were to establish a uniform regulatory scheme and to ensure that a plan’s provisions would be enforced.95 The contrasting legal values are clear: a narrow, historical reading of “equity” really does lead to the conclusion the majority reached. A broader reading of “equity” combined with consideration of the statute’s purpose really does lead to the conclusion that the dissenters reached. What these cases have in common is that the language is sufficiently indeterminate to allow different interpretations that are linguistically licensed. When there is room for people to differ, courts are often willing to approach problems pragmatically, as Posner suggests, especially when it is easy enough to ascertain what the legislature was trying to accomplish. Notwithstanding the fears Scalia expressed in Knudson, such reference to purpose should actually serve to constrain judges. It is no accident that the five justices who voted in the majority were the conservative members of the Court, with the liberals dissenting. Limiting private lawsuits by civil litigants to enforce rights granted by regulatory statutes is part of the conservative legal agenda. The purpose behind the statute, in contrast, appeared to favor the dissenters. By finding the vague language plain and ignoring the legislative purpose, the majority was able to further an agenda without ever mentioning it. Perhaps Justice Scalia could have made a good argument for why certain

146

chapter 5

legislative goals would be best served by the result that was reached. Taking on that burden would have reduced the set of legitimate interpretations by requiring additional justification. This is not to say that such analysis will always lead to agreement. There often are competing values embedded in a law, with hard cases placing them in conflict. Which values a judge considers the most important will likely influence her vote. Nonetheless, when a statute is susceptible to more than one interpretation, a candid debate about what the law was intended to accomplish and how to best achieve that result does not appear to be any more of a threat to rule-of-law values than pretending that the law is clear and avoiding the substantive issues. The Legislative Process It is possible to make the situation more difficult for those who would consider a statute’s purpose. Sometimes the legislature has a particular goal in mind and then writes a law whose substance does not fulfill its purpose at all because the legislature was mistaken about various facts. That is, the legislature intended to write the law it wrote but would have written a different law if it had gotten the facts straight. The legislative errors discussed in chapter 4 all involved what appear to be statements that the legislature did not likely intend. In Locke, the legislature seems to have accidentally set a deadline at December 30 rather than December 31; in Granderson, it referred to “the original sentence” in a law revoking probation when there was no original sentence; in Green, the legislature wrote an absurd rule of evidence, if the rule is read literally; and in X-Citement Video, the legislature made a grammatical error.96 The Court corrected the errors in the last three of these cases and elegantly avoided injustice in the first. In contrast, sometimes the legislature gets the legislative facts wrong and then writes a law based on these erroneous findings.97 Courts are far more reluctant to correct such errors. In these situations, courts tend to place respect for the enactment process above fulfilling the law’s intended purpose. Consider the following history of a law based on mistaken facts. The law was repealed before any cases made their way through the courts. What is interesting for our purposes is that no one took the position that the law should not be enforced according to its letter because literal interpretation would not further the law’s purpose. On January 1, 1992, the New Jersey Department of Health promulgated a regulation that became known as the “runny egg law.”98 It required

stability, dynamism, and other values

147

restaurants to cook eggs to a minimum temperature of 140 degrees Fahrenheit. In essence, this banned the serving of sunny-side-up eggs and eggs over easy, because by the time an egg gets to 140 degrees, the yolk congeals. It also banned such dishes as Caesar salad, whose dressing includes a raw egg yolk. The result, not surprisingly, was that New Jersey became a laughingstock. Late-night talk shows made fun of the law. Johnny Carson remarked, “There’s something wrong with a state in which I can buy an Uzi, but there’s a 10-day waiting period to get a Caesar salad.”99 Governor Jim Florio did not think Carson’s joke was funny (Uzis actually were not legal in New Jersey) but still called the law “unenforceable, intrusive and silly.”100 He made a point of being seen in a New Jersey diner flouting the law by eating an egg cooked sunny-side up. In February, the law was amended to require restaurants to continue handling eggs as a potentially hazardous food in general but allowing them to resume serving undercooked eggs. In June, it was repealed altogether.101 The law really was silly. Just walk into any diner in New Jersey and watch one person after another eating eggs that would have been illegal to serve. They do not get sick, and they do not want to have the government telling them what to eat for breakfast. But the law was triggered by a reality. The Federal Food and Drug Administration had recommended that eggs be treated as a “potentially hazardous” foodstuff because outbreaks of salmonella poisoning had been attributed to eggs. Salmonella grows on eggs in temperatures between 45 and 140 degrees Fahrenheit. No one really disputed any of that. What they disputed was the cause of the outbreaks. In arguing for the law’s immediate repeal, the New Jersey Restaurant Association pointed out that many restaurants crack dozens of eggs in advance of the morning rush and store them in a vat for making such dishes as scrambled eggs, omelets, and French toast. One bad egg can contaminate the entire batch. It is in that phase, or even earlier when egg producers fail to refrigerate the eggs, that the risk of bacteria growth is greatest. The Restaurant Association had a good point. The other potentially hazardous foods subject to this regulation included raw meats, poultry, and fish, well known to be the source of salmonella outbreaks. Apart from requiring that these foods be cooked well enough to kill the bacteria, the regulation required that other foods not be permitted to touch them during preparation, and that restaurants have appropriate utensils and equipment, which they should keep both clean and separate from other foods. It seems, then, that when eggs were declared “potentially hazardous,” they

148

chapter 5

got swept into a set of requirements tailored for the cooking and serving of meat. Many of them, such as the use of clean, segregated equipment, fit eggs as well as meat and would not have provoked ridicule. To the contrary, it was a good idea to include eggs in the rule. But the rule had to be tailored to permit eggs to be served in ordinary preparations, while outlawing their being handled in ways that could jeopardize people’s health. Absent during all the debate and ridicule was the argument that courts would construe the law to exclude, say, sunny-side-up eggs, because their inclusion would not serve the purposes of the regulation. It was clear to everyone, it seems, that the regulation was enacted with the intention of treating eggs and meat alike. Absent an “absurd-results” argument, the language and intent would trump considerations of the rule’s broader purpose. Now let us look at a well-studied case of this nature that did make it to court, United States v. Marshall, a case in which Judge Easterbrook wrote the majority opinion and in which Judge Posner dissented.102 Congress had enacted a statute setting minimum penalties for the distribution of illegal drugs depending on the weight of the drugs distributed. It called for a sentence of five years in prison for selling more than one gram of a “mixture or substance containing a detectable amount” of LSD, and additional time for additional grams.103 Marshall was sentenced to twenty years for distributing more than ten grams of LSD. The problem with the statute has to do with the weight of the carrier. LSD is made as a liquid, diluted with alcohol, spread over a piece of blotter paper, and eventually cut into one-dose squares. Comparatively, the blotter paper weighs a great deal; the LSD, very little. Because the ratio of the weight of the carrier to the weight of the drug is much greater for LSD than for the other drugs covered in the statute (e.g., heroin, PCP), the sentence per dose of LSD is far greater. For example, “to have received a comparable sentence for selling heroin, Marshall would have had to sell ten kilograms, which would yield between one and two million doses.”104 Moreover, LSD can be distributed on different carriers, including sugar cubes and orange juice. It makes no sense for the length of a prison sentence to depend upon how thirsty a user or distributor was at the time of sale. The majority affirmed the sentence based largely on the language of the statute and the right of Congress, through the legislative process, to set whatever penalties it wishes to set. Arguing that the carrier should be excluded notwithstanding the literal language of the statute, Posner noted: “The positivist [majority] view, applied unflinchingly to this case,

stability, dynamism, and other values

149

commands the affirmance of prison sentences that are exceptionally harsh by the standards of the modern Western world, dictated by an accidental, unintended scheme of punishment nevertheless implied by the words (taken one by one) of the relevant enactments. The natural law or pragmatist view leads to a freer interpretation, one influenced by norms of equal treatments.”105 Posner argued further that excluding the medium would lead to a rational system of sentencing under the statute, far more in line with what the legislature intended to accomplish. But Posner’s position did not prevail, and this approach usually does not prevail in these circumstances.106 Congress erred. But the mistake it made was not in failing to draft a statute that adequately communicated its intent. Rather, the mistake was in its understanding of the underlying facts that led it to decide on this legislation. When the language is as clear as it was in this case, courts are not always willing to give the process a back seat, even if it means undermining the obvious purpose of the legislation. Courts take seriously the obligation to respect the results of the legislative process. When more than one result is possibly consistent with statutory language, they employ a host of values to decide a case, including the statute’s purpose. But when the language makes the intent clear, courts are reluctant to put purpose over process in order to fix a bad law. Coherence In chapter 4 we saw that courts sometimes look at coherence as a surrogate for intent. We assume that legislatures intend to write laws that work in harmony with each other. When an interpretive issue arises, the law should be construed in a manner that makes it coherent with other laws, not for the purpose of imposing a better legal order from the bench, but simply in order to effectuate what the legislature intended in the first place. As Justice Scalia put it in Green v. Bock Laundry Machine Co., Inc.: The meaning of terms on the statute books ought to be determined, not on the basis of which meaning can be shown to have been understood by a larger handful of the Members of Congress; but rather on the basis of which meaning is (1) most in accord with context and ordinary usage, and thus most likely to have been understood by the whole Congress which voted on the words of the statute (not to mention the citizens subject to it), and (2) most compatible with the surrounding body of law into which the provision must be integrated—a compatibility which, by a benign fiction, we assume Congress always has in mind.107

150

chapter 5

But this is not the only reason for a court to value coherence. Consider Scalia’s opinion in West Virginia University Hospitals v. Casey.108 The issue there was whether the fees paid to expert witnesses should count as part of “a reasonable attorney’s fee” with respect to a statute that allows the prevailing party in certain actions to recover attorney’s fees. Writing for a majority of six justices, Scalia argued that they should not. The most compelling argument was that Congress had enacted other statutes that shift fees, some of which mention expert fees specifically. Therefore, if Congress had intended the statute in question to provide for expert fees, it would have said so. Arguments like this are frequently stated in intentionalist terms even by devoted textualists, as we saw in chapter 4. But they need not be. One can say: regardless of what Congress had in mind, the Court should, in order to contribute productively to the legislative process, interpret statutes in such a way as to keep codes maximally coherent. Dan Simon has argued that there is a strong psychological impulse toward coherence that drives a great deal of legal decision making, including inferences about state of mind.109 This principle would be a case in point. In fact, Scalia made his point in just that way in Casey: “Where a statutory term presented to us for the first time is ambiguous, we construe it to contain that permissible meaning which fits most logically and comfortably into the body of both previously and subsequently enacted law. . . . We do so not because that precise accommodative meaning is what the lawmakers must have had in mind (how could an earlier Congress know what a later Congress would enact?), but because it is our role to make sense rather than nonsense out of the corpus juris.”110 This interpretive methodology no doubt has merit in many cases. But the quest for coherence must serve some purpose for it to be justified. One such purpose, asserted frequently by judges, is effectuating the intentions of the lawmakers. Absent evidence to the contrary, it is fair enough to infer that legislators want to write coherent codes and would be happy enough if courts maximized that value when a case arises involving some issue that they never noticed. Scalia’s stated purpose for promoting coherence in Casey is more radical. He justifies the practice as good judicial lawmaking, regardless of what the legislature might have intended. It shows less respect for legislative primacy than does the intentional approach. It more resembles Ronald Dworkin’s approach to statutory interpretation, which requires decision makers to decide statutory cases in a manner that “follows from the best interpretation of the legislative process as a whole.”111 It is at least argu-

stability, dynamism, and other values

151

able that working toward coherence in the interpretation of legal codes serves the goal of “making sense of the corpus juris.” Yet Scalia’s approach is less constrained than Dworkin’s, because Dworkin espouses looking at history as an important means of reducing the options available to the interpreter at any given time. Whether the focus on coherence serves any other purpose, how well it serves as a proxy for intent is an open matter. William Buzbee suggests that coherence arguments are not very good surrogates for intent.112 If he is right, then heavy reliance on coherence may at times undermine the principle of legislative primacy, not at all what Scalia attempts to accomplish. In reality, legislatures do not always pay attention to how a word in, say, the criminal code might have been used in a statute that regulates mining or in some other remote law. Casey, in fact, may be just such a case. As Justice Stevens’s dissent makes clear, in enacting the fee-shifting statute, Congress had intended to override a parsimonious Supreme Court decision that refused to shift fees in civil rights cases absent legislation requiring it.113 Congress responded by enacting the disputed statute. The Court in Casey then construed this new statute narrowly, and Congress had to respond once again by amending the statute to make it clear that it was to apply to expert fees as well.114 This dynamic may explain why coherence arguments are, by and large, also justified by the goal of enforcing the will of the legislature. Einer Elhauge presents a somewhat different argument, to the effect that coherence is a poor surrogate for enforcing the will of the legislature.115 Although he directs his remarks against Dworkin’s reliance upon coherence, his critique has far more general application.116 Elhauge points out that dictionary acts that are parts of the codes of many states often contain rules to the effect that specific legislative directives be given priority over general ones when conflicts occur. As Elhauge observes, the fact that a legislature instructs courts to ignore general principles in favor of specific, inconsistent rules provides evidence that the legislature does not itself value coherence; coherence cannot in such cases be supported on grounds of legislative intent. As for Scalia’s argument that intent is irrelevant because the Court should concern itself instead with coherence, his opinion does not test Elhauge’s proposition, because coherence is evidence of intent as well as a value in its own right. To argue against the relevance of intent in such cases, Scalia would have to take the position that coherence values must prevail even if they thwart the clear intent of the legislature. Although the

152

chapter 5

Casey dissent in fact accuses him of doing just that, Scalia never made that argument in Casey, and I doubt that he would make it in any case in which coherence is the principal argument. In fact, Scalia argued in Casey that Congress uses specific language to call for the shifting of expert fees in certain circumstances “when a shift is intended.”117 I know of no case in which a court has expressed a willingness to thwart the intentions of the legislature for the sake of adding coherence to the corpus juris, to use Scalia’s term. Certainly, interpreting a code to enhance coherence makes good sense as a default position. Absent evidence of intent to the contrary, why not promote coherence as a value, especially within a statute or a closely related cluster of statutes? Moreover, coherence can at least arguably serve as a proxy for intent in some cases. But coherence on its own is not a strong enough value to justify rejecting the contrary will of the legislature. Promoting Constitutional Values Consider the rule that statutes are construed to avoid constitutional questions. This canon is often stated in intentionalist terms: the court assumes that the legislature did not intend to enact an unconstitutional statute and therefore resolves difficulties in favor of a reading that survives constitutional challenge. For example, in United States v. X-Citement Video, Inc., the Supreme Court observed: “[W]e do not impute to Congress an intent to pass legislation that is inconsistent with the Constitution as construed by this Court.”118 In that case, the Court corrected what appeared to be a legislative error in the federal child pornography statute: it failed to require that the person who committed the crime knew that he was trafficking in child pornography. If interpreted literally, the statute might have applied, say, to the Federal Express driver who knew he was delivering films but had no idea that they involved child pornography. To solve the problem, the Court inserted a state-of-mind requirement on the assumption that Congress had most likely intended to do so but drafted the statute incorrectly. Contrast this approach with that of National Labor Relations Board v. Catholic Bishop of Chicago, in which the Supreme Court held that the National Labor Relations Act did not give religious school teachers the right to unionize.119 The Court set a high hurdle for the National Labor Relations Board to vault: because application of the labor laws to religious school teachers could, on some future occasion, cause the courts to

stability, dynamism, and other values

153

become enmeshed in questions of religious freedom, the Court required some “clear expression” by Congress that it intended to include this group. Otherwise, the Court would assume the group not to be included: “There is no clear expression of an affirmative intention of Congress that teachers in church-operated schools should be covered by the Act. Admittedly, Congress defined the Board’s jurisdiction in very broad terms; we must therefore examine the legislative history of the Act to determine whether Congress contemplated that the grant of jurisdiction would include teachers in such schools.”120 Notably, the Court did not claim that such values can survive the Marshall Test. That is, the Court did not say that the constitutional values are sufficiently important that the judiciary is willing to override the clear intention of the legislature. But a fair reading of the opinion suggests that is exactly what happened, as Eskridge correctly notes.121 Congress had actually considered and rejected the insertion of a provision that would exempt religious schools from the labor laws, just as they had considered and rejected amending the antitrust laws to include baseball.122 What Catholic Bishop reflects, then, is an effort to protect parochial schools from governmental interference into the religious aspects of their day-to-day affairs. What would happen, for example, if Catholic-school teachers claimed that the church had committed an unfair labor practice by refusing to negotiate in good faith over work rules that gave teachers time off just when religious activities were likely to occur? These are important concerns. But it is a stretch to claim that Congress intended them to trump the right to bargain collectively. The Court did not admit that it was balancing legislative primacy against other values that it thought more important. Supporting Law Enforcement Consider the following dilemma: From a fair reading of the legislative history and other contemporaneous data, such as old dictionaries, it appears that the legislature had intended to write a narrow statute some time ago. Now, however, the statute is most easily understood as including conduct that the legislature did not contemplate. Moreover, that conduct is bad on its face. Should a court subordinate the value of legislative primacy to good law enforcement and hold that the statute applies? That is what happened in Moskal v. United States. The statute in question banned the interstate transportation of “falsely made” securities. As Justice Scalia argued

154

chapter 5

convincingly in his dissent, at the time the statute was enacted “falsely made” was largely a synonym for “counterfeit.”123 In Moskal, the defendant was a car dealer who had been rolling back odometers, sending the false information to the motor vehicle authorities, and then receiving clean titles procured with the false information. There was nothing counterfeit about any of the documents, but the new titles certainly contained misinformation. The majority found that the clear notice of the statute was enough to trump the more limited mission of the enacting legislature and affirmed Moskal’s conviction. Over time, the “counterfeit” meaning of the expression seems to have become less salient, and we are likely to understand it in terms of the meanings of its components: “made to be false.” I have conducted a Lexis search of articles from major newspapers that use the words “falsely made” during a five-year period beginning in 2000. There were forty instances. Several report indictments that used the expression “falsely made securities,” quoting statutory language. One article reports that “District Attorney Terrence Parker said Mary Colletta ‘falsely’ made a ‘written statement, which was calculated to become, when completed,’ the surgeon’s will.”124 Interestingly, in that case, it appears that the prosecutor making the statement intended the expression to mean a forgery but that the reporter did not understand it that way. The quotation marks around the single word “falsely” suggest that the reporter thought this particular part of the story was about the defendant’s having lied about being married to the person whose will she later forged. None of the instances quoting anything other than legal documents, however, appears to use the expression as a synonym for “counterfeit.” Instead, we read of falsely made exit poll data containing bogus information,125 a “falsely-made” confession (referring to its content),126 and falsely made claims for repayment.127 From this, it seems fair to conclude that considerations of law enforcement have trumped legislative intent. Whether one agrees with the Court’s balancing of values in Moskal, it is not radical in principle. Imagine the result if McBoyle v. United States were decided today.128 Recall from chapter 3 that McBoyle, decided by a unanimous Supreme Court in 1931, held that an airplane should not count as a vehicle for purposes of construing a statute that made it a crime to transport stolen vehicles across state lines. In all likelihood, the 1919 legislature that enacted the statute would not have had airplanes in mind, as Justice Holmes wrote in his opinion. But nearly a century later, with commercial air traffic having been a part of daily life for about fifty years,

stability, dynamism, and other values

155

it would be strange for a court to let a McBoyle off the hook, no matter what the enacting legislature had in mind. Our prototype of a vehicle has changed over time, and people are on adequate notice that if they steal an airplane they have stolen a vehicle. It would not be surprising if fair notice is enough to trump legislative intent, as it was in Moskal, especially when the statute is an old one. Political Ideology There are too many cases in which the conservatives vote for the conservative outcome and the liberals for the liberal outcome to deny that a judge’s individual political values play a role in statutory interpretation. We saw this in the discussion of Circuit City Stores v. Adams in chapter 2.129 The federal arbitration act can be interpreted as permitting lawsuits to be brought in court by employees who would have been deemed to engage in interstate commerce when the law was passed early in the twentieth century or as permitting such lawsuits by those deemed to be engaged in interstate commerce whenever their claims arise. The five conservative justices prevailed, choosing the first alternative; the four liberals opted for the second. The statute is susceptible to both meanings, and both the majority and the dissenting opinions contain legitimate argumentation that most judges employ at one time or another: making use of ordinary meaning, nuances in the way related statutory provisions are worded, the context in which the law was enacted, and so on. Is it plausible that the political views of the justices played no role in the outcome of this case? Or consider Ledbetter v. Goodyear Tire and Rubber Co., decided by the Supreme Court in 2007.130 Ledbetter claimed that she was being paid less because of her sex. A jury agreed with her, but the case was reversed on appeal131 and then went to the Supreme Court. The Civil Rights Act requires that claims of discrimination be brought within 180 days of a discriminatory act. The statute makes it an “unlawful employment practice” to discriminate “against any individual with respect to his compensation . . . because of such individual’s . . . sex.”132 Ledbetter claimed that her case was timely because each paycheck that she received constituted a discriminatory act, since each paycheck reflected the intention to pay her less because of her sex. Goodyear argued that if there was a discriminatory act, it happened when Goodyear first set her pay as below the pay of men and that, therefore, the 180-day period had expired before she filed her claim.

156

chapter 5

Ultimately, Goodyear prevailed in the Supreme Court, with Justice Alito, a conservative, writing for the majority, and Justice Ginsburg, a liberal, writing the dissent. It is possible to conceptualize the phrase “discriminatory act” either way. The majority is correct in its argument that the clearest act of discriminatory intent occurred when Goodyear’s management decided to pay her less than they paid men. After that, the issuance of paychecks at a big company like Goodyear is merely ministerial and thus occurred without the intent needed to prove employment discrimination. At the same time, the dissent is correct that the decision creates such huge hurdles for those who actually have suffered wage discrimination that it is unlikely that Congress would have at the same time outlawed wage discrimination and made it almost impossible to assert one’s rights to be free of such discrimination. Compensation decisions, unlike, say, promotion decisions, are typically “hidden from sight”133 and come to light, if at all, only after their effects have continued for a lengthy period of time. Both sides relied opportunistically on the Court’s earlier decisions, with most of the disagreement over which earlier cases were more analogous with the case at hand. If I were a judge, I would have voted with the dissenters. I do not believe that Congress would have intended the law to be construed so narrowly. We saw in chapters 3 and 4 that effectuating the intent of the legislature is a major goal of just about all statutory interpreters, whatever their political stripe. The language permits the interpretation I prefer, and principles of interpretation make my perspective fully legitimate, perhaps even the better of the two options given the strong role that legislative primacy plays in statutory interpretation. Yet it is self-evident to me that my own political values play a role. I not only would construe the statute as permitting such claims but would be glad that I had done so, because doing so helps to further values that I consider important. For the same reasons, I was gratified to learn that, in the first days of the Obama administration, Congress passed and the president signed the Lilly Ledbetter Fair Pay Act of 2009, which overrode the Supreme Court decision and enacted into law the rule espoused by the dissenters.134 Yet by the same token, I have little doubt that those conservative justices who voted to limit the scope of such claims considered such facts as the importance of providing to employers certainty of their litigation risks to enable them to thrive. They, too, were faced with a statute that permitted an interpretation they would prefer and were in a position to do what they considered to be the right thing without doing significant damage to the vocabulary of statutory interpretation.

stability, dynamism, and other values

157

Thus, the question is not whether the political values of the interpreter play a role in how statutes are construed but rather whether statutory interpretation is so political that it has no substance at all other than as a vehicle for judges to do as they wish. Karl Llewellyn demonstrated more than half a century ago that the canons of construction are so subject to manipulation that for many canons there is a matched canon that tells the court to do the opposite of the first one.135 Since then, others have made similar observations.136 Contrast the plain-language and absurd-results rules to see one of many such examples. And as we saw earlier in this chapter, Lindquist and Cross’s study convincingly shows that the political affiliation of appellate judges correlates with their decisions in civil rights cases.137 A recent study by James Brudney and Corey Ditslear confirms that there is a relationship between a justice’s ideology and the strategic use of various interpretive devices. They examined 632 “workplace law” cases decided by the Supreme Court over the thirty-five-year period ending in 2003. (Both of the cases discussed above would fit into their data set, but only Circuit City Stores was decided within their period of study.) They found that canons of construction related to the language of a statute are used more or less equally to justify both liberal (pro-worker) and conservative (pro-employer) decisions.138 Moreover, depending on the issue and the level of statutory indeterminacy, justices sometimes vote contrary to their political leanings. Nonetheless, the liberal justices are far more likely to use these canons to justify a liberal decision than to justify a conservative one, while just the reverse holds true for the conservative justices. The same distribution of canon usage holds for substantive canons of construction. From this, the authors conclude “that the canons are not having an independent, constraining effect on the Justices’ decisionmaking—in particular, they are not functioning as a set of overarching ‘neutral principles’ in the hands of either liberal or conservative Justices. Put differently, the canons’ self-evident persuasiveness and logical force are not leading liberal, or conservative, Justices whose opinions rely on the canons closer to the Court’s ideological center.”139 Thus, liberals typically use the canons to justify liberal results, and conservatives use them to justify conservative results. How much this infiltration of ideology into statutory interpretation should concern us depends both upon how deep the infiltration goes and upon our initial expectations. Because of the precedential effect of Supreme Court decisions, it matters that the political leanings of appellate

158

chapter 5

judges influence, say, the breadth with which civil rights statutes and employment statutes designed to protect workers are construed. At the margins, workers will have more rights (and employers more burden) if a liberal court resolves disputes in close cases than they will after a conservative court does so. Nonetheless, as I argued earlier, the impact of these decisions must be measured against the overall applicability of the statutes, and for the most part laws are clear enough to give us a sense of what we must, may, and cannot do. Appellate decisions in close cases constitute only a small fraction of all cases decided, a much smaller fraction of cases brought, and an even smaller fraction of instances in which the law applies, including those many times it does so without controversy. Even less indicative are the cases decided by the Supreme Court, which are preselected for their controversial nature, often because of conflicts among the courts of appeals. One partial solution to the problem is to eliminate — or at least reduce—the precedential value of judicial decisions interpreting statutes. A court deciding a hard case will certainly know how other courts—including the Supreme Court—has ruled in the past, but it need not follow earlier cases if the judges believe that doing so would undermine the value of legislative primacy. That, in fact, is how most civil-law countries approach statutory interpretation. This solution, however, would so undermine the value of stability and be so at odds with our common-law tradition that it would have little chance of being accepted. We will, therefore, continue to accept, whether we like it or not, and whether we deny it or not, some degree of political ideology in the resolution of difficult statutory cases. Our ability to make laws is simply not so crisp as to avoid this.

How Should Courts Decide Statutory Cases? To the extent that the law requires certainty, the kinds of cases discussed in this chapter are not comforting. As we have seen, courts strongly favor legislative primacy as a value, although they disagree over how best to be faithful to the lawmakers. In earlier chapters, I showed why these problems are virtually intractable, explaining why they do not go away. It is no accident that cases more than a century old resonate today. The same holds true for cases in which competing values are at stake. How old must a statute be before nuances of legislative intent seem less important when a defendant today seems to be on notice? How much

stability, dynamism, and other values

159

should a reasonable sense of the relationship between the Constitution and the statute books influence interpretation when the law is itself constitutional? How wedded to old decisions should courts remain when times have changed and the earlier decisions were weakly argued even on their own terms? These questions all require judgments in which conflicting evidence and values compete. How these judgments are made is complicated and a matter of controversy. Nonetheless, work by such scholars as Dan Simon and by Gilbert Harman and Sanjeev Kulkarni suggest that in close cases decision makers, over time, form mental models that tend to reinforce that evidence that tends to support even small initial preferences and to minimize evidence that runs contrary to initial preferences.140 The result is that decision makers become more and more confident in decisions that they earlier recognized as hard ones. Experimentation shows that this is the case, whichever side the decision maker ultimately takes and whether it involves jury decisions or the interpretation of statutes.141 How much uncertainty this brings to the legal system depends largely on the extent to which decision makers are faced with difficult problems in which they must weigh conflicting evidence and adduce inconsistent values. I have argued both in this chapter and in chapter 2 that such cases are relatively rare in the context of a baseline of clear cases. If I am right, this chapter suggests once again that we cannot dispense with decision makers who must use their discretion and that judges are not “activists” simply because they have exercised judgment. But it also suggests that the application of laws is clear enough most of the time to prevent this residue of uncertainty from having a crippling effect on the rule of law.

chapter six

Who Should Interpret Statutes?

J

udges are not the only ones who interpret statutes. The other branches of government get into the act as well. Sometimes the effort is cooperative, but at other times a sort of tug-of-war develops, with the various branches vying for power. In this chapter I describe some of these battles and identify some winners and losers. What we will see is the various branches of government doing their jobs in conventional ways and some struggles for power among the branches. Indeed, the legislature passes laws, the executive branch carries them out, and the judiciary referees disputes over the applicable scope of statutes. But we will also see that the president tries to wrest away some of the legislature’s influence with the courts; the legislature tries to manage the courts more tightly; and agencies, given free rein under current doctrine, use their power aggressively.1 The courts are very much part of the game as well, creating rules of the road that legislatures must follow to accomplish goals that the courts disfavor. Moreover, what courts give, they can take away. The delegation doctrines that have been so much a part of the jurisprudence of statutory interpretation for a quarter century are applied with a political thumb on the scale, as researchers have demonstrated convincingly. Although judges may be rightly criticized for many of the turns they take in interpreting statutes, in this chapter I suggest that the courts should remain the principal institution engaged in statutory interpretation. Stripping judges of the power to make bad decisions almost always strips them

who should interpret statutes?

161

of the ability to make good ones. What is hard about judging is not that there are too many tools. Rather, it is the difficulty in determining what the legislature wished to accomplish and then balancing legislative primacy against other values.2 Telling judges that they may no longer judge will not make this central task any easier.

Should the Executive Play a Greater Role? Interpreting Statutes in the Administrative State In conducting its everyday business, an agency must interpret the statute delegating power to it in order to understand its obligations to regulate and the limits of its power to regulate. Since the Supreme Court’s 1984 landmark decision in Chevron USA, Inc. v. Natural Resources Defense Council, Inc., courts have been instructed to defer to agencies’ interpretations of their own power.3 The Chevron doctrine, advocated by the Reagan administration as part of a more general effort to limit regulation,4 calls for a two-step analysis. First, the court must decide “whether Congress has directly spoken to the precise question at issue,” including an inquiry into whether the congressional delegation to the agency is clear and unambiguous.5 Second, the court reviewing an agency interpretation must decide whether the agency has acted permissibly within the conceptual space allowed by the statute. The agency is given leeway to act within any reasonable interpretation of the law, even if the reviewing court would have interpreted the statute differently.6 The Chevron doctrine, now in its third decade, seems firmly planted in American jurisprudence. Supporters applaud the fact that agency interpretation places policy decisions closer to the elected representatives, rendering this aspect of statutory interpretation both more expert and more democratic. As Cass Sunstein has put it: “For the resolution of ambiguities in statutory law, technical expertise and political accountability are highly relevant, and on these counts the executive has significant advantages over courts. Changed circumstances, involving new values and new understandings of fact, are relevant too, and they suggest further advantages on the part of the executive.”7 In fact, some have argued that the doctrine does not go far enough. Dan Kahan, for example, has suggested that the courts defer to the government’s interpretation of criminal statutes, abandoning the rule of lenity if the Department of Justice has already announced its position.8 Adrian

162

chapter 6

Vermeule goes even further, advocating that judges should refrain from most interpretive activity other than to apply the plain language of a statute to a case before a court, deferring to the executive branch when the language is not clear.9 The Chevron doctrine fosters a dynamic legal order, a theme we explored in chapter 5. By making agency decisions relatively impervious to judicial review, the doctrine allows agencies to be more responsive to the nuances of their statutes’ domain without fear of being overruled by the courts. The Chevron doctrine should also lead to a more unified set of regulatory directives. For if all of the federal courts of appeals defer to agency decisions, then the courts are less likely to make inconsistent rulings—rulings that create the familiar chaos of federal law differing from place to place unless the Supreme Court takes up the matter to resolve the differences.10 But along with this uniformity, the Chevron doctrine also brings with it two kinds of instability. The first is obvious enough: courts, as we saw earlier, are wedded to a strong principle of stare decisis when it comes to the interpretation of statutes. Judicial decisions construing statutes are hard to dislodge even if they did not make much sense in the first place. This is not true of the interpretation of statutes by administrative agencies. On the contrary, as administrations change, so do their regulatory policies. The Chevron case itself illustrates this dynamic. The regulation upheld in that case was a Reagan administration rule weakening certain clean-air regulations initiated by the Carter administration.11 To take a more recent example, during the Clinton administration, the Environmental Protection Agency (EPA) pursued an aggressive interpretation of the law requiring power plants to upgrade their pollution controls when upgrading a facility. Some power companies capitulated, but others chose to litigate. In Environmental Defense v. Duke Energy Corp., decided in 2007, the Supreme Court unanimously upheld the more aggressive regulatory scheme.12 The legal question hinged on how much flexibility is contained in the word “modification.” Two statutory schemes, enacted in the 1970s to control pollution emanating from power plants, regulated both newly constructed plants and those undergoing “modification.” The goal of each was to require new and modified plants to adhere to far stricter output standards, while grandfathering the old plants until they had to be replaced or upgraded. The 1970 (Nixon administration) statute enacting the first of the schemes defines “modification” as follows: “(4) The term ‘modification’ means any physical change in, or change

who should interpret statutes?

163

in the method of operation of, a stationary source which increases the amount of any air pollutant emitted by such source or which results in the emission of any air pollutant not previously emitted.”13 The 1977 (Carter administration) statute enacting the other scheme adopts the same definition of “modification”.14 The definition leaves open an important question: how should “increases [in] the amount of any air pollutant” be measured? If measured on the basis of the amount of pollutant emitted per hour, then plant upgrades that increase the amount of time that a plant can operate but that do not affect the hourly output would not count as modifications. Thus, an old plant that emits a large amount of pollution could expand its operation without having to improve its pollution control system. If measured, in contrast, on an annual basis, an upgrade that permits a plant to run more and therefore pollute more would be considered a statutory modification, falling within the statutes’ regulatory apparatus. With Congress having left this much uncertainty in the meaning of “modification,” it became the EPA’s job to define the term. It did so twice. In 1975, during the Ford administration, when only the first statutory scheme was in place, the EPA defined “modification” in terms of kilograms per hour, an industry-favorable definition.15 As for the second statutory scheme, enacted in 1977, the EPA defined “modification” for purposes of interpreting that law in 1980, the last year of the Carter administration, in terms of “tons per year,” which would require power plants renovated to increase production to also enhance their pollution controls.16 The Supreme Court was left to decide what happens when an ambiguous statutory definition is used as part of two different schemes to control pollution, and the administrative agency fails to resolve the ambiguity in a uniform manner. The answer—a unanimous one, just as Chevron itself was unanimous—was that the EPA had every right to do so. It is one thing to assume that Congress has in mind the same understanding when it uses a word more than once in the same statute. It is another when Congress has left to an administrative agency the task of filling in gaps in the meaning of the same word in two separate statutory schemes. The EPA had the right to enforce the more demanding of the two standards if it chose to do so. This is about as deferential as a court can be to administrative action. Yet, as the Supreme Court demonstrated in Massachusetts v. Environmen­ tal Protection Agency the same day it decided Environmental Defense, deference is not limitless.17 There, the Court decided that the George W.

164

chapter 6

Bush administration lacked the authority under the Clean Air Act to refuse to regulate the emission of greenhouse gases affecting climate change. The administration had taken the position that the statute does not mention climate change, and carbon dioxide is not a pollutant under the act. The law provides: “The [EPA] Administrator shall by regulation prescribe (and from time to time revise) in accordance with the provisions of this section, standards applicable to the emission of any air pollutant from any class or classes of new motor vehicles or new motor vehicle engines, which in his judgment cause, or contribute to, air pollution which may reasonably be anticipated to endanger public health or welfare.”18 The administration argued that climate change is so important that, had the legislature wanted the legislation to deal with gases that are benign in some circumstances but whose emission can cause the atmospheric temperature to increase, it would have said so. It further argued that being forced to regulate greenhouse gases could affect foreign-policy positions that were inconsistent with such a requirement. The Court rejected all such arguments, focusing on the statutory language: “While the Congresses that drafted § 202(a)(1) might not have appreciated the possibility that burning fossil fuels could lead to global warming, they did understand that without regulatory flexibility, changing circumstances and scientific developments would soon render the Clean Air Act obsolete. The broad language of § 202(a)(1) reflects an intentional effort to confer the flexibility necessary to forestall such obsolescence.”19 In a strong dissent, Justice Scalia argued that the words “in his judgment” in the statute allow the administration the flexibility not to regulate greenhouse gases.20 Yet the position did not garner Justice Kennedy’s necessary fifth vote. As in Food & Drug Administration v. Brown & Wil­ liamson Tobacco Corp.,21 the case involving the Clinton administration’s effort to regulate tobacco (discussed later in this chapter), the majority of justices found the gap between the enacting Congress’s intent and the administration’s policy too great to tolerate. Interestingly enough, it was the failure to regulate that brought down the Bush administration’s efforts and the attempt to regulate too aggressively that doomed the Clinton policy. Notwithstanding these limits, strong deference to the interpretations of administrative agencies expands the range of legitimate policies that may be pursued from one administration to the next. The cost is the inability to rely upon a court’s interpretation as providing a stable set of parameters within which individuals and businesses may operate. Of course, there is always the risk that the legislature will change the law. But the risk that

who should interpret statutes?

165

a new administration will interpret the same law differently is far greater when business decisions need to be made. To the extent that statutory interpretation concerns itself with reasonable reliance on earlier interpretive decisions, that goal is sacrificed in favor of more flexibility and, perhaps, political accountability. In some instances, application of the Chevron doctrine has actually led to increases in criminal liability without the enactment of new legislation. Chapter 5 discussed the dynamic of statutory inflation, by which expansive interpretations of regulatory statutes in civil cases become the basis of broadened criminal prosecutions without further legislative action. That phenomenon reaches its limit when a court in a criminal case defers not to an earlier judicial decision but to an agency’s expansive interpretation of a statute under the Chevron doctrine. This is just what happened in an important insider-trading case. In United States v. O’Hagan, decided in 1997, the Supreme Court endorsed the misappropriation theory of securities fraud.22 The Court based its decision on civil precedents, two law review articles, an extremely narrow reading of the Court’s earlier precedents, and the Chevron doctrine. James O’Hagan was a Minneapolis lawyer whose firm represented Grand Metropolitan PLC, a company that was planning to make a tender offer to buy shares of Pillsbury Company. Before the merger was publicly announced, O’Hagan began buying shares and options of Pillsbury. Once the merger was announced, the price of Pillsbury shares shot up. O’Hagan sold and made more than four million dollars. O’Hagan was indicted for securities fraud. The government first alleged that he had violated section 10 of the Securities Exchange Act of 193423 and Rule 10b-5 of the Securities and Exchange Commission (SEC).24 The prosecution was aggressive. In prototypical insider-trading cases, the defendant takes material information from a company to which he has a fiduciary duty and uses that information for improper personal gain. But O’Hagan’s firm did not represent Pillsbury, the company whose stock he was trading. The alleged fraud was based on the theory that O’Hagan had misappropriated information from his law firm and had used that misappropriated information in the purchase and sale of securities without disclosing his activities. At the time of the trial, the Supreme Court had not passed on the misappropriation theory, and lower courts remained divided. 25 The government also alleged that O’Hagan had violated section 14(e) of the Exchange Act26 and Rule 14e-3(a) of the SEC, which prohibits fraudulent conduct in connection with tender offers.27 O’Hagan was convicted

166

chapter 6

of securities fraud and other crimes. The Eighth Circuit reversed the convictions,28 and the Supreme Court reversed the circuit. O’Hagan illustrates how the relationship between fidelity to the legislature and deference to administrative agencies is a complex one. On the one hand, deference is itself licensed by the assumption that the legislature has intentionally delegated interpretive power to the agency, as the Supreme Court made clear in O’Hagan. On the other hand, deference robs courts of even considering whether an agency’s views are at odds with the enacting legislature’s intent when the language of the enabling statute appears to contain flexibility. Thus, deference to administrative agencies may pass the Marshall Test set forth in chapter 5: a court may recognize that the enacting legislature might have favored a particular agency interpretation, yet defer to the agency on the theory that the express delegation of authority outweighs unexpressed legislative intent, even when that intent is fairly easy to infer. Nonetheless, applying a controversial regulation criminally in the first instance is a challenge to traditional rule-of-law values. The standard model of statutory interpretation requires that criminal statutes be applied narrowly to give fair notice and to ensure appropriate application of separation-of-powers principles. Here, in contrast, the SEC’s reading was expansive and “prophylactic,” consistent with the policy of construing remedial statutes broadly but inconsistent with the jurisprudence governing the interpretation of criminal laws. As Neil Komesar points out, the institutional setting in which laws are implemented has profound substantive ramifications for how laws are carried out.29 Coupled with the instability that comes from empowering agencies to interpret statutes is another uncertainty: judges are not consistent in deferring to the interpretations of agencies. Richard Pierce observed this in an early study in which he found a surprising number of cases holding agency interpretations out of bounds in the teeth of what courts held to be unambiguous statutes.30 We saw an example of this, MCI Telecommuni­ cations Corp. v. AT&T, in chapter 3.31 That case also involved the interpretation of the word “modify.” A statute allowed the Federal Communications Commission to modify the rule governing the promulgation of telephone tariffs. The case asked whether that power permitted the agency to eliminate the requirement entirely for small operators. A majority of Supreme Court justices said no. While the Court’s interpretation is certainly a possible one, perhaps even a preferred one, it is entirely possible to modify a scheme by changing

who should interpret statutes?

167

the rules that specify to whom the scheme applies, even if that means eliminating it entirely for a subset of players. For example, if a high school has a rule allowing only seniors to leave the campus for lunch, it would not seem strange to say that the school “modified” the rule to extend the privilege to juniors. The Court, it seems, is unwilling to divest itself of the power to construe statutes even if this comes at the expense of murkiness in the application of the Chevron doctrine. And while it is certainly possible to harmonize the approaches to “modification” in MCI and Environ­ mental Defense, a jurisprudence that permits such judicial control in one case and extraordinary deference in another is surely not a jurisprudence whose primary values are stability and predictability. Thomas Miles and Cass Sunstein have examined both Supreme Court and court of appeals decisions reviewing agency interpretation of statutes. They looked at 84 cases between 1989 and 2005 in which the Supreme Court purported to apply the Chevron doctrine (or at least a dissenter said it should have done so) and 253 court of appeals cases from 1990 through 2004 that reviewed agency decisions made by the EPA and the National Labor Relations Board. Miles and Sunstein analyzed how individual justices voted. The answer could not be clearer: politics matter in predicting whether and when a justice defers to an agency interpretation. Their summary of the results merits quoting in full: The simplest finding is that on both the Supreme Court and the courts of appeals, the application of the Chevron framework is greatly affected by the judges’ own convictions. Whatever Chevron may say, the data reveal a strong relationship between the justices’ ideological predispositions and the probability that they will validate agency determinations. The most conservative justices are 30 percentage points more likely to vote to validate agency interpretations that are coded as conservative than to validate agency interpretations coded as liberal. By contrast, the more liberal justices are 27 percentage points more likely to vote to validate agency interpretations coded as liberal than to validate those coded as conservative. Moreover, the most conservative justices on the Supreme Court show a validation rate 19 percentage points lower when applying the Chevron framework to the interpretations of the Clinton administration than to those of the two Bush administrations—while the more liberal justices show a validation rate 6 percentage points higher under the Clinton administration than under the two Bush administrations. If judicial decisions under the Chev­ ron framework are assessed in crudely political terms, the voting patterns of Supreme Court justices fit with the conventional groupings of the justices along

168

chapter 6

political lines—a clear signal that the Chevron framework is not having the disciplining effect that it is supposed to have.32

Much the same holds true for the courts of appeals.33 The significance of these findings depends upon how these decisions fit into the overall landscape, as we saw in our discussion of the bribery statute in chapter 5. Without doubt, the 84 Supreme Court decisions constitute only a tiny fraction of reviewable agency decisions, most of which are so uncontroversial that no one would think of challenging them at all. The court of appeals decisions are somewhat more troublesome, but not enormously so, if the base rate is high enough. What this means is that, as we saw in chapter 5, decision making at the margins may be legally unpredictable and more loaded with politics than we would wish if we envision the rule of law as a concept largely independent of politics. However, if the failure of the system to meet this ideal is limited to a small number of difficult, although perhaps highly salient, cases, the system can be said to be working more or less as designed and indeed to be based heavily on rule-of-law values. At the same time, the assertion that judges in the position to exercise discretion in hard cases must do so without regard to their own values begins to resemble a fairy tale. They cannot avoid bringing their values into the process, and the law in these cases is not certain enough to give an answer in any event.34 Both of these dynamics—shifts in policy from one administration to another and the politics of judges trumping the Chevron doctrine—are transparent in the case of Food & Drug Administration v. Brown & Williamson Tobacco Corp.35 The case involved a Clinton administration rule permitting the federal Food and Drug Administration (FDA) to regulate the cigarette industry. The Supreme Court held that Congress did not intend, when it wrote the Food, Drug, and Cosmetics Act, to give the FDA authority to regulate tobacco, despite language in the act that arguably supported such authority. The Court was closely divided, with the five more conservative justices, including Justice Scalia, voting in the majority.36 The problem facing the majority was that the case was governed by the Chevron doctrine. The underlying statute permitted the FDA to regulate not only drugs but also “device[s]” through which drugs are delivered.37 It is easy enough to argue that cigarettes are not what one would ordinarily consider a device, and therefore not a proper subject of regulation. But the Chevron doctrine is supposed to take that argument away

who should interpret statutes?

169

from the courts. Surely a cigarette can reasonably be considered a device to deliver nicotine. If so, the courts lack the power to substitute their judgment for that of the administrative agency that Congress empowered to enforce the statute. Justice Breyer, writing on behalf of himself and three others, made this point forcefully in his textually oriented dissent.38 At the same time, based on its analysis of the statute’s legislative his­ tory, other statutes that regulate tobacco, and other contextual information, the majority was correct that Congress never intended to permit the FDA to regulate tobacco. Efforts had been made to legislate such authority over the years, and they invariably had failed. With such an array of arguments on both sides, it is difficult to deny that politics played a substantial role. The conservative majority’s concern for legislative intent and its lack of concern for both the Chevron doctrine and the language of the statute are not likely to have been accidents. The same can be said for the Court’s more liberal justices, who ignored legislative intent in favor of a wooden application of the Chevron doctrine. The alignment of judges in this decision can be sensibly explained only with reference to their political leanings, providing a good illustration of the results of Miles and Sunstein’s study of how the Chevron doctrine is actually implemented. Presidential Signing Statements The Chevron doctrine is not the only effort to shift interpretive power from the courts to the executive. President George W. Bush controversially increased the use of Presidential Signing Statements as a challenge to courts to take the president’s interpretation of a statute seriously. But signing statements have been issued routinely by presidents since President Carter,39 and it was President Reagan, under the initiative of Attorney General Edwin Meese, who first advocated their increased use to enhance the executive’s influence over statutory interpretation.40 Samuel Alito, now a Supreme Court justice, worked as an attorney in Meese’s Department of Justice and authored a memorandum entitled “Using Presidential Signing Statement to Make Fuller Use of the President’s Constitutionally Assigned Role in the Process of Enacting Law.” In it, Alito argued that because a bill becomes a law only after it has been approved by both houses of Congress and the president, “it seems to follow that the President’s understanding of the bill should be just as important as that of Congress.”41 A more recent article by Steven Calabresi and Daniel Lev argues similarly.42

170

chapter 6

Alito conjectured that presidents, “unlike Congress, do not customarily comment on their understanding of Bills.” He suggested filling this gap in order to “increase the power of the Executive to shape the law” and to “help to curb some of the prevalent abuses of legislative history.”43 After suggesting a strategy for piloting the new program of signing statements, Alito noted: “The most important step will be approval of this project by the President. Obviously there can be no project unless the President wishes to sign interpretive statements of the type we envision. For the purpose of presenting this issue to the President, it may be helpful if we draft a sample of a new-style signing statement either for a bill that is now pending before Congress or one that was recently enacted. Also, as a first step, the proposal should be discussed with White House counsel.”44 Signing statements have become controversial largely because the second President Bush used them just as Alito suggested: in an effort to reduce legislative power and expand the scope of the president’s executive power upon his signing bills into law. The most prominent instance came after he signed into law in December 2005 the Detainee Treatment Act, which outlaws “cruel, inhuman, or degrading treatment or punishment” of anyone in U.S. custody.45 His signing statement asserted that he would interpret this provision “in a manner consistent with the constitutional authority of the President to supervise the unitary executive branch and as Commander in Chief and consistent with the constitutional limitations on the judicial power.”46 Earlier, he had opposed the ban on torture in the bill, but he relented and signed the bill into law, with its torture ban, under strong pressure from Senator John McCain, a war hero, who vigorously opposed the use of torture.47 Issuing the signing statement after agreeing to the bill could be seen as an end run around the veto and an abuse of power. As a formal matter, the president did nothing controversial. He merely stated that he would construe the law consistent with the structure of the Constitution and provided information about how he understood that structure. That is what many signing statements by many presidents have done. For this reason, I agree with defenders of signing statements that there is nothing unconstitutional and, for that matter, nothing at all improper about their issuance.48 But in this instance, the president actually agreed to a compromise on a bill and then, at least arguably, undermined that compromise by disclaiming his obligation to abide by the terms that he did not like. Were a court to determine whether the president retained the right to torture detainees notwithstanding the statute, I do not believe that the

who should interpret statutes?

171

Presidential Signing Statement would have much of a role to play. Either the president has that power, or he does not. The fact that he claimed to have it after the statute’s enactment does not seem to be of much moment in deciding the constitutional question. Presidential Signing Statements are also issued with the goal of influencing courts in their interpretation of statutes. There, too, they have not had much bite. Consider the Sarbanes-Oxley Act,49 which imposed increased reporting requirements on corporations in the wake of a host of corporate scandals at the beginning of the millennium, highlighted by the Enron scandal. The bill was signed into law by President George W. Bush on July 30, 2002, in a White House ceremony in which he praised Congress, including sponsors from both parties, for passing such extensive reforms. Among other things, the law contains a provision protecting whistleblowers who inform, among others, “any Member of Congress or any committee of Congress” of various violations of the act.50 Later that same day, the White House posted a signing statement asserting that the executive branch would construe this provision “as referring to investigations authorized by the rules of the Senate or the House of Representatives and conducted for a proper legislative purpose.”51 Thus, after praising the sponsors of the bill for shutting down corporate corruption, the president said that whistleblowers would be protected only if they blew the whistle by notifying a member of Congress after a congressional investigation had already begun for a “proper legislative purpose.” Nothing in the act supports such a limitation, and as far as I know, since its enactment no one has claimed in a case that it should be so limited. The administration itself subsequently abandoned this position. Notwithstanding the attention that they received in the press during the George W. Bush administration, courts do not pay much attention to Presidential Signing Statements, and it is unlikely that they will. Most importantly, Presidential Signing Statements are issued after a bill is signed into law. Therefore, they cannot possibly have influenced the enactment process. Statements and actions taken by an administration during the enactment process can do so, but not a statement that comes only after the legislature has finished doing its work. Just as postenactment statements by members of Congress as to preenactment thinking have no serious status as legislative history, Presidential Signing Statements suffer from the same problem of timing. I argued in chapter 4 that what legitimizes the use of legislative history, committee reports in particular, is the preenactment agreement, whether

172

chapter 6

tacit or explicit, that a subgroup of the legislature will act on behalf of the whole body in working out the details of laws that the entire body could not possibly take the time and effort to understand in full detail. That deference also applies to the executive branch when the legislature has determined to enact an initiative proposed by that branch of government. I also argued that Justice Scalia and others are correct in their assertions that legislative history may not always be good evidence of legislative intent and that it must be used prudently by those who would use it at all. When a president issues a statement after the fact with its only purpose to influence judges if disputes later arise, that statement does not yield credible evidence of what the law was intended to accomplish. Were presidents to issue their statements in advance as a warning to legislators before they cast their final votes, the situation might be entirely different. Compare the sparse use of presidential signing statements to the more robust use of presidential veto statements. To take one example, section 1001 of the federal Criminal Code made it a crime to make false statements “in any matter within the jurisdiction of a federal agency.”52 In United States v. Yermian, the issue before the Supreme Court was whether the defendant had to know that the false statement was in connection with the business of a federal agency.53 The language of the statute as it existed at the time of the case (1983) was not entirely clear on the matter, with the more natural reading favoring the requirement that the defendant did have such knowledge.54 What was clear, however, was the fact that President F. D. Roosevelt had vetoed an earlier enactment, complaining that the law was too weak on such matters. The Supreme Court inferred that Congress’s redrafting of the law prior to resubmission to the president was intended to reconcile the earlier difference and thus construed the statute accordingly. The point is that presidential action during the legislative process is fair game for those employing legislative history, even if Presidential Signing Statements are not. Article I of the Constitution vests the “Legislative Power” in the Congress. In fact, after the preamble, that is the very first thing that the Constitution says. Notwithstanding that laws do require the president’s agreement, and that the president therefore has leverage to affect a law’s content by threatening a veto, it is the Congress that has the legislative power — not the president. When courts pay more attention to the legislature’s preenactment intent than they do to the president’s postenactment interpretation, they are honoring this distribution of governmental responsibility.

who should interpret statutes?

173

Thus, it should not be surprising that although, as I noted in chap­ ter 4, the federal courts have used “intent” words in close proximity with “Congress” or “Legislature” more than 3,000 times every year from 1988 through 2007, for a total of more than 60,000 times during that period, they have made reference to Presidential Signing Statements only 29 times during that same period.55 In many of those cases, the court mentions the signing statement either as additional legislative history supporting the language of a committee report or the like or as a succinct and seemingly noncontroversial statement of the law’s purpose.56 This pattern is not likely to change significantly, and it should not change. Prosecutorial Discretion as Statutory Interpretation In chapter 1, I discussed the problem of an individual accused of violating a transit authority rule prohibiting movement between subway cars. The law was clear, but signs telling passengers what was prohibited were misleading. I suggested that the best solution would be for the police to be lenient and for prosecutors to act similarly. This prudence would serve to avoid compelling judges to choose between enforcing an unambiguous law that had been applied thousands of times before and taking seriously the right to fair notice. In fact, by not bringing cases in which the law’s applicability can be challenged, prosecutors themselves simulate the rule of lenity. When government brings only cases in which a clear violation of the law has occurred, the system operates within boundaries that bring less controversy and more respect for legal institutions. Dan Kahan argues that the exercise of prosecutorial discretion should be seen as an important element of statutory interpretation through which the executive branch already makes important decisions about how statutes apply to events.57 I agree. In chapter 5, and earlier in this chapter, I discussed statutory inflation: the process though which criminal liability grows without the need for new legislation. But not all statutes exhibit this dynamic. For example, prosecutors most frequently bring criminal bankruptcy charges and criminal copyright charges only in the most egregious cases.58 Limited resources, the availability of civil remedies, and the absence of an agency pushing for aggressive enforcement of the laws conspire to make crimes in these areas less of a priority than those aggrieved by violations would optimally wish. The result is that the set of convictions under these statutes simulates a regime in which the rule of lenity plays a significant role. Convictions

174

chapter 6

occur only in the clearest cases, but it is not the judge who rejects a statute’s expansive interpretation: it is the prosecutor, and the decision is made more on economic grounds than on jurisprudential grounds. Although such restraint indeed may have the stabilizing effect of limiting the application of statutes to situations in which they most unequivocally apply, not all prosecutorial discretion enhances rule-of-law values in this way. For example, a report from the Government Accountability Office issued in July 2008 accused the Wage and Hours Division of the U.S. Department of Labor of systematically failing to investigate adequately complaints of not paying workers their last paycheck, failing to comply with minimum-wage laws, and failing to pay workers overtime.59 In many instances, the department closed investigations after about one year without assigning an investigator at all; in others, it closed investigations after calling an accused employer and taking the employer’s word on the matter; and in still others, it reported that it had not located the employer and so it closed the file. When an agency charged with the enforcement of a law does not enforce it, the result is the same as a judge’s denying that the law applies in the situations in which the agency has decided not to act.60 Thus, prosecutorial discretion can lead not only to the law’s application in the most solid cases, enhancing the rule of law, but also to the failure to act in clear cases, diminishing the strength of the rule of law as a value. The situation becomes more complex still when prosecutors prudently do not apply antiquated laws, such as those prohibiting marriage between the races, at all, thus averting both social and constitutional crises.

Should the Legislature Tell Courts How to Interpret the Law? Nicholas Rosenkranz has suggested that the legislature can do far more than it now does to control judicial interpretation of the statutes it writes.61 Should the legislature put a stop to interpretive practices of which it disapproves? At first glance, it seems like a good idea for legislatures to issue interpretive instructions to the courts. The legislature writes the laws. Principles of legislative primacy suggest that courts should defer to the will of the legislature and not impose their own values. What better way to ensure that the legislative intent makes its way into statutory interpretation than for the legislature to tell the courts how to resolve disputes if and when

who should interpret statutes?

175

they arise. This section explores legislative efforts to instruct courts on interpretive questions. Among the questions asked are whether rules of interpretation prescribed by legislatures are likely to ensure fidelity to the goals of the statute; how far the legislature can use this technique without impinging on the judicial function; and whether some methods of legislative instruction are better than others. Robert Courchesne’s Murderous Night On the night of December 15, 1998, Robert Courchesne murdered Demetris Rodgers by stabbing her to death. At the time, Ms. Rodgers was pregnant with her daughter, Antonia, who was delivered by cesarean section after her mother’s death. Antonia lived for forty-two days and then died from a deprivation of oxygen to the brain.62 In Connecticut, where the murder took place, the death penalty is available if a capital crime has been committed and if certain aggravating factors are present. Among the capital offenses is “murder of two or more persons at the same time or in the course of a single transaction.”63 Among the aggravating factors is that “the defendant committed the offense in an especially heinous, cruel or depraved manner.”64 Courchesne had agreed to a bench trial on the question of his guilt, but a jury would later have to decide whether to impose the death penalty if he was convicted of a capital crime. Courchesne was convicted of capital murder. At that point, the government and the defendant disagreed over whether the prosecution would have to prove that all of the murders making up the capital offense must have been committed “in an especially heinous, cruel or depraved manner” or whether his crime met the statutory requirements if only one of the murders was so performed. The trial court had ruled that the state must prove both murders were performed in the aggravated manner. An appeal was allowed to the Supreme Court of Connecticut, which reversed the trial court’s ruling in March 2003. The case was sent back for a trial in which the jury would be entitled to impose the death penalty if either murder was found to be aggravated as the statute requires. On December 17, 2003, a jury returned a death sentence.65 As of this writing, Courchesne remains on death row in Connecticut. Among the most interesting aspects of this story is a political incongruity that has had significant consequences. At the very least, the Connecticut statute can be read as Courchesne’s lawyers had proposed and the trial

176

chapter 6

court had ruled. The statute defines “the offense” that is eligible for the death penalty as the “murder of two or more persons . . . in the course of a single transaction,” but it also says that an offence is aggravated if “the defendant committed the offense in an especially heinous, cruel or depraved manner.” If the offense includes both murders, then only if both murders were committed in an especially heinous manner could “the offense” have been committed that way. Aware of the logic of this argument, Justice David Borden of the Supreme Court of Connecticut wrote a lengthy, erudite majority opinion justifying a contextual approach to statutory interpretation in which the legislature’s purpose can sometimes trump a statute’s plain language. Examples from the case law and the scholarly literature were adduced and used in argument. Perhaps ironically, the most liberal use of context was brought out to justify a broad interpretation of a death penalty statute. To illustrate, “murder by a kidnapper of a kidnapped person during the course of a kidnapping or before such person is able to return or be returned to safety” is a capital offence in Connecticut.66 The court asked rhetorically why the legislature would decide to impose the death penalty when a heinous murder is committed in the context of a kidnapping but not when a heinous murder is committed in the context of another murder. It answered its own question: We can conceive of no reason why the legislature, in adding the two additional capital felonies that it did in 1980, would have intended that, with respect to a murder committed in the course of a kidnapping or sexual assault, it would be sufficient, with regard to the imposition of the death penalty, for the state to prove the aggravating factor with respect to just one of the underlying felonies, but with respect to the murder of two persons in the course of a single transaction, the state would be required to prove the aggravating factor with respect to both of the underlying felonies.67

The argument is one of legislative intent based on the presumption that the legislature is most likely to have intended to create a coherent statutory scheme. I examined this core argument of statutory interpretation in chapter 4. In my opinion, the decision was wrong notwithstanding the intellectual force of the argument. As we saw in chapters 2 and 5, the rule of lenity requires that criminal statutes be construed narrowly when disputes over statutory meaning arise. The application of this rule is notoriously uneven.

who should interpret statutes?

177

But one line is almost always drawn: the linguistic wall.68 Judges do not always give criminal statutes the narrowest reading consistent with the language in each case, but they rarely interpret a criminal statute so broadly that the interpretation falls outside the range of any reasonable understanding of the statutory language. The court in Courchesne did just that. The Connecticut legislature did not like the Supreme Court’s opinion either. It gave too little deference to the language that the legislators actually used and served to aggrandize the power of the court in legislative matters. In reaction, the legislature enacted a statute that, in effect, made textualism the law of Connecticut and banned the excessive use of context when statutory language is clear. The statute reads: “The meaning of a statute shall, in the first instance, be ascertained from the text of the statute itself and its relationship to other statutes. If, after examining such text and considering such relationship, the meaning of such text is plain and unambiguous and does not yield absurd or unworkable results, extratextual evidence of the meaning of the statute shall not be considered.”69 The statute was effective as of October 1, 2003, leaving no room to argue that it applied to the Courchesne decision itself, which had been handed down some seven months earlier. Was the legislature’s decision wise? Was it an unconstitutional usurpation of the judicial power? If Congress were to pass a similar law, should the Supreme Court uphold it? Justice Borden, who wrote the majority decision in Courchesne, has suggested that the statute might be an unconstitutional usurpation of the judicial power under principles of separation of powers.70 As we will see, other courts have also shown concern with legislative efforts to limit the judicial power to interpret statutes. As a practical matter, the enactment of this statute in Connecticut has probably not changed judicial activity very much. Judges continue to argue about where plain meaning is found, just as we saw in chapter 3, and continue to consider ordinary usage as an appropriate standard, notwithstanding that the statute makes no reference to it. In one case, the majority of the Supreme Court of Connecticut chastised the dissent for deviating “from our well established rule pursuant to which we ascertain the ordinary meaning of a term by consulting the dictionary.”71 Instead, the dissent attempted to rely upon earlier judicial decisions con­stru­ing the word in question.72 In another, the court began with the premise that the goal of statutory interpretation is to “ascertain and give effect to the apparent intent of the legislature,” using the statutory plain-language rule as a principle of evidence, much as we saw in chapter 4.73

178

chapter 6

Moreover, courts applying this statute typically use various canons of construction, notwithstanding that such canons are not statutorily authorized.74 And in one instance, the court held that the principle of stare de­ cisis trumped the statute, refusing to reconsider a decision it had made ten years earlier that did not construe a law in conformance with its plain meaning.75 What the statute has done is to require the courts to structure their argument around an initial finding of plain language and to declare a statute ambiguous before looking toward such aids as legislative history. Abbe Gluck, in an important new study of state court jurisprudence on statutory interpretation, regards such uniformity in approach— even if accompanied by some level of divergence in actual application—as a positive development in statutory interpretation, a development in which state courts and legislatures are taking the lead.76 Legislative Rules of Statutory Interpretation A great deal has been written about how judges sometimes appear to overstep their authority to engage in legislating, which is not their domain.77 The argument often arises when the Supreme Court strikes down a popular statute as unconstitutional and can be heard from both ends of the political spectrum depending upon whose values have been offended in a particular case. For example, scholars, especially (but not exclusively) those with liberal leanings, have been critical of Supreme Court decisions striking down a number of acts of Congress for their inconsistency with principles of federalism.78 Perhaps for conservatives, such rulings feel like “payback” for many years of what they considered judicial activism from the left in which the Court had held invalid various laws, including laws prohibiting abortion, to focus on the most prominent example.79 Much less has been said about the ways that legislatures intrude into the business of the judiciary. The Constitution vests the “Judicial Power” of the United States in “one Supreme Court, and in such inferior courts as the Congress may from time to time ordain and establish.”80 State constitutions often have similar provisions. The Constitution does not define “Judicial Power.” For that matter, the Constitution does not define “Legislative Power” either.81 Whatever the boundaries of the judicial power, surely the interpretation of statutes is part of it. Marbury v. Madison established the courts’ power to review statutes to determine their constitutionality.82 But the job of determining whether a statute applies in a

who should interpret statutes?

179

particular case was considered part of what judges do from the beginning of this country’s history. How courts should interpret statutes has been debated for centuries. Whether courts should interpret statutes has not really been an issue. By the late eighteenth century, the judicial interpretation of statutes had long been the practice in England, and it was inherited without controversy. Law students continue to read and argue about the current relevance of “the Rule in Heydon’s Case,” in which Lord Coke, in 1584, described the task of the judge in terms of both the purpose and the intent of the legislature: “[T]he office of all the judges is always to make such construction as shall suppress the mischief, and advance the remedy, and to suppress subtle inventions and evasions for continuance of the mischief, and pro privato commodo, and to add force and life to the cure and remedy, according to the true intent of the makers of the Act, pro bono publico.”83 Whether one subscribes to either the substance of this approach or its being a fair description of English law at that time, it clearly shows that judges believed themselves to be in the business of interpreting statutes. An interesting debate has arisen between William Eskridge and John Manning over the attitude toward statutory texts that American judges inherited from Britain in the late eighteenth century. Eskridge argues that English judges exercised a great deal of flexibility at the time of America’s infancy, especially flexibility to shrink statutory scope in an “ameliorative” way to prevent their application in unintended circumstances.84 According to Eskridge, at least some of that flexibility was inherited. Manning, in contrast, focuses on a particular device, “the equity of the statute,” used by English judges to avoid harsh results that would seemingly be mandated by strict application of a statute’s language. He then argues that the separation of powers contained in the U.S. Constitution led judges to abandon this judicial power to legislate.85 Whatever flexibility to deviate from textual language American judges inherited, there is no disputing that statutory interpretation was part of the job of judging. But leaving interpretation to the judges is not the only option for a legal system. If the duty of the court is to effectuate the intent of the legislature, then, at least in principle, the legislature can establish rules for the courts to follow that are likely to accomplish this goal. In fact, that is how the civil-law countries of continental Europe and elsewhere work. For example, chapter II of the Italian Civil Code contains the following provision:

180

chapter 6

Interpretation of Statutes. In applying statutes no other meaning can be attributed to them than that made clear by the actual significance of the words, according to the connection between them, and by the legislative intent. If a controversy cannot be decided by a precise provision, consideration is given to provisions that regulate similar cases or analogous matters; if the case still remains in doubt, it is decided according to the general principles of the legal order of the state.86

American legislatures (in addition to Connecticut’s) also frequently enact laws that tell courts how to interpret statutes, and some look remarkably like the Italian Civil Code. A number of states have enacted the following statute: If a statute is ambiguous, the court, in determining the intention of the legislation, may consider among other matters: 1. The object sought to be attained. 2. The circumstances under which the statute was enacted. 3. The legislative history. 4. The common law or former statutory provisions, including laws upon the same or similar subjects. 5. The consequences of a particular construction. 6. The administrative construction of the statute. 7. The preamble.87

Other states have enacted similar statutes, containing slightly different lists of considerations for courts to take into account. 88 These laws are quite general and ultimately not terribly helpful when it comes to deciding cases that present difficult interpretive issues. What can be said for them is that in their generality they appear to give courts great leeway to use whatever information they see fit to make the best decisions they can. For that reason, these enactments do not test the limits of legislative intrusions into judicial prerogatives. Not all statutes governing interpretation are that general, however. Nicholas Rosenkranz has suggested that legislatures do not take as much advantage as they might to guide courts by instructing them on the intended interpretation of the statutes they have enacted. While I agree that legislatures have more power to do so than they exercise, I do not believe

who should interpret statutes?

181

that the value of legislative primacy is well served by their being more expansive in their instructions to judges on the interpretation of statutes. To begin, the boundaries between the legislative power and the judicial power in this domain are not at all clear. Consider the Illinois antitrust statute that instructed courts as follows: “When the language of this Act is the same or similar to the language of a Federal Antitrust Law, the courts of this state in construing this Act shall follow the construction given to the Federal Law by the Federal Courts.”89 The Supreme Court of Illinois held this provision to be an unconstitutional legislative usurpation of the judicial power, which the state constitution confers upon the courts: “The judicial power is vested solely in the courts, and the legislative branch is without power to specify how the judicial power shall be exercised under a given circumstance; it is without authority to state explicitly how the judiciary shall construe a statute.”90 More recently, the Supreme Court of Delaware struck down a statute intended to override a ruling of that court in a criminal case by legislating a textualist approach to statutory interpretation, an approach that would favor the government in the underlying prosecution.91 In Evans v. State, the court held that it violated the state constitution for the legislature to pass a law mandating a particular decision in a particular case.92 But the court then went further. Citing Marbury v. Madison, the Delaware court held: House Bill No. 31 is unconstitutional in other respects as well. The United States Supreme Court has recognized that the “essential balance” created by the separation of governmental authority in the existing state constitutions and the proposed United States Constitution “was a simple one. The Legislature would be possessed of the power to ‘prescribe the rules by which the duties and rights of every citizen are to be regulated,’ but the power of ‘the interpretation of the laws’ would be ‘the proper and peculiar province of the courts.’ ” House Bill No. 31 states that the General Assembly asserts its “right and prerogative to be the ultimate arbiter of the intent, meaning, and construction of its laws and to vigorously defend them.” House Bill No. 31 also establishes specific standards for judicial officers to apply when interpreting or construing Delaware law. Those provisions in House Bill No. 31 attempt to confer upon the General Assembly fundamental judicial powers. Consequently, those provisions in sections 5402 and 5403 of House Bill No. 31 also violate Article IV of the Delaware Constitution.93

182

chapter 6

Nonetheless, as Rosenkranz argues persuasively, surely it cannot be unconstitutional per se for a legislature to instruct courts on the interpretation of statutes. For example, the federal Dictionary Act tells interpreters of statutes that “words importing the singular include and apply to several persons, parties, or things” and that “words importing the plural include the singular.”94 Should courts find that provision objectionable? Not ordinarily. Dictionary acts typically formalize conventions that are easy enough for the legislature to follow and not terribly controversial. They are not often referred to but at least have the potential to make it hard for lawyers to make language-based arguments that have no chance of prevailing as a practical matter. It is not difficult, however, to imagine cases where serious questions may arise over whether a court should follow this statute. In an interesting article that discusses both the Connecticut and Delaware statutes that I have discussed here, Linda Jellum analyzes which legislative actions are most likely to violate constitutional separation-of-powers principles.95 According to Jellum, legislative directives to courts that define terms are almost sure to be held to be constitutional. Statutory definitions are certainly acceptable. Consider, for example, Brogan v. United States, decided by the Supreme Court of the United States, in which the Court addressed a federal statute that prohibited making false statements to a federal agency.96 In that case, Brogan falsely said “no” to an FBI agent who came to his house in the middle of the night and asked him if he had accepted payments from a company whose employees were members of the union in which he was an officer. It is illegal to accept such payments. However, the statute of limitations had expired on his illegal conduct, so he could no longer be prosecuted. Instead, he was prosecuted for making a false statement to a federal official about not having committed the crime for which he could no longer be prosecuted. The relevant statute at the time read: “Whoever, in any matter within the jurisdiction of any department or agency of the United States knowingly and willfully falsifies, conceals or covers up by any trick, scheme, or device a material fact, or makes any false, fictitious or fraudulent statements or representations, or makes or uses any false writing or document knowing the same to contain any false, fictitious or fraudulent statement or entry, shall be fined not more than $10,000 or imprisoned not more than five years, or both.”97 Brogan asked the Court to apply what had been known as the “exculpatory no” doctrine. Most courts of appeals had declined to apply this statute to cases in which an individual did no more than deny

who should interpret statutes?

183

guilt to a federal agent. But the Second Circuit rejected this theory in Brogan’s case, and the Supreme Court took the case to resolve conflicting positions of various courts of appeals. Notice that the words “statements or representations” is in the plural, while “false writing” is singular. Read literally, the statute really does require proof that the defendant made more than one false statement, which is consistent with the plain language of the statute. However, the very first section of the United States Code, part of the Dictionary Act, states that singular words “include and apply to several persons, parties, or things,” while plural words “include the singular.”98 The Supreme Court had no trouble ruling in favor of the government and did not even consider the constitutionality of the federal Dictionary Act, which brought that holding within the statute’s language. Excluding Legislative History: Rules of Evidence for Judges Only Our legal system relies on rules of evidence to ensure orderly trials in which reliable evidence is put before the jury and unreliable, confusing, or unnecessarily cumulative evidence is not put before the jury. Although evidence rules apply to trials held judges as well as before juries, it is the presence of a group of lay citizens as decision makers that motivates the rules. When a judge rules a piece of evidence inadmissible at a jury trial, the jury never sees it. When a judge rules that same piece of evidence inadmissible at a bench trial over which he presides, he of course does see the evidence and orders himself, as a legal matter, to ignore it. The concerns reflected in the evidence rules are reasonable enough. For example, evidence must be authenticated.99 For samples of blood and other physical evidence, chains of custody must be established. For recordings, someone must testify as to how the recording was made and how it has been maintained since then. For documents, authorship must be established. And so on. The goal is to keep jurors from drawing inferences from evidence that is not what it purports to be. Along these same lines, reliability is a concern. Witnesses who testify must be competent.100 The various rules against hearsay are designed to reduce the volume of unreliable evidence at a trial.101 Sensibly enough, the system does not consider it fair for me to testify that Bill told me that John stole some cheese as proof that John really did steal the cheese. It would be too difficult for the defendant to test the reliability of the statement unless Bill is there to say why he said John stole the cheese.

184

chapter 6

Other evidence rules deal with inferences that jurors may draw from evidence that is authentic but that may mislead the unwary. The most versatile is Rule 403 of the Federal Rules of Evidence, which permits courts to exclude evidence if its prejudicial effect is likely to outweigh its probative value: “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.”102 It is up to the trial judge to decide how to apply this rule. On appeal, the standard is very deferential, so the trial judge basically has the final word. Some kinds of evidence are considered so inherently inflammatory that the rules ban their admission without the usual balancing of probative value against prejudicial effect. A rule bans the introduction of evidence to prove that the defendant is a person of bad character and thus must have acted in conformity with these negative traits.103 Evidence of prior crimes is especially prejudicial. Once a juror knows that the defendant is a possible repeat offender, she is likely to infer that the defendant had it in him to do it again. The rules of evidence, therefore, bar such evidence, except in limited circumstances.104 For the most part, though, not much is excluded per se. With this as background, it seems extreme to impose on appellate judges an absolute obligation to ignore legislative history as evidence of the intent of the enacting legislature. For one thing, in jury trials, judges typically have discretion to evaluate evidence to determine whether it is likely to do more harm than good. It seems only consistent with the judicial function that appellate judges should have that much discretion to decide whether reliance on a particular piece of legislative history is likely to do more harm than good. To impose a per se rule against using legislative history is to liken it to the few dangerous types of evidence— unauthenticated evidence, prior convictions, the defendant’s character, and so on—that are excluded per se. Can we really say that a Supreme Court justice’s propensity to rely upon a Senate report is as much a threat to the system as a juror’s propensity to focus on the fact that this defendant had committed crimes in the past? Those concerned with the issue seem to be saying: “Some of us may sometimes use evidence of legislative intent that is unreliable. To protect the system from this happening, no appellate judge may consider legislative history, whether or not the judge can defend its use in that particular case.”

who should interpret statutes?

185

Yet it would be a mistake to base an argument on the notion that courts have free rein to decide what to consider and what not to consider in making legal decisions generally. A close look at federal law and the laws of many states shows that legislatures really do enact statutes designed to keep judges from considering possibly relevant information in making legal decisions. For example, in deciding whether an arbitration agreement should be enforced, some judges may tend to overstep their authority and to say “no” if it looks like the claimant is making nonsensical claims. The Uniform Arbitration Act, adopted in a number of states, contains a provision that prohibits judges from looking at the merits of the case in determining arbitrability.105 Other statutes set similar limitations in a variety of circumstances. California’s Civil Procedure Code requires mediation in cases with less than $50,000 in controversy. The statute warns judges that “in making this determination [concerning the amount in controversy], the court shall not consider the merits of questions of liability, defenses, or comparative negligence.”106 And an Illinois domestic-relations statute prohibits courts from considering “conduct of a present or proposed custodian that does not affect his relationship to the child” in determining the best interest of the child in custody disputes.107 The goal of these laws is to prohibit the consideration of evidence that is not relevant to the task at hand but that a court is likely to take into account anyway. In that sense, they are similar to the rules of evidence that prohibit jurors from considering prior convictions and other such things. Other statutory provisions exclude evidence to prevent judges from being too soft on parties who miss deadlines. A federal statute authorizes an agency to take disciplinary action against customs brokers, subject to a right of judicial review. However, the reviewing court “shall not consider any objection to the decision . . . unless that objection was raised before the hearing officer in suspension or revocation proceedings unless there were reasonable grounds for failure to do so.”108 Similar statutes apply to judicial review within other administrative schemes.109 Still other statutes contain evidentiary rules that appear designed to affect the outcome of litigations brought under the statute. A federal statute allows people to bring lawsuits against cable television operators that fail to make available the minimally required channel capacity on reasonable terms. The statute instructs the trial courts: “In any such action, the court shall not consider any price, term or condition established between an operator and an affiliate for comparable services.”110 Clearly, this will

186

chapter 6

disadvantage plaintiffs who attempt to claim that the terms offered them are unreasonable. What can be more relevant than evidence of comparable services? Somewhat less controversial, the Delaware Child Protection from Domestic Violence Act makes sure that courts do not penalize parents who violate custody or visitation orders to protect themselves or their children from domestic violence.111 There are many other evidentiary laws directed at limiting the way that courts make legal decisions, a significant number of which involve laws governing domestic relations. To take just one, Ohio encourages adoption by prohibiting courts from considering the age of the potential adoptive parent in determining whether the interest of the person sought to be adopted is being served.112 Thus, legislatures often impose rules of evidence on judges to regulate the way that fact-based legal decisions are to be made. Does this mean that legislatures ought to consider ruling out the use of legislative history by judges attempting to discern legislative intent? As we have seen, at least one state Supreme Court has held such laws to be unconstitutional intrusions into the judicial power. In contrast, Jellum, in her thoughtful article, characterizes such laws as “interpretive directives” that do not unconstitutionally restrict the courts.113 Putting aside whether these laws would be constitutional, there is a very sharp difference between the evidentiary laws I described earlier and the broader directive we are considering now. Just about all of the evidentiary statutes discussed above are applicable in very specific situations. They anticipate reliance upon a particular sort of evidence that a judge may unadvisedly consider in a limited factual setting. Many reduce the judge’s discretion to be lenient with those who miss deadlines. Others prevent a judge from considering the merits of a case when her job is to make rulings on procedural matters only. Still others, especially in the domain of domestic-relations law, make it easier or harder for parties to prevail in particular cases as part of the substantive policy of the statute. What these statutes do not do is to tell a judge how to be a judge with respect to questions that come up routinely in the course of deciding cases. Jellum distinguishes between specific and general directives as well, and I agree with her position. In fact, a number of states in addition to Connecticut do have laws that refer to the use of legislative history by courts interpreting statutes. They typically support the procedure. In most instances, legislative history is one of the tools that the legislature proposes that courts use. The only

who should interpret statutes?

187

mention of any limit on the use of legislative history is contained in the Oregon statute, which says: “A court may limit its consideration of legislative history to the information that the parties provide to the court. A court shall give the weight to the legislative history that the court considers to be appropriate.”114 The statute seems to pay lip service to legislators concerned about judicial activism, without having any substantive bite, because the statute does not require that courts limit their consideration of legislative history and certainly does not eliminate its use entirely. Interestingly, as Gluck points out, the Supreme Court of Oregon had earlier ruled that legislative history may be used only after analysis leaves legislative intent unclear, leaving somewhat uncertain the relationship between the judicial and legislative perspectives.115 Legislatures can and do limit the evidence that courts may consider when making legal decisions, and legislatures can and do instruct courts with respect to the use of legislative history. But these laws do not bar the use of legislative history the way Rule 404 of the Federal Rules of Evidence bars the introduction of character evidence except under limited circumstances. Limits on the use of legislative history do exist, however, as part of individual statutes. Consider the following federal statute that sets up an appeals process for those denied certain grants from the Department of Housing and Urban Development: “Following notice of intent to withhold, deny, or cancel assistance under paragraph (1) or (2), the Secretary shall provide a period of not less than 90 days in which the applicant can appeal. . . . Notwithstanding any other provision of this section, nothing in this section or in any legislative history related to the enactment of this section [enacted August 22, 1974] may be construed to permit an inference or conclusion that the policy of the Congress in the urban development action grant program is to facilitate the relocation of businesses from one area to another.”116 The statute itself tells courts not to try to squeeze from the legislative history inference of a policy at odds with the actual purpose of the legislature. The Civil Rights Act of 1991 contains a similar admonition: “No statements other than the interpretive memorandum appearing at Vol. 137 Congressional Record S 15275 (daily ed. Oct. 15, 1991) shall be considered legislative history of, or relied upon in any way as legislative history in construing or applying, any provision of this Act . . . that relates to Wards Cove —Business necessity/cumulation /alternative business practice.”117 Even such small limits on the judicial function are not easy to find.

188

chapter 6

Slightly more common are statutes that make reference to legislative history as a source of information that the executive will need in performing the tasks that the statute authorizes or requires. These statutes actually encourage its use in applying the law. Consider the federal statute that deals with the procedure for displaying the military budget, which requires the secretary to identify and explain “any inconsistencies between (A) the execution, during the previous fiscal year and the current fiscal year, of programs and activities of the Department of Defense combating terrorism program, and (B) the funding and specification for such programs and activities for those fiscal years in the manner provided by Congress (both in statutes and in relevant legislative history).”118 Or consider this statute from the state of Washington, which deals with capital projects: “For the purposes of this section, the legislature intends that each project be defined as proposed to the legislature in the governor’s budget document, unless the legislative history demonstrates that the legislature intended to define the scope of a project in a different way.”119 Both of these laws require government to use legislative history to determine legislative intent. Thus, although there are laws that act as rules of evidence for judges making legal decisions, and there are legislative pronouncements about the use of legislative history, most evidentiary rules apply to specific situations and do not tell judges to stop referring to it globally. To the contrary, almost every one that mentions legislative history assumes that courts and others will make use of this information, and some instruct courts to use it globally, as the judges see fit. When a legislature, in a particular statute, tells the courts not to draw specific inferences about its intent from the legislative history, it does not usurp the judicial function. It merely makes its intent clearer. Were the legislature, in contrast, to tell judges that they must never look at information that they have traditionally considered useful in gleaning the intent of the legislature, the legislature would be intruding on the judicial function. Legislatures, from time to time, set limits on the information that a court may consider in making legal decisions. But, with very few exceptions, they do not go so far as to bar the use of legislative history. Whether or not they can legally do so, given constitutional separation of powers, it would be a bad idea in any event. Legislative history, as we saw in chapter 4, is most often used to confirm the legitimacy of a particular reading of a statute that is largely justified on other grounds. Of course, appellate courts may use their hierarchical superiority to tell trial courts that they are not to look at legislative history in making

who should interpret statutes?

189

decisions about statutory interpretation. But it is not clear what this would mean. Trial court judges cannot be disciplined for thinking about legislative history. The worst that would happen is that they would be reversed. To avoid even that slight embarrassment, all they would have to do is to look at the legislative history to confirm their sense of legislative intent and just not tell anyone about it. And look they will, given the psychological attraction of historical argument, as we have already seen. Like evidence of character or of a prior criminal record, evidence of legislative history can be misused. We saw examples of the unconvincing use of legislative history in chapter 4. But unlike character and criminal records, which are banned per se under the law of evidence except in limited situations, legislative history cannot be misused by a jury, and sophisticated judges are certainly talented enough to use the information wisely at least much of the time. Many of Justice Scalia’s opinions that disapprove of the use of legislative history are concurring opinions—not dissents.120 Even he, then, usually agrees with the result that the Court reaches when it uses this information. As a consequence of all these facts, I think it is a better idea to trust judges to do their jobs with care than it is to impose upon them hard and fast rules requiring them to ignore evidence that they might otherwise have found helpful.

Courts Fight Back While both the executive and legislative branches attempt to wrest interpretive power from the courts—sometimes with the active cooperation of the courts themselves—the judges do not always sit idly by. As we will see, although some legislatures have removed the rule of lenity legislatively, courts apply it anyway. They also impose drafting rules on the legislature, requiring clear statements if the legislature is to accomplish otherwise legitimate legislative goals. Finally, the Chevron doctrine, discussed at the beginning of this chapter, is not as hard and fast as it first appeared in light of subsequent judicial decisions. Lenity and Due Process The rule of lenity, as we have seen, tells judges that when a criminal statute is unclear in its scope, the uncertainty should be resolved in favor of the defendant.121 Beginning in the late nineteenth century, a number of states

190

chapter 6

eliminated the rule of lenity by statute. In New York, for example, the Penal Law contains the following provision, which has been part of New York law since 1881: “The general rule that a penal statute is to be strictly construed does not apply to this chapter, but the provisions herein must be construed according to the fair import of their terms to promote justice and effect the objects of the law.”122 This change to the code, enacted as a reaction to what was perceived as judicial activism, has not made much difference. Given that (1) the values of fair notice and legislative primacy are deeply embedded in our system, (2) language is by its nature vague and ambiguous at times, (3) context is sometimes, but only sometimes, able to remediate vagueness and ambiguity, and (4) lenity is applied only in cases in which the legislative purpose with respect to a series of events is not clear after investigation, we should expect that (5) lenity is a virtual necessity. A review of the experience in New York suggests that this is so. In case after case, the courts continue to construe criminal statutes narrowly when context does not make it obvious that the legislature intended disputed conduct to fit within the scope of the statute. This was the case when Livingston Hall observed the practice disapprovingly seventy-five years ago, and it is still true today.123 For example, in People v. Phyfe, an 1893 case decided shortly after the legislature enacted the interpretive rule, the New York Court of Appeals made it clear that defendants were still entitled to fair notice: “The citizen is entitled to an unequivocal warning before conduct on his part, which is not malum in se, can be made the occasion of a deprivation of his liberty or property.”124 Seventy-three years later, the courts were still saying the same thing. In People v. Sansanese, the New York Court of Appeals held that a false driver’s license should not be deemed an “instrument” for purpose of a statute prohibiting the filing or recording of forged instruments. The court reasoned: “While on the one hand we must not be overly technical in interpreting penal provisions, on the other hand ‘Penal responsibility . . . cannot be extended beyond the fair scope of the statutory mandate.’ ”125 Moreover, New York courts read the statute to require them to look outside the words of the statute, to “the policy considerations underlying the statute and the ultimate results sought by the Legislature.”126 Using this approach, the courts have found, notwithstanding section 5.00 of the Penal Law, that “it is impossible to obtain a conviction under an ambiguous statute unless the defendant is clearly guilty under all reasonable interpretations of that statute.”127

who should interpret statutes?

191

When Courts apply the statute, they often use it to reject unnaturally narrow readings of laws whose applicability is relatively transparent in the case under dispute. Thus, despite a long history with a penal code that rejects lenity, New York courts continue to impose lenity when they have nothing better to say about how a statute should be interpreted. This may not occur often, but it does occur periodically, as it has for decades. California, which abolished strict construction of its criminal code in 1871, has seen a similar response from the courts.128 Although Hall listed California as among the enlightened states whose courts abided by its liberal construction statute, the actual record is mixed.129 Thus, in 1890, the Supreme Court of California noted in the course of ruling against the state on a criminal appeal: “While it is true the rule of the common law that penal statutes are to be strictly construed has been abrogated by the Code . . . it is also true that the defendant is entitled to the benefit of every reasonable doubt, whether it arises out of a question of fact, or as to the true interpretation of words, or the construction of language used in a statute.”130 Decades later, in 1932, a California appellate court held that a courtappointed receiver’s embezzlement of an estate was not a violation of a statute prohibiting the embezzlement of public monies, notwithstanding the receiver’s status as an officer of the court. The court explained: “Where the language employed is ambiguous or doubtful in its intent, a construction of the statute should always be favorable, rather than unfavorable, to any person accused of a violation of the law. In other words, although the common-law rule that penal statutes be strictly construed has no application to the provisions contained in the Penal Code, nevertheless a criminal statute should be construed according to the fair import of its terms.”131 This language is double-talk. The court was trying, unsuccessfully, to reconcile its common-law handling of this case with the dictates of a statute that demanded the contrary. Probably the most famous case in which the California courts have refused to construe a criminal statute liberally is Keeler v. Superior Court, in which the Supreme Court of California decided a fetus is not a “human being” for purposes of construing California’s homicide statute.132 Keeler, who had discovered that his estranged wife was pregnant, kicked her in the abdomen during a confrontation, cracking the fetus’s skull. Justice Stanley Mosk, who wrote the majority opinion, resorted to lenity. He noted that, throughout California history, “human being” had meant a human being born alive for purposes of the statute: “It is the policy of this state to

192

chapter 6

construe a penal statute as favorably to the defendant as its language and the circumstances of its application may reasonably permit; just as in the case of a question of fact, the defendant is entitled to the benefit of every reasonable doubt as to the true interpretation of words or the construction of language used in a statute.”133 After quoting Marshall’s warning in Wiltberger that courts should not expand criminal statutes to criminalize all instances of the evil with which the legislature had concerned itself, Mosk concluded: “Whether to thus extend liability for murder in California is a determination solely within the province of the Legislature. For a court to simply declare, by judicial fiat, that the time has now come to prosecute under section 187 one who kills an unborn but viable fetus would indeed be to rewrite the statute under the guise of construing it.”134 Thus, Mosk not only presented the rule of lenity as a basis for the court’s decision but resorted to its underlying separation-of-powers rationale, Marshall and all. John Jeffries agrees with the holding in Keeler but would reach the result by a different route. He argues that the holding in Keeler is attractive not because it promotes strict construction but because it furthers the rule of law as a value.135 The two may not be separable, however. Notice and legislative primacy are important values to us precisely because they promote the rule of law. For this reason, Jeffries’s suggestion that lenity can be replaced by canons that call for decisions consistent with rule-of-law values may not fully solve these interpretive issues. The bottom line is that courts sometimes do not know what to do when asked to interpret a statute. Lenity best promotes deeply held values when that situation arises. This is not to say that California courts never give criminal cases the liberal construction that the penal code requires. In recent years, California’s Supreme Court has been a conservative one. It has not been afraid to interpret criminal statutes liberally and has even challenged the boundaries of the linguistic wall.136 A broader survey of the California judiciary, however, shows that in other cases courts analyze the facts in terms of the narrow rule of lenity. In some of these cases, the court rejects the narrowest possible interpretation after engaging in the analysis.137 In other cases, courts apply lenity to resolve the case in favor of the defendant.138 In neither New York nor California have the courts been willing to give up lenity, especially when it is applied narrowly. The courts continue to feel that a tiebreaker is needed despite legislative action seemingly to the contrary. Our concept of the rule of law combines with limits in our linguistic competence to make the rationales that underlie lenity robust to the point of being virtually indestructible.

who should interpret statutes?

193

Clear-Statement Rules and Judicial Aggrandizement Just as legislatures write rules by which courts are expected to abide, courts write rules instructing legislatures how they must behave. Perhaps the best example is the use of “clear-statement rules,” through which the courts require a strong indication of legislative intent before they will rule in a particular direction. A well-studied case is Gregory v. Ashcroft, decided by the Supreme Court in 1991.139 The issue there was whether the federal Age Discrimination in Employment Act applies to state judges, making it illegal for states to impose mandatory retirement ages on their own judiciaries.140 The question was an important one, because some states actually have constitutionally based retirement ages for judges. The statute exempts certain government employees by its own terms: “The term ‘employee’ means an individual employed by any employer except that the term ‘employee’ shall not include any person elected to public office in any State or political subdivision of any State by the qualified voters thereof, or any person chosen by such officer to be on such officer’s personal staff, or an appointee on the policymaking level or an immediate adviser with respect to the exercise of the constitutional or legal powers of the office.”141 The Supreme Court held that the Missouri judges involved in the case were “appointees on the policymaking level” and thus not covered by the act. It would have been easy enough for the Court to have used conventional reasoning about legislative intent — including arguments about the traditional relationship between the state and federal governments—to support the inference that Congress would not have wished to upset the manner in which states organized their judiciaries. But the Court went further. Citing an earlier decision, it held that “Congress should make its intention ‘clear and manifest’ if it intends to preempt the historic powers of the States.”142 This plain statement rule sets conditions on Congress’s power to legislate. Nothing in Article I, section 7, of the Constitution suggests any such requirement, whether in the context of federalism or in any other context. A concurring opinion by Justice White made this point strongly, as does some scholarly commentary.143 Clear-statement rules have stayed in the forefront in recent years in debates over the executive’s power in conducting the war on terror. In Hamdan v. Rumsfeld, the Court held that the traditional forms of adjudication should be presumed unless Congress has legislated exceptional procedures in clear language.144 This, too, has triggered debate in the scholarly literature about the presidential power in a time of emergency.145

194

chapter 6

I will not develop these arguments here, other than to observe that the courts continue to dictate to the executive and legislative branches the language that the legislature must use to accomplish certain goals. In the case of Hamdan, it may well be that the Court is more concerned with an excessively broad interpretation of a legitimate statute by the executive than it is with legislative overreaching.146 Where Chevron Stands By definition, judicial deference to the executive branch cedes interpretive territory from one branch of government to the other. Earlier, we saw that this deference is tempered to some extent by political expediency. It has also been limited by the Supreme Court itself. Most significant is the Supreme Court’s 2001 decision in United States v. Mead Corp. There, the Court held that letter rulings by the Customs Service are not entitled to deference under Chevron. The Court ruled “that administrative implementation of a particular statutory provision qualifies for Chevron deference when it appears that Congress delegated authority to the agency generally to make rules carrying the force of law, and that the agency interpretation claiming deference was promulgated in the exercise of that authority. Delegation of such authority may be shown in a variety of ways, as by an agency’s power to engage in adjudication or notice-and-comment rulemaking, or by some other indication of a comparable congressional intent.”147 That did not, however, end the story. The Court further held that while they were not entitled to the full deference that Chevron would afford them, the letter rulings were entitled some deference. After all, the Customs Service is highly experienced in categorizing goods for the purpose of appropriately assessing customs duty. Before Chevron, this was the deference given agencies in general under the 1944 case Skidmore v. Swift & Co.148 Thus, there is now a hierarchy of deference that the courts may afford administrative agencies, depending on whether the legislature has delegated rule-making authority to the agency carrying the force of law, whether the agency action in dispute was taken pursuant to any such authority, the expertise of the agency, and other factors. To use Peter Strauss’s expression, there is a “rulemaking hierarchy,” and it is up to the courts to decide where a particular executive action fits on that hierarchy.149 How do courts exercise this broad discretion? A study by Eskridge and Baer examining over one thousand agency-interpretation cases shows that

who should interpret statutes?

195

more than half the time (53.6 percent) the Supreme Court does not apply any deference principles at all.150 It simply rules in favor of or against the agency on an ad hoc basis, never even discussing the various deference principles. More generally, when the Court does rely on deference principles, it defers on a continuum ranging from no deference in criminal cases because of the rule of lenity to superstrong deference in cases involving issues of national security.151 Kristin Hickman and Matthew Krueger’s study of the Court’s Skidmore deference practices since Mead reached a similar conclusion: the Court treats deference on a sliding scale.152 What is inescapable is the conclusion that the Supreme Court has not allowed itself to become a passive player in all of this. Whether formally or in practice, the judiciary has maintained its power to construe statutes even in the teeth of administrative action when it believes that it is appropriate to do so.

What Should Happen Next? Statutory interpretation is not a matter left to a single branch of government. All three branches get into the act: the executive, through the actions of agencies, prosecutors, and presidential statements; the legislature, through its enacting laws that tell judges what they must consider, what they may consider, and what they may not consider; and the judiciary, whose job it has traditionally been to interpret statutes. Legislatures, with a few notable exceptions, have directed most of their attention either to general, liberal rules of interpretation permitting courts to act with discretion or to bits of specific evidence relating to particular fact-finding pursuant to specific statutes. I see no reason to believe that, at this point, judges have too much power in this complex scheme and that more of it should be stripped from them. I thus disagree with those who argue such a position. There is no convincing evidence that other branches or agencies will routinely do more to enforce the will of the legislature than judges do now. As for deference to the executive branch, there is certainly room for far more candor and perhaps the development of analytical rigor in the distribution of interpretive control of statutory law.

chapter seven

Jurors as Statutory Interpreters

W

e have just seen how the different branches of government fight over ownership of statutory language and the right to construe it in equivocal cases.1 Here, we look at a different fight over ownership of statutory words, a fight that is played out in courtrooms around the United States every day: how much is statutory interpretation a matter for the judge, and to what extent should the jury be permitted to decide whether the facts of a case establish the elements of a crime? In the United States, criminal law is entirely statutory: there are no common-law crimes.2 At least in principle, it is up to the court to decide as a legal matter whether particular conduct falls within a statute’s prohibitions. All the jury need do, again in principle, is to decide whether the government has proven beyond a reasonable doubt that the defendant did all of the things that make up the various elements of the crime. This was not always so. From the time of the country’s founding until the end of the nineteenth century, criminal juries were empowered to determine the law as well as to find the facts.3 The practice was considered an important part of the structure of democratic government, dating back to seventeenth-century England.4 During the colonial period of American history, it took on special significance. One way to make sure that the English government did not subject the colonists to oppressive laws was to continue the common-law tradition of placing twelve citizens as a buffer between Parliament and the criminal-justice system. What we now

jurors as statutory interpreters

197

call nullification was considered an important protection against tyranny.5 In fact, one of the injuries listed in the Declaration of Independence was deprivation, “in many cases, of the benefit of trial by jury.”6 Upon the founding of the Republic the practice stood and was taken seriously as one of the institutions that protected democratic values in the context of governmental structure that allowed for little participation from the citizenry. Alexander Hamilton wrote in The Federalist: “The friends and adversaries of the plan of the convention, if they agree in nothing else, concur at least in the value they set upon the trial by jury; or if there is any difference between them it consists in this: the former regard it as a valuable safeguard to liberty; the latter represent it as the very palladium of free government.”7 If the legislature and executive became intoxicated with their own power and passed oppressive laws, they would still have to get past a jury to obtain a conviction—no easy feat, it was predicted.8 During the first hundred years of this country’s history, as judges became more professional and as the value placed in uniform treatment by the law came to surpass concerns about tyranny, the role of the jury shrank. By the end of the nineteenth century, judges were instructing the jury in detail on the law, and jurors were expected to abide by these instructions.9 As the Supreme Court put it in 1895, “it is the duty of juries in criminal cases to take the law from the court and apply that law to the facts.”10 Thus ended the jury’s official role as statutory interpreter. Nonetheless, the maxim that jurors find the facts while judges articulate the law is an overstatement. Jurors are not limited to responding to special verdict questionnaires, which a judge or some functionary uses to determine whether the law was broken. That may be true with such matters as adding court costs or interest on to a judgment, but it does not characterize the main event of the trial process. Rather, juries enter verdicts, first by finding the facts and then by determining whether the facts as they found them constitute a violation of the law. Deciding whether a statute applies to a given set of facts is what judges do when they interpret statutes, and it is part of the jury’s job as well. Scholars have taken note of this similarity in function between judge and jury, especially in the context of jury nullification. Darryl Brown has suggested that examining the deliberations of nullifying juries provides evidence that jurors today engage in a great deal of statutory interpretation, notwithstanding the official story to the contrary.11 Relying on filmed deliberations in a small sample of actual cases, Brown argues that what may appear superficially to constitute nullification by jurors of the legal

198

chapter 7

standards they have been instructed to apply may instead reflect jurors deciding whether guilt in a particular case would promote the justice values for which the statute was enacted. Such discussion closely resembles the discourse of statutory interpretation in which courts and legal academics engage. Brown proposes a model of practical reasoning to account both for the decisions that courts make about a statute’s applicability under various circumstances and for the decisions of jurors to convict or acquit. Approaching the role of the jury from a more historical perspective, Nancy Marder argues similarly.12 Marder compares the current conventional model of the jury with what she calls a process model and concludes that what is typically seen as nullification may be recast as jurors participating actively in a system designed to give them authority to determine that a law is inapplicable in a given situation. In this chapter I look more closely at how the legal system deals with jurors as interpreters of statutes. Judges use a host of tools in interpreting statutes, many of which are not available to jurors. For example, to the extent that legislative history is useful in determining the intent of the legislature, jurors are clearly not the ones to conduct the relevant analysis. Nonetheless, just as the jury can be seen as an institution that filters legal proceedings through a “communal sense of right and wrong” before convicting a person of a crime,13 some aspects of statutory interpretation ordinarily performed by judges do precisely the same thing. Among them are the rule that requires judges to interpret statutes to avoid absurd results and the rule that the words in a statute are to be interpreted according to their ordinary meaning. Each of these canons of construction is designed to make sure that the substantive law as written does not become so overreaching as to defeat reasonable understandings of the citizenry — just what the jury is supposed to do. Moreover, jurors do not enter the jury room as blank slates. Rather, they perform their duty with a wealth of prior knowledge about what typically characterizes crimes to which they are exposed in our culture, whether from newspapers, books, magazines, or television. Jurors are typically asked in voir dire whether they or members of their families have been victims of the types of crime for which the defendant is on trial, but even without such experience people develop a sense of the characteristics of a prototypical rape, burglary, robbery, etc. Experimentation shows that it is hard for people to put these notions aside and to replace them with the

jurors as statutory interpreters

199

statutory definitions contained in the instructions that the judge reads to them.14 Judges, in contrast, with their legal training and experience, may not succumb to precisely the same set of influences as jurors. This suggests that the jury should be especially suited to determine a statute’s “ordinary meaning” and to decide whether the defendant’s conduct fits within that meaning.15 In addition, how much interpretation is delegated to jurors depends in large part on how they are instructed. Some instructions appear to leave a great deal of interpretive discretion to jurors, while others attempt to control the process with such precision that the instructions are virtually incomprehensible in their length and detail. The type of instruction that a jury is likely to hear is, in turn, partly a function of how the legislature drafted the law that the instruction is intended to describe. A detailed statute that contains its own glossary, for example, is likely to lead to instructions that reflect this legislative choice. Statutes drafted more skeletally leave room for different styles of instruction. The system appears to be ambivalent about the jury’s role in a scheme that purports to seek uniform treatment through precise and sometimes complex statu­ tory law.

Modern Jurors as Statutory Interpreters: How Issues of Statutory Interpretation Arise Statutory interpretation is traditionally seen as the business of the court. Judicial statements like the following are commonplace: “The interpretation of a statute is a question of law which this Court reviews de novo.”16 Jurors are not part of this picture. But the lines are not so clear once one considers how questions of statutory interpretation make their way to the appellate courts that ultimately decide the questions. One procedural context intended to remove the jury from the process is a motion to dismiss an indictment or civil complaint. When a court grants a motion to have the charges dropped because the statute does not properly apply to the defendant’s alleged conduct, the jury never hears the matter. In fact, taking the case away from the jury is the defense lawyer’s goal in making the motion. When the defendant loses the motion to dismiss, or never makes such a motion, the case may ultimately go to trial, and the judge decides how to

200

chapter 7

instruct the jury. At that point, the court has two alternatives. It can simply instruct the jury on what the statute says and allow the jurors to decide whether the facts as established violate the statute, or it can give specific instructions on the meanings of the statutory terms that specifically establish a relationship between the evidence in the case and the statute. In one instance, the jury is given considerable latitude not only to find the facts but also to determine whether the statute covers the facts as the jury found them. In the other, the court attempts to keep the jury on a “short leash.” 17 Whichever approach the court takes, a subsequent appellate opinion that deals with the issue of statutory interpretation is procedurally about the instructions to the jury in the case. We saw many examples of this in chapter 2’s discussion of the federal bribery statute. Both of these very different approaches to instructing the jury can be found in the case law on statutory interpretation. Consider Smith v. United States, the case discussed in chapter 3, in which the defendant was in the process of trading an unloaded machine gun for some illegal narcotics when he got cold feet and decided to run.18 The appeal was about the jury instructions at the trial. Smith argued that the trial court should have instructed the jury that trading a gun for drugs does not constitute “use” for purposes of the statute.19 The court of appeals rejected this contention as a matter of statutory interpretation. The jury was not, however, instructed in the opposite direction either: that use in trade does constitute “use” for purposes of the statute. Rather, at trial it was up to the jury to determine whether Smith had violated the statute by his conduct. That made jurors interpreters of the statute as well as finders of fact. On the other hand, since the Smith decision was announced, prosecutors have been able to argue that they are entitled to an instruction that “using” a firearm during a drug-trafficking crime includes trading it for illegal drugs. Smith suggests that jurors play a significant role as statutory interpreters unless and until an appellate court rules in a published opinion on a particular issue concerning the statute’s applicability. Since there are many issues that are never resolved in that way, jurors remain statutory interpreters with respect to portions of many statutes. Moreover, even when an appellate court does decide an issue, it is possible to construe the decision as permitting the jury to find that the statute applies rather than as requiring the jury to so find. Even on the weaker alternative, Smith shows, at least indirectly, that the juror’s role as statutory interpreter has not been entirely eliminated, notwithstanding the modern rule that statutory interpretation is up to the court—not the jury.

jurors as statutory interpreters

201

Contrast Smith with another statutory case, Bryan v. United States, discussed in chapter 2.20 In that case, the defendant was accused of “willfully” dealing in firearms without a license.21 At issue was what it means to violate the statute willfully. The defendant asked for an instruction that jurors “must be persuaded that with the actual knowledge of the federal firearms licensing laws Defendant acted in knowing and intentional violation of them.”22 Instead, the trial judge told the jury that “the person need not be aware of the specific law or rule that his conduct may be violating. But he must act with the intent to do something that the law forbids.”23 The Supreme Court affirmed the conviction, holding that knowledge that the defendant was in some manner acting wrongly met the “willfulness” requirement under the statute. Neither the parties nor the trial judge considered leaving it to the jury to determine whether the defendant’s state of mind constituted willful violation of the statute. Now that Bryan has been decided, that approach to interpreting the statute is very unlikely to occur in the future. Taken together, these two cases illustrate the ambivalence the legal system continues to have about the role that jurors should play in the criminal-justice system. In Smith, the Court left a great deal up to the jury. It was up to the jury to determine what counted as “using a firearm.” But in Bryan, the trial court, urged by both sides, left little to chance, providing a legal definition for the state-of-mind requirement. I do not argue that the court was wrong to do so. As we have already seen, there is a substantial body of law interpreting the mens rea requirements of criminal statutes,24 and uniformity in the application of the criminal laws is a goal worth pursuing. But these cases lead to two very different pictures of the role that the jury is expected to play in the trial process.

How Judges and Juries Engage in Statutory Interpretation Jurors as Discoverers of Ordinary Meaning Statutes define the circumstances under which legal consequences will obtain. Consider the New Jersey statute that makes stalking a crime: “A person is guilty of stalking, a crime of the fourth degree, if he purposefully or knowingly engages in a course of conduct directed at a specific person that would cause a reasonable person to fear for his safety or the safety of a third person or suffer emotional distress.”25 It is easy enough to divine from the statute the “elements” of stalking—that is, the conditions

202

chapter 7

that are individually necessary and together sufficient for stalking to occur. Jury instructions are, at least on their surface, designed around this structure. Standard instructions lay out the elements of a crime and tell the jurors that the government must prove each of the elements beyond a reasonable doubt. But, as we have seen, sorting out the elements of a crime is not the task that judges find difficult when they are confronted with problems of statutory interpretation. Rather, the problem is almost always a question of how broadly individual words in a statute should be interpreted. And word meaning can be addressed either in terms of definitions (i.e., in a rulelike manner) or in terms of ordinary meaning (based on typical experience). For example, the stalking statute does not define “course of conduct.” How should the legal system decide when an alleged stalker’s actions meet that requirement, and what should a judge tell the jurors? The experimental literature supports the notion that jurors tend to concern themselves with the ordinary meaning of a statute. In one set of studies, Vicki Smith presented subjects with the names of a number of crimes, including assault, burglary, kidnapping, murder, and robbery, and asked them to list what they considered the most common characteristics of those crimes.26 She then presented a second set of subjects with scenarios of crimes that met the statutory definitions but did not contain the prototypical features of the crime. For example, a ransom demand and a youthful victim were the two most prevalent features of kidnapping as defined by Smith’s first group of subjects, but those characteristics play no role in the definition of the crime in kidnapping statutes. The results were that, when presented with fictional cases, the second group’s conviction rates were much higher when the cases concerned prototypical scenarios than when they concerned atypical scenarios, even though both scenarios fit the statutory definitions without controversy.27 What does this say about the role that jurors play as statutory interpreters in the criminal-justice system? Judges are ambivalent about dismissing prosecutions that are based on facts that fit within a statute’s literal language but are outside its ordinary meaning. Certainly they have done so, sometimes in very prominent cases, such as Church of the Holy Trinity, but often enough, as in Smith, judges opt for a broad interpretation if the statutory language easily enough supports it.28 Jurors do not always opt for ordinary usage either, as the guilty verdict in Smith illustrates. Yet the combination of the two events—trial motions (to dismiss or to instruct

jurors as statutory interpreters

203

the jury narrowly) and jury deliberation—serves to keep the application of statutes somewhere within the meanings that ordinary people would find plausible. Jurors as Discoverers of Absurd Results By the same token, both judges and jurors engage in analysis under the “absurd-results rule,” discussed in chapter 4. Appellate courts refer to the rule often in cases involving statutory interpretation. A classic example is United States v. Kirby, a case decided in 1868. Kirby was a local sheriff in Kentucky who had a warrant for the arrest of Farris for murder. Farris was a letter carrier. Kirby caught up with him while Farris was delivering the mail on his route and arrested him. In a postbellum display of federal power, Kirby was himself arrested, tried, and convicted of interfering with the delivery of the U.S. mail. In reversing the conviction, the Supreme Court had this to say: “All laws should receive a sensible construction. General terms should be so limited in their application as not to lead to injustice, oppression, or an absurd consequence. It will always, therefore, be presumed that the legislature intended exceptions to its language, which would avoid results of this character. The reason of the law in such cases should prevail over its letter.”29 The absurd-results rule is both stronger and weaker than the ordinarymeaning canon. On the one hand, absurd results need not arise from an excessively broad interpretation of a particular word. In Kirby, it was the whole situation that made the result indefensible as a fair dispensation of justice. On the other hand, absurdity is not a very exacting standard for a system of justice. There may be prosecutions that a community would regard as disproportionate that do not rise to the level of absurdity. In fact, Brown discusses jurors making exactly that point when deliberating in a case in which they regarded the charges as unfairly harsh.30 It is here that jurors provide defendants with an extra level of protection beyond that given by appellate courts applying the absurd-results rule. If a prosecution is too aggressive and the defendant is a sympathetic person, jurors have the right not to convict. Marder discusses this category of “nullification” in terms of her process model of the jury.31 The jury serves as a buffer against prosecutors who attempt to take advantage of linguistic nuances that might put conduct within the scope of a statute that was never intended to apply to a particular situation. Appellate judges are charged with performing a similar task, but there is no guarantee that their

204

chapter 7

assessment of the fit between statutory language and ordinary meaning captures the sense of the community at large. In fact, there is good reason to believe that people—including judges— are not always aware of the range of interpretations that others might assign to statutory language. Consider the following vignette created by linguists Linda Coleman and Paul Kay: “John and Mary have recently started going together. Valentino is Mary’s ex-boyfriend. One evening John asks Mary, ‘Have you seen Valentino this week?’ Mary answers, ‘Valentino’s been sick with mononucleosis for the past two weeks.’ Valentino has in fact been sick with mononucleosis for the past two weeks, but it is also the case that Mary had a date with Valentino the night before. Did Mary lie?”32 The question is whether Mary lied when she made a truthful but misleading statement. Subjects participating in Coleman and Kay’s study disagreed sharply about whether such a statement constituted a lie. In a study of statutory language I have used the expression “pernicious ambiguity” to refer to situations in which language is unclear but the lack of clarity goes unnoticed until it causes harm.33 When language is perniciously ambiguous, a system that requires deliberation among many is more likely to uncover the lack of clarity. An important study by Ward Farnsworth and his colleagues shows further that people do not agree whether statutory language—including the language of statutes discussed in this book—is clear or ambiguous.34 Significantly, judgments of ambiguity correlate with strength of commitment to a particular outcome: the more committed a person is to the outcome, the clearer the statutory language seems to be. Furthermore, despite this finding, people are fairly open-minded about recognizing that others might not be in consensus about the meaning of the language. My research on the interpretation of contractual language, conducted with Terri Rosenblatt and Daniel Osherson, shows that people recognize that others might disagree about meaning, but they significantly overestimate the extent to which other people share their own interpretation when the language is not clear. Significantly, judges are just as prone to this “false-consensus bias” as are laypeople.35 From a linguistic perspective, jurors who acquit when the language permits the prosecutor’s interpretation, but does so unnaturally, may merely be applying the ordinary-meaning rule to the language as it relates to the events in the case rather than to particular statutory words. In this sense, jurors are participating in the task of statutory interpretation, using tools very much like those used by judges.

jurors as statutory interpreters

205

Jury Instructions and Statutory Fidelity Let us see how, in typical cases, the system attempts to guide jurors in their task of determining whether a statute applies to the facts before them. When judges decide whether particular conduct violates a statute, they engage in minute analysis, looking closely at the statutory language and at extrinsic sources, such as legislative history, dictionaries, other sections of the statute, and so on. Jurors have neither the time nor the resources to do this. In fact, the material that would allow them to do so is not available to them. Thus, the legal system must decide what to tell jurors about the law. Peter Tiersma writes of cases in which jurors were held to have committed misconduct by using a dictionary to look up words that they did not understand.36 Jury instructions are supposed to accomplish this task. But statutes are very difficult to understand. Thus, the system is torn between promoting fidelity to statutory language and writing instructions that jurors can follow. Both concerns are clearly visible in the array of instructions that courts use, with compromises often apparent within a particular instruction. But the decision about how to instruct a jury creates higher stakes than that. The more detailed the instruction, the less the risk, if the jurors indeed follow the minutiae, that jurors will use their everyday experience to reach a verdict. If every relevant word is defined, and every contingency is the subject of an instruction, then the task of the jury becomes more mechanical, at least if everything works as designed.37 In this sense, the decision about how to instruct a jury is at least in part a decision about how much of the task of interpreting the statute is to be left to the jury. Below we see how this dynamic plays out by looking first at the federal mail fraud statute, which is a relatively short law, and then at the federal moneylaundering statute, which is a very long and complex law designed to leave little to chance. Finally, we examine judicial pronouncements concerning the relationship between statutory language and jury instructions. Instructing on Mail Fraud: Presenting the Ideas of a Simple Statute The federal mail fraud statute reads in relevant part: Whoever, having devised or intending to devise any scheme or artifice to defraud, or for obtaining money or property by means of false or fraudulent pretenses, representations, or promises . . . for the purpose of executing such

206

chapter 7

scheme or artifice or attempting so to do . . . knowingly causes to be delivered by mail or such carrier according to the direction thereon, or at the place at which it is directed to be delivered by the person to whom it is addressed, any such matter or thing . . . shall be fined not more than $1,000,000 or imprisoned not more than 30 years, or both.38

Some of the language, like the language of the bribery statute discussed in chapter 2, is syntactically complex. For example, the statute is one long sentence. To understand it, one must in some sense “know” the following: 1. The subject of the sentence is a clause that begins “whoever,” and the predicate of the sentence begins with “shall be fined.” 2. The subject of the clause that is the sentence’s subject is the word “whoever” and a clause that modifies that word, and the predicate of the clause that is the subject of the sentence begins with “knowingly.” The main verb is “causes.” 3. The clause beginning “having devised” modifies “whoever” and is embedded within the clause that is the sentence’s subject. 4. The predicate of the clause that is the sentence’s subject is complex, with the phrase “to be delivered by mail” embedded in the predicate. It is a passive clause, whose subject is missing. Thus, the sentence does not specify who must do the delivering, just that the defendant must cause the delivery. 5. The clause beginning “having devised” has no subject of its own, but we interpret “whoever” as its subject.

There are many more complexities in the structure of this language, just about all of which make it more difficult to process.39 Yet once one parses one’s way through the statute, just about all of these relationships among its clauses and phrases fall into place. One interpretive problem in the mail fraud statute, however, does appear to have arisen from syntactic ambiguity. Courts have held, consistent with the language of the statute, that the statute does not require that the person doing the mailing be the defendant.40 Any lack of clarity results from the use of the passive construction without an agent, “causes to be delivered by mail.” The statute does not specify by whom. Otherwise, arguments over the applicability of this statute are largely over the scope of its various terms. How much of the interpretation of the statute should be left to the jury? If the jury were to decide on a case-by-case basis whether the defendant has to be the one doing the mailing, it would lead to inconsistent

jurors as statutory interpreters

207

outcomes in cases with similar facts. Two juries with different linguistic intuitions could decide differently even if they both would have acted the same had they understood the statute the same way. The system does not accept this risk of inconsistent verdicts today. Rather, courts rule as a matter of law in such cases and instruct the jury accordingly. In contrast, interpretation of the statute’s terms requires decisions to be made about borderline cases: What counts as a scheme to defraud? How clear does the defendant’s statement have to be to count as a representation or a promise? These are questions that may be resolved by courts, but they are the kinds of conceptual questions that people grapple with every day and upon which people may differ as a result of different perspectives and experiences. The issue comes down to the tension between democratic participation and uniform treatment. The conflicting vision of the jury’s role as interpreter comes through both in a leading Supreme Court case interpreting the statute and in the various approaches to jury instructions that have been proposed. In Schmuck v. United States, discussed in chapter 5, the defendant had been operating a business in which he would roll back the odometers of used cars and then sell the cars to dealers, who would resell them to customers at prices inflated to reflect the inaccurately low mileage resulting from Schmuck’s scheme.41 No mailing was involved until after the customer bought the car and the dealer used the mail to register the newly purchased vehicle with the state motor vehicle department in Wisconsin. The question raised in Schmuck was whether any mailing occurred “for the purpose of executing” a fraudulent scheme, given that the fraud had been completed before the mailing occurred. In a five-to-four decision, the Supreme Court affirmed the conviction. The Court held that the trial court had properly denied the defendant’s motion to dismiss the indictment. That led to a trial in which the jury issued a guilty verdict. As for the jurors’ role as interpreters of the statute, the Court held: “Thus, Schmuck’s was not a ‘one-shot’ operation in which he sold a single car to an isolated dealer. His was an ongoing fraudulent venture. A rational jury could have concluded that the success of Schmuck’s venture depended upon his continued harmonious relations with, and good reputation among, retail dealers, which in turn required the smooth flow of cars from the dealers to their Wisconsin customers.”42 The Court’s assumption seems correct. It is at least possible to draw the inference that Schmuck was advantaged by the mailing.

208

chapter 7

This approach leaves a great deal of discretion to the jury. The chain of thought that the Court deems “possible” permits the jury to convict a defendant of allowing others to use the mail for the sake of facilitating some future fraudulent scheme with which the defendant was not charged, as Justice Scalia points out forcefully in his dissent.43 At least in principle, jurors are permitted to convict only when the government has proven the crime for which the defendant was indicted beyond a reasonable doubt. In contrast, the Court appears to have left room for a common-law crime of mail fraud.44 But not all mail fraud instructions give jurors such broad discretion. While many instructions make it clear that, as in Schmuck, the person using the mail as part of the scheme may be someone other than the defendant,45 others explain the mens rea requirements. For example, in United States v. Lennartz, the Seventh Circuit affirmed the use of the “ostrich,” or “conscious avoidance,” instruction. The case involved a commercial ambulatory service that routinely overcharged Medicare for the distances it drove and for waiting time while patients underwent kidney dialysis. The owner of the company claimed ignorance of the drivers’ corrupt practices, although there was evidence to the contrary. The court instructed the jury: When the word “knowingly” is used in these instructions, it means that the defendant realized what he was doing and was aware of the nature of the conduct and did not act through ignorance, mistake, or accident. Knowledge may be proved by the defendant’s conduct and by all the facts and circumstances surrounding the case. Knowledge may be inferred from a combination of suspicion and indifference to the truth. If a person is found to have had a strong suspicion that things were not what they seemed or that someone had withheld some important facts yet shut his eyes for fear of what he would learn, that person may be inferred to have acted knowingly or with knowledge, as I have used these terms.46

Such tugs-of-war are a typical part of the trial. For almost every issue, one party will want the jury to be told in no uncertain terms that its discretion is limited, while the other party will want to rely on the jury’s discretion. It is instructive in this context to compare two model jury instructions with radically different approaches. They are set out in the appendix. The Federal Judicial Center (FJC), the research arm of the federal judiciary,

jurors as statutory interpreters

209

has opted for a plain-language approach that summarizes the statute in words that are relatively easy to understand.47 The instruction is 225 words long and makes certain tradeoffs between fidelity to the legislature’s choice of language and comprehensibility. It does not mention that the defendant need not be the one who did the mailing, although that issue is not relevant in run-of-the-mill mail fraud cases, for which the instruction is designed. Contrast this instruction with a lengthy one derived from the case law in various federal circuits and contained in a practice volume called Modern Federal Jury Instructions (MFJI).48 It is more than 2,000 words long, including alternative paths that the instruction might take depending upon the facts. The shortest version that I could compile still has more than 1,500 words. It leaves nothing to chance except the possibility that the instruction is so long and complicated that it cannot reasonably serve to inform jurors of the subtleties that its very structure is intended to convey. Both instructions divide the statute into three elements. I will not look closely at all of the language, but let us compare how each instruction handles the first element: that the defendant has devised “any scheme or artifice to defraud.”49 The FJC instruction states this element succinctly, but deviates from the statutory language: First, that _________________ made a plan [e.g.: to obtain money based on giving false information about the Apex Corporation to Sarah Stone and Rubin Ross].50

The instruction has substituted “plan” for “scheme” and, instead of using the word “defraud,” describes what the fraud is alleged to be. Elsewhere the instruction says that it is against the law to “cheat” someone out of property instead of its being illegal to “devise a plan to defraud. . . .” The use of the word “cheat” is especially interesting. Not only is that word absent from the mail fraud statute, but Congress rarely uses it at all. I could find only one occurrence in the entire United States Code: a statute that regulates commodities exchanges makes it a crime “to cheat or defraud or attempt to cheat or defraud such other person.”51 Yet not only do people use the word frequently in everyday speech (beginning at a very young age), but judges use it fairly often to describe a defendant’s conduct or to define fraudulent behavior.52 I focus on the word “cheat” because we all know it so well that we are likely to have mental models of prototypical cheating based on various

210

chapter 7

experiences. This means that the language of the FJC instruction, by including terms that jurors can relate to their own experience, may make it easier for them to exercise judgment in determining whether alleged conduct is actually serious enough to warrant a conviction. In other words, the language of the instruction invites the jury to play its traditional part in the adjudication of criminal cases. If the defendant’s conduct does not match all jurors’ sense of what it means to cheat someone, the best the government should be able to expect is a hung jury. The first element in the mail fraud instruction from MFJI, in contrast, edited to exclude optional provisions that do not routinely apply, comes much closer to what a jury is likely to hear: The first element that the government must prove beyond a reasonable doubt is that there was a scheme or artifice to defraud [the victim] of money or property . . . by means of false or fraudulent pretenses, representations or promises. This first element is almost self-explanatory. A “scheme or artifice” is merely a plan for the accomplishment of an object. A scheme to defraud is any plan, device, or course of action to obtain money or property . . . by means of false or fraudulent pretenses, representations or promises reasonably calculated to deceive persons of average prudence. “Fraud” is a general term which embraces all the various means by which human ingenuity can devise and which are resorted to by an individual to gain an advantage over another by false representations, suggestions or suppression of the truth, or deliberate disregard for the truth. Thus, a “scheme to defraud” is merely a plan to deprive another of money or property . . . by trick, deceit, deception or swindle. The scheme to defraud is alleged to have been carried out by making false (or fraudulent) statements (representations) (claims) (documents). A statement, representation, claim or document is false if it is untrue when made and was then known to be untrue by the person making it or causing it to be made. A representation or statement is fraudulent if it was falsely made with the intention to deceive. Deceitful statements of half truths or the concealment of material facts, and the expression of an opinion not honestly entertained may also constitute false or fraudulent statements under the statute. The deception need not be premised upon spoken or written words alone. The arrangement of the words, or the circumstances in which they are used may

jurors as statutory interpreters

211

convey the false and deceptive appearance. If there is deception, the manner in which it is accomplished is immaterial. . . . The false or fraudulent representation (or failure to disclose) must relate to a material fact or matter. A material fact is one which would reasonably be expected to be of concern to a reasonable and prudent person in relying upon the representation or statement in making a decision (e.g., with respect to a proposed investment). This means that if you find a particular statement of fact to have been false, you must determine whether that statement was one that a reasonable person or investor might have considered important in making his or her decision. The same principle applies to fraudulent half truths or omissions of material facts. . . . In addition to proving that a statement was false or fraudulent and related to a material fact, in order to establish a scheme to defraud, the government must prove that the alleged scheme contemplated depriving another of money or property (or of the intangible right of honest services). However, the government is not required to prove that the defendant himself originated the scheme to defraud. Furthermore, it is not necessary that the government prove that the defendant actually realized any gain from the scheme or that the intended victim actually suffered any loss. . . . A scheme to defraud need not be shown by direct evidence, but may be established by all of the circumstances and facts in the case.53

In actual cases, an instruction of this sort will be tailored to the facts of the particular case. Nonetheless, the instruction contains some typical features that reflect the legal system’s long-standing concerns about the jury system. Putting to one side the almost entertaining remark “This first element is almost self-explanatory,” the instruction actually uses more or less the same language of the shorter, FJC instruction. It explains: “Thus, a ‘scheme to defraud’ is merely a plan to deprive another of money or property . . . by trick, deceit, deception or swindle.”54 Nonetheless, the MFJI instruction attempts to keep the jurors’ analysis of the case under tight control to ensure that individuals accused of committing the same crime will be treated more or less the same and in accordance with established legal principles. Definitions of the elements of the crime that result from various judicial decisions make their way into the instructions. As the corpus of published opinions expands, so do the instructions that the jury must digest. Uniformity of treatment is what motivated the reduction of the jury’s law-finding function in the nineteenth

212

chapter 7

century.55 More than a hundred years later, it has resulted in complex, linguistically loyal instructions that are difficult to understand. What they give lawyers and trial judges is the peace of mind that comes with deference to the legislature and to appellate judges put in charge, respectively, of enacting and clarifying the criminal law. Taken together, the two mail fraud instructions illustrate the end points of the different perceptions of the modern American criminal jury. The FJC instruction treats the jury as an institution whose job it is to exercise its judgment wisely once it is introduced to the governing legal principles. It says little about the law in the abstract, opting instead to describe as illegal the facts as the government alleges them. The MFJI instruction does just opposite. Money Laundering: Legislative Control of Jury Deliberations My focus has thus far been on how courts control the scope of jury deliberations, whether by deciding issues as a matter of law or by presenting jurors with instructions designed to minimize their discretion in reaching a verdict. But the analysis should begin before the courts get involved. The starting point should really be the legislature and its decision to draft laws that are likely to lead to one vision of the jury versus another. In both criminal and regulatory legislation, statutes are becoming longer and more complex. Whether as a result of negotiation and compromise56 or of a perception that judges and juries are not to be trusted,57 Congress in particular has been leaving less and less to the discretion of others, especially in the realm of the criminal law. The mail fraud statute may give courts some room to determine the wording, length, and tone of the instruction. But long, complicated statutes, as a practical matter, do not. Consider the federal money-laundering statute, presented below in relevant part: § 1956. Laundering of monetary instruments (a) (1) Whoever, knowing that the property involved in a financial transaction represents the proceeds of some form of unlawful activity, conducts or attempts to conduct such a financial transaction which in fact involves the proceeds of specified unlawful activity— (A) (i) with the intent to promote the carrying on of specified unlawful activity; or

jurors as statutory interpreters

213

(ii) with intent to engage in conduct constituting a violation of section 7201 or 7206 of the Internal Revenue Code of 1986; or (B) knowing that the transaction is designed in whole or in part— (i) to conceal or disguise the nature, the location, the source, the ownership, or the control of the proceeds of specified unlawful activity; or (ii) to avoid a transaction reporting requirement under State or Federal law, shall be sentenced to a fine of not more than $500,000 or twice the value of the property involved in the transaction, whichever is greater, or imprisonment for not more than twenty years, or both.58

The syntax of the statute is rather convoluted, but the elements of the crime are clear. To commit the crime of money laundering the defendant must (1) conduct or attempt to conduct a financial transaction, (2) knowing that the property involved represents the proceeds of some form of unlawful activity, (3) with the intent to promote the carrying on of the unlawful activity. In addition, like many statutes, this one contains a set of statutory definitions.59 Legal scholars are accustomed to reading (and sometimes criticizing) such definitions. But laypeople are not. How should a jury be instructed when the legislature has taken the care to define statutory terms but has done so in language that is very difficult to understand? Consider the following example. The statute is entitled “Laundering of monetary instruments.” Jurors are not likely to be familiar with the term “monetary instruments.” This may give a court the opportunity to define the term in everyday language that jurors might understand. For example, let us say that the defendant is accused of laundering cash and personal checks. A court may say something like: “The statute makes it a crime to launder ‘monetary instruments.’ You may not be familiar with that term. For our purposes, it includes cash and personal checks.” But Congress defines the term in the statute as follows: “(5) the term ‘monetary instruments’ means (i) coin or currency of the United States or of any other country, travelers’ checks, personal checks, bank checks, and money orders, or (ii) investment securities or negotiable instruments, in bearer form or otherwise in such form that title thereto passes upon delivery.”60 This definition is largely irrelevant, and not easy to understand because it contains the names of instruments with which many jurors are likely to be unfamiliar. Hearing it may confuse some jurors and, because the language can be alienating, cause others to stop listening carefully. But to instruct

214

chapter 7

the jury on only the potentially applicable portions of this section misses an important point that Congress wrote into the statute: the definition of “monetary instruments” is written broadly in order to catch as many such instruments as possible. Unless the judge reads the definition, the jury will not be made privy to that fact, but if the judge does read it, the language is likely to be confusing. Standard jury instructions resolve this problem with a compromise. The judge gives a number of examples but leaves out the exotic ones: “The term ‘monetary instrument’ includes, among other things, coin or currency of the United States or any other country, personal checks, traveler’s checks, cashier’s checks, bank checks, money orders, and investment securities or negotiable instruments in bearer form or otherwise in such form that title thereto passes upon delivery.”61 This strikes me as a perfectly reasonable jury instruction, yet it is a far cry from the FJC’s approach to instructing the jury on mail fraud. The top-down approach to jury instruction has won. The structure of the statute makes it difficult to do otherwise. For other terms, the instructions succumb completely to the language of the statute, making the terms very difficult to understand. Consider, for example, the MFJI definition of “financial transaction”: The term “financial transaction” means a transaction involving a financial institution which is engaged in, or the activities of which affect, interstate or foreign commerce in any way or degree, or a transaction which in any way or degree affects interstate or foreign commerce and involves the movement of funds by wire or other means, or involves one or more monetary instruments, or involves the transfer of title to any real property, vehicle, vessel or aircraft.62

This language closely tracks the language of the statute: (4) the term “financial transaction” means (A) a transaction which in any way or degree affects interstate or foreign commerce (i) involving the movement of funds by wire or other means or (ii) involving one or more monetary instruments, or (iii) involving the transfer of title to any real property, vehicle, vessel, or aircraft, or (B) a transaction involving the use of a financial institution which is engaged in, or the activities of which affect, interstate or foreign commerce in any way or degree.63

When the statute’s glossary does not define a term but the instruction does, the instruction is written more comprehensibly than the statute. Philosopher

jurors as statutory interpreters

215

Roy Sorensen calls such descriptions “precisifying definitions,” since they purport to define but elaborate beyond the term they are defining.64 In contrast, consider the following MFJI definition: “Proceeds can be any kind of property, not just money.”65 This shows that the drafters of these instructions are capable of writing in plain English when the opportunity presents itself. Mixed Loyalties: The Text Usually Wins Most often, jury instructions based on particular criminal statutes simply track the language of the statute. However, this is not always so. The U.S. Supreme Court reversed a death sentence because the trial court allowed the jury to rely only on the language of a statute without further guidance. In Godfrey v. Georgia, the defendant was sentenced to death under a Georgia statute that permitted the jury to impose the death penalty for a murder that “was outrageously or wantonly vile, horrible or inhuman in that it involved torture, depravity of mind, or an aggravated battery to the victim.”66 The instruction to the jury was the statute itself, without further explanation. The Supreme Court of Georgia affirmed the death sentence, but the Supreme Court of the United States reversed. Even though the Georgia death penalty statute had been found constitutional, and even though the instruction tracked the language of the statute, the Court found that the instruction gave the jury no guidance: In the case before us, the Georgia Supreme Court has affirmed a sentence of death based upon no more than a finding that the offense was “outrageously or wantonly vile, horrible and inhuman.” There is nothing in these few words, standing alone, that implies any inherent restraint on the arbitrary and capricious infliction of the death sentence. A person of ordinary sensibility could fairly characterize almost every murder as “outrageously or wantonly vile, horrible and inhuman.” Such a view may, in fact, have been one to which the members of the jury in this case subscribed. If so, their preconceptions were not dispelled by the trial judge’s sentencing instructions. These gave the jury no guidance concerning the meaning of any of § (b)(7)’s terms. In fact, the jury’s interpretation of § (b)(7) can only be the subject of sheer speculation.67

In recent years the Court has had a mixed record in deciding whether to strike down unclear jury instructions in death penalty cases.68 Yet Godfrey remains a good illustration of one approach to the relationship between

216

chapter 7

jury instructions and a statute’s words: the instruction must be loyal to the meaning of the statute; however, if the statutory language is either hard to understand or is simply too vague to be of much help, the instruction must go beyond the words of the statute itself. Courts sometimes recognize the need for clarification of statutory language simply as a matter of sound judicial management. Consider, for example, State v. Alexander, a case decided by the Supreme Court of New Jersey in 1994. The defendant was accused of violating various drug laws, including New Jersey’s “drug kingpin” statute. Alexander had hired another individual, named Harewood, to sell crack cocaine to Alexander’s customers. Alexander would supply Harewood with cocaine, and Harewood would sell the cocaine and collect the money. Alexander would then get 70 percent of the receipts; Harewood got the other 30 percent.69 Below is the relevant part of the drug kingpin statute: A person is a leader of a narcotics trafficking network if he conspires with others as an organizer, supervisor, financier or manager, to engage for profit in a scheme or course of conduct to unlawfully manufacture, distribute, dispense, bring into or transport in this State methamphetamine, lysergic acid diethyl­amide, phencyclidine or any controlled dangerous substance classified in Schedule I or II, or any controlled substance analog thereof. Leader of narcotics trafficking network is a crime of the first degree and upon conviction thereof . . . a person shall be sentenced to an ordinary term of life imprisonment during which the person must serve 25 years before being eligible for parole.70

Its language is opaque at best, even though the Supreme Court of New Jersey had earlier held that it survived attack for being unconstitutionally vague.71 Affirming a decision of an intermediate appellate court, the New Jersey Supreme Court in Alexander held that explanation of the statutory terms was required. It held that the trial court “should instruct the jury that it must find that the defendant occupies a high-level position, that is, a position of superior authority or control over other persons, in a scheme or organization of drug distribution (or manufacture or dispensing or transporting), and that in that position the defendant exercised supervisory power or control over others engaged in an organized drugtrafficking network.”72 This additional language was needed, the court explained, to make the instruction consistent with the purpose of the statute, which was set forth in a statement of purpose that was part of the statute itself. Moreover, the court explained:

jurors as statutory interpreters

217

This Court has made abundantly clear that correct jury instructions are at the heart of the proper execution of the jury function in a criminal trial: “ ‘[a]ppropriate and proper charges to a jury are essential for a fair trial.’ ” A court’s obligation properly to instruct and to guide a jury includes the duty to clarify statutory language that prescribes the elements of a crime when clarification is essential to ensure that the jury will fully understand and actually find those elements in determining the defendant’s guilt. For the purpose of instructing and guiding juries, courts regularly explain and define statutory language consistent with legislative intent. Courts commonly clarify statutory language to give more precise meaning to statutory terms to effect the legislative intent and to make sure that juries carry out that intent in determining criminal culpability.73

Thus, at least in New Jersey, courts sometimes take seriously the role of the juror as statutory interpreter and provide jurors with the kind of explanatory material that judges themselves use when they perform that task. For the most part, though, appellate courts do not reverse convictions when the instructions track the language of the statute. As the Eighth Circuit has put it: “The best way to comply with [the Federal Death Penalty Act] is to actually use the language of the statute in the jury instruction.”74 Thus, statutory language can offer a safe harbor for trial courts concerned about being reversed, just as statutory language provides a safe harbor for prosecutors who draft indictments. Of course, such a tactic can lead to an incomprehensible description of the law. It is hard enough to describe subtle legal principles to laypeople in terms that they are likely to understand and remember. Judge Easterbrook made the point dramatically in affirming the death penalty of serial killer John Wayne Gacy.75 Shortly before Gacy’s scheduled execution, another Illinois death row prisoner, James P. Free, had engaged the late sociologist Hans Zeisel to review the Illinois death penalty jury instructions to determine whether they wrongly gave jurors the misimpression that they were to impose the death penalty in a greater range of circumstances than either the statute or constitutional rulings by the courts permitted.76 Zeisel conducted a study, which appeared to show that prospective jurors indeed would apply the death penalty excessively. Gacy attempted to use the Zeisel study to argue that he had been improperly sentenced to death. Rejecting Gacy’s argument, Judge Easterbrook distinguished between jury instructions that are difficult to understand because they are not

218

chapter 7

written as clearly as they should be and jury instructions that are difficult to understand because they describe complex concepts with which jurors are unfamiliar. He reasoned that poorly written instructions might constitute a legitimate complaint. As for instructions that are difficult to understand because the law governing capital punishment is complex, Easterbrook wrote: “That subjects being peppered with questions about a complex concept they have encountered for the first time will not give answers satisfactory to lawyers is no surprise. It cannot be that the Constitution, which requires judges to tell juries to use these elusive concepts, is selfdestructive because lay persons will experience difficulty in answering questions about what they have been told. Confusion and misunderstanding attributable to the Constitution of the United States does not yield a violation of that document.”77 Because the Zeisel study did not demonstrate that the language of the instructions—and not their substance—is what made them difficult to comprehend, the court rejected the Zeisel study. Once that happened, the court of appeals in Free followed suit. Both men were executed. Perhaps ironically given Free’s fate, Shari Seidman Diamond (a psychologist) and Judith N. Levi (a linguist) subsequently conducted a study of the Illinois death penalty instructions that tested the clarity of the language.78 They found that rewriting the instructions to make them more comprehensible increased understanding, indicating that the results of the Zeisel study had to be explained, at least in part, by the way the instructions were written. State courts typically also approve instructions that track a statute’s exact words, regardless of how comprehensible they are. A Minnesota statute says: “No employer or agent thereof shall directly or indirectly solicit or require a polygraph, voice stress analysis, or any test purporting to test the honesty of any employee or prospective employee.”79 At a trial over an employer’s use of a polygraph, the Supreme Court of Minnesota approved the trial court’s refusal to define “solicit,” since the trial court “used the exact words of the statute.”80 This does not mean that the exact words are required. They are not. The Supreme Court of Hawaii has made it clear that the “trial court is not required to instruct the jury in the exact words of the applicable statute but to present the jury with an understandable instruction that aids the jury in applying that law to the facts of the case.”81 But using a statute’s exact words generally shields trial judges.82 It is rare to find a reversal, whether or not the language of the statute used in the instruction is comprehensible to the average person.

jurors as statutory interpreters

219

The legal system is thus of two minds when it comes to letting the jury rely on the exact words of a statute. Sometimes the system is satisfied with a “read the statute and run” approach to the jury. At other times, it takes the role of the jury seriously in deciding whether the exact words are the best words. In contrast, although “[a] defendant is entitled to an instruction on his theory of the case if the instruction is a correct statement of the law and if he has offered sufficient evidence for the jury to find in his favor,” defendants are not entitled to have the judge read the exact language of the instruction they suggest.83 As long as the judge “substantially covered the point that [the defendant] sought to get across,” the instruction will be affirmed on appeal.84 The point of these cases is clear enough to trial judges: you will rarely get into trouble if you explain a statute to jurors by reading its language to them. We are willing, for the most part, to accept the legal fiction that jurors understand what they are told, even when we know they do not. In fact, it is a nearly unrebuttable principle of law that jurors do understand the instructions that are read to them. As Chief Justice Rehnquist wrote in deciding that it was proper for a trial judge simply to repeat an instruction in a death penalty case that jurors said they did not understand: “A jury is presumed to follow its instructions. Similarly, a jury is presumed to understand a judge’s answer to its question. Weeks’ jury did not inform the court that after reading the relevant paragraph of the instruction, it still did not understand its role. To presume otherwise would require reversal every time a jury inquires about a matter of constitutional significance, regardless of the judge’s answer.”85 A jury wishing to be better instructed must state its failure to understand and then argue with the judge about his solution in order to be heard. It is not realistic to expect this to happen. Thus, not only does the opinion accept the legal fiction, but it works to keep juror comprehension fictional.

Communal Justice and Statutory Interpretation In their important studies, Paul Robinson and John Darley compare a set of criminal-law doctrines with the views of individuals concerning crime and punishment. They find that “often the legal codes and the community standards reflect similar rules in assigning liability to a case of wrongdoing;

220

chapter 7

but also often they do not.”86 In the introductory chapter, they argue that this comparison is important: [R]ecent empirical evidence suggests that criminal law’s most effective mechanism of compliance is not the deterrent threat of sanction; it is its capacity to authoritatively describe the moral and proper rules of conduct. People will follow those rules not because they fear punishment for violating them but because they see themselves as good and law-abiding people who are inclined to obey the law because it is the right thing to do. Most important, the compliance power of criminal law is directly proportional to its moral credibility. If the criminal law is seen as unjust in one instance, its moral credibility and its concomitant compliance power are, accordingly, incrementally reduced.87

Robinson and Darley focus their attention on particular legal doctrines that do not match community standards of justice. To the extent that they are correct that this gap diminishes the moral force of the law, one solution is to change the doctrines, which requires legislative action. This solution, however, is not accomplished easily. Consider, for example, California’s “three-strikes” law, which calls for a sentence of twenty-five years to life for a third offense.88 In a 2003 case involving a series of three thefts over a period of years, the U.S. Supreme Court has ruled that the statute does not violate the Eighth Amendment’s prohibition of cruel and unusual punishment. The last theft, which triggered the statute’s application, was of three golf clubs.89 A companion case involved a similar sentence, where the triggering crime was the theft of videotapes from two department stores.90 The Court based its decision in large part on its “traditional deference to legislative policy choices” in determining how long a sentence to impose.91 The application of the three-strikes statute in such circumstances may well offend some people’s sense of justice. Many would consider a sentence of up to life in prison for shoplifting extreme, no matter what illegal acts a person had committed years earlier. The vigorous dissent of four justices certainly demonstrated that there is no consensus that the law as applied in such cases even meets the most minimal standards of proportionality that constitutional doctrine requires.92 But the law is not likely to be repealed, leaving neither the legislature nor the courts to protect against prosecutions that are more aggressive than community values would permit.

jurors as statutory interpreters

221

A second approach to the gap between statutory law and community standards of justice is to embed into the system procedures that diminish the likelihood that people will actually be convicted for conduct that is not deemed by the community to be antisocial, or at least not as antisocial as the law would have it. As Marder points out, juries in some California counties are unlikely to convict an individual of an offense that will trigger the three-strikes law if the jury believes the application of that law to a particular case is unjust.93 This reluctance, in turn, affects decisions to prosecute.94 By reducing the rate of conviction for prosecutions that would result in disproportionate punishment, juries, at least in some counties, are limiting the statute’s application to the kinds of cases that led to the statute’s enactment in the first place. Prosecutions for violations that stray too far from the prototype are less likely to be successful, an outcome that is consistent with the experimental literature discussed earlier in this chapter.95 Therefore, they are less likely to occur at all, since prosecutors’ offices have no reason to waste their resources. Examples like these have led commentators to look favorably on the role that juries play in such cases. But such examples do not fully address the larger questions about the proper role of jurors as statutory interpreters. To see why, let us return to Mr. Schmuck, who was convicted of mail fraud. The applicability of the statute to the facts of the case was questionable. Yet the jury convicted. In other cases, courts have reversed convictions by applying the rule of lenity and ruling that a prosecution was improper.96 There are thus instances in which the community values expressed by jurors lead to conviction when the language of the statute does not apply, at least not clearly so. One way to handle this problem is by giving the jurors greater discretion but instructing them accordingly. Brown has suggested that when disputes arise over the wording of jury instructions, and both versions are acceptable as a matter of law, principles of lenity dictate that the judge should give the defendant the benefit of the doubt and read his proposed instruction.97 Brown’s proposal, while well intentioned, is not likely to win over many courts. Not all defense-proposed instructions are the most faithful to the overall meaning of the statute, even if they are legally acceptable. Courts will want to continue to make that analysis. My proposal is a more modest one. Courts should continue to permit jurors to interpret statutes to allow their verdicts to reflect the standards of the community. This process was originally contemplated as a unidirectional

222

chapter 7

filter and should remain so. Thus, courts should continue applying such canons as the absurd-results rule and the ordinary-meaning rule to weed out cases that appear to exceed the intended scope of a statute. If a case survives that level of scrutiny and makes its way to the jury, a defendant will have yet another chance for acquittal if the prosecution is more aggressive than the ordinary meaning and the jury’s sense of decency permit. Whenever possible, instructions should be drafted so that they do not undermine the jury’s ability to exercise this level of discretion. Providing even this much role for the jury is controversial, at least in some instances. Some laws make it harder for jurors to bring community values into their deliberations. For example, California law permits bifurcated trials in three-strikes cases.98 First, a jury determines guilt or innocence with respect to the crime that constitutes the third strike. Then that same jury determines whether the defendant was earlier convicted of other crimes that make up the first two strikes.99 The court pronounces a sentence based on these verdicts. This procedure is consistent with the general notion that jurors are not supposed to take sentencing considerations into account in deciding guilt or innocence.100 The function of this bifurcation, of course, is to reduce the chance that during the first phase of the trial jurors know that they are participating in a three-strikes case and thus apply community values of lenity and proportionality to the decision-making process. Judges and prosecutors will not always succeed in keeping from the jury the fact that they are sitting on a three-strikes case.101 Yet the goal is clear enough, and it is very much at odds with the traditional role of the jury as a counterbalance to excessively harsh prosecutions. Despite such examples, even in an era seemingly far removed from one in which the jury was the finder of the law as well as the finder of fact, jurors can continue to play a positive role in determining the applicability of statutory law. To the extent that their contribution to the judicial process helps to narrow the gap between the community’s views of justice and the outer reaches of the law, things have not changed much in over two hundred years.

chapter eight

Legislatures, Judges, and Statutory Interpretation

Let us look at some conclusions that have emerged thus far: 1. Laws generally work quite well. They are written with certain recurrent problems in mind, and they typically perform as intended when those recurrent problems predictably arise. Easy cases predominate in the legal landscape. Moreover, as unexpected scenarios begin to surface, the courts resolve them. To the extent that these also recur, statutory interpretation becomes more stable over time. 2. Our cognitive capacities are not structured to enable us to produce laws that are, in all situations, both flexible enough to absorb new situations and precise enough so that one’s rights and obligations are always clear. Most significantly, statutes are written in words, and words designate concepts. Our conceptualizations produce not only clear cases but borderline cases as well. Most of the difficult cases of statutory interpretation involve borderline cases. 3. There is sharp disagreement about how the legal system should dispose of difficult cases. Judges of all persuasions typically agree that language should play a central role. But when the language does not give a definitive answer, judges disagree about what kinds of extrinsic evidence to consider. The central debate over the past quarter century has been over the use of legislative history as evidence of legislative intent. Moreover, because we can look at statutory language alternatively in terms of its ordinary meaning or in terms of the outer limits of its meaning, disagreements can arise even when courts limit themselves to the statutory language.

224

chapter 8

4. To the extent that values other than deference to the legislature are relevant, there is loose agreement on what these values should be and, more importantly, no agreement about what weight to give these values. Among these values are coherence with the larger body of statutory and case law, responsiveness to changing times, constitutional values, and the need to enable law enforcement agencies to do their jobs. 5. When the resolution of a hard case engages political values, arguments adduced by particular judges tend to lead to a result consistent with their political orientation. This fact may create the impression that statutory interpretation is both indeterminate and political. With respect to these cases, especially when they reach the Supreme Court, that impression is accurate. But with respect to the wide range of cases in which a statute may apply, the impression can be significantly overstated. 6. While judges and scholars argue over the kinds of evidence that should be used in statutory interpretation, other branches involve themselves as well. The executive branch receives deference to its interpretations under the Chevron doctrine, the legislature attempts to limit judicial discretion by enacting rules of interpretation, and courts fight back both by setting limits on the amount of deference they are willing to give and by creating hurdles for legislatures that wish to enact laws that are inconsistent with values that the courts hold to be important aspects of governmental structure.

For the most part, this situation is basically healthy. A small subset of cases is so up in the air that their results become politicized, but most of the time statutory interpretation works relatively well. There are, however, a few areas of concern, and some possible solutions. As we have seen, a number of institutions play a role in the interpretation of statutes. These include the courts (including both judges, who interpret statutes as a matter of law, and juries, who interpret statutes as part of their verdict), various parts of the executive branch, and the legislature. I conclude this book with some brief remarks about how each of these institutions might be more responsive to the problems summarized above.

The Courts A great deal of the debate over statutory interpretation is couched as a competition among judges in disagreement to demonstrate that one judge has deferred more to the legislature than the other. This, by and large,

legislatures, judges, and statutory interpretation

225

reflects a healthy restraint in that it limits the range of legitimate argumentation. Nonetheless, judicial debate takes on a sterile tone when it fails even to mention the values lurking in the background of a dispute. Federal judges are appointed for life under the Constitution. They should be less timid about admitting that they sometimes must exercise discretion in deciding a statutory dispute. To take an example discussed in earlier chapters, consider Circuit City Stores v. Adams, the case that asked whether the Federal Arbitration Act requires that employment disputes be arbitrated.1 The five conservative Supreme Court justices in the majority did not acknowledge that they exercised their discretion in favor of employers, and the four liberal justices did not acknowledge that they exercised their discretion in favor of employees. That is exactly what happened, but no one on the Court talked about it. It would have added legitimacy to the process if the majority had written more about their long-standing effort to move dispute resolution toward arbitration and, given the statute’s ambiguity, their refusal to interrupt that trend, even though they realized that employees would often rather litigate in court. Consistent with Judge Richard Posner’s pragmatic approach, the Court might have debated not only the language and the history of the statute’s enactment but also the consequences for the litigation system from choosing one result over another. Given the statute’s ambiguity, there is no certainty of the outcome in any event. Of course, this smacks of legislating from the bench. But that is exactly what the Court did without acknowledging it. When judges are forced to defend the consequences of their decisions overtly, it can only serve to reduce the range of arguments that are deemed legitimate, thus making the exercise of judicial discretion less of a problem—not more of one. The obligation to defend one’s decisions on more than one set of grounds results in fewer choices. This reduction will not present itself in the contentious cases that make up the Supreme Court’s docket. Rather, it will manifest itself in the number of cases that present themselves as “easy” cases in that multiple arguments point to the same conclusion. One might also hope that such candor would reduce the instances in which judges sneak their own values into an opinion at the expense of intellectual rigor, but that seems unlikely to me— especially in cases in which multiple interpretations are possible, even if arguments for one side might seem stronger to the disinterested observer. Also, as we saw in chapter 7, jurors play a larger role in statutory interpretation than the standard discourse would lead us to conclude. How

226

chapter 8

much of a role they play depends upon the instructions that the judge gives them. Instructions vary on two different axes. The first is their complexity; the second, their fidelity to the legislature. As the instructions in the appendix illustrate, instructions that merely reiterate a long and complicated statute have little chance of being understood in all their detail. On the other hand, the more simplification in which a court engages, the less the statutory language enacted by the legislature will form the basis of liability. This is very important in the criminal context. If courts adhere to lenity, at least to some extent, then why should jurors be given a synopsis of a statute whose language might be subject to nuanced differences in interpretation? Despite this concern, courts must err to some extent on the side of comprehensibility. Although they should not abdicate their responsibility to construe statutes, they should not make the juror’s task into an obvious charade either. These suggestions are not designed to lead to radical change. This book’s principal contribution to judges—if any judges indeed read the book— is to bring some level of insight and comfort to them about the difficulty of their task. By recognizing both the strengths and the limitations of our cognitive capacities to make laws that are at once crisp and flexible, judges should be able to get a better sense of when their exercise of discretion becomes an absolute necessity. Whether one considers this exercise of discretion a necessary evil or a welcomed degree of flexibility depends upon one’s political perspective. But the need for such discretion is inevitable. I hope this book has succeeded in making it clear that judges need not apologize for doing their jobs.

Legislatures Legislators can do more to help judges decide close cases. For various reasons, it is not always easy to know how the enacting legislature would have wanted a particular dispute to be decided. To some extent, the problem arises from the fact that the most difficult questions of statutory interpretation often involve details that the legislature might not have resolved in its own mind, a point that Justice Scalia emphasizes.2 When this happens, judges frequently resort to coherence arguments to predict what the legislators would have considered the appropriate result.3 Because all such arguments require that inferences of intent be drawn from indirect evidence, there is often room for vigorous dispute.

legislatures, judges, and statutory interpretation

227

Legislators can be more helpful to judges by giving them more information about what they are trying to accomplish. One way of heading off fights over judicial reference to statutory intent is for the legislature to make a statement of its intent right in the statute itself. When the legislature expresses its intent, there can be no serious argument that statutory language must trump it. Congress has incorporated statements of purpose into statutes from time to time. For the most part, though, these statements have played only an occasional role in statutory interpretation. The main reason for the absence of such statements arises from the legislative process itself. Just about any complex statute can be enacted only after a great deal of debate and compromise. At that point, it would be very difficult for legislators to come to agreement over a statement of purpose, knowing that courts will use that statement to resolve disputes about the statute’s scope in one direction or the other. The result is that most statements of purpose, at least when it comes to federal statutes, concern laws that are fairly straightforward. For example, a federal law contains the following provision: Statement of purpose. It is the purpose of this title [section] to encourage and support educational programs for criminal offenders in correctional institutions.4

The statute goes on to do just that. Or consider a law creating the Women’s Rights National Historical Park: Statement of purposes. It is the purpose of this section to preserve and interpret for the education, inspiration, and benefit of present and future generations the nationally significant historical and cultural sites and structures associated with the struggle for equal rights for women and to cooperate with State and local entities to preserve the character and historic setting of such sites and structures.5

The statute then authorizes the creation of the park in upstate New York. Moreover, it would be strange to find a statement of purpose attached to a criminal statute. Criminal laws are passed to punish those who violate the statute. Perhaps such a statement would be useful in cases that pit the rule of lenity and respect for legislative intent against each other by making clear the latter, but such cases are not common.

228

chapter 8

There are a few notable cases, however, in which the Supreme Court has relied on a statutory statement of purpose. Perhaps the most prominent of these is Babbitt v. Sweet Home Chapter of Communities for a Great Oregon, often known as the Northern Spotted Owl Case.6 The Endangered Species Act makes it illegal to “take” a member of an endangered species. It defines “take” as follows: “The term ‘take’ means to harass, harm, pursue, hunt, shoot, wound, kill, trap, capture, or collect, or to attempt to engage in any such conduct.”7 A regulation issued by the Department of the Interior defined the statutory word “harm” as follows: “Harm in the definition of ‘take’ in the Act means an act which actually kills or injures wildlife. Such act may include significant habitat modification or degradation where it actually kills or injures wildlife by significantly impairing essential behavioral patterns, including breeding, feeding, or sheltering.”8 Loggers had been cutting down trees in which the northern spotted owl bred, a practice that was depleting the population and threatening the species. The question before the Court was the legitimacy, under the Chevron doctrine, of a definition of “harm” that includes “significant habitat modification.” A divided Court ruled in favor of the Department of the Interior (Babbitt, the named party, was then the secretary). Justice Stevens’s majority opinion is long and complex, as is Justice Scalia’s dissent. Together, they resemble fatigued boxers at the end of a long bout throwing as many punches as they still can. Among the arguments raised by the majority was the law’s statement of purpose, enacted as part of the statute itself. The Court noted: “As stated in § 2 of the Act, among its central purposes is ‘to provide a means whereby the ecosystems upon which endangered species and threatened species depend may be conserved.’ ”9 Nonetheless, statutory statements of purpose are not widely cited by courts. When they are, it is often to confirm an analysis that the Court seems likely to have reached on other grounds. There are exceptions.10 Courts are somewhat more likely, though, to refer to legislative findings that are part of the enacted statute. In a series of cases beginning with United States v. Lopez in 1995, the Supreme Court has strongly encouraged Congress to make legislative findings sufficient to demonstrate that laws enacted pursuant to the commerce clause of the Constitution have a sufficient connection to the regulation of interstate commerce to pass muster.11 Legislative findings are potentially relevant to statutory interpretation as evidence of the legislature’s broader purpose in enacting a law. As a general matter, judges resist referring to a statute’s purpose because,

legislatures, judges, and statutory interpretation

229

as Justice Scalia has put it, “the purpose of a statute includes not only what it sets out to change, but also what it resolves to leave alone.”12 In contrast, judges appear to be more comfortable referring to the facts that gave rise to the legislation and then drawing inferences of purpose and intent from those facts.13 Thus, if a legislature wishes to give judges a better idea of what concerned the drafters at the time of enactment, they can say so in a set of legislative findings, enacted as part of the law itself to avoid the concerns with legislative history discussed in chapter 4, and courts may well be willing to draw inferences from this information. What legislatures should not do, in contrast, is to attempt to dictate to judges the precise principles of interpretation they should employ. As discussed in chapter 6, the tools of statutory interpretation are broad, and there is no accurate way to predict which one will best uncover the legislative will in a specific case. Efforts by the Connecticut legislature to impose a regime of textualist interpretation have served the positive function of structuring the courts’ analysis but have otherwise accomplished very little. Judges now declare a law’s language either plain or ambiguous and as a result rule whether the interpretive statute applies at all. I have examined a great many cases in this book and found that only rarely do courts engage in interpretation that is so out-of-bounds that it appears that they need a special set of rules to rein them in.

The Executive Finally, we have seen that the executive branch plays a substantial role in statutory interpretation. This includes the interpretation and implementation of regulatory statutes by agencies and the interpretation of criminal laws by prosecutors, who must decide whether to bring criminal charges against a defendant. Some have argued that the executive branch should play a greater role in statutory interpretation, and judges a smaller one. As discussed in chapter 6, Adrian Vermeule has taken a strong position on this question.14 To the extent that statutory interpretation remains dominated by the value of legislative primacy, I see little reason to extend the role of the executive at the expense of the courts. No doubt the sense that an agency’s interpretation— even if it is a plausible one that should pass muster under the Chevron doctrine—falls outside the legislative will is a principal

230

chapter 8

reason for judicial decisions not to apply the doctrine, as we have seen. And prosecutorial discretion, to the extent that it is tempered by questions of resource allocation, tends to keep prosecutors from bringing cases that ask for undue expansion of the law. The bottom line, then, is that traditional institutional roles perform an adequate job as a regime of statutory interpretation. Problems of construal are inevitable because of mismatches between the goal of lawmaking and our cognitive makeup. Borderline cases also exist in decisions about how much deference to yield to an agency with delegated interpretive authority. Nonetheless, a principal theme of this book has been that the existence of hard cases, even hard cases that have become opportunistically politicized by the judges, should not lead to the conclusion that interpretation is arbitrary in the bulk of cases. Evidence from which judges and others may draw inferences ranges from the clear to the indeterminate. We should not be so concerned that interpreting laws is a human function that we fight to deny aspects of our psychology that we use every day to cope successfully with life’s problems.

Appendix: Two Mail Fraud Jury Instructions

Federal Judicial Center, Pattern Criminal Jury Instructions The defendant, _________________, is accused of [e.g.: planning to get money by giving false information to Sarah Stone and Rubin Ross] and using the mail in connection with this plan. It is against federal law to cheat someone if the mail is used. For you to find _________________ guilty of this crime, you must be convinced that the government has proved each of these things beyond a reasonable doubt: First, that _________________ made a plan [e.g.: to obtain money based on giving false information about the Apex Corporation to Sarah Stone and Rubin Ross]. Second, that when _________________ made the plan, he knew the information he was giving was false. Third, that _________________ mailed something (caused another person to mail something) for the purpose of carrying out this plan. It does not matter whether this plan succeeded, or whether _________________ made money from this plan. Nor does it matter whether the false information was contained in the material that was mailed. However, for you to decide that ______ ___________ is guilty, you must find, beyond a reasonable doubt, that __________ _______ made this plan intending to deceive [Stone and Ross] and to make money from the plan and that the mail was used to carry out the plan. Each separate use of the mail during the carrying out of a scheme to defraud is a separate offense.

232

appendix

Leonard B. Sand et al., Modern Federal Jury Instructions In order to sustain this charge, the government must prove each of the following elements beyond a reasonable doubt: First, that there was a scheme or artifice to defraud or to obtain money or property (if applicable: or the intangible right of honest services) by materially false and fraudulent pretenses, representations or promises, as alleged in the indictment; Second, that the defendant knowingly and willfully participated in the scheme or artifice to defraud, with knowledge of its fraudulent nature and with specific intent to defraud (if applicable: or that he knowingly and intentionally aided and abetted others in the scheme); and Third, that in execution of that scheme, the defendant used or caused the use of the mails (or a private or commercial interstate carrier or interstate wires) as specified in the indictment. The first element that the government must prove beyond a reasonable doubt is that there was a scheme or artifice to defraud [the victim] of money or property (if applicable: or the intangible right of honest services) by means of false or fraudulent pretenses, representations or promises. This first element is almost self-explanatory. A “scheme or artifice” is merely a plan for the accomplishment of an object. A scheme to defraud is any plan, device, or course of action to obtain money or property (or the intangible right of honest services) by means of false or fraudulent pretenses, representations or promises reasonably calculated to deceive persons of average prudence. “Fraud” is a general term which embraces all the various means by which human ingenuity can devise and which are resorted to by an individual to gain an advantage over another by false representations, suggestions or suppression of the truth, or deliberate disregard for the truth. Thus, a “scheme to defraud” is merely a plan to deprive another of money or property (or of the intangible right to honest services) by trick, deceit, deception or swindle. The scheme to defraud is alleged to have been carried out by making false (or fraudulent) statements (representations) (claims) (documents). A statement, representation, claim or document is false if it is untrue when made and was then known to be untrue by the person making it or causing it to be made. A representation or statement is fraudulent if it was falsely made with the intention to deceive. Deceitful statements of half truths or the concealment of material facts, and the expression of an opinion not honestly entertained may also constitute false or fraudulent statements under the statute.

two mail fraud jury instructions

233

The deception need not be premised upon spoken or written words alone. The arrangement of the words, or the circumstances in which they are used may convey the false and deceptive appearance. If there is deception, the manner in which it is accomplished is immaterial. (If applicable: The failure to disclose information may also constitute a fraudulent representation if the defendant was under a legal, professional or contractual duty to make such a disclosure, the defendant actually knew such disclosure ought to be made, and the defendant failed to make such disclosure with the intent to defraud.) The false or fraudulent representation (or failure to disclose) must relate to a material fact or matter. A material fact is one which would reasonably be expected to be of concern to a reasonable and prudent person in relying upon the representation or statement in making a decision (e.g., with respect to a proposed investment). This means that if you find a particular statement of fact to have been false, you must determine whether that statement was one that a reasonable person or investor might have considered important in making his or her decision. The same principle applies to fraudulent half truths or omissions of material facts. (If applicable: The representations which the government charges were made as part of the scheme to defraud are set forth in paragraph ___ of the indictment, which I have already read to you. It is not required that every misrepresentation charged in the indictment be proved. It is sufficient if the prosecution proves beyond a reasonable doubt that one or more of the alleged material misrepresentations were made in furtherance of the alleged scheme to defraud.) In addition to proving that a statement was false or fraudulent and related to a material fact, in order to establish a scheme to defraud, the government must prove that the alleged scheme contemplated depriving another of money or property (or of the intangible right of honest services). However, the government is not required to prove that the defendant himself originated the scheme to defraud. Furthermore, it is not necessary that the government prove that the defendant actually realized any gain from the scheme or that the intended victim actually suffered any loss. (If applicable: In this case, it so happens that the government does contend that the proof establishes that persons were defrauded and that the defendant profited. Although whether or not the scheme actually succeeded is really not the question, you may consider whether it succeeded in determining whether the scheme existed.) A scheme to defraud need not be shown by direct evidence, but may be established by all of the circumstances and facts in the case. If you find that the government has sustained its burden of proof that a scheme to defraud, as charged, did exist, you next should consider the second element. The second element that the government must prove beyond a reasonable doubt is that the defendant participated in the scheme to defraud knowingly, willfully and with specific intent to defraud.

234

appendix

“Knowingly” means to act voluntarily and deliberately, rather than mistakenly or inadvertently. “Willfully” means to act knowingly and purposely, with an intent to do something the law forbids; that is to say, with bad purpose either to disobey or to disregard the law. “Intent to defraud” means to act knowingly and with the specific intent to deceive, for the purpose of causing some financial or property loss to another (if applicable: or of depriving another of the intangible right of honest services). The question of whether a person acted knowingly, willfully and with intent to defraud is a question of fact for you to determine, like any other fact question. This question involves one’s state of mind. Direct proof of knowledge and fraudulent intent is almost never available. It would be a rare case where it could be shown that a person wrote or stated that as of a given time in the past he committed an act with fraudulent intent. Such direct proof is not required. The ultimate facts of knowledge and criminal intent, though subjective, may be established by circumstantial evidence, based upon a person’s outward manifestations, his words, his conduct, his acts and all the surrounding circumstances disclosed by the evidence and the rational or logical inferences that may be drawn from them. Circumstantial evidence, if believed, is of no less value than direct evidence. In either case, the essential elements of the crime must be established beyond a reasonable doubt. Since an essential element of the crime charged is intent to defraud, it follows that good faith on the part of the defendant is a complete defense to a charge of mail (or wire) fraud. A defendant, however, has no burden to establish a defense of good faith. The burden is on the government to prove fraudulent intent and the consequent lack of good faith beyond a reasonable doubt. Under the mail fraud statute, even false representations or statements, or omissions of material facts, do not amount to a fraud unless done with fraudulent intent. However misleading or deceptive a plan may be, it is not fraudulent if it was devised or carried out in good faith. An honest belief in the truth of the representations made by a defendant is a good defense, however inaccurate the statements may turn out to be. (If applicable: There is another consideration to bear in mind in deciding whether or not defendant acted in good faith. You are instructed that if the defendant participated in the scheme to defraud, then a belief by the defendant, if such belief existed, that ultimately everything would work out so that no one would lose any money does not require a finding by you that the defendant acted in good faith. If the defendant participated in the scheme for the purpose of causing some financial or property loss to another, then no amount of honest belief on the part of the defendant that the scheme would (e.g., ultimately make a profit for the investors) will excuse fraudulent actions or false representations by him.)

two mail fraud jury instructions

235

As a practical matter, then, in order to sustain the charges against the defendant, the government must establish beyond a reasonable doubt that he knew that his conduct as a participant in the scheme was calculated to deceive and, nonetheless, he associated himself with the alleged fraudulent scheme for the purpose of causing some loss to another. (If applicable: The government can also meet its burden of showing that the defendant had knowledge of the falsity of the statements if it establishes beyond a reasonable doubt that he acted with deliberate disregard of whether the statements were true or false, or with a conscious purpose to avoid learning the truth. If the government establishes that the defendant acted with deliberate disregard for the truth, the knowledge requirement would be satisfied unless the defendant actually believed the statements to be true. This guilty knowledge, however, cannot be established by demonstrating that the defendant was merely negligent or foolish.) To conclude on this element, if you find that the defendant was not a knowing participant in the scheme or that he lacked the specific intent to defraud, you should acquit him. On the other hand, if you find that the government has established beyond a reasonable doubt not only the first element, namely the existence of the scheme to defraud, but also this second element, that the defendant was a knowing participant and acted with specific intent to defraud, and if the government also establishes the third element, as to which I am about to instruct you, then you have a sufficient basis upon which to convict the defendant. The third and final element that the government must establish beyond a reasonable doubt is the use of the mails in furtherance of the scheme to defraud. (For conduct occurring after September 13, 1994, add: The use of the mails as I have used it here includes material sent through either the United States Postal Service or a private or commercial interstate carrier.) The mailed matter need not contain a fraudulent representation or purpose or request for money. It must, however, further or assist in the carrying out of the scheme to defraud. It is not necessary for the defendant to be directly or personally involved in the mailing, as long as the mailing was reasonably foreseeable in the execution of the alleged scheme to defraud in which the defendant is accused of participating. In this regard, it is sufficient to establish this element of the crime if the evidence justifies a finding that the defendant caused the mailing by others. This does not mean that the defendant must specifically have authorized others to do the mailing. When one does an act with knowledge that the use of the mails will follow in the ordinary course of business or where such use of the mails can reasonably be foreseen, even though not actually intended, then he causes the mails to be used. (If applicable: The government contends that it was reasonably foreseeable that the mails would be used in the ordinary course of business (e.g., to send confirmations of purchases), and therefore that the defendant caused the mailings.)

With respect to the use of the mails, the government must establish beyond a reasonable doubt the particular mailing charged in the indictment. However, the government does not have to prove that the mailings were made on the exact date charged in the indictment. It is sufficient if the evidence establishes beyond a reasonable doubt that the mailing was made on a date substantially similar to the date charged in the indictment.

Notes Chapter One 1. See William Blackstone, 1 Commentaries on the Laws of England, intro­ duction, § 2 at *59. 2. Antonin Scalia, A Matter of Interpretation: Federal Courts and the Law 13 (1997). 3. Adrian Vermeule, Judging under Uncertainty: An Institutional Theory of Legal Interpretation (2006). 4. Vermeule’s provocative proposal, which would take away far more judicial power than would most other theorists, including Scalia and other textualists, has generated critical response. See Jonathan R. Siegel, Judicial Interpretation in the Cost-Benefit Crucible, 92 Minn. L. Rev. 387 (2007); Caleb E. Nelson, Statutory Interpretation and Decision Theory, 74 U. Chi. L. Rev. 329 (2007). 5. Guido Calabresi, A Common Law for the Age of Statutes 164 (1982); this book publishes Calabresi’s Oliver Wendell Holmes Lectures, delivered at Harvard in 1977. 6. William N. Eskridge Jr., Dynamic Statutory Interpretation (1994). 7. This last point is made by Deborah Widiss, Shadow Precedents and the Separation of Powers: Statutory Interpretation of Congressional Overrides, 84 Notre Dame L. Rev. 511 (2009). 8. Richard Posner, How Judges Think 214–15 (2008). 9. Circuit City Stores v. Adams, 532 U.S. 105 (2001). 10. Food & Drug Administration v. Brown & Williamson Tobacco Corp., 529 U.S. 120 (2000). 11. Judge Posner argues, in fact, that the most important thing for a judge to do is to take seriously the consequences of a decision for future actors. See Richard A. Posner, Law, Pragmatism, and Democracy (2003). 12. N.J. Stat. Ann. § 48:12-152(b) (2009). 13. N.J. Stat. Ann. § 48:12-100 (2009).

238

notes to pages 5–16

14. N.J. Stat. Ann. § 48:12-152(a) (2009). 15. See http://www.mta.info/nyct /rules/index.html (accessed June 24, 2008). 16. Letter from New York City Transit Authority dated June 30, 2008. I am grateful to the Transit Authority’s office of legal counsel for providing this information. 17. William Neuman, Between-Cars Crackdown Adds to Arrests on Subway, N.Y. Times, March 27, 2007, http://www.nytimes.com/2007/03/27/nyregion/27crime .html?scp=42&sq=March%2027,%202007&st=cse (accessed August 27, 2008). 18. The sign is illustrated in Sewell Chan, Watch Those Changing Rules: Finish Sodas on the Platform, N.Y. Times, June 28, 2005, http://www.nytimes .com/2005/06/28/nyregion/28rules.html (accessed June 24, 2008). 19. See http://www.mta.info/nyct /safety/index.html (accessed June 24, 2008). 20. That is the case, for example, for tax evasion. See Cheek v. United States, 498 U.S. 192 (1991) (construing 26 U.S.C. § 7201). 21. See Tom Tyler, Why People Obey the Law (2006). 22. Cheryl Boudreau, Arthur Lupia, Matthew D. McCubbins, & Daniel B. Rodriguez, What Statutes Mean: Interpretive Lessons from Positive Theories of Communication and Legislation, 44 San Diego L. Rev. 957, 965–67 (2007). 23. For a recent case, see Volvo Trucks of North America, Inc. v. United States, 367 F.3d 204, 211–12 (4th Cir. 2004). 24. 471 U.S. 84 (1985). 25. 551 U.S. 205 (2007). 26. 28 U.S.C. § 2107(c) (2009). 27. Bowles, 551 U.S. at 215 (Souter, J., dissenting). 28. Aristotle, Nicomachean Ethics bk. 5, ch. 10 (F. H. Peters trans.). For dis­ cussion, see Frederick Schauer, Profiles, Probabilities and Stereotypes (2004). 29. Aristotle, Nicomachean Ethics bk. 5, ch. 10 (F. H. Peters trans.). 30. For some interesting discussion, see Paul Bloom, Descartes’ Baby: How the Science of Child Development Explains What Makes Us Human (2004). 31. See Muscarello v. United States, 524 U.S. 125 (1998), discussed in chapter 3. 32. 143 U.S. 457 (1892). 33. For discussion of some linguistic strategies that arise in the context of tax shelters, see Steven A. Dean & Lawrence M. Solan, Tax Shelters and the Code: Navigating between Text and Intent, 26 Va. Tax Rev. 879 (2007). 34. See chapter 5 for further discussion.

Chapter Two 1. Antonin Scalia, A Matter of Interpretation: Federal Courts and the Law 23 (1997). See also Antonin Scalia, The Rule of Law as a Law of Rules, 56 U. Chi. L. Rev. 1175 (1989).

notes to pages 16–18

239

2. William K. Eskridge, The New Textualism, 37 UCLA L. Rev. 621 (1990). 3. Benjamin Cardozo, The Nature of the Judicial Process 143 (1921). 4. Richard Posner makes a similar point: “Most cases are not even appealed, because the outcome of the appeal is a foregone conclusion, usually because the case really is ‘controlled’ by precedent or clear statutory language. For the same reason, many potential cases are never even filed.” Richard Posner, How Judges Think 44–45 (2008). 5. George L. Priest & Benjamin Klein, The Selection of Disputes for Litigation, 13 J. Legal Stud. 1, 17 (1984). See also Kevin M. Clermont & Theodore Eisenberg, Appeal from Jury or Judge Trial: Defendants’ Advantage, 3 Am. L. & Econ. Rev. 125 (2001) (analyzing empirical data to show that civil defendants appealing their loss at trial obtain a reversal in 31 percent of cases, while plaintiffs succeed in just 13 percent of their appeals). 6. Peter Siegelman & Joel Waldfogel, Toward a Taxonomy of Disputes: Evi­ dence through the Prism of the Priest /Klein Model, 28 J. Legal Stud. 101, 104 (1999). The authors attribute the differing rates of success to such factors as the litigation experience of the parties, which might create asymmetries in their levels of certainty of victory. 7. For evidence of this dynamic, see William J. Stuntz, Plea Bargaining and Criminal Law’s Disappearing Shadow, 117 Harv. L. Rev. 2548 (2004). See also David Lynch, The Impropriety of Plea Agreements: A Tale of Two Counties, 19 Law & Soc. Inquiry 115 (1994) (recounting the author’s own experience with plea bargains as prosecutor and defense lawyer in state courts). For a historical perspective, see George Fisher, Plea Bargaining’s Triumph: A History of Plea Bargaining in America (2003). 8. U.S. Sentencing Commission, 2003 Sourcebook of Federal Sentencing Statistics 22, table 10, available at http://www.ussc.gov/ANNRPT/2003/SBTOC03. htm (reproducing data for federal criminal plea bargains from 1997 through 2003). The seventeen state courts participating in the National Association for Court Management report a similar trend: the vast majority of criminal defendants plead guilty, and just 3 percent of state felony cases went to trial in 2003. Cynthia G. Lee & Robert C. LaFountain, Felony Trial Rates Are Declining, 12 Caseload Highlights (National Center for State Courts, Williamsburg, VA, October 2005), http://www.ncsconline.org/D_Research/csp/Highlights/ V0112N01.pdf. For a thorough account of the empirical data from the 1960s through 2002, see Marc Galanter, The Vanishing Trial: An Examination of Trials and Related Matters in Federal and State Courts, 1 J. Empirical Legal Stud. 459, 495 (2004). See also Brian J. Ostrom et al., Examining Trial Trends in State Courts: 1976–2002, 1 J. Empirical Legal Stud. 755, 766 (2004) (examining data from state courts that show a decrease in trial rates: in 1976, 9 percent of cases went to a full trial on the merits; in 2002, just 3 percent of cases reached trial). 9. U.S. Sentencing Commission, 2007 Sourcebook of Federal Sentencing

240

notes to pages 18–23

Sta­tistics, fig. C, available at http://www.ussc.gov/annrpt /2007/SBtoc07.htm. The cases are Blakely v. Washington, 542 U.S. 296 (2004), and United States v. Booker, 543 U.S. 220 (2005). For discussion, see Stephanos Bibas, White-Collar Plea Bargaining and Sentencing after Booker, 47 Wm. & Mary L. Rev. 721, 732 (2005) (predicting an increase in the rate of criminal cases going to trial “unless, as seems likely, prosecutors offer even more generous plea bargains to compensate, driving sentences down”). 10. 18 U.S.C. § 1621 (2009). 11. But, of course, there are difficult perjury cases, including the allegations made against President Clinton. For discussion of language issues in the law of perjury, see Lawrence M. Solan & Peter M. Tiersma, Speaking of Crime: The Lan­ guage of Criminal Justice ch. 11 (2005). 12. See Antonin Scalia, The Rule of Law as a Law of Rules, 56 U. Chi. L. Rev. 1175 (1989). The importance of this article as a reflection of Scalia’s approach to the law more generally is recognized and discussed extensively by Steven L. Winter, A Clearing in the Forest: Life, Law, and Mind 186 (2001). 13. See Philip N. Johnson-Laird, Mental Models (1983). 14. See, e.g., Andrew Goliszek, In the Name of Science: A History of Secret Programs, Medical Research, and Human Experimentation (2003); Alfred W. McCoy, A Question of Torture: CIA Interrogation from the Cold War to the War on Terror (2006); John Marks, The Search for the Manchurian Candidate: The CIA and Mind Control (1979). 15. The Roman historian Procopius recorded the first account of a “wild” child raised without human language. Procopius named the boy Aegisthus and claimed to have met him in person. Procopius, History of the Wars xvii (H. B. Dewing trans., 1914). See also Michael Newton, Savage Girls and Wild Boys: A History of Feral Children (2002); Roger Shattuck, The Forbidden Experiment: The Story of the Wild Boy of Aveyron (1994). 16. 18 U.S.C. § 201 (2009). 17. In 1986, Congress amended the statute’s corrupt-intent language. Previously, the statute read: “Whoever . . . corruptly asks, demands, exacts, solicits, seeks, accepts, receives, or agrees to receive anything of value for himself. . . .” In the course of amending the law to remove gender-specific pronouns, Congress changed this language to read: “Whoever . . . corruptly demands, seeks, receives, accepts, or agrees to receive or accept anything of value personally. . . .” Thus, Congress eliminated the verbs “asks” and “exacts” from the law. These two verbs were also dropped from (b)(4), the section dealing with bribing witnesses. Criminal Law and Procedure Technical Amendments Act of 1986, Pub. L. 99-946, § 46(c), 100 Stat. 3592 (codified as amended at 18 U.S.C. § 201 (2009)). 18. 18 U.S.C. § 201 (2009). 19. 18 U.S.C. § 201(b)(3)–(4) (2009). 20. 18 U.S.C. § 201(a) (2009).

notes to pages 23–28

241

21. 18 U.S.C. § 201(a)(1) (2009). 22. Dixson v. United States, 465 U.S. 482 (1984). 23. Cf. United States v. Griffin, 401 F. Supp. 1222 (S.D. Ind. 1975) (corporate president is a public official), and United States v. Del Toro, 513 F.2d 656 (2d Cir. 1975) (city employee administering federal program is not a public official). 24. Krichman v. United States, 256 U.S. 363 (1921). 25. 18 U.S.C. § 201(b)(1) (2009). 26. Linda Greenhouse, Justices Refuse to Hear Appeals in 7 ABSCAM Cases, N.Y. Times, June 1, 1983, A1. For linguistic analysis of state-of-mind issues in the Abscam scandal, see Roger Shuy, Language Crimes: The Use and Abuse of Language Evidence in the Courtroom (1993). 27. For a useful account of the history and scope of the entrapment defense, see Ronald J. Allen et al., Clarifying Entrapment, 89 J. Crim. L. & Criminology 407 (1999). 28. United States v. Williams, 705 F.2d 603, 622–23 (2d Cir. 1983). 29. United States v. Gorman, 807 F.2d 1299, 1304–5 (6th Cir. 1986). Gorman was prosecuted under the gratuity statute, which contains language similar to that in the bribery statute and is presently codified at 18 U.S.C. § 201(c). For the differences between a bribe and a gratuity, see United States v. Sun-Diamond Growers of California, 526 U.S. 398 (1999). 30. United States v. Myers, 692 F.2d 823 (2d Cir. 1982). 31. Ian Ayres & Gregory Klass, Insincere Promises: The Law of Misrepresented Intent (2005). 32. What the court did not discuss in the Myers case, however, is how the separate requirement of corruption fits into this analysis. Even assuming that Myers accepted money in exchange for an insincere promise, doing so may not be a corrupt act if he never intended to follow through. The point never arose in that case because the trial judge had permitted the jury to hear Myers’s “I was just play-acting” defense, and the jury rejected it. But the issue remains to be resolved if the defense is used again and rejected based on the court’s linguistic analysis in Myers. 33. 18 U.S.C. § 201(c)(2) (2009). 34. See, e.g., United States v. Singleton, 165 F.3d 1297 (10th Cir. 1999); United States v. Anty, 203 F.3d 305 (4th Cir. 2000). 35. United States v. Tomblin, 46 F.3d 1369 (5th Cir. 1995). 36. Id. at 1379 n. 17. 37. Id. 38. United States v. Barash, 365 F.2d 395 (2d Cir. 1966). 39. Id. at 401–2. 40. Id. at 402. 41. United States v. Brewster, 506 F.2d 62 (D.C. Cir. 1974). 42. 18 U.S.C. § 201(b)(2)(A) (2009). 43. 18 U.S.C. § 201(c)(1)(B) (2009).

242

notes to pages 28–35

44. Brewster, 506 F.2d at 81. 45. Id. at 82. 46. For accessible discussion, see Paul Bloom, Descartes’ Baby: How the Science of Child Development Explains What Makes Us Human (2004). 47. For other examples of how pragmatic inferences make their way into the system of criminal justice, see Lawrence M. Solan & Peter M. Tiersma, Speaking of Crime: The Language of Criminal Justice (2005). 48. United States v. Biaggi, 909 F.2d 662 (2d Cir. 1990). A federal statute now prohibits members of Congress and their senior staff from serving as private legal counsel or in other such “fiduciary” roles. Ethics in Government Act of 1978, Pub. L. 101-194, § 601, 103 Stat. 1716, 1761 (codified at 5 U.S.C.A. app. 4, § 502 (2008)). 49. Biaggi was not the only public official who fell in the Wedtech scandal. Others included a former Bronx Borough president, Stanley Simon, and the New York regional administrator of the Small Business Administration, Peter Neglia. Neglia was convicted of permitting Wedtech to be awarded contracts despite illegal activity. He had been offered a lucrative job at Wedtech after his government service ended. 50. Biaggi, 909 F.2d at 681. 51. Id. at 685. 52. See, e.g., United States v. Chen, 754 F.2d 817 (9th Cir. 1985); United States v. Strand, 574 F.2d 993, 995 (9th Cir. 1978); United States v. Fenster, 449 F. Supp. 435 (E.D. Mich. 1978). 53. George Miller, Contextuality, in Mental Models in Cognitive Science: Essays in Honour of Phil Johnson-Laird 1 (Jane Oakhill & Alan Garnham eds., 1996). 54. 18 U.S.C. § 201(a) (2009) (emphasis added). 55. United States v. Gjieli, 717 F.2d 968 (6th Cir. 1983). 56. Id. at 972. 57. Id. 58. Dixson, 465 U.S. at 493 n. 12, citing Federal Conflict of Interest Legislation: Hearings on H.R. 302, H.R. 3050, H.R. 3411, H.R. 3412, and H.R. 7139 before the Antitrust Subcommittee of the House Committee on the Judiciary, 87th Cong., 1st sess., 36 (1961). 59. Committee reports from both houses of Congress emphasized that the 1962 revision of the bribery laws would make “no significant changes of substance” and “would not restrict the broad scope of the present bribery statutes as construed by the courts.” Dixson, 465 U.S. at 494, citing S. Rep. 87-2213 at 4 (1962) and H.R. Rep. 87-748 at 17 (1961). 60. See, e.g., United States v. Neville, 82 F.3d 1101 (D.C. Cir. 1996). 61. 532 U.S. 105 (2001). 62. 9 U.S.C. § 2 (2009).

notes to pages 35–38

243

63. 9 U.S.C. § 1 (2009). 64. For an excellent discussion of this linguistic phenomenon in the context of the Americans with Disabilities Act, see Jill C. Anderson, Just Semantics: The Lost Readings of the Americans with Disabilities Act, 117 Yale L.J. 992 (2008). 65. Stephen Breyer, Active Liberty: Interpreting Our Democratic Constitution 91–95 (2005). 66. See, e.g., 14 Penn Plaza LLC v. Pyett, 129 S. Ct. 1456 (2009) (collective bargaining agreement calling for arbitration of age discrimination suits enforceable, requiring dismissal of class action); Buckeye Check Cashing, Inc. v. Cardegna, 546 U.S. 440 (2006) (holding that determination of whether a contract containing an arbitration clause is an illegal contract is for the arbitrators to decide); AlliedBruce Terminix Companies, Inc. v. Dobson, 513 U.S. 265 (1995) (applying Federal Arbitration Act to state court proceedings); Southland Corp. v. Keating, 465 U.S. 1 (1984) (claims brought pursuant to state statutory law subject to arbitration). 67. 524 U.S. 125 (1998). 68. There is a great deal of discussion of these concepts in the philosophical and legal literatures. For excellent discussion, see Ralf Poscher, Ambiguity and Vagueness in Legal Interpretation, in The Oxford Handbook of Language and Law (L. Solan & P. Tiersma eds., forthcoming, 2011). See also Lawrence M. Solan, Vagueness and Ambiguity in Legal Interpretation, in Vagueness in Normative Texts (V. K. Bhatia, J. Enberg, M. Gotti, & D. Heller eds., 2005); Roy Sorensen, Vagueness and Contradiction (2001). 69. Portions of this section are excerpted from my essay, Why Laws Work Pretty Well, but Not Great: Words and Rules in Legal Interpretation, 26 Law & Soc. Inquiry 243 (2001), reviewing Steven Pinker’s book Words and Rules: The Ingredients of Language (1999). On brain function, see, e.g., Oliver Sacks, The Case of the Colorblind Painter, in An Anthropologist on Mars 3 (1995). A particular part of the brain, the V4 region, is responsible for color processing. In this story, traumatic injury to this region caused an artist to become colorblind. 70. There is a great deal of literature on the issue of mental modularity. Some of it is quite polarized, but much of it, sensibly enough, directs itself to the question of the extent to which the mind is modular and the extent to which it is plastic (i.e., able to use general properties for functions that only appear different on the surface). For a seminal work, see Jerry A. Fodor, The Modularity of Mind (1983). 71. Steven Pinker, Words and Rules: The Ingredients of Language (1999). 72. See Peter W. Culicover & Ray Jackendoff, Simpler Syntax 25–26 (2005); Ray Jackendoff, Construction after Construction and Its Theoretical Challenges, 84 Language 8 (2008). 73. This is the connectionist position that Pinker rebuts — successfully in my opinion. See, e.g., Jeffrey L. Elman, Elizabeth A. Bates, Mark H. Hohnson, Annette Karmiloff-Smith, Domenico Parisi, & Kim Plunkett, Rethinking Innateness (1996).

244

notes to pages 38–46

For a somewhat polemical, but insightful, reaction to this and similar research in the connectionist framework, see Gary Marcus, The Algebraic Mind: Integrating Connectionism and Cognitive Science (2003). 74. See Charles Yang, Dig-Dug, & Think-Thunk, London Review of Books, August 24, 2000 (reviewing Pinker’s Words and Rules). 75. Noam Chomsky, Aspects of the Theory of Syntax 21 (1965). 76. 471 U.S. 419 (1985). I discuss this case in more detail in Lawrence M. Solan, The Language of Judges (1993). 77. 7 U.S.C. § 2024(b)(1) (1985). 78. Church of the Holy Trinity v. United States, 143 U.S. 457 (1892). 79. McBoyle v. United States, 283 U.S. 25 (1931). 80. Smith v. United States, 508 U.S. 223 (1993). 81. Jerry A. Fodor, Concepts: Where Cognitive Science Went Wrong 137 (1998). 82. For discussion of how the dissipation of consensus can have an effect on legal judgments, see Lawrence M. Solan, Terri Rosenblatt, & Daniel Osherson, False Consensus Bias in Contract Interpretation, 108 Columbia L. Rev. 1268 (2008). 83. Eleanor Rosch, Cognitive Representations of Semantic Categories, 104 J. Experimental Psychol.: Gen. 192 (1975). 84. Portions of this section are adapted from Lawrence M. Solan, Law, Language and Lenity, 40 Wm. & Mary L. Rev. 57 (1998). 85. 18 U.S. (5 Wheat.) 76 (1820). 86. Id. at 79. 87. Id. at 95 (emphasis added). 88. William Blackstone, 1 Commentaries on the Laws of England at *88. 89. 28 F. Cas. 733 (C.C.D. Mass. 1838). 90. Id. at 734. 91. Id. at 735. 92. See, e.g., United States v. R.L.C., 503 U.S. 291, 306 n. 6 (1992). 93. 529 U.S. 667 (2000). 94. 18 U.S.C. § 666(b) (2009). 95. Fischer, 529 U.S. at 677–78. 96. 524 U.S. 184 (1998). 97. 18 U.S.C. § 924 (1998) (the language has since beem amended). 98. Bryan, 524 U.S. at 191. 99. 515 U.S. 593, 598 (1995). 100. Id. at 600 (quoting McBoyle v. United States, 283 U.S. 25, 27 (1931)). 101. Id. at 612 (Scalia, J., concurring in part and dissenting in part). 102. Jones v. United States, 529 U.S. 848 (2000); 18 U.S.C. § 844(i) (2009). 103. “The Congress shall have Power . . . To regulate Commerce with foreign Nations, and among the several States, and with the Indian Tribes.” U.S. Const. art. I, § 8.

notes to pages 46–51

245

104. Jones, 529 U.S. at 856. 105. 511 U.S. 39 (1994). 106. 18 U.S.C. § 3565(a) (1994). 107. Granderson, 511 U.S. at 45. 108. Pub. L. 103-322, 108 Stat. 1796, 2017 (1994). 109. 510 U.S. 135 (1994). 110. 31 U.S.C. §§ 5322(a), 5324 (2009). 111. 976 F.2d 1280 (9th Cir. 1992). 112. 532 U.S. 105 (2001). 113. See United States v. Brewster, 408 U.S. 501 (1972). In fact, only after this question was resolved by the Supreme Court could the prosecution of Senator Brewster, discussed above, proceed. 114. The Supreme Court’s most recent decision under that statute, United States v. Sun-Diamond Growers of California, 526 U.S. 398 (1999), held that to prove that a government official has accepted an illegal gratuity the government must show a relationship between the exchange and a particular official act that the giver hopes to influence.

Chapter Three 1. This chapter is adapted from my article The New Textualists’ New Text, 38 Loy. L.A. L. Rev. 2027 (2005). 2. For a useful summary of some of this literature, see Michael H. Koby, The Supreme Court’s Declining Reliance on Legislative History: The Impact of Justice Scalia’s Critique, 36 Harv. J. on Legis. 369 (1999). 3. Antonin Scalia, A Matter of Interpretation: Federal Courts and the Law (1997); John F. Manning, The Absurdity Doctrine, 116 Harv. L. Rev. 2387 (2003); John F. Manning, Textualism and the Equity of the Statute, 101 Colum. L. Rev. 1 (2001); William T. Mayton, Law among the Pleonasms: The Futility and Aconstitutionality of Legislative History in Statutory Interpretation, 41 Emory L.J. 113 (1992); Adrian Vermeule, Legislative History and the Limits of Judicial Competence: The Untold Story of Holy Trinity Church, 50 Stan. L. Rev. 1833 (1998). 4. See, e.g., Stephen Breyer, On the Uses of Legislative History in Interpreting Statutes, 65 S. Cal. L. Rev. 845 (1992); Patricia M. Wald, The Sizzling Sleeper: The Use of Legislative History in Construing Statutes in the 1988–89 Term of the United States Supreme Court, 39 Am. U. L. Rev. 277 (1990). 5. See, e.g., Richard Posner, How Judges Think (2008); Richard A. Posner, The Problems of Jurisprudence (1990); William N. Eskridge Jr., Legislative History Values, 66 Chi.-Kent L. Rev. 365 (1990); Daniel B. Rodriguez & Barry R. Weingast, The Positive Political Theory of Legislative History: New Perspectives on the 1964 Civil Rights Act and Its Interpretation, 151 U. Pa. L. Rev. 1417 (2003); Jonathan R.

246

notes to pages 51–55

Siegel, The Use of Legislative History in a System of Separated Powers, 53 Vand. L. Rev. 1457 (2000); Jonathan R. Siegel, What Statutory Drafting Errors Teach Us about Statutory Interpretation, 69 Geo. Wash. L. Rev. 309 (2001); Charles Tiefer, The Reconceptualization of Legislative History in the Supreme Court, 2000 Wis. L. Rev. 205. 6. United States v. Fisher, 6 U.S. (2 Cranch) 358, 386 (1805). For recent citations by the Supreme Court, see Koons Buick Pontiac GMC, Inc. v. Nigh, 543 U.S. 50, 66 n. 1 (2004) (Stevens, J., concurring); Lamie v. U.S. Trustee, 540 U.S. 526, 543 n. 1 (2004) (Stevens, J., concurring). 7. Jonathan T. Molot, The Rise and Fall of Textualism, 106 Colum. L. Rev. 1, 36–48 (2006). 8. Munaf v. Green, 128 S. Ct. 2207, 2217 (2008) (Roberts, C.J.) (“We begin, as always, with the language of the statute”) (quoting Duncan v. Walker, 533 U.S. 167, 172 (2001)). 9. Leocal v. Ashcroft, 543 U.S. 1, 8 (2004) (Rehnquist, C.J.) (“Our analysis begins with the language of the statute”). 10. Intel Corp. v. Advanced Micro Devices, Inc., 542 U.S. 241, 255 (2004) (Ginsburg, J.) (“As ‘in all statutory construction cases, we begin [our examination of § 1782] with the language of the statute’ ”) (quoting Barnhart v. Sigmon Coal Co., 534 U.S. 438, 450 (2002)). 11. See Conroy v. Aniskoff, 507 U.S. 511, 519 (1993) (Scalia, J., concurring) (“The greatest defect of legislative history is its illegitimacy”); Scalia, A Matter of Interpretation at 31–32; Manning, Absurdity at 2431. 12. Caleb Nelson, What Is Textualism? 91 Va. L. Rev. 347, 349 (2005). 13. Manning, Absurdity at 2392–93; see also Manning, Textualism at 69 (discussing the constitutional danger in the judiciary’s performance of legislative functions). 14. See Smith v. United States, 508 U.S. 223, 241–47 (1993), discussed below, in which Justice Scalia dissents from an opinion affirming the conviction of the defendant for “using” a firearm during a drug-trafficking crime, where the defendant attempted to trade a weapon for illegal drugs. 15. 143 U.S. 457 (1892). 16. Act of February 26, 1885, ch. 164, 23 Stat. 332–33. 17. Church of the Holy Trinity, 143 U.S. at 458, 459, 463. 18. See Scalia, A Matter of Interpretation at 18–23; Adrian Vermeule, Judging under Uncertainty: An Institutional Theory of Legal Interpretation (2006); Vermeule, Legislative History at 1835–36. 19. Actually, Justice Brewer himself slipped up in writing the opinion, as my colleague Bailey Kuklin has pointed out to me. The opinion says: “Where the mind labors to discover the design of the legislature, it seizes everything from which aid can be derived.” 143 U.S. at 462.

notes to pages 55–59

247

20. See, e.g., Tennessee Valley Authority v. Hill, 437 U.S. 153 (1978); Committee for Hand Gun Control, Inc. v. Consumer Product Safety Commission, 388 F. Supp. 216 (1974). 21. 283 U.S. 25 (1931). See H. L. A. Hart, Positivism and the Separation of Law and Morals, 71 Harv. L. Rev. 593, 606–15 (1958); Lon L. Fuller, Positivism and Fidelity to Law—A Reply to Professor Hart, 71 Harv. L. Rev. 630 (1958). For an interesting analysis, see Steven L. Winter, Transcendental Nonsense, Metaphoric Reasoning, and the Cognitive Stakes for Law, 137 U. Pa. L. Rev. 1105, 1172–98 (1989). 22. McBoyle, 283 U.S. at 26. 23. Ludwig Wittgenstein, Philosophical Investigations vol. 1, ¶ 139 (G. E. M. Anscombe trans., 1958). 24. 546 U.S. 481 (2006). 25. 28 U.S.C. § 1346(b)(1) (2006). 26. 28 U.S.C. § 2680(b) (2009). 27. Dolan, 546 U.S. at 486. 28. Smith, 508 U.S. 223 (1993). 29. 18 U.S.C. § 924(c)(1)(A) (2009). 30. 18 U.S.C. at § 924(c)(1). The sentence length under the statute has since been changed. 31. Smith, 508 U.S. at 224. 32. Id. at 228. 33. Id. at 228–29. 34. Black’s Law Dictionary 1541 (6th ed. 1990). 35. See Arthur v. Morgan, 112 U.S. 495, 500 (1884); Astor v. Merritt, 111 U.S. 202, 213 (1884). 36. Smith, 508 U.S. at 241 (Scalia, J., dissenting). 37. Id. at 242–43 (citations omitted). 38. See id. at 241–42. 39. Id. at 230. 40. See Scalia, A Matter of Interpretation at 18–23. 41. Id. at 23–25. 42. 242 U.S. 470, 485 (1917). 43. See Laime v. U.S. Trustee, 540 U.S. 526, 534 (2004); Hartford Underwriters Insurance Co. v. Union Planters Bank, N.A., 530 U.S. 1 (2000); Abdul-Akbar v. McKelvie, 239 F.3d 307, 313 (3d Cir. 2001) (“[W]e have adopted what is called the American Plain Meaning Rule”). 44. Peter M. Tiersma, A Message in a Bottle: Text, Autonomy, and Statutory Interpretation, 76 Tul. L. Rev. 431 (2001). 45. See William N. Eskridge Jr., The New Textualism, 37 UCLA L. Rev. 621 (1990); William N. Eskridge Jr., Textualism, the Unknown Ideal? 96 Mich. L. Rev. 1509 (1998).

248

notes to pages 59–62

46. 544 U.S. 385 (2005); 18 U.S.C. § 922(g)(1) (2009) (emphasis added). 47. Small, 544 U.S. at 390. 48. Id. at 396 (Thomas, J., dissenting). 49. See Manning, Textualism at 71–72. 50. See generally Eskridge, New Textualism at 623 (discussing Scalia’s newtextualism approach). 51. Morales v. Trans World Airlines, Inc., 504 U.S. 374, 383 (1992) (quoting Park’N Fly, Inc. v. Dollar Park & Fly, Inc., 469 U.S. 189, 193–94 (1985)). 52. Statements of this sort are inconsistent with Scalia’s stronger position that legislative intent is, in principle, irrelevant to the process of statutory interpretation because legislatures can have no intent. For detailed discussion, see Lawrence M. Solan, Private Language, Public Laws: The Central Role of Legislative Intent in Statutory Interpretation, 93 Geo. L.J. 427, 433–36 (2005). 53. Chisom v. Roemer, 501 U.S. 380, 404 (1991) (Scalia, J., dissenting). 54. 74 U.S. (7 Wall.) 482 (1868); Church of the Holy Trinity v. United States, 143 U.S. 457, 460 (1892). 55. Kirby, 74 U.S. at 482. 56. See id. at 487. 57. See generally Miranda Oshige McGowan, Do as I Do, Not as I Say: An Empirical Investigation into Justice Scalia’s Ordinary Meaning Method of Statutory Interpretation, 78 Miss. L.J. 129 (2008); Philip P. Frickey, Revisiting the Revival of Theory in Statutory Interpretation: A Lecture in Honor of Irving Younger, 84 Minn. L. Rev. 199 (1999) (arguing for a “critical pragmatism” that candidly acknowledges that the Court goes beyond the plain meaning of text during statutory analysis); Frederick Schauer, The Practice and Problems of Plain Meaning: A Response to Aleinikoff and Shaw, 45 Vand. L. Rev. 715, 728 (1992) (arguing that courts “use various information-limiting decisional devices of which plain meaning is a preeminent example to allocate scarce decisional resources and to achieve a degree of agreement . . . which might be harder to achieve were decisions based on a wider set of decisionally relevant information”). 58. Ardestani v. Immigration and Naturalization Service, 502 U.S. 129, 135 (1991); see also PUD No. 1 v. Washington Department of Ecology, 511 U.S. 700, 725 (1994) (Thomas, J., dissenting) (discussing the “plain and ordinary meaning” of “discharge” in connection with the applicability of an environmental statute). 59. See, e.g., Auguste v. Ridge, 395 F.3d 123, 145 (3d Cir. 2005) (“we have noted that ‘congressional intent is presumed to be expressed through the ordinary meaning of the statute’s plain language’ ”); United States v. Lachman, 387 F.3d 42, 50 (1st Cir. 2004) (“if the language of a statute or regulation has a plain and ordinary meaning, courts need look no further and should apply the regulation as it is written”) (quoting Textron Inc. v. Commissioner, 336 F.3d 26, 31 (1st Cir. 2003)); Carter v. Tennant Co., 383 F.3d 673, 682 (7th Cir. 2004) (declaring the “plain and ordinary meaning” of the statutory language clear).

notes to pages 63–65

249

60. For further discussion of this point, see Lawrence M. Solan, Language and Law: Definitions in Law, in Elsevier Encyclopedia of Language and Linguistics (2006). 61. See J. A. Fodor et al., Against Definitions, 8 Cognition 263 (1980). 62. This example has a long history in American legal literature. See H. L. A. Hart, Positivism and the Separation of Law and Morals, 71 Harv. L. Rev. 593, 607 (1958); Lon L. Fuller, Positivism and Fidelity to Law—a Reply to Professor Hart, 71 Harv. L. Rev. 630, 663 (1958). For recent discussion, see Frederick Schauer, Formalism, 97 Yale L.J. 509, 526 (1988); Pierre Schlag, No Vehicles in the Park, 23 Seattle Univ. L. Rev. 381 (1999); Steven L. Winter, An Upside/Down View of the Countermajoritarian Difficulty, 69 Tex. L. Rev. 1881, 1885–1905 (1991). 63. Charles J. Fillmore, Towards a Descriptive Framework for Spatial Deixis, in Speech, Place, and Action 31 (Robert J. Jarvella & Wolfgang Klein eds., 1982). 64. See Ludwig Wittgenstein, Philosophical Investigations § 67 (G. E. M. Anscombe trans., 3d ed. 1968). 65. Examples abound. For a good example, see Daniel Solove, Conceptualizing Privacy, 90 Calif. L. Rev. 1087, 1096 (2002). 66. See Eleanor Rosch, Cognitive Representations of Semantic Categories, 104 J. Experimental Psychol.: Gen. 192 (1975). 67. See generally Fodor et al., Against Definitions (discussing alternatives to definitional treatments of language and mind). 68. Rosch, Cognitive Representations at 233. 69. Sharon Lee Armstrong, Lila R. Gleitman, & Henry Gleitman, What Some Concepts Might Not Be, 13 Cognition 263, 267 (1983). 70. For recent discussion, see Andrew C. Connolly, Jerry A. Fodor, Lila R. Gleitman, & Henry Gleitman, Why Stereotypes Don’t Even Make Good Defaults, 103 Cognition 1 (2006). See also Jerry A. Fodor & Ernest Lepore, The Red Herring and the Pet Fish: Why Concepts Still Can’t Be Prototypes, 58 Cognition 253 (1996); Jerry A. Fodor, Concepts: Where Cognitive Science Went Wrong 88–108 (1998). 71. Douglas L. Medin, William D. Wattenmaker, & Sarah E. Hampson, Family Resemblance, Conceptual Cohesiveness, & Category Construction, 19 Cognitive Psychol. 242 (1987). 72. Steven A. Sloman, The Empirical Case for Two Systems of Reasoning, 119 Psychol. Bull. 3, 19 (1996). 73. For more detailed discussion of these issues, including suggestions for how broader theories make their way into our concepts, see Gregory L. Murphy, The Big Book of Concepts (2002). 74. For discussion of this dynamic, see Robert W. Gordon, Legalizing Outrage, in Aftermath: The Clinton Impeachment and the Presidency in the Age of Political Spectacle 97 (Leonard V. Kaplan & Beverly I. Moran eds., 2001).

250

notes to pages 65–68

75. See id. For discussion of gradations of lying in the context of perjury prosecutions, see Richard A. Posner, An Affair of State: The Investigation, Impeachment, and Trial of President Clinton 147–48 (1999). 76. 18 U.S.C. § 1621 (1) (2009). For discussion, including the vocabulary for true but misleading statements, see Steven L. Winter, A Clearing in the Forest: Law, Life, and Mind (2001). 77. For fuller discussion of this issue, see Lawrence M. Solan, Perjury and Impeachment: Rule of Law or Rule of Lawyers? in Aftermath at 199, 203; Lawrence M. Solan & Peter M. Tiersma, Speaking of Crime: The Language of Criminal Justice ch. 11 (2005). 78. CNN.com, Poll: Most Americans Think Clinton Committed Perjury (reporting on a CNN/USA Today/Gallup poll conducted in December 1998, while the House Judiciary considered filing Articles of Impeachment), http://www.cnn .com/ALLPOLITICS/stories/1998/12/10/poll/ (accessed June 22, 2008). 79. For a study showing that people do not always draw the same conclusions about what constitutes a lie, see Linda Coleman & Paul Kay, Prototype Semantics: The English Word Lie, 57 Language 26 (1981). 80. See Vermeule, Judging under Uncertainty at 93–102. 81. 501 U.S. 380 (1991). 82. 42 U.S.C. § 1973(b) (2009). 83. 412 U.S. 755, 766 (1973). 84. Chisom, 501 U.S. at 398. 85. For discussion of Scalia’s selective use of dictionaries in this case, see Ellen P. Aprill, The Law of the Word: Dictionary Shopping in the Supreme Court, 30 Ariz. St. L.J. 275, 317 (1998). 86. Chisom, 501 U.S. at 410 (Scalia, J., dissenting). As Aprill points out, the third edition of Webster’s had come out prior to the opinion, but it appears that it was less helpful to Scalia’s argument than was the second, which he cited in support of his analysis of word meaning. See Aprill, The Law of the Word at 317. 87. Chisom, 501 U.S. at 410 (Scalia, J., dissenting). 88. McGowan, Do as I Do. 89. See, e.g., Rush Prudential HMO, Inc. v. Moran, 536 U.S. 355, 365 (2002) (“In trying to extrapolate congressional intent in a case like this, when congressional language seems simultaneously to preempt everything and hardly anything, we ‘have no choice’ but to temper the assumption that ‘the ordinary meaning . . . accurately expresses the legislative purpose’ ” (citations omitted)); Johnson v. United States, 595 U.S. 694, 707 n. 9 (“As Justice Scalia remarks, in relying on an uncommon sense of the word, we are departing from the rule of construction that prefers ordinary meaning. . . . But this is exactly what ought to happen when the ordinary meaning fails to fit the text and when the realization of clear congressional policy . . . is in tension with the result that customary interpretive rules would deliver”).

notes to pages 68–73

251

90. See Lawrence M. Solan, Statutory Inflation and Institutional Choice, 44 Wm. & Mary L. Rev. 2209 (2003), for discussion of the broad use of this principle by the courts. 91. Padilla v. Lever, 429 F.3d 910, 919 (9th Cir. 2005). 92. 432 F.3d 161 (3d Cir. 2005). 93. Id. at 162. 94. 42 U.S.C. § 9607(a)(4)(A), (B) (2009) (emphasis added). 95. Dupont De Nemours, 432 F.3d at 164. 96. See Aprill, The Law of the Word at 277. 97. See id. at 277 n. 2; Samuel A. Thumma & Jeffrey L. Kirchmeier, The Lexicon Has Become a Fortress: The United States Supreme Court’s Use of Dictionaries, 47 Buff. L. Rev. 227, 256 (1999) (“Justice Scalia has relied on the dictionary more times than any other Justice in the history of the Court”); Note, Looking It Up: Dictionaries and Statutory Interpretation, 107 Harv. L. Rev. 1437, 1439 n. 14 (1994). 98. 512 U.S. 218 (1994); 47 U.S.C. §§ 203(a), 203(b)(2) (2001). 99. See MCI Telecommunications Corp., 512 U.S. at 226 (citing Chevron USA, Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 843 (1984)). 100. See MCI Telecommunications Corp., 512 U.S. at 225–28. 101. Id. at 227. 102. My thanks to Jed Rubenfeld for this important observation. 103. Smith, 508 U.S. 223 (1993). 104. 516 U.S. 137 (1995). 105. Id. at 143. 106. Clark D. Cunningham & Charles J. Fillmore, Using Common Sense: A Linguistic Perspective on Judicial Interpretation of “Use a Firearm,” 73 Wash. U. L.Q. 1159 (1995). 107. For an analysis of situations in which phrasal analysis might lead to more nuanced interpretation than reliance on the dictionary, see Craig Hoffman, Parse the Sentence First: Curbing the Urge to Resort to the Dictionary When Interpreting Legal Texts, 6 N.Y.U. J. Legis. & Pub. Pol. 401 (2003). 108. 128 S. Ct. 579 (2007). 109. Id. at 583 (citations omitted). 110. 524 U.S. 125 (1998). 111. 18 U.S.C. § 924(c)(1) (2009). 112. Muscarello, 524 U.S. at 127–28. 113. See id. at 128–39. 114. Id. at 130 (quoting Black’s Law Dictionary 214 [6th ed. 1990]). 115. See id. at 130. 116. Id. at 128. 117. Id. at 129 (quoting 2 Kings 9:28 and Isaiah 30:6 [King James]). 118. Id.

252

notes to pages 73–78

119. Id. 120. Id. at 139–44 (Ginsburg, J., dissenting). 121. Id. at 142 n. 4 (citing different translations of the Bible, other than King James, including the New English Bible, the New American Bible, and Jewish Publication Society, Tanakh, the Holy Scriptures: The New JPS Translation according to Traditional Hebrew Text [1998]). In a conversation with Professor Aaron Twerski, a biblical and legal scholar, he stated that, in the first example, the original is translated literally as “cause to ride” and refers to a dead body. In the second, a better translation is “lift.” 122. Muscarello, 524 U.S. at 143 (citation omitted). 123. Id. at 130. 124. United States v. Sheldon, 15 U.S. 119, 121 (1817). The next instance was Buel v. Van Ness, 21 U.S. 312, 323 (1823). 125. 524 U.S. 184 (1998). 126. Williams v. Taylor, 529 U.S. 420, 434 (2000) (“When the words of the Court are used in a later statute governing the same subject matter, it is respectful of Congress and of the Court’s own processes to give the words the same meaning in the absence of specific direction to the contrary”). 127. See Jonathan Siegel, The Polymorphic Principle and the Judicial Role in Statutory Interpretation, 84 Tex. L. Rev. 339 (2005), for an argument that me­ chanical application of this presumption is not justifiable. 128. William W. Buzbee, The One-Congress Fiction in Statutory Interpretation, 149 U. Pa. L. Rev. 171 (2000). 129. Lexis search of Supreme Court library: “dictionary w/20 ordinary meaning” (conducted on February 24, 2009). 130. See Federal Express Corp. v. Holowecki, 128 S. Ct. 1147, 1161 (2008) (Thomas, J., dissenting) (construing the word “charge”); Lopez v. Gonzales, 549 U.S. 47 (2006) (Thomas, J., dissenting) (construing “illicit trafficking”); Dolan v. United States Postal Service, 546 U.S. 481, 493–94 (2006) (Thomas, J., dissenting) (con­ struing “negligent transmission”). 131. Bailey, 516 U.S. at 143. 132. Cunningham & Fillmore, Using Common Sense at 1186–87. 133. See Muscarello, 524 U.S. at 125. 134. Id. 135. Tanakh, 581, 678. 136. 18 U.S.C. § 924(c)(1) (1998). 137. Muscarello, 524 U.S. at 146–47 (Ginsburg, J., dissenting). 138. 18 U.S.C. § 924(b) (1998) (“Whoever, with intent to commit . . . an offense . . . ships, transports, or receives a firearm. . . .”). 139. See Muscarello, 524 U.S. at 143 (Ginsburg, J., dissenting). 140. See, e.g., Tony McEnery & Andrew Wilson, Corpus Linguistics (1996).

notes to pages 78–82

253

141. See, e.g., Malcolm Coulthard, On the Use of Corpora in the Study of Forensic Texts, 1 Forensic Linguistics 27 (1994). 142. See William S. Blatt, Interpretive Communities: The Missing Element in Statutory Interpretation, 95 Nw. U. L. Rev. 629 (2001). 143. 149 U.S. 304 (1893). 144. Id. at 307. 145. Id. 146. Minor v. Mechanics’ Bank, 26 U.S. (1 Pet.) 46, 64 (1828) (“The ordinary meaning of the language, must be presumed to be intended, unless it would manifestly defeat the object of the provisions”). 147. See William N. Eskridge Jr., Textualism and Original Understanding: Should the Supreme Court Read The Federalist but Not Statutory Legislative History? 66 Geo. Wash. L. Rev. 1301, 1323 (1998); Jane S. Schacter, The Confounding Common Law Originalism in Recent Supreme Court Statutory Interpretation: Implications for the Legislative History Debate and Beyond, 51 Stan. L. Rev. 1 (1998). 148. See, e.g., Brian H. Bix, Can Theories of Meaning and Reference Solve the Problem of Legal Determinacy? 16 Ratio Juris 281 (2003); Michael Green, Dworkin’s Fallacy, or What the Philosophy of Language Can’t Teach Us about the Law, 89 Va. L. Rev. 1897 (2003); Dennis Patterson, Fashionable Nonsense, 81 Tex. L. Rev. 841 (2003).

Chapter Four 1. This chapter is adapted from my article Private Language, Public Laws: The Central Role of Legislative Intent in Statutory Interpretation, 93 Geo. L.J. 427 (2005). 2. For interesting argument along these lines, see John F. Manning, Textualism as a Nondelegation Doctrine, 97 Colum. L. Rev. 673 (1997). For arguments to the contrary, see Jonathan R. Siegel, The Use of Legislative History in a System of Separated Powers, 53 Vand. L. Rev. 1457 (2000). See also Manning’s response, John F. Manning, Putting Legislative History to a Vote: A Response to Professor Siegel, 53 Vand. L. Rev. 1529 (2000); and Siegel’s reply to Manning’s response, Jonathan R. Siegel, Timing and Delegation: A Reply, 53 Vand. L. Rev. 1543 (2000). 3. See Blanchard v. Bergeron, 489 U.S. 87, 98–99 (1989) (Scalia, J., concurring) (arguing that legislative history is sometimes inserted “to influence judicial construction”); Manning, Textualism at 684–88, 731–33; Kenneth W. Starr, Observations about the Use of Legislative History, 1987 Duke L.J. 371, 376–77. 4. Antonin Scalia, A Matter of Interpretation: Federal Courts and the Law 29–37 (1997).

254

notes to pages 82–85

5. See, e.g., id. at 29–30; Frank H. Easterbrook, The Role of Original Intent in Statutory Construction, 11 Harv. J.L. & Pub. Pol. 59 (1988). 6. Oliver Wendell Holmes, The Theory of Legal Interpretation, 12 Harv. L. Rev. 417, 419 (1899). See, e.g., Hain v. Mullin, 324 F.3d 1146, 1151 (10th Cir. 2003) (Lucero, J., dissenting); St. Charles Investment Co. v. Commissioner, 232 F.3d 773, 776 (10th Cir. 2000); Public Citizen v. Carlin, 184 F.3d 900, 904 (D.C. Cir. 1999); Eubanks v. Stengel, 28 F. Supp. 2d 1024, 1033 (W.D. Ky. 1998). 7. Cf. William N. Eskridge Jr., Dynamic Statutory Interpretation 14–21 (1994) (providing liberal critiques), with Scalia, A Matter of Interpretation (providing a conservative critique). 8. Max Radin, Statutory Interpretation, 43 Harv. L. Rev. 863, 870 (1930). 9. See, e.g., Patricia M. Wald, The Sizzling Sleeper: The Use of Legislative History in Construing Statutes in the 1988–89 Term of the United States Supreme Court, 39 Am. U. L. Rev. 277, 282 (1990). 10. Stephen Breyer, On the Uses of Legislative History in Interpreting Statutes, 65 S. Cal L. Rev. 845 (1992). The arguments are also made in Breyer’s book Active Liberty: Interpreting Our Democratic Constitution (2005). 11. Examples abound. See, e.g., Bayer AG v. Housey Pharmaceuticals, Inc., 340 F.3d 1367, 1372 (Fed. Cir. 2003) (“Under these circumstances the text is ambiguous, and we must look beyond the particular language being construed”); United States v. Calor, 340 F.3d 428, 431 (6th Cir. 2003); United States v. Martin, 332 F.3d 827, 835 (5th Cir. 2003) (Dennis, J., concurring in part and dissenting in part). 12. See Eskridge, Dynamic Statutory Interpretation at 14; Scalia, A Matter of Interpretation at 31–32. 13. See Kenneth A. Shepsle, Congress Is a “They,” Not an “It”: Legislative Intent as Oxymoron, 12 Int’l. Rev. L. & Econ. 239, 254 (1992). 14. Such motivations for voting are consistent with public-choice theory. For a clear exposition of this perspective on the legislative process, see Daniel A. Farber & Philip P. Frickey, Law and Public Choice: A Critical Introduction (1991). 15. Patricia M. Wald, Some Observations on the Use of Legislative History in the 1981 Supreme Court Term, 68 Iowa L. Rev. 195, 214 (1983). 16. See, e.g., Blanchard v. Bergeron, 498 U.S. 87, 98–99 (1989) (Scalia, J., concurring); Manning, Textualism at 687–88. 17. This is a major theme among those who oppose the use of legislative history. See, e.g., Adrian Vermeule, Legislative History and the Limits of Judicial Competence: The Untold Story of Holy Trinity Church, 50 Stan. L. Rev. 1833, 1896 (1998). Eskridge regards this issue as an empirical question that, notwithstanding all the rhetoric, has not been adequately studied. William N. Eskridge Jr., Should the Supreme Court Read The Federalist but Not Statutory Legislative History? 66 Geo. Wash. L. Rev. 1301, 1323 (1998). I agree with Eskridge’s position. 18. This argument has, until very recently, resonated with English judges, who opposed the use of legislative history in legal argument. The House of Lords finally

notes to pages 85–93

255

permitted the use of legislative history under limited circumstances in Pepper v. Hart, [1992] 3 W.L.R. 1032. 19. Scalia, A Matter of Interpretation at 29–30. 20. Id. at 31–32. 21. 507 U.S. 511 (1993). 22. 50 U.S.C. app. §§ 501, 561 (2009). 23. 507 U.S. at 519 (Scalia, J., concurring) (citations omitted). 24. Wisconsin Public Intervenor v. Mortier, 501 U.S. 597, 610 n. 4 (1991) (cita­ tions omitted). 25. United States v. Fisher, 6 U.S. (2 Cranch) 358, 386 (1805). 26. Breyer, On the Uses of Legislative History at 864–65. 27. In re Sinclair, 870 F.2d 1340, 1343 (7th Cir. 1989). 28. See Frank Easterbrook, Statutes’ Domains, 50 U. Chi. L. Rev. 533, 545 (1983); Antonin Scalia, The Rule of Law as a Law of Rules, 56 U. Chi. L. Rev. 1175 (1989). 29. See Fed. R. Evid. 404, 609 (2009). 30. Margaret Gilbert, Sociality and Responsibility: New Essays in Plural Subject Theory (2000). My vignettes are modeled after hers but have been altered to make them more relevant to the points that I wish to make about the interpretation of statutes. 31. According to Wikipedia, SoHo is bounded roughly by Houston Street on the north, Lafayette Street on the east, Canal Street on the south, and Sixth Avenue on the west. 32. See Michael E. Bratman, Shared Intention, in Faces of Intention: Selected Essays on Intention and Agency 109 (1999). Although I have added Bratman’s insights to Gilbert’s for purposes of exposition, the two have somewhat different perspectives on group intent. For discussion, see Christopher Kutz, Complicity: Ethics and Law for a Collective Age 91 (2000); J. David Velleman, How to Share an Intention, 58 Phil. & Phenomenological Res. 29, 29–39 (1997). 33. See Hilary Putnam, The Meaning of “Meaning,” in Minnesota Studies in the Philosophy of Science: Language, Mind, and Language vol. 7, 131, 139 (Keith Gunderson ed., 1975). 34. Bratman makes this point by discussing plans to paint a house. Bratman, Shared Intention at 120–21. 35. Richard A. Posner, The Problems of Jurisprudence 269 (1990). 36. T. Alexander Aleinikoff, Updating Statutory Interpretation, 87 Mich L. Rev. 20, 21 (1988). 37. Ronald Dworkin, Law’s Empire 228–29 (1986). 38. Donald T. Campbell, Common Fate, Similarity, and Other Indices of the Status of Aggregates as Social Entities, 3 Behavioral Sci. 14, 17 (1958). 39. Robert P. Abelson et al., Perceptions of the Collective Other, 2 Personality & Soc. Psychol. Rev. 243, 246 (1998).

256

notes to pages 93–96

40. Jennifer L. Welbourne, The Impact of Perceived Entitivity on Inconsistency Resolution for Groups and Individuals, 35 J. Experimental Soc. Psychol. 481, 499 (1999). 41. Matthew J. O’Laughlin & Bertram F. Malle, How People Explain Actions Performed by Groups and Individuals, 82 J. Personality & Soc. Psychol. 33, 34 (2002). 42. Id. at 40–41. 43. For an extremely interesting discussion of this dynamic, see Daniel B. Rodriguez & Barry R. Weingast, The Positive Political Theory of Legislative History: New Perspectives on the 1964 Civil Rights Act and Its Interpretation, 151 U. Pa. L. Rev. 1417 (2003). 44. More complex statutes typically have a more complicated history. Yet the general outline — a small number of people heavily involved in the details — is necessarily true for almost all statutes, given the enormous demands on members’ time. For a legislative history relevant to my hypothetical story, see Arnold W. Reitze, The Legislative History of U.S. Air Pollution Control, 36 Hous. L. Rev. 679, 715–25 (1999). 45. It is not difficult to find almost exactly those words in judicial opinions. See, e.g., Jones v. United States, 527 U.S. 373, 419 (1999) (Ginsburg, J., dissenting) (“The House Report suggests that Congress understood and approved that construction”). 46. 516 U.S. 264, 276–77 (1996) (Stevens, J., concurring). 47. See Charles Tiefer, The Reconceptualization of Legislative History in the Supreme Court, 2000 Wisconsin L. Rev. 206, 209, 252–53. Among the cases Tiefer cites are United States v. Romani, 523 U.S. 517, 533–34 (1996) (Stevens, J., using legislative history), and Lewis v. United States, 523 U.S. 155, 170 (1998) (Breyer, J., using legislative history). 48. Breyer, On the Uses of Legislative History at 865–66. 49. See John C. Roberts, Are Congressional Committees Constitutional? Radical Textualism, Separation of Powers, and the Enactment Process, 52 Case W. Res. L. Rev. 489, 543–46 (2001). 50. For discussion of one such example, see Michelle S. Marks, The Legislative History of the “Equitable Remuneration” Provision Granted in the New Patent Term of 35 U.S.C. § 154(c), 4 U. Balt. Intell. Prop. L.J. 33, 52 (1995) (describing statements made by the administration in the context of the treaty negotiations that set the policy underlying the statute). 51. See, e.g., George A. Costello, Average Voting Members and Other “Benign Fictions”: The Relative Reliability of Committee Reports, Floor Debates, and Other Sources of Legislative History, 1990 Duke L.J. 39, 43–60; David Nimmer, Appreciating Legislative History: The Sweet and Sour Spots of the DMCA’s Commentary, 23 Cardozo L. Rev. 909, 921–27 (2002); Rodriguez & Weingast, Positive Political Theory at 1464–73. These complicated moments that threaten

notes to pages 96–101

257

the passage of a bill are frequently referred to as “veto gates” in the literature. See McNollgast, Legislative Intent: The Use of Positive Political Theory in Statutory Interpretation, 57 Law & Contemp. Probs. 3, 11 (1994). 52. See William N. Eskridge Jr., Overriding Supreme Court Statutory Interpretation Decisions, 101 Yale L.J. 331, 350, table 8 (1991). 53. Reed Dickerson, Statutory Interpretation: Dipping into Legislative History, 11 Hofstra L. Rev. 1125, 1131 (1983) (citations omitted). 54. 508 U.S. 223 (1993). 55. Id. at 240. 56. See George A. Miller, The Science of Words 237–39 (1991); Susan Carey, The Child as Word Learner, in Linguistic Theory and Psychological Reality 264 (Morris Halle et al. eds., 1978). 57. See, e.g., Gregory L. Murphy, The Big Book of Concepts 240–41 (2002); Gregory L. Murphy & Douglas L. Medin, The Role of Theories in Conceptual Coherence, in Concepts: Core Readings 425 (Eric Margolis & Stephen Laurence eds., 1999). 58. Hilary Putnam uses this observation in his approach to meaning. See Putnam, The Meaning of “Meaning” at 144–45. Putnam uses “gold” to illustrate this point. 59. See, e.g., Utah v. Evans, 536 U.S. 452, 467 (2002) (examining “technical literature” to come to grips with the meaning of “sampling” in conducting a census). 60. See Paul Bloom, How Children Learn the Meanings of Words (2000). 61. Alison Gopnik, Andrew N. Meltzoff, & Patricia K. Kuhl, The Scientist in the Crib: Minds, Brains, and How Children Learn 34 (1999). 62. Betty M. Repacholi, Infants’ Use of Attentional Cues to Identify the Referent of Another Person’s Emotional Expression, 34 Dev. Psychiatry 1017, 1019–24 (1998), discussed in Gopnik et al., Scientist in the Crib at 33. 63. Bloom, How Children Learn at 61; see also Paul Bloom, Descartes’ Baby: How the Science of Child Development Explains What Makes Us Human (2004). 64. Steven Pinker, The Blank Slate: The Modern Denial of Human Nature 62 (2002). 65. For a summary of some of this work, see Andrea D. Rosati et al., The Rocky Road from Acts to Dispositions: Insights for Attribution Theory from Developmental Research on Theories of Mind, in Intentions and Intentionality: Foundations of Social Cognition 287 (Bertram F. Malle et al. eds., 2001). 66. 538 U.S. 580, 588 (2003) (emphasis added) (citation omitted). 67. A LexisNexis search, conducted in both the federal court and state court databases on August 25, 2008, yielded more than 3,000 hits for each year of the search terms ((legislature or Congress) w/6 inten!) (the words “intense” and “intensive” account for a minuscule portion of those hits). LexisNexis stops counting at 3,000.

258

notes to pages 101–107

68. Holmes, Theory of Legal Interpretation at 419. 69. United States v. P. Lorillard Co., 267 U.S. 471, 473 (1925). 70. New York ex rel. Clyde v. Gilchrist, 262 U.S. 94, 98 (1923). 71. American Bank & Trust Co. v. Federal Reserve Bank, 256 U.S. 350, 357 (1921). 72. See Nicholas S. Zeppos, Legislative History and the Interpretation of Statutes: Toward a Fact-Finding Model of Statutory Interpretation, 76 Va. L. Rev. 1295, 1374 (1990) (showing that Holmes, despite his famous statement, was not devoted to textualism when it appeared that reliance on the plain language of a statute would lead to results inconsistent with broader principles of justice). 73. United States v. Mead Corp., 533 U.S. 218, 241 (2001) (Scalia, J., dissenting). 74. Holloway v. United States, 526 U.S. 1, 20 (1999) (Scalia, J., dissenting). 75. Sullivan v. Everhart, 494 U.S. 83, 90 (1990) (first emphasis added). 76. Morales v. Trans World Airlines, Inc., 504 U.S. 374, 383 (1992) (quoting FMC Corp. v. Holliday, 498 U.S. 52, 56 (1990)) (quoting Park ‘N Fly, Inc. v. Dollar Park & Fly, Inc., 469 U.S. 189, 193–94 (1985)). See discussion in chapter 3. 77. See Larry Alexander & Saikrishna Prakash, “Is That English You’re Speaking?” Some Arguments for the Primacy of Intent in Interpretation, 41 San Diego L. Rev. 967 (2004); Kent Greenawalt, Are Mental States Relevant for Statutory and Constitutional Interpretation? 85 Cornell L. Rev. 1609, 1619 (2000) (both drawing similar conclusions). 78. Green v. Bock Laundry Machine Co., 490 U.S. 504, 528 (1989) (Scalia, J., concurring). 79. Caleb Nelson, What Is Textualism? 91 Va. L. Rev. 347, 355 (2005). 80. For an excellent study of the contemporary use of this rhetoric, see James J. Brudney & Corey Ditslear, Canons of Construction and the Elusive Quest for Neutral Reasoning, 58 Vand. L. Rev. 1 (2005). 81. 511 U.S. 39 (1994). 82. 18 U.S.C. § 3565(a) (1990). This statute was amended in 1994 to correct the problem uncovered in Granderson; however, the 1990 version was still in force at the time of this decision. 83. See John F. Manning, The Absurdity Doctrine, 116 Harv. L. Rev. 2388 (2003). 84. Id. at 2476–85. 85. 513 U.S. 64, 67–68 (1994) (referring to 18 U.S.C. § 2252 (1988 & Supp. V)). 86. See, e.g., 513 U.S. at 73 (“We do not impute to Congress an intent to pass legislation that is inconsistent with the Constitution as construed by this Court”). 87. Id. at 80–81 (Scalia, J., dissenting). 88. 490 U.S. 504 (1989). 89. Fed. R. Evid. 609(a) (1989). 90. Green, 490 U.S. at 527 (Scalia, J. concurring).

notes to pages 107–112

259

91. For discussion, see Jonathan R. Siegel, What Statutory Drafting Errors Teach Us about Statutory Interpretation, 69 Geo. Wash. L. Rev. 309, 329–32 (2001). 92. 471 U.S. 84 (1985). 93. 43 U.S.C. § 1744 (2009). 94. Locke, 471 U.S. at 90 n. 7. By the time that Locke made his one-day error, the BLM had corrected its brochure. 95. Id. at 95–96 (citations omitted). 96. Richard A. Posner, The Problems Of Jurisprudence 267–68 (1990). See William D. Popkin, Law-Making Responsibility and Statutory Interpretation, 68 Ind. L.J. 865, 876–77 (1993); Nicholas S. Zeppos, Justice Scalia’s Textualism: The “New” New Legal Process, 12 Cardozo L. Rev. 1597, 1614–18 (1991). 97. Locke, 471 U.S. at 128–29 (Stevens, J., dissenting). 98. Id. at 90 n. 7. 99. Id. at 111–12 (O’Connor, J., concurring). 100. Harold Swafford, Esq., telephone interview with author, August 1996. Mr. Swafford represented the Locke family. 101. 501 U.S. 380 (1991). 102. 42 U.S.C. § 1973(b) (2009). 103. Id. at 410 (Scalia, J., dissenting). For discussion of Scalia’s preference for the second edition of Webster’s as a source of ordinary meaning, see Ellen P. Aprill, The Law of the Word: Dictionary Shopping in the Supreme Court, 30 Ariz. St. L.J. 275, 315–18 (1998). 104. Chisom, 501 U.S. at 410 (Scalia, J., dissenting). 105. Arthur Unger, Watergate—a Greek Tragedy of Our Times—Relived on PBS, Christian Sci. Monitor, July 22, 1983, at 17. 106. See Noam Chomsky, Knowledge of Language: Its Nature, Origin and Use 21–24 (1986); Noam Chomsky, New Horizons in the Study of Language and Mind (2000). For an accessible discussion, see Steven Pinker, The Language Instinct: How the Mind Creates Language (1994). 107. This perspective is consistent with that of those scholars who discuss the interpretation of statutes in terms of signal theory. See Cheryl Boudreau, Arthur Lupia, Matthew D. McCubbins, & Daniel B. Rodriguez, What Statutes Mean: Interpretive Lessons from Positive Theories of Communication and Legislation, 44 San Diego L. Rev. 957, 965–67 (2007). 108. Chomsky, Knowledge of Language at 21–24. 109. The distinction between knowledge of language in this sense and skills in the use of this knowledge in various social settings has been a matter of some controversy in the linguistic literature. For a valuable discussion, see Frederick J. Newmeyer, Grammar Is Grammar and Usage Is Usage, 79 Language 682 (2003). 110. See chapter 2 for discussion. 111. 18 U.S.C. § 201(b)(1) (2009).

260

notes to pages 112–116

112. For a very clear discussion of the difference, see M. B. W. Sinclair, Legislative Intent: Fact or Fabrication? 41 N.Y. L. Sch. L. Rev. 1329, 1358–60 (1997) (reviewing Eskridge, Dynamic Statutory Interpretation). 113. H. P. Grice, Utterer’s Meaning and Intention, 78 Phil. Rev. 147 (1969). For helpful discussion of these and related concepts, see Brian H. Bix, Can Theories of Meaning and Reference Solve the Problem of Legal Determinacy? 16 Ratio Juris 281 (2003); Heidi M. Hurd, Sovereignty in Silence, 99 Yale L.J. 945, 962–67 (1990). For discussion of these perspectives in the context of expressive theories of law, see Matthew Adler, Expressive Theories of Law: A Skeptical Overview, 148 U. Pa. L. Rev. 1363 (2000). 114. My thanks to Caleb Mason for this observation. 115. 242 U.S. 470, 485 (1917). See also Lamie v. United States Trustee, 540 U.S. 526, 532 (2004) (“It is well established that when the statute’s language is plain, the sole function of the courts—at least where the disposition required by the text is not absurd—is to enforce it according to its terms,” quoting Hartford Underwriters Insurance Co. v. Union Planters Bank, N.A., 530 U.S. 1, 6 (2000)). 116. United States v. Lavin, 942 F.2d 177, 184 (3d Cir. 1991) (boldface added) (additional citations omitted). 117. See, e.g., Miller v. Carlson, 768 F. Supp. 1331, 1334 (N.D. Cal. 1991) (“In construing statutory provisions, courts must first consider the text of the statute. It is well-settled that the plain language of a statute provides the best evidence of legislative intent” [citation omitted].). 118. See, e.g., In re Order on Prosecution of Criminal Appeals by the Tenth Judicial Circuit Pub. Defender, 561 So.2d 1130, 1137 (Fla. 1990) (“The best evidence of the intent of the legislature is generally the plain meaning of the statute”); State v. Martinez, 52 P.3d 1278, 1278 (Utah 2002) (“We discern legislative intent and purpose by first looking to the ‘best evidence’ of its meaning, which is the plain language of the statute itself”). 119. Conroy v. Aniskoff, 507 U.S. 511, 528 (1993) (Scalia, J., concurring). 120. Id. 121. Eskridge, Should the Supreme Court Read The Federalist? at 1323. For an excellent example of an empirically driven account of the use of legislative history, see Jane S. Schacter, The Confounding Common Law Originalism in Recent Supreme Court Interpretation: Implications for the Legislative History Debate and Beyond, 51 Stan. L. Rev. 1 (1998). 122. 143 U.S. 457 (1892). 123. Act of February 26, 1885, ch. 164, 23 Stat. 332–33. 124. Adrian Vermeule, Judging under Uncertainty: An Institutional Theory of Legal Interpretation 93–98 (2005); Vermeule, Legislative History at 1837. See also United States v. Rector, etc., of the Church of the Holy Trinity, 36 F. 303 (C.C.S.D.N.Y. 1888) (lower-court decision).

notes to pages 116–124

261

125. See Pepper v. Hart, [1993] A.C. 593, 1 All E.R. 42. For discussion of this issue, see Michael Sinclair, The Proper Treatment of “Interpretive Choice” in Statutory Decision-Making, 45 N.Y. L. Sch. L. Rev. 389, 427–30 (2002). 126. See Carol Chomsky, Unlocking the Mysteries of Holy Trinity: Spirit, Letter, and History in Statutory Interpretation, 100 Colum. L. Rev. 901 (2000). 127. Chisom, 501 U.S. 380. 128. 334 F.3d 670 (7th Cir. 2003). 129. 18 U.S.C. § 922(d) (2000). 130. Jefferson, 334 F.3d at 675 (citation omitted). 131. Radin, Statutory Interpretation at 870.

Chapter Five 1. Richard Posner, How Judges Think 93–121 (2008). 2. Richard Primus, When Should Original Meanings Matter? 107 Mich. L. Rev. 165 (2008). 3. United States v. Lanier, 520 U.S. 259, 266 (1997). 4. 18 U.S. (5 Wheat.) 76, 101 (1820). 5. 283 U.S. 25, 27 (1931). 6. United States v. Aguilar, 515 U.S. 593, 600 (1995). 7. See William N. Eskridge Jr., Public Values in Statutory Interpretation, 137 U. Pa. L. Rev. 1007 (1989). 8. See Toolson v. New York Yankees, Inc., 346 U.S. 356 (1953). 9. 18 U.S.C. § 201 (2009). 10. The information comes from annual data sets made available by the Administrative Office of the Courts, part of the U.S. Department of Justice. The government sorts cases into categories, including bribery. However, that sorting includes a number of bribery statutes, only one of which is being studied here. Thus, analysis of the data required the separation of bribery prosecutions under 18 U.S.C. § 201. The data are available from the Inter-university Consortium for Political and Social Research (ICPSR), http://www.icpsr.umich.edu (search for “Federal Court Cases: Integrated Data Base”). 11. The rules governing the publication of appellate opinions are not uniform among the federal circuit courts and have not been entirely stable during the period studied. See Deborah Jones Merritt & James J. Brudney, Stalking the Secret Law: What Predicts Publication in the United States Court of Appeals, 54 Vand. L. Rev. 71 (2001). However, the slope of the curve is sufficiently uniform over time to suggest that it is reliable in general. 12. Sentencing Act of 1987, Pub. L. 100-182, § 2, 101 Stat. 1266 (stating that the Sentencing Guidelines took effect on November 1, 1987).

262

notes to pages 124–132

13. U.S. Sentencing Guidelines Manual 3.E.1.1(a) (2009). 14. For discussion, see Marc L. Miller, Domination and Dissatisfaction: Prosecutors as Sentencers, 56 Stan. L. Rev. 1211, 1252 (2004). 15. The figures in this graph are actually running three-year averages. Presenting them in this way smooths the curves, making trends easier to see. Had each year been presented without such averaging, the overall trend of the total appeals would be exactly the same, but the graph more jagged. Moreover, in some instances a prosecution brought in one year might not result in an appeal until the following year. This fact should make only a small difference in the numbers presented here, since the total number of prosecutions increased slowly and linearly. 16. Stefanie A. Lindquist & Frank B. Cross, Empirically Testing Dworkin’s Chain Novel Theory: Studying the Path of Precedent, 80 N.Y.U. L. Rev. 1156 (2005). 17. Ronald Dworkin, Law’s Empire 229 (1986). 18. Lindquist & Cross, Empirically Testing at 1204. 19. Steven L. Winter, A Clearing in the Forest: Law, Life, and Mind (2001). 20. Steven Shepard, Settled Questions: The Interpretation of Federal Criminal Statues in the Circuit Courts of Appeals, unpublished paper, Yale Law School (2006) (on file with the author). The statutes he examined are 21 U.S.C. § 848 (2006), 18 U.S.C. § 1951 (2006), 18 U.S.C. § 1029 (2006), and 18 U.S.C. § 1956 (2006). 21. Money Laundering Control Act of 1986, Pub. L. 99-570, § 1352(a), 100 Stat. 3207 (1986). 22. 283 U.S. 25 (1931). 23. 407 U.S. 258 (1972); Federal Baseball Club of Baltimore v. National League of Professional Baseball Clubs, 259 U.S. 200 (1922). 24. www.mlb.com. 25. Toolson, 346 U.S. at 357. 26. Flood, 407 U.S. at 285–86 (Burger, C. J., concurring) (citation omitted). The question of statutory interpretation and institutional choice is an important one. For recent work that relies heavily on the notion of institutional competence, see Adrian Vermeule, Judging under Uncertainty (2006). For a broad, interesting study of questions of institutional choice, see Neil K. Komesar, Law’s Limits: The Rule of Law and the Supply and Demand of Rights (2001). I return below to the question of how institutional choice can affect the course of statutory interpretation. 27. William N. Eskridge, Dynamic Statutory Interpretation (1994). 28. Id. at 48–80. 29. Guido Calabresi, A Common Law for the Age of Statutes (1982). 30. Dworkin, Law’s Empire at 229. 31. Peter M. Tiersma, The Textualization of Precedent, 82 Notre Dame L. Rev. 1187 (2007). 32. 763 N.E.2d 38 (Mass. 2002). 33. 318 N.E.2d 478,481 (Mass. 1974).

notes to pages 133–137

263

34. 18 U.S.C. § 1341 (2009). 35. Dan M. Kahan, Lenity and Federal Common Law Crimes, 1994 Sup. Ct. Rev. 345. See also John Calvin Jeffries Jr., Legality, Vagueness, and the Construction of Penal Statutes, 71 Va. L. Rev. 189 (1985). 36. 489 U.S. 705, 707 (1989). 37. Id. at 711–12. 38. Id. at 723 (Scalia, J., dissenting). 39. For a stimulating discussion of this distinction, with attention given to the tax system, see Leo Katz, Ill-Gotten Gains: Evasion, Blackmail, Fraud, and Kindred Puzzles of the Law (1996). Steven Dean and I discuss some problems that tax shelters pose for statutory interpreters in Steven A. Dean & Lawrence M. Solan, Tax Shelters and the Code: Navigating between Text and Intent, 26 Va. Tax Rev. 879 (2007). 40. This discussion is modified from Lawrence M. Solan, Should Criminal Statutes Be Interpreted Dynamically? 2002 Issues in Legal Scholarship, Dynamic Statutory Interpretation, article 8, http://www.bepress.com /ils /iss3/art8. For further discussion, see Peter M. Tiersma, A Message in a Bottle: Text, Autonomy, and Statutory Interpretation, 76 Tul. L. Rev. 431, 462–71 (2001). 41. California Penal Code § 12276(a)–(c) (2009). 42. Id. § 12276.1(a) (2009). 43. Id. § 12276.5 (2009), amended Stats. 2006 ch. 793 (AB 2728). 44. Tiersma, Textualization. 45. Richard A. Posner, The Federal Courts: Crisis and Reform 285–86 (1985); Richard A. Posner, Legal Formalism, Legal Realism, and the Interpretation of Statutes and the Constitution, 37 Case W. Res. L. Rev. 179 (1987). 46. Hotchkiss v. Greenwood, 52 U.S. (11. How.) 248 (1851). 47. Cuno Engineering Corp. v. Automatic Devices Corp., 314 U.S. 84, 91 (1941) (emphasis added). 48. See, e.g., Koochook Co. v. Barrett, 158 F.2d 463, 466 (8th Cir. 1946) (denying patent for new method of making brake shoes for automobiles). 49. 35 U.S.C. § 103(a) (2009) (emphasis added). 50. 383 U.S. 1 (1966). 51. Id. at 3. Note that the question of whether Deere’s plow also infringed Graham’s first 1950 patent was not presented. 52. Id. at 17. 53. Id. at 18. 54. For readers unfamiliar with the structure of the U.S. appellate court system, the U.S. Court of Appeals for the Federal Circuit is a specialized appellate court with jurisdiction to hear appeals of patent cases (among other things) from trial courts throughout the United States. It is unlike the other circuit courts in that they are regional and not so tightly limited in the subject matter of the appeals they hear.

264

notes to pages 137–141

55. Greenwood v. Hattori Seiko Co., 900 F.2d 238, 241 (Fed. Cir. 1990) (emphasis added). 56. For a brief but useful account of this debate, see Robert P. Merges et al., Intellectual Property in the New Technological Age 194–96 (3d ed. 2003). 57. The search was conducted in the federal courts library, October 14, 2009, and revealed 782 cases using that expression, including 5 cases decided by the Supreme Court. 58. 15 U.S.C. § 1125(a) (2009). 59. Robert P. Merges et al., Intellectual Property in the New Technological Age 634 (4th ed. 2006). 60. 505 U.S. 763 (1992). 61. Id. at 765. 62. The district court’s opinion is unreported. The questions are reproduced in the Brief for Respondent, id. at 763. 63. Id. at 764 n. 1 (quoting the district court’s instructions to the jury). 64. Id. at 764–65. 65. Id. at 768. 66. 15 U.S.C. § 1052 (2009). 67. 15 U.S.C. § 1064(3) (2009). 68. 537 F.2d 4 (2d Cir. 1976). 69. Well before the Two Pesos litigation, the Court had acknowledged that a “secondary meaning” would generally cause a “merely descriptive” mark to become “distinctive,” and therefore registrable under § 2(f). See Inwood Laboratories, Inc. v. Ives Laboratories, Inc., 456 U.S. 844, 851 (1982). 70. See, e.g., Wal-Mart Stores, Inc. v. Samara Brothers, Inc., 529 U.S. 205 (2000). A Lexis search of federal court cases conducted on October 1, 2009 (“Abercrombie w/25 generic or descriptive or suggestive or fanciful and date aft 1997 and date bef 2008”), yielded 188 hits for the decade between 1998 and 2007. Thus, Judge Friendly’s language is referred to by federal judges in published opinions about 18 times per year more than three decades after he wrote the opinion. 71. Portions of this section are extracted from my article Statutory Inflation and Institutional Choice, 44 Wm. & Mary L. Rev. 2209 (2003). 72. Antitrust: 15 U.S.C. §§ 1–7 (2009) (criminal). Securities: 15 U.S.C. § 15 (2009) (civil); 15 U.S.C. §§ 77e, 77q (2009) (civil); 15 U.S.C. § 78ff(a) (2009) (criminal). Environmental (Clean Water Act): 33 U.S.C. § 1319(c)(2) (2009) (criminal); 33 U.S.C. § 1319(b) (2009) (civil). Copyright Act: 17 U.S.C. § 504 (2009) (civil); 17 U.S.C. § 506 (2009) (criminal). 73. See, e.g., Leocal v. Ashcroft, 543 U.S. 1 (2004); Clark v. Martinez, 543 U.S. 371, 380–81 (2005); United States v. Thompson /Center Arms Co., 504 U.S. 505 (1992); Crandon v. United States, 494 U.S. 152 (1990). For discussion of this interpretive technique, see Kristin Hickman, Of Lenity, Chevron, and KPMG, 26 Va. Tax L. Rev. 905 (2007); Margaret V. Sachs, Harmonizing Civil and Criminal

notes to pages 141–143

265

Enforcement of Federal Regulatory Statutes: The Case of the Securities Exchange Act of 1934, 2001 U. Ill. L. Rev. 1025, 1029. 74. 17 C.F.R. § 240.10b-5 (2009). That rule states: It shall be unlawful for any person, directly or indirectly, by the use of any means or instrumentality of interstate commerce, or of the mails or of any facility of any national securities exchange, (a) To employ any device, scheme, or artifice to defraud, . . . (c) To engage in any act, practice, or course of business which oper­ ates or would operate as a fraud or deceit upon any person, in connec­ tion with the purchase or sale of any security. 75. See, e.g., In re Investors Mgmt. Co., 1971 SEC LEXIS 992 (July 29, 1971); In re Cady, Roberts & Co., 40 S.E.C. 907 (1961). The history discussed in this section is derived in large part from Roberta S. Karmel, Outsider Trading on Confidential Information—a Breach in Search of a Duty, 20 Cardozo L. Rev. 83, 87–91 (1998). 76. 401 F.2d 833, 855 (2d Cir. 1968) (citations and footnotes omitted). 77. 495 F.2d 228, 237 (2d Cir. 1974). Subsequently, the Supreme Court endorsed the “fraud on the market” theory of Rule 10b-5 liability. See Basic, Inc. v. Levinson, 485 U.S. 224 (1988). 78. 445 U.S. 222 (1980). 79. Roberta Karmel has recognized the tension between the broad interpretation of remedial statutes and the rule of lenity in the context of insider-trading cases. See Transcript, Roundtable on Insider Trading: Law, Policy, and Theory after O’Hagan, 20 Cardozo L. Rev. 7, 15 (1998). 80. It is not unusual for civil and criminal cases to arise from the same set of circumstances, which may lead to complicated intragovernmental issues. See Thomas C. Newkirk & Ira L. Brandriss, The Advantages of a Dual System: Parallel Streams of Civil and Criminal Enforcement of U.S. Securities Laws, 2 Int’l. & Comp. L.J. 29 (2000). For a thorough and insightful discussion of the history of insider-trading doctrine, see William K. S. Wang & Marc I. Steinberg, Insider Trading ch. 4 (2d ed. 2005). 81. 143 U.S. 457, 464 (1892). 82. See, e.g., Aharon Barak, Judicial Discretion 61–64 (1989). 83. See, e.g., Massimo La Torre, Enrico Pattaro, & Michele Taruffo, Statutory Interpretation in Italy, in Interpreting Statutes: A Comparative Study 213 (D. Neil MacCormick and Robert S. Summers eds., 1991). 84. Lexis search conducted on July 12, 2008: “(Congress or legislature) w/5 purpose and (date aft 1999 and date bef 2006)” in the federal and state cases library. 85. 553 U.S. 137 (2008). 86. 18 U.S.C. § 924(e)(2)(B)(ii) (2009) (emphasis added). 87. 553 U.S. at 147. 88. See Richard Posner, How Judges Think 230–65 (2008).

266

notes to pages 144–152

89. Jonathan R. Siegel, The Inexorable Radicalization of Textualism, 158 U. Pa. L. Rev. 117, 153 (2009). 90. Walton v. United Consumers Club, Inc., 786 F.2d 303, 310 (7th Cir. 1986). 91. 512 U.S. 218, 231 (1994). 92. 534 U.S. 204 (2002). 93. 29 U.S.C. § 1132(a)(3) (2009). 94. Knudson, 534 U.S. at 220. 95. Id. at 227 (Ginsburg, J., dissenting). 96. United States v. Locke, 471 U.S. 84 (1985); United States v. Granderson, 511 U.S. 39, 41 (1994); Green v. Bock Laundry Machine Co., Inc., 490 U.S. 504 (1989); United States v. X-Citement Video, Inc., 513 U.S. 64 (1994). 97. For interesting discussion of the role of legislative facts in adjudication, see Robert Keeton, Legislative Facts and Similar Things: Deciding Disputed Premise Facts, 73 Minn. L. Rev. 1 (1988). 98. N.J. Admin. Code § 8:24-3.3(d) Supp. 1-3-84 (1992). 99. Quoted in United Press International, January 16, 1992. 100. Off and Runny-ing, Newsday, January 18, 1992. 101. N.J. Admin. Code § 8:24-3.3(d). 102. 908 F.2d 1312 (7th Cir. 1990), aff’d sub nom. Chapman v. United States, 500 U.S. 453 (1991). 103. 21 U.S.C. § 841(b)(1)(A)(v) (2009). 104. 908 F.2d at 1334 (Posner, J., dissenting). 105. Id. at 1335 (Posner, J., dissenting). 106. See, e.g., Lamie v. U.S. Trustee, 540 U.S. 526 (2004). 107. 490 U.S. 504, 528 (1989) (Scalia, J., concurring) (emphasis added). 108. 499 U.S. 83 (1991). 109. See Dan Simon, Freedom and Constraint in Adjudication: A Look through the Lens of Cognitive Psychology, 67 Brook. L. Rev. 1097 (2002). 110. Casey, 499 U.S. at 100–101 (citations omitted). 111. Dworkin, Law’s Empire at 337. 112. William W. Buzbee, The One-Congress Fiction in Statutory Interpretation, 149 U. Pa. L. Rev. 171, 245–46 (2000). 113. Casey, 499 U.S. at 108–9 (Stevens, J., dissenting). 114. Civil Rights Act of 1991, Pub. L. 102-166, 105 Stat. 1071 (codified at 42 U.S.C. § 1988(c) (2000)). 115. Einer Elhauge, Preference-Estimating Statutory Default Rules, 102 Colum. L. Rev. 2027, 2045–46 (2002). These ideas are further developed in Einer Elhauge, Statutory Default Rules: How to Interpret Unclear Statutes (2008). 116. Dworkin, Law’s Empire at 329–37. 117. Casey, 499 U.S. at 90. 118. X-Citement Video, 513 U.S. at 73. 119. 440 U.S. 490 (1979).

notes to pages 153–160

267

120. Id. at 504. 121. Eskridge, Public Values at 1066. 122. Catholic Bishop, 440 U.S. at 512–18 (Brennan, J., dissenting). 123. 498 U.S. 103, 119–20 (1990) (Scalia, J., dissenting). 124. Bob Weigland, Wellsville Woman Given Probation, Fine in Forging of Will, Buffalo News, April 1, 2000, at 5C. 125. Don Wycliff, Executive Decisions in Putting Out a Newspaper, Chicago Tribune, November 30, 2000, at 31. 126. Art Barnum, Doubt Cast on Confession in Girl’s Death, Chicago Tribune, April 26, 2001, at 6. 127. Law & Order, Tampa Tribune, October 26, 2001, at 6 Metro. 128. 283 U.S. 25 (1931). 129. 532 U.S. 105 (2001). 130. 550 U.S. 618 (2007). 131. 421 F.3d 1169 (11th Cir. 2005). 132. 42 U.S.C. § 2000e-2(a)(1) (2009). 133. Ledbetter, 550 U.S. at 649 (Ginsburg, J., dissenting). 134. Lily Ledbetter Fair Pay Act of 2009, Pub. L. 111-2, 123 Stat. 5 (January 29, 2009). See S. 181, 111th Cong. (2009) (passed by Senate, January 22, 2009, by a vote of 61 to 36); H.R. 11, 111th Cong. (2009) (passed by House of Representatives, January 27, 2009, by a vote of 250 to 177). 135. Karl N. Llewellyn, Remarks on the Theory of Appellate Decision and the Rules or Canons about How Statutes Are to Be Construed, 3 Vand. L. Rev. 395, 401–6 (1950). 136. See, e.g., Bradford C. Mank, Textualism’s Selective Canons of Statutory Construction: Reinvigorating Individual Liberties, Legislative Authority, and Deference to Executive Agencies, 86 Ky. L.J. 527 (1998); Max Radin, Statutory Interpretation, 43 Harv. L. Rev. 863 (1930). 137. See Lindquist & Cross, Empirically Testing. 138. James J. Brudney & Corey Ditslear, Canons of Construction and the Elusive Search for Neutral Reasoning, 58 Vand. L. Rev. 1, 55–56 (2005). 139. Id. at 59. 140. See Simon, Freedom and Constraint; Gilbert Harman & Sanjeev Kulkarni, The Problem of Induction, 72 Phil. & Phenomenological Res. 559 (2007). 141. See Dan Simon, A Third View of the Black Box: Cognitive Coherence in Legal Decision Making, 71 U. Chi. L. Rev. 511 (2004).

Chapter Six 1. Perhaps too aggressively. See Peter Strauss, Legislation That Isn’t—Attending to Rulemaking’s “Democracy Deficit,” 97 Calif. L. Rev. (forthcoming 2010).

268

notes to pages 161–165

2. For a judicial perspective on this process, see Aharon Barak, Foreword: A Judge on Judging: The Role of a Supreme Court in a Democracy, 116 Harv. L. Rev. 16 (2002). 3. 467 U.S. 837 (1984). 4. For interesting discussion of the history of the Chevron litigation, see William N. Eskridge Jr. & Lauren E. Baer, The Continuum of Deference: The Supreme Court Treatment of Agency Statutory Interpretations from Chevron to Hamdan, 96 Geo. L.J. 1083 (2008). 5. Chevron, 467 U.S. at 842–43. 6. Later in this chapter I will argue that courts have, in fact, kept some of their authority, notwithstanding the Chevron regime. 7. Cass R. Sunstein, Beyond Marbury: The Executive’s Power to Say What the Law Is, 115 Yale L.J. 2580, 2583 (2006). 8. Dan M. Kahan, Lenity and Federal Law Crimes, 1994 Sup. Ct. Rev. 345 (1994). 9. Adrian Vermeule, Judging under Uncertainty: An Institutional Theory of Legal Interpretation (2006). 10. This point is made by Richard J. Pierce Jr., Administrative Law vol. 1, § 3.4 (4th ed. 2002). 11. Thanks to Claire Kelly for bringing this important observation to my attention. 12. 549 U.S. 561 (2007). 13. 42 U.S.C. § 7411(a)(4) (2009). 14. 42 U.S.C. § 7479(2)(C) (2009). 15. 40 C.F.R. § 60.14(b) (2009). 16. 40 C.F.R. §§ 51.166(b)(21)(ii), 51.166(b)(23)(i) (2009). 17. 549 U.S. 497 (2007). 18. 42 U.S.C. § 7521(a)(1) (2009). 19. Massachusetts, 549 U.S. at 532. 20. Id. at 549–53 (Scalia, J., dissenting). 21. 529 U.S. 120 (2000). 22. 521 U.S. 642 (1997). 23. 15 U.S.C. § 78j (2009). The statute reads: It shall be unlawful for any person, directly or indirectly, by the use of any means or instrumentality of interstate commerce or of the mails, or of any facility of any national securities exchange . . . (b) To use or employ, in connection with the purchase or sale of any security registered on a national securities exchange or any security not so registered, . . . any manipulative or deceptive device or contriv­ ance in contravention of such rules and regulations as the [Securities and Exchange] Commission may prescribe as necessary or appropri­ ate in the public interest or for the protection of investors.

notes to pages 165–169

269

24. See 17 C.F.R. § 240.10b-5 (2009). 25. For discussion of the state of the law at that time, see O’Hagan, 521 U.S. at 649–50. 26. According to the Exchange Act: “It shall be unlawful for any person . . . to engage in any fraudulent, deceptive, or manipulative acts or practices, in connection with any tender offer. . . . The [SEC] shall, for the purposes of this sub-section, by rules and regulations define, and prescribe means reasonably designed to prevent, such acts and practices as are fraudulent, deceptive, or manipulative” (15 U.S.C. § 78n(e) (2009)). 27. 17 C.F.R. § 240.14e-3(a) (2009). 28. United States v. O’Hagan, 92 F.3d 612 (8th Cir. 1996). 29. Neil K. Komesar, Law’s Limits: The Rule of Law and the Supply and Demand of Rights (2004). Along these lines, for an interesting study of how the courts and the relevant agency differ in their interpretations of the same statute, see Margaret Lemos, The Consequences of Congress’s Choice of Delegate: Judicial and Agency Interpretations of Title VII, Vand. L. Rev. (forthcoming), available at www.ssrn .com/abstract=1474873. 30. Richard J. Pierce, The Supreme Court’s New Hypertextualism: An Invitation to Cacophony and Incoherence in the Administrative State, 95 Colum. L. Rev. 749 (1995). 31. 512 U.S. 218 (1994). 32. Thomas J. Miles & Cass R. Sunstein, Do Judges Make Regulatory Policy? An Empirical Investigation of Chevron, 73 U. Chi. L. Rev. 823, 825–26 (2006). 33. Id. at 826–27. 34. For discussion of the psychological process through which judges tend to focus attention only on arguments that support the position they have decided to take, see Dan Simon, A Third View of the Black Box: Cognitive Coherence in Legal Decision Making, 71 U. Chi. L. Rev. 511 (2004). 35. Brown & Williamson Tobacco, 529 U.S. 120 (2000). 36. These positions are reflected in the contemporaneous academic debate as well. Cf. Cass R. Sunstein, Regulations: Is Tobacco a Drug? Administrative Agencies and Common Law Courts, 47 Duke L.J. 1013 (1998); Richard A. Merrill, The FDA May Not Regulate Tobacco Products as “Drugs” or as “Medical Devices,” 47 Duke L.J. 1071 (1998). 37. 21 U.S.C. § 353(g)(1) (2000). 38. See Brown & Williamson Tobacco, 529 U.S. at 170–71 (Breyer, J., dis­ senting). 39. See Curtis A. Bradley & Eric A. Posner, Presidential Signing Statements and Executive Power, 23 Constitutional Commentary 307, 322 (2006). 40. For a brief history, see Phillip J. Cooper, George W. Bush, Edgar Allan Poe, and the Use and Abuse of Presidential Signing Statements, 35 Presidential Stud. Q. 515 (2005).

270

notes to pages 169–174

41. Memorandum from Samuel A. Alito to the Litigation Strategy Working Group (February 5, 1986), at 1, http://www.archives.gov/news /samuel-alito/ accession-060 – 89 – 269/Acc060 – 89 – 269-box6-SG-LSWG-AlitotoLSWGFeb1986.pdf (accessed September 30, 2008). 42. Steven G. Calabresi & Daniel Lev, The Legal Significance of Presidential Signing Statements, 2 The Forum, article 8 (2006), http://www.bepress.com /cgi/ viewcontent.cgi?article=1131&context=forum. 43. Alito Memorandum at 1, 2. 44. Id. at 5. 45. 42 U.S.C. § 2000dd (2009). 46. 41 Weekly Comp. Pres. Doc. 1918 (December 30, 2005), www.whitehouse .gov/news/releases/2005/12/20051230 – 8.html (accessed November 30, 2008). 47. See Josh White, President Relents, Backs Torture Ban, Wash. Post, De­ cember 16, 2005, A1. 48. See Bradley & Posner, Presidential Signing Statements; Calabresi & Lev, Legal Significance. 49. 107 Pub. L. 204, 116 Stat. 745 (2002). 50. 18 U.S.C. § 1514A(a)(1)(B) (2009). 51. Statement on Signing the Sarbanes-Oxley Act of 2002, July 30, 2002, Weekly Comp. Pres. Doc. 1286 (August 2, 2002). 52. 18 U.S.C. § 1001 (1983). 53. 468 U.S. 63 (1983). 54. The statute has been amended since the Yermian decision. I will not discuss those changes in statutory language here. 55. The search was of the Lexis combined federal courts library: “president! w/7 signing statement and date aft 1989 and date bef 2008” (visited February 19, 2009). A sampling of some dates without the word “president” yielded no additional entries, suggesting that this search captured all or almost all the relevant citations. 56. See, e.g., United States v. Perlaza, 439 F.3d 1149, 1163 (9th Cir. 2006) (referring to a signing statement of President Clinton to reinforce congressional committee report in upholding criminal conviction). 57. Kahan, Lenity and Federal Law Crimes. 58. See Lawrence M. Solan, Statutory Inflation and Institutional Choice, 44 Wm. & Mary L. Rev. 2209 (2003). See also Komesar, Law’s Limits. 59. U.S. Government Accountability Office, Department of Labor: Case Studies from Ongoing Work Show Examples in Which Wage and Hour Division Did Not Adequately Pursue Labor Violations, testimony before the Committee on Education and Labor, House of Representatives, July 15, 2008, http://www.gao .gov/new.items/d08973t.pdf (accessed October 19, 2008). 60. Others have noted this relationship. See Michael Edmund O’Neill, When Prosecutors Don’t: Trends in Federal Prosecutorial Declinations, 79 Notre Dame L. Rev. 221, 232 (2003) (“Just as judges often refine legislative pronouncements

notes to pages 174–178

271

through statutory interpretation, so do prosecutors when they exercise prosecutorial discretion in deciding whether or not to pursue a particular case. The inevitability of prosecutorial discretion is thus taken as axiomatic.”). 61. Nicholas Quinn Rosenkranz, Federal Rules of Statutory Interpretation, 115 Harv. L. Rev. 2085 (2002). 62. State v. Courchesne, 816 A.2d 562, 567–68 (Conn. 2003). 63. Conn. Gen. Stat. § 53a-54b(7) (2009). 64. Conn. Gen. Stat. § 53a-46a(i)(4) (2009). 65. Lynne Tuohy, Killer’s Jury Decides on Death; Convicted Killer Stabbed Pregnant Woman in 1998, Hartford Courant, December 18, 2003, A1. 66. Conn. Gen. Stat. 53a-54b(5) (2009). 67. Courchesne, 816 A.2d at 572. 68. See Lawrence M. Solan, Law, Language, and Lenity, 40 Wm. & Mary L. Rev. 57, 85 (1998). 69. Conn. Gen. Stat. § 1-2z (2009). 70. Genesky v. East Lyme, 881 A.2d 114, 132 n. 5 (Conn. 2005) (Borden, J., concurring). 71. Rivers v. City of New Britain, 950 A.2d 1247, 1259 (Conn. 2007). 72. Id. at 1263–64 (Schaller, J., dissenting). 73. Clark v. Commissioner of Correction, 917 A.2d 1 (Conn. 2007). 74. See Ames v. Commissioner of Motor Vehicles, 839 A.2d 1250 (Conn. 2004), relying on purpose and context to construe suretyship statute. 75. Hummel v. Marten Transport, Ltd., 923 A.2d 657 (Conn. 2007). 76. Abbe Gluck, Statutory Interpretation beyond the Supreme Court: State Courts, Methodological Consistency, and the Emerging “Modified Textualism,” Yale L.J. (forthcoming). 77. To take just a few examples, see, e.g., Charles Gardner Geyh, When Courts and Congress Collide: The Struggle for Control of America’s Judicial System (2006); Rick Santorum, It Takes a Family: Conservatism and the Common Good (2005); Mark Tushnet, A Court Divided: The Rehnquist Court and the Future of Constitutional Law (2005); Vincent Bugliosi, The Betrayal of America: How the Supreme Court Undermined the Constitution and Chose Our President (2001); Malcolm M. Feeley & Edward L. Rubin, Judicial Policy Making and the Modern State: How the Courts Reformed America’s Prisons (2000); Daniel Lazare, The Velvet Coup: The Constitution, the Supreme Court, and the Decline of American Democracy (2001); Bruce G. Peabody, Congress, the Court, and the “Service Constitution”: Article III Jurisdiction Controls as a Case Study of the Separation of Powers, 2006 Mich. St. L. Rev. 269. 78. John T. Noonan Jr., Narrowing the Nation’s Power: The Supreme Court Sides with the States (2002). See, e.g., United States v. Morrison, 529 U.S. 598 (2000) (holding that the regulation of gender-motivated violent crime is not properly under the scope of the commerce clause); United States v. Lopez, 514

272

notes to pages 178–181

U.S. 549 (1995) (holding that federal regulation of the possession of guns in school zones is not properly under the scope of the commerce clause); Cass R. Sunstein, Radicals in Robes: Why Extreme Right-Wing Courts Are Wrong for America (2005). 79. Reactions to the Supreme Court’s famous abortion decision in Roe v. Wade, 410 U.S. 113 (1973), is the paradigm example. See Christopher Wolfe, The Rise of Modern Judicial Review: From Constitutional Interpretation to Judge-Made Law 307 (1986) (describing the Court’s decision in Roe v. Wade as “the most raw exercise of judicial power since Dred Scott. . . . The Court imposed its own approbation of abortion on the nation at large.”); Gregory C. Sisk, The Willful Judging of Harry Blackmun, 70 Mo. L. Rev. 1049, 1055 (2005) (“Justice Blackmun offered extraneous non-legal observations of questionable accuracy to justify judicial activism”). See also testimony of Robert Bork, Hearings on S. 158 before the Subcommittee on Separation of Powers of the Committee on the Judiciary, 97th Cong., 1st sess., 310–11 (1981) (describing Roe as “an unconstitutional decision, a serious and wholly unjustified usurpation of state legislative authority”). 80. U.S. Const. art. III, § 1. 81. U.S. Const. art. I, § 1. 82. 5 U.S. 137 (1803). 83. 76 ER 637 (1584). 84. William N. Eskridge Jr., All about Words: Early Understandings of the “Judicial Power” in Statutory Interpretation, 1776–1806, 101 Colum. L. Rev. 990 (2001). 85. John F. Manning, Textualism and the Equity of the Statute, 101 Colum. L. Rev. 1, 57 (2001). 86. Italian Civil Code, ch. II, provision 12. 87. See N.D. Cent. Code § 1-02-39 (2009); Iowa Code § 4.6 (2009). 88. See, e.g., Colo. Rev. Stat. § 2-4-203 (2009); Ohio Rev. Code Ann. § 1.49 (2009); 1 Pa. Cons. Stat. § 1921 (2009); Tex. Gov’t Code Ann. § 311.023 (2009). 89. Ill. Rev. Stat. 1969, ch. 38, ¶ 60-11; current version at 740 Ill. Comp. Stat. Ann. 10/11 (2009) (“When the wording of this Act is identical or similar to that of a federal antitrust law, the courts of this State shall use the construction of the federal law by the federal courts as a guide in construing this Act. However, this Act shall not be construed to restrict the exercise by units of local government or school districts of powers granted, either expressly or by necessary implication, by Illinois statute or the Illinois Constitution.”). 90. People v. Crawford Distributing Co., 291 N.E.2d 648, 652 (Ill. 1972). My thanks to Abbe Gluck for bringing this case to my attention. 91. 75 Del. Laws, ch. 1, § 1 (H.B. 31, section 1 § 5402 (Del. 2004)). 92. 872 A.2d 539, 549 (Del. 2005). See United States v. Klein, 80 U.S. (13 Wall.) 128 (1872). 93. Evans, 872 A.2d at 550.

notes to pages 182–189

273

94. 1 U.S.C. § 1 (2009). 95. Linda D. Jellum, “Which Is to Be Master,” the Judiciary or the Legislature? When Statutory Directives Violate Separation of Powers, 56 U.C.L.A. L. Rev. 837 (2009). 96. 522 U.S. 398 (1998). 97. 18 U.S.C. § 1001 (1988). 98. 1 U.S.C. § 1 (2009). 99. Fed. R. Evid. 901. 100. Fed. R. Evid. 601. 101. Fed. R. Evid. 802. 102. Fed. R. Evid. 403. 103. Fed. R. Evid. 404(a). There are exceptions, especially if the defendant first raises issues of character and the prosecution wishes to rebut this evidence. 104. See Fed. R. Evid. 404(b), 609. 105. See, e.g., N.Y. C.P.L.R. § 7501 (2009) (“In determining any matter arising under this article, the court shall not consider whether the claim with respect to which arbitration is sought is tenable, or otherwise pass upon the merits of the dispute”). 106. Cal. Civ. Proc. Code § 1775.5 (Deering 2009). 107. 750 Ill. Comp. Stat. Ann. 5/602(b) (2009). 108. 19 U.S.C. § 1641(e)(2) (2009). 109. 42 U.S.C. § 3545 (2009) (relating to penalties imposed by the Department of Housing and Urban Development); 12 U.S.C. § 1735f-14(d)(2) (2009) (regulating participants in FHA loan program). 110. 47 U.S.C. § 532(d) (2009). 111. Del. Code Ann. tit. 13, § 709A (2009); see also N.Y. Dom. Rel. Law § 76-g (2009). 112. Ohio Rev. Code Ann. § 3107.14 (2009). 113. Jellum, Which Is to Be Master at 848. 114. Or. Rev. Stat. § 174.020(3) (2009). 115. See Gluck, Statutory Interpretation, discussing the statute in light of Portland General Electric v. Bureau of Labor and Industries, 859 P.2d 1143 (Or. 1993). 116. 42 U.S.C. § 5318(h)(4) (2009). 117. Construction of [Civil Rights] Act of Nov. 21, 1991, in relation to Wards Cove—Business necessity/cumulation/alternative business practice, Pub. L. 102166, § 105(b), 105 Stat. 1071 (2003). 118. 10 U.S.C. § 229(c)(2) (2009). 119. Wash. Rev. Code § 43.88.145(2) (2009). 120. See, e.g., Flores-Figueroa v. United States, 129 S. Ct. 1886, 1894 (2009) (Scalia, J., concurring in part and concurring in the judgment); Zedner v. United States, 547 U.S. 489, 509–10 (2006) (Scalia, J., concurring in part and concurring

274

notes to pages 189–192

in the judgment); United States v. Estate of Romani, 523 U.S. 517, 535–36 (1998) (Scalia, J., concurring in part and concurring in the judgment); Posters ‘N’ Things, Ltd. v. United States, 511 U.S. 513, 530 (1994) (Scalia, J., concurring in the judgment); Thunder Basin Coal Co. v. Reich, 510 U.S. 200, 219 (1994) (Scalia, J., concurring in part and concurring in judgment); Conroy v. Aniskoff, 507 U.S. 511, 518–28 (1993) (Scalia, J., concurring in judgment); United States v. Thompson /Center Arms Co., 504 U.S. 505, 521 (1992) (Scalia, J., concurring in judgment); Wisconsin Public Intervenor v. Mortier, 501 U.S. 597, 617 (1991) (Scalia, J., concurring in judgment); Blanchard v. Bergeron, 489 U.S. 87, 98–100 (1989) (Scalia, J., concurring in part and concurring in judgment); United States v. Taylor, 487 U.S. 326, 344 (1988) (Scalia, J., concurring in part). 121. This section of the chapter is adapted from my article Law, Language, and Lenity, 40 Wm. & Mary L. Rev. 57, 85 (1998). 122. N.Y. Penal Law § 5.00 (McKinney 2009). The statute was first enacted as section 11 of the Penal Code of 1881 and was first proposed in 1864 as part of the Field Code. 123. See Livingston Hall, Strict or Liberal Construction of Penal Statutes, 48 Harv. L. Rev. 748 (1935). 124. 32 N.E. 978, 979 (N.Y. 1893). 125. 217 N.E.2d 660 (N.Y. 1966) (citing People v. Wood, 167 N.E.2d 736, 738 (N.Y. 1960)). 126. People v. Santiago, 506 N.Y.S.2d 136, 138 (N.Y. Sup. Ct. 1986). 127. People v. Familo, 516 N.Y.S.2d 416, 419 (Oswego City Ct. 1987). See also People v. Hill, 830 N.Y.S.2d 33 (App. Div., 1st Dep’t 2007) (applying rule of lenity); People v. Scott, 258 N.E.2d 206, 210 (N.Y. 1970) (ignoring Penal Law 5.00 in favor of the principle that “[a] penal enactment must not only be strictly construed, especially where, as here, the act is a malum prohibitum and not a malum in se, but it must be reasonable and pellucid as well” [citation omitted]). 128. See Cal. Penal Code § 4 (West 2009). The statute reads: “The rule of the common law, that penal statutes are to be strictly construed, has no application to this Code. All its provisions are to be construed according to the fair import of their terms, with a view to effect its objects and to promote justice” (id.). For a discussion of the contemporaneous enactment of the Civil Code and the jurisprudential ramifications of the codification of interpretation, see William S. Blatt, The History of Statutory Interpretation: A Study in Form and Substance, 6 Cardozo L. Rev. 799, 819 (1985). 129. See Hall, Strict or Liberal Construction. 130. In re Rosenheim, 23 P. 372, 373 (Cal. 1890). 131. People v. Showalter, 14 P.2d 1034, 1035 (Cal. Dist. Ct. App. 1932). 132. 470 P.2d 617, 623 (Cal. 1970). 133. Id. at 624.

notes to pages 192–195

275

134. Id. at 625–26. 135. See John Calvin Jeffries Jr., Legality, Vagueness, and the Construction of Penal Statutes, 71 Va. L. Rev. 189 (1985). 136. See, e.g., People v. Cruz, 919 P.2d 731 (Cal. 1996). The court interpreted “inhabited dwelling house” to include boats for purposes of a sentence enhancement statute in order to further legislative intent. Id. at 733. The burglary statute in effect at that time listed boats, but the statute in question did not. See id. at 734–35. The case resembles Wiltberger in many respects, but with the opposite result. 137. See People v. Salas, 38 Cal. Rptr. 3d 624 (2006) (acknowledging and discussing rule of lenity as a principle of California law); People v. Phelps, 48 Cal. Rptr. 2d 855, 858 (Cal. Ct. App. 1996) (acknowledging a narrow version of lenity as California policy but holding it inapplicable because the restitution statute in issue was clear); People v. Ramon A., 47 Cal. Rptr. 2d 59, 62 (Cal. Ct. App. 1995) (analyzing case under narrow rule of lenity, then holding that statute was clear enough; rejecting argument of youth who stole a car and allowed a passenger carrying a gun to ride with him that he was not guilty of permitting firearm into car because he did not know whether gun was loaded). 138. See, e.g., People v. Rosalio S., 41 Cal. Rptr. 2d 534, 537–38 (Cal. Ct. App. 1995) (holding that statute that made reference to the length of a knife’s blade should be construed to require measurement of only the sharpened portion of the instrument); People v. Bartlett, 276 Cal. Rptr. 460, 465 (Cal. Ct. App. 1990) (holding that defendant convicted of “selling or transporting cocaine” should not have probation revoked, because revocation statute applies to those convicted of “selling” controlled substance; this is a prototypical case of lenity applied to an ambiguous statute). 139. 501 U.S. 452 (1991). 140. 29 U.S.C. §§ 621–34 (2009). 141. 29 U.S.C. § 630(f) (2009). 142. Gregory, 501 U.S. at 461 (internal citations omitted). 143. See William Eskridge Jr. & Philip Frickey, Quasi-constitutional Law: Clear Statement Rules as Constitutional Lawmaking, 45 Vand. L. Rev. 593 (1992). 144. 548 U.S. 557 (2006). 145. Cf. Cass Sunstein, Clear Statement Principles and National Security: Hamdan and Beyond, 2006 Sup. Ct. Rev. 1; Curtis A. Bradley & John L. Goldsmith, The War on Terrorism: International Law, Clear Statement Requirements and Constitutional Design, 118 Harv. L. Rev. 2683 (2005). 146. Thanks to Heidi Kitrosser for pointing this out. 147. 533 U.S. 218, 226–27 (2001). 148. 323 U.S. 134 (1944). 149. Peter L. Strauss, The Rulemaking Continuum, 41 Duke L.J. 1463 (1992). 150. Eskridge & Baer, Continuum of Deference at 1099.

276

notes to pages 195–199

151. Id. at 1100. 152. Kristin E. Hickman & Matthew D. Krueger, In Search of the Modern Skidmore Standard, 107 Colum. L. Rev. 1235 (2007).

Chapter Seven 1. This chapter is adapted from my article Jurors as Statutory Interpreters, 78 Chi.-Kent L. Rev. 1281 (2003). 2. At least under federal law, this has been true since the early nineteenth century. See United States v. Hudson & Goodwin, 11 U.S. (7 Cranch) 32 (1812). 3. See Matthew P. Harrington, The Law-Finding Function of the American Jury, 1999 Wis. L. Rev. 377, 377–79; Stephen C. Yeazell, The New Jury and the Ancient Jury Conflict, 1990 U. Chi. Legal F. 87. 4. For the history of the development of this doctrine in England, see Thomas Andrew Green, Verdict according to Conscience: Perspectives on the English Criminal Trial Jury, 1200–1800 153–99 (1985). For an overview of the history, see Irwin A. Horowitz et al., Jury Nullification: Legal and Psychological Perspectives, 66 Brook. L. Rev. 1207 (2001). 5. Harrington, Law-Finding Function at 393. 6. Declaration of Independence ¶ 20 (U.S. 1776). 7. The Federalist no. 83 (Alexander Hamilton) (Clinton Rossiter ed., 1999) at 467. 8. See Harrington, Law-Finding Function at 398–400. 9. This transformation was gradual. See Renee B. Lettow, New Trial for Verdict against Law: Judge Jury Relations in Early Nineteenth Century America, 71 Notre Dame L. Rev. 505 (1996). 10. Sparf & Hansen v. United States, 156 U.S. 51, 102 (1895). 11. See Darryl K. Brown, Plain Meaning, Practical Reason, and Culpability: Toward a Theory of Jury Interpretation of Criminal Statutes, 96 Mich. L. Rev. 1199 (1998). 12. Nancy S. Marder, The Myth of the Nullifying Jury, 93 Nw. U. L. Rev. 877, 879–80 (1999). 13. Yeazell, New Jury at 91. 14. Vicki L. Smith, Prototypes in the Courtroom: Lay Representations of Legal Concepts, 61 J. Personality & Soc. Psychol. 857 (1991). 15. For discussion of how jurors use prototypes, see Dan M. Kahan, Lay Perceptions of Justice vs. Criminal Law Doctrine: A False Dichotomy? 28 Hofstra L. Rev. 793 (2000). For an interesting discussion of prototypes in the context of jury nullification, see Katherine T. Kixmiller, Jury Nullification: A Morally Appropriate Response or a License for Bias? unpublished undergraduate thesis, Princeton University (2001) (on file with the author).

notes to pages 199–206

277

16. King v. Moore, 312 F.3d 1365, 1366 (11th Cir. 2002); see also Lozen International, L.L.C. v. Sea-Land Service, Inc., 285 F.3d 808, 813 (9th Cir. 2002); Fritz v. Principi, 264 F.3d 1372, 1374 (Fed. Cir. 2001). 17. The less legalistic, more expansive alternative is more consistent with the traditional role of the jury prior to the twentieth century. See Nancy S. Marder, Juries and Technology: Equipping Jurors for the Twenty-First Century, 66 Brook. L. Rev. 1257, 1261 (2001). 18. 508 U.S. 223 (1993). 19. United States v. Smith, 957 F.2d 835, 836 n. 4 (11th Cir. 1992). 20. 524 U.S. 184 (1998). 21. 18 U.S.C. § 924(a)(1)(D) (2009). 22. Bryan, 524 U.S. at 190 n. 10. 23. Id. at 190. 24. For insightful discussion of these terms as used in statutes, see Joseph E. Kennedy, Making the Crime Fit the Punishment, 51 Emory L.J. 753 (2002). 25. N.J. Stat. Ann. § 2C:12-10(b) (2009). For further discussion of this statute, see Lawrence M. Solan, Finding Ordinary Meaning in the Dictionary, in Language and Law: Proceedings of a Conference 255 (M. Robinson ed., 2003). 26. Smith, Prototypes in the Courtroom. 27. Id. at 864; see also Brown, Plain Meaning; Kahan, Lay Perceptions (for further discussion of the ramifications). 28. See chapter 3 for discussion of Church of the Holy Trinity v. United States, 143 U.S. 457 (1892). 29. 74 U.S. (7 Wall.) 482, 486–87 (1868). 30. Brown, Plain Meaning at 1264–65. 31. Marder, Nullifying Jury at 879–80. 32. Linda Coleman & Paul Kay, Prototype Semantics: The English Verb Lie, 57 Language 26, 31 (1981). 33. Lawrence M. Solan, Pernicious Ambiguity in Contracts and Statutes, 79 Chi.-Kent L. Rev. 859 (2004). 34. Ward Farnsworth, Dustin F. Guzior, & Anup Malani, Ambiguity into Ambiguity: An Empirical Inquiry into Legal Interpretation, 93 Minn. L. Rev. 1535 (2009). 35. Lawrence M. Solan, Terri Rosenblatt, & Daniel Osherson, False Consensus Bias in Contract Interpretation, 108 Colum. L. Rev. 1268 (2008). 36. Peter Meijas Tiersma, Reforming the Language of Jury Instructions, 22 Hofstra L. Rev. 37 (1993). 37. See Kahan, Lay Perceptions; Smith, Prototypes in the Courtroom (showing that it is sometimes very difficult to dislodge people’s prior understanding of a crime and to replace it with the statutory definition contained in the jury instructions). 38. 18 U.S.C. § 1341 (2009). This is an abbreviated version of the statute, for purposes of discussion.

278

notes to pages 206–215

39. For discussion of how syntax and word choice affect the comprehensibility of jury instructions, see Robert P. Charrow & Veda R. Charrow, Making Legal Language Understandable: A Psycholinguistic Study of Jury Instructions, 79 Colum. L. Rev. 1306 (1979); see also Amiram Elwork et al., Making Jury Instructions Understandable (1982). 40. See, e.g., United States v. Castor, 558 F.2d 379, 385 (7th Cir. 1977). 41. 489 U.S. 705, 707 (1989). 42. Id. at 711–12 (emphasis added). 43. Id. at 723 (Scalia, J., dissenting). 44. See Dan M. Kahan, Lenity and Federal Law Crimes, 1994 Sup. Ct. Rev. 345, 373–78 (arguing that the breadth of some statutes, including the mail fraud statute, leaves so much judicial discretion as to in effect create common-law crimes). 45. See, e.g., United States v. Voss, 787 F.2d 393, 401 (8th Cir. 1986). 46. 948 F.2d 363, 368–69 (7th Cir. 1991). 47. Federal Judicial Center, Pattern Criminal Jury Instructions § 85 (1987). 48. Leonard B. Sand et al., Modern Federal Jury Instructions vol. 2, §§ 44-4 to 44-6 (2009). 49. 18 U.S.C. § 1341 (2009). 50. Federal Judicial Center, Jury Instructions § 85. 51. 7 U.S.C. § 6b(a)(1) (2009). 52. For a recent example, see United States v. Leahy, 464 F.3d 773, 781 (7th Cir. 2006) (“We now shift our attention to Duff’s endeavors to cheat his insurers, a fraud of even longer duration”). 53. Sand et al., Modern Federal Jury Instructions vol. 2, §§ 44-4. 54. Id. (emphasis added). 55. See Harrington, Law-Finding Function at 380. 56. This position is consistent with the public-choice model of legislation. See generally Daniel A. Farber & Philip P. Frickey, Law and Public Choice: A Critical Introduction (1991). 57. Consider, e.g., the federal Sentencing Guidelines: U.S. Sentencing Guidelines Manual ch. 1, pt. A (2007). 58. 18 U.S.C. § 1956(a)(1) (2009). 59. 18 U.S.C. § 1956(c) (2009). 60. 18 U.S.C. § 1956(c)(5) (2009). 61. Sand et al., Modern Federal Jury Instructions vol. 3, § 50A-3 (2009). 62. Id. 63. 18 U.S.C. § 1956(c)(4) (2009). 64. See Roy Sorensen, Vagueness and the Desiderata for Definition, in Definitions and Definability: Philosophical Perspectives 71, 71–72 (James H. Fetzer et al. eds., 1991). 65. Sand et al., Modern Federal Jury Instructions vol. 3, § 50A-3. 66. 446 U.S. 420, 422 (1980) (quoting Ga. Code Ann. § 27-2534.1(b)(7) (1978)).

notes to pages 215–219

279

67. 446 U.S. at 428–29. 68. Cf. Weeks v. Angelone, 528 U.S. 225, 227 (2000) (even when jurors claim not to have understood a term in an instruction, courts are under no obligation to offer a meaningful explanation), with Kelly v. South Carolina, 534 U.S. 246 (2002) (most recent in a series of cases holding that jurors must be informed when life without parole is the only alternative to the death penalty and government puts defendant’s future dangerousness into issue). See generally Peter M. Tiersma, Dictionaries and Death: Do Capital Jurors Understand Mitigation? 1995 Utah L. Rev. 1. 69. 643 A.2d 996, 997 (N.J. 1994). 70. Id. at 998 (quoting N.J. Stat. Ann. § 2C:35-3 (West 1987)). 71. State v. Afanador, 631 A.2d 946, 950 (N.J. 1993). 72. Alexander, 643 A.2d at 1000. 73. Id. (citations omitted). 74. United States v. Paul, 217 F.3d 989, 997 (8th Cir. 2000); see also Harjo v. Gibson, no. 99-7041, 2000 U.S. App. Lexis 14381, at *11 (10th Cir. June 21, 2000) (“Instructing the jury according to the statutory language of the aggravator, as the trial court did, meets constitutional standards”); Farrington v. Senkowski, 214 F.3d 237, 244 (2d Cir. 2000) (“These instructions on larceny mirror the statutory language, and were therefore not ‘clearly constitutionally deficient’ ” [citations omitted]). 75. Gacy v. Welborn, 994 F.2d 305 (7th Cir. 1993). 76. See Free v. Peters, 12 F.3d 700 (7th Cir. 1993). 77. Gacy, 994 F.2d at 312. 78. Shari Seidman Diamond & Judith N. Levi, Improving Decisions on Death by Revising and Testing Jury Instructions, 79 Judicature 224 (1996). For discussion of the history of these cases, including detailed analysis of the language of the instructions, see Peter M. Tiersma, The Rocky Road to Legal Reform: Improving the Language of Jury Instructions, 66 Brook. L. Rev. 1081 (2001). 79. Minn. Stat. Ann. § 181.75(1) (West 2009). 80. Kamrath v. Suburban National Bank, 363 N.W.2d 108, 111 (Minn. Ct. App. 1985). 81. State v. Apao, 586 P.2d 250, 263 (Haw. 1978); see also State v. Barbour, 258 S.E.2d 475, 481 (N.C. Ct. App. 1979). 82. See, e.g., State v. Booth, 276 N.W.2d 673, 677–78 (Neb. 1979). 83. United States v. Dozal, 173 F.3d 787, 796 (10th Cir. 1999) (quoting United States v. McIntosh, 124 F.3d 1330, 1337 (10th Cir. 1997)) (internal quotations omitted). 84. United States v. Smith, no. 93–5953, 1994 U.S. App. Lexis 991417598, at *13 (6th Cir. April 28, 1994); see also Riverwoods Chappaqua Corp. v. Marine Midland Bank, N.A., 30 F.3d 339, 346 (2d Cir. 1994) (“The district court is not obligated to charge the jury using the exact words proposed by a party. . . . While a more specific instruction might have been helpful, there is no basis for concluding that the jury was given a misleading or inaccurate impression of the law.”).

280

notes to pages 219–228

85. Weeks v. Angelone, 528 U.S. 225, 234 (2000) (citations omitted). For discussion of the Court’s dismissive attitude toward juror understanding, see Tiersma, Rocky Road to Legal Reform at 1085–89. 86. Paul H. Robinson & John M. Darley, Justice, Liability, & Blame: Community Views and the Criminal Law 2 (1995). 87. Id. at 6 (citing Tom Tyler, Why People Obey the Law (1990)) (citation omitted). 88. Cal. Penal Code § 667(e)(2)(A) (2009). 89. Ewing v. California, 538 U.S. 11, 17–18 (2003). 90. Lockyer v. Andrade, 538 U.S. 63, 66–68 (2003). 91. Ewing, 538 U.S. at 25 (Kennedy, J., concurring). 92. Id. at 32–62 (Stevens & Breyer, JJ., dissenting). 93. See Marder, Nullifying Jury at 895–99; Nancy S. Marder, Juries, Drug Laws, and Sentencing, 6 J. Gender, Race, & Justice 337, 341–52 (2002). 94. Marder, Nullifying Jury at 897. Kixmiller, Jury Nullification, also makes this point. 95. See Smith, Prototypes in the Courtroom. 96. For examples that deal largely with linguistic issues in such cases, see chapters 2 and 6. 97. Darryl K. Brown, Regulating Decision Effects of Legally Sufficient Jury Instructions, 73 S. Cal. L. Rev. 1105, 1124 (2000). 98. Cal. Penal Code §§ 1025, 1044 (2009); see People v. Cline, 71 Cal. Rptr. 2d 41 (Cal. App. 1998) (holding it proper for court to grant prosecutor’s motion for bifurcation in three-strikes case). 99. See People v. Tindall, 102 Cal. Rptr. 2d 533, 536 (2000) (holding that same jury must decide both issues). 100. See People v. Nichols, 62 Cal. Rptr. 2d 433, 434 (Cal. App. 1997); Marder, Juries, Drug Laws, and Sentencing at 345. 101. See Marder, Juries, Drug Laws, and Sentencing at 346–47.

Chapter Eight 1. 532 U.S. 105 (2001). 2. See Antonin Scalia, A Matter of Interpretation 32 (1997). 3. See Dan Simon, A Psychological Model of Judicial Decision Making, 30 Rutgers L.J. 1, 111 (1998). 4. 20 U.S.C. § 3423a(b) (2009). 5. 16 U.S.C. § 41011(b) (2009). 6. 515 U.S. 687 (1995). 7. 16 U.S.C. § 1532(19) (1994). 8. 50 C.F.R. § 17.3 (1994).

notes to pages 228–229

281

9. 515 U.S. at 698 (quoting 16 U.S.C. § 1531(b)). 10. See Cipollone v. Liggett Group, 505 U.S. 504, 511–21 (1991) (using statutory statement of purpose to draw inferences of legislative intent concerning preemption of state law by federal statute). 11. 514 U.S. 549 (1995). 12. West Virginia University Hospital v. Casey, 499 U.S. 83, 98 (1991). 13. For recent examples, see District of Columbia v. Heller, 128 S. Ct. 2783, 2819 (2008); and Gonzalez v. Carhart, 550 U.S. 124, 157–58, 165 (2007). 14. See Adrian Vermeule, Judging under Uncertainty: An Institutional Theory of Legal Interpretation (2006).

Cases Abercrombie & Fitch Co. v. Hunting World, Inc., 139 Babbitt v. Sweet Home Chapter of Communities for a Great Oregon, 228 Bailey v. United States, 70–71, 74, 76 Bank One Chicago, N.A. v. Midwest Bank & Trust Co., 95 Begay v. United States, 143–44 Bowles v. Russell, 8–9 Brogan v. United States, 182–83 Bryan v. United States, 45, 75, 201 Caminetti v. United States, 59, 113 Chevron USA, Inc. v. Natural Resources Defense Council, Inc., 161–69, 194, 224, 228, 229 Chiarella v. United States, 142 Chisom v. Roemer, 67–68, 109–10, 117 Church of the Holy Trinity v. United States, 11, 54, 55, 57, 59, 66, 75, 116, 142, 202 Circuit City Stores v. Adams, 35–37, 47, 155, 157, 225 Commonwealth v. Balthazar, 132 Conroy v. Aniskoff, 86, 114–15 Dolan v. United States Postal Service, 55–56 Environmental Defense v. Duke Energy Corp., 162–63 Evans v. State, 181 Fischer v. United States, 44 Flood v. Kuhn, 129–30 Food & Drug Administration v. Brown & Williamson Tobacco Corp., 164, 168 Gay & Lesbian Advocates & Defenders v. Attorney General, 132 Godfrey v. Georgia, 215

Graham v. John Deere Co., 136–37 Great-West Life Assurance & Annuity Co. v. Knudson, 144–45 Green v. Bock Laundry Machine Co., Inc., 107, 146, 149 Greenwood v. Hattori Seiko Co., 137 Gregory v. Ashcroft, 193 Hamdan v. Rumsfeld, 193–94 Keeler v. Superior Court, 191–92 Ledbetter v. Goodyear Tire and Rubber Co., 155–56 Liparota v. United States, 40, 46 Marbury v. Madison, 178, 181 Massachusetts v. Environmental Protection Agency, 163–64 McBoyle v. United States, 55, 122, 128–29, 154 MCI Telecommunications Corp. v. AT&T, 69, 144, 166 Moskal v. United States, 153–55 Muscarello v. United States, 37, 72, 75–78 National Labor Relations Board v. Catholic Bishop of Chicago, 152–53 Nix v. Hedden, 78–79 People v. Phyfe, 190 People v. Sansanese, 190 Ratzlaf v. United States, 47 Roell v. Winthrow, 100 Schmuck v. United States, 133–34, 207–8, 221 Securities and Exchange Commission v. Texas Gulf Sulphur Co., 141 Shapiro v. Merrill Lynch, Inc., 141 Skidmore v. Swift & Co., 194–95 Small v. United States, 59–60, 69–70

284 Smith v. United States, 57–59, 69–71, 98, 200–2 State v. Alexander, 216–17 Two Pesos, Inc. v. Taco Cabana, Inc., 138–40 United States v. Aguilar, 45 United States v. E. I. Dupont De Nemours & Co., 68 United States v. Granderson, 46–47, 104–6, 146 United States v. Jefferson, 117–18 United States v. Kirby, 61–62, 203 United States v. Lennartz, 208 United States v. Locke, 8–9, 108–9, 146

cases United States v. Lopez, 228 United States v. Marshall, 148–49 United States v. Mead Corp., 194–95 United States v. O’Hagan, 165–66 United States v. Wiltberger, 41–43, 122, 192 United States v. Winn, 43–44 United States v. X-Citement Video, Inc., 106–7, 146, 152 United States v. Yermian, 172 Watson v. United States, 71 West Virginia University Hospitals v. Casey, 150–52 White v. Regester, 67

Index Abelson, Robert, 93 absurd result rule, 4, 47, 52, 61, 84, 104–7, 146, 148, 157, 177, 198, 203, 222; and jurors, 203–4 Aleinikoff, Alexander, 92 Alito, Samuel ( justice), 169–70 ambiguity, 22–23, 32–41, 46–49, 180, 190–91, 204, 206, 225; of reference 21–22, 104–5; pernicious, 204; syntactic (see syntax and meaning) Aristotle, 10–11 Armstrong, Sharon, 64 Ayres, Ian, 25 Baer, Lauren E., 194 Blackstone, William, 42 Bloom, Paul, 99 Bratman, Michael, 90–91 Brewer, David ( justice), 54–59 Breyer, Stephen ( justice), 36, 37, 60, 72, 78, 84, 88, 95–96, 169 Brown, Daryl, 197–98 Brudney, James, 157 Burger, Warren (chief justice), 129–30 Bush, George W., 169–71 Buzbee, William, 76, 151 Calabresi, Guido, 2, 131 Calabresi, Steven, 169 Campbell, Donald, 92 canons of construction, 3, 36, 61, 68, 102, 104, 106, 110, 114, 152, 157, 178, 192, 198, 203, 222; avoidance of constitutional questions, 46, 152–53; ejusdem ge-

neris 36, 55; ordinary meaning rule (see ordinary meaning); plain language rule (see plain meaning); remedial statutes construed broadly 68, 140–41, 166; rule against construing language as surplusage 104; rule of lenity (see lenity, rule of) Cardozo, Benjamin ( justice), 16 Chevron doctrine, 161–69, 189, 194, 224, 228–29 Chomsky, Carol, 116 Chomsky, Noam, 39, 111 clear statement rules, 68, 189, 193–94 Clinton, Bill, 65, 162, 164, 167–68 coherence, 11, 14, 35, 52, 103, 149–52, 224, 226 Coleman, Linda, 204 common law, 1–3, 51, 130–40, 158, 191, 196, 208 common law crimes, 196, 208 common law statutes, 135–40 Constitution of the U.S., Art I, sec. V, 62, 96; Art. I, sec. VII, 193; judicial power, 178–81, 186; legislative power 172, 178 constitutional values, 14, 51, 106, 152–53, 224 corpus linguistics, 78 Cross, Frank, 125–27, 157 Cunningham, Clark, 71, 76–77 Darley, John, 219–20 definitional meaning, 19, 23, 38, 53–70, 74, 79–81, 98, 116, 134 Diamond, Shari Seidman, 218 Dickerson, Reed, 97

286 dictionaries, 18, 40, 43, 52–53, 56–60, 63, 65–76, 79–80, 151, 153, 177, 182–83, 205, dictionary acts, 151, 182–83 Ditslear, Corey, 157 Dworkin, Ronald, 92, 125–26, 131, 150–51 dynamic statutory interpretation, 2, 130–32 Easterbrook, Frank ( judge), 88, 111, 114, 144, 148, 217–18 ejusdem generis. See canons of construction Elhauge, Einer, 151 equity of the statute, 179 Eskridge, William N. Jr., 2, 83, 97, 123, 130–32, 153, 179, 194, executive branch: legislative history and, 96; as statutory interpreter, 161–74, 229–30 fair notice, 4, 12, 14, 21, 51, 118, 120–22, 131, 134, 155, 166, 173, 190 Farnsworth, Ward, 204 Federal Arbitration Act, 35–37, 155, 225 Federal Sentencing Guidelines, 18, 46–47, 104, 124–25, Fillmore, Charles, 63, 71, 77, Finneran, Kimberly, 124 Fodor, Jerry, 41, 63–64 Friendly, Henry ( judge), 139–40 Fuller, Lon, 55 Gilbert, Margaret, 89, 96, 159 Ginsburg, Ruth Bader ( justice), 37, 51, 73, 145, 156 Gleitman, Henry, 64 Gleitman, Lila, 64 Gluck, Abbe, 178, 187 Gopnik, Alison, 99 group intent, 89–94 Hall, Livingston, 190–91 Hampson, Sarah, 64 Harman, Gilbert, 159 Hart, H. L. A., 55 Heydon’s Case, rule in, 179 Hickman, Kristin, 195 Holmes, Oliver Wendell ( justice), 55, 82, 101–2, 110, 112, 122–23, 154 ideology of judges, 36, 125–27, 155–58, 168–69 institutional choice, 262

index intent of legislature, 14–15, 44, 52, 54, 66, 82–88, 94–97, 101–6 passim, 110, 113–17, 122–23, 144, 151, 154–55, 158, 166, 169, 172–80 passim, 184–89, 193, 217, 223, 227 interpretive communities, 78–89 Italian Civil Code, 179 Jeffries, John, 192 Jellum, Linda, 182, 186 jury: instructions, 205–19, 231–36; nullification, 197–98, 203; as statutory interpreter, 133–34, 196–222 Kahan, Dan, 133, 161, 173 Kay, Paul, 204 Kennedy, Anthony ( justice), 56, 60, 164 Klass, Gregory, 25 Klein, Benjamin, 17 Komesar, Neil, 166 Krueger, Matthew, 195 Kuhl, Patricia, 99 Kulkarni, Sanjeev, 159 law enforcement 153–55, 224 legislation: legislative errors, 4, 104–9, 146, 152; legislative findings, 228–29; legislative rules of interpretation, 178–83; statements of purpose 227–28 legislative history, 3, 14, 34, 50–55 passim, 59, 74, 80, 82–88, 95–98, 105, 108, 115–19, 121, 153, 170–73, 183–86, 189, 198, 205, 223, 229; committee reports, 82, 86, 94–98, 171, 173; executive branch participation, 96; floor debates, 96–97; regulated by rules of evidence for appellate judges, 183–89 legislative intent. See intent of legislature legislative primacy, 3–5, 13–14, 51, 68, 84, 119, 130–32, 150–61, 174, 181, 190, 192, 229 legislative process, 83, 92, 96, 109, 119, 130, 146–50, 172, 227 legislative purpose. See purpose of legislation legislature as entity with intent, 94–98, 101 lenity, rule of, 3, 21, 37, 41–47, 68, 74, 80, 105, 117 passim, 133, 140–42, 173, 186, 189–92, 195, 221–22, 226–27 Lev, Daniel, 169 Leventhal, Harold ( judge), 85

index Levi, Judith N., 218 Lindquist, Stefanie, 125–27, 157 linguistic wall, 42, 118, 177, 192 Llewellyn, Karl, 157 Malle, Bertram, 93–94 Manning, John, 52, 60, 105–6, 179, Marder, Nancy, 198, 203, 221 Marshall test, 121–23, 153, 166 Marshall, John (chief justice), 41–44, 51, 87, 121–23, 153, 166, 192 Marshall, Thurgood ( justice), 108–9 McCain, John (senator), 170 McGowan, Miranda, 68 Medin, Douglas, 64 Meese, Edwin, 169 Meltzoff, Andrew N., 99 Miles, Thomas, 167, 169 modular mind, 37–38 Molot, Jonathan, 51 Mosk, Stanley ( justice), 191–92 Nelson, Caleb, 52, 104, 169 new textualism. See textualism O’Connor, Sandra Day ( justice), 57–59, 69, 71, 109 O’Laughlin, Matthew, 93–94 opacity, 47 ordinary meaning, 11, 44, 46, 53–81, 98, 103–6, 109–10, 114–17, 121, 155, 177, 198–204, 222–23; and jurors 201–3 Osherson, Daniel, 204 pernicious ambiguity. See ambiguity Pierce, Richard, 166 Pinker, Steven, 38, 99; words vs. rules, 38, 40, 41, 46, 48, 49, 53, 60, 65, 202 plain language. See plain meaning plain meaning ,11, 51, 55, 59, 62, 113, 177–78, plain statement rules. See clear statement rules politics of judges. See ideology of judges Posner, Richard ( judge), 3, 91, 120, 135, 143, 145, 148–49, 225 precedent in statutory cases, 49, 125–27, 131, 135–40, 158 presidential signing statements, 169–73, 270 Priest, George, 17

287 Primus, Richard, 121 prosecutorial discretion, 173–75, 230, 271 prototypes and meaning, 19, 41, 43, 63–70, 76–80, 83, 135, 155, 198, 209, 221 purpose of legislation, 61, 103, 142–46, 171, 190, 225 Radin, Max, 83–84, 118 Rehnquist, William (chief justice), 45, 51, 106, 219 remedial statutes. See canons of construction Roberts, John (chief justice), 51 Roberts, John, 96 Robinson, Paul, 219–20 Rosch, Eleanor, 41, 63–64 Rosenblatt, Terri, 204 Rosenkranz, Nicholas, 180, 182 rule in Heydon’s Case. See Heydon’s Case, rule in rule of lenity. See lenity, rule of Scalia, Antonin ( justice), 2, 16, 19, 37, 45, 52, 57–61, 67–70, 75–76, 83, 85–87, 102–7, 110, 114–16, 133, 144–45, 149–53, 164, 168, 172, 189, 208, 226, 228–29 Shepard, Steven, 127–28 Siegel, Jonathan, 144 Siegelman, Peter, 17 Simon, Dan, 150, 159, 269 Sloman, Steven, 64 Smith, Vicki, 202 Solan, Lawrence, 204 Sorensen, Roy, 215 Souter, David ( justice), 9, 100 stability of interpretation, 49, 123–31, 162, 166–67 stare decisis. See precedent in statutory cases state of mind, 18–19, 106–7, 150, 152, 233–34; corrupt, 21, 22, 26–32, 85; knowing, 40, 45, 61, 182, 201, 206, 208; willful, 18–19, 45, 47, 71, 75, 140, 182, 201 statutes: antitrust laws, 123, 128–29, 140, 153, 181; bribery, 13, 20–37, 44, 48–49, 112–13, 124–28, 168, 200, 206 (see also statutes, bribery); civil rights laws, 127–28, 131, 151, 155, 157–58, 187; Employee Retirement Income Security Act (ERISA), 144; environmental laws, 68, 162–65; Federal Arbitration Act, 35, 155, 225; Food, Drug and Cosmetics Act, 140, 168; insider

288 statutes (cont.) trading, 141–42, 165, 265; mail fraud, 133, 205–12, 221, 231–36; money laundering, 212–15; National Labor Relations Act, 152–53, 167–68; Patent Act, 135–38; perjury, 18–19, 65; runny egg law, 146–48; Sarbanes-Oxley Act, 171; securities laws, 141–42, 165–66 (see also statutes, insider trading); “three strikes” laws, 220–22; trademark law, 137–40; Voting Rights Act, 67–68, 109–10 statutes, bribery: Abscam scandal, 24–25; Wedtech scandal, 29–30 statutory inflation, 140–42, 165, 173 Stevens, John Paul ( justice), 95, 108–9, 151, 228, Story, Joseph ( justice), 43–44 Strauss, Peter, 194 Sunstein, Cass, 161, 167, 169 syntax and meaning, 21–23, 33, 38–40, 48–49 textualism, 13–16, 36, 50–55, 59, 62, 70, 75, 79–80, 98, 102–7, 110–11, 118, 130,

index 143–44, 150, 177, 181, 229; and canons of construction, 102–4 theory of mind, 98–100 Thomas, Clarence ( justice), 9, 60, 69, 76 Tiefer, Charles, 95 Tiersma, Peter, 59, 131, 135, 205 Tiersma, Peter textualization of precedent, 131, 135–40 passim transparency, 47 vagueness, 35, 37, 39–40, 42, 46, 48, 121, 134, 190, 216, 243 Vermeule, Adrian, 2, 66, 116, 162, 229 Wald, Patricia ( judge), 85 Waldfogel, Joel, 17 Wattenmaker, William, 64 White, Byron ( justice), 87, 139, 193 Wittgenstein, Ludwig, 55, 63 word meaning. See definitional meaning; ordinary meaning; prototypes Zeisel, Hans, 217–18

the chicago series in law and society Edited by John M. Conley and Lynn Mather Series titles, continued from front matter: rights of inclusion: law and identity in the life stories of americans with disabilities by David M. Engel and

Frank W. Munger

the internationalization of palace wars: lawyers, economists, and the contest to transform latin american states by Yves Dezalay and Bryant G.

Garth

free to die for their country: the story of the japanese american draft resisters in world war ii by Eric L. Muller

overseers of the poor: surveillance, resistance, and the limits of privacy by John Gilliom

pronouncing and persevering: gender and the discourses of disputing in an african islamic court by Susan F. Hirsch

the common place of law: stories from everyday life by Patricia Ewick and

Susan S. Silbey

the struggle for water: politics, rationality, and identity in the american southwest by Wendy Nelson

Espeland

dealing in virtue: international commercial arbitration and the construction of a transnational legal order by Yves Dezalay and Bryant G. Garth

rights at work: pay equity reform and the politics of legal mobilization by Michael W. McCann

the language of judges by Lawrence M. Solan

reproducing rape: domination through talk in the courtroom by Gregory M. Matoesian

getting justice and getting even: legal consciousness among workingclass americans by Sally Engle Merry rules versus relationships: the ethnography of legal discourse

by John M. Conley and William M. O’Barr