Chapter 1
ELVIS IS DEAD. Tabloid headlines to the contrary, Elvis Presley really did die on August 16, 1977, and he has remained dead ever since.
For purposes of this book, the importance of Elvis’s death is that it is a fact. And as a fact, its existence is distinct from people’s personal or policy preferences. As we will see, even issues as seemingly straightforward as the death of Elvis are not nearly so simple, but it is still crucial at the outset to distinguish the fact of Elvis’s death from the preferences of many people that he be alive and perhaps even the preferences of some people that he be dead. Such preferences exist apart from the fact that Elvis is dead.1
For all its triviality, the Elvis example highlights three important distinctions about the place of facts both in public policy and in personal decision making. First is this distinction between the empirical reality and what some or many people prefer or wish that empirical reality to be. Disliking anchovies is a preference. Believing that anchovies do not exist is an empirical mistake. And taking a dislike for anchovies as a reason for denying their empirical existence is a fallacy.
The second important distinction is between actual empirical reality and what some or many people believe that empirical reality to be. It is a fact that vaccination does not cause autism, even though many people believe otherwise.2 It is a fact that the earth is warming, although some people and political parties are committed to denying that fact. And recently, and to some extent still, major government policies regarding public expenditures, regulation of businesses, and personal freedom have turned on whether the rates of incidence, hospitalization, and death from Covid-19 are rising, falling, or disappearing. But these rates, whatever they are, are facts. So are conclusions about whether these rates are higher than, lower than, or the same as they were a month ago, at least assuming a common scale of measurement. Accordingly, when a president or anyone else asserts that the rate is lower than it was at some time in the past, we can check the assertion against the facts. And we can also investigate whether some drug—such as hydroxychloroquine, whose effectiveness was often and erroneously touted in the early days of the Covid-19 pandemic—is or is not effective against Covid-19 and whether it does or does not cause side effects of a certain type.3 These are facts, and they exist apart from whatever may be asserted by a president, the Centers for Disease Control and Prevention, or anyone else.
The Covid-19 pandemic highlights the third important distinction—the distinction between (a) what the facts are and (b) what ought to be done about them. Describing the world and prescribing policy are two different things. The fact that the rate of Covid-19 infection is falling, whenever it is in fact falling, does not itself tell government whether to relax restrictions that are in place, nor does that fact tell people whether to start eating in restaurants, participating in live political rallies, or attending church services. And the fact that rates of infection are rising, whenever that is the fact, does not itself tell universities to cease holding in-person classes or advise individuals to stop taking public transportation. All these decisions about what to do are based on facts, but they require something more than facts alone. It is wise—indeed essential—to follow the science, and thus to follow the facts, in making policy and personal decisions. But following the facts and following the science means basing policies on actual facts—and not incorrect factual beliefs or “alternative facts.” It means using rather than contradicting what the science tells us. But following the facts and following the science does not, and cannot, mean—despite occasional pronouncements by scientists and government officials to the contrary—that science and facts alone can determine what some policy ought to be. Making policy decisions requires using the facts to reach what are irreducibly normative and value-laden conclusions. And these policy decisions usually require trade-offs whose resolution cannot be determined solely by the facts. What to do about the facts is not solely a factual inquiry, just as what to do with science is not solely a scientific task. However important it is to follow the science and follow the facts, it is also important to recognize that science, and the facts, can only take us so far.
Facts alone cannot indicate what people or governments should do, but facts remain the foundation for sound public and personal decision making. In memorably remarking that people are entitled to their own opinions but not their own facts, the late senator Daniel Patrick Moynihan reminded us that making correct decisions depends on getting the facts right.4 As he put it on the same occasion, “First, get your facts straight.” And although Moynihan was talking largely about governmental policy decisions, questions of fact provide the basis for many of the more personal decisions we make on a daily basis. Some people, for example, refuse to patronize establishments owned by members of the Ku Klux Klan or by others who hold racist or other opinions that the potential customers abhor. But the initial question facing the potential customer is whether the proprietor is or is not a Klan member, or does or does not harbor racist views. These are questions of fact, and they are part of our everyday lives. It may (or may not) be a morally good thing to refuse to patronize a shop whose owner is a member of the Klan, but it is plainly not a morally good thing to refuse to patronize a shop on the belief that the owner is a Klan member if in fact he is not. Our lives and our decisions involve numerous choices that are driven by values—normative political or moral or personal preferences. But acting on those preferences requires making an initial judgment that is factual and not normative. Doing the right thing matters, but getting the facts right is the first step.
The forgoing is obvious to the point of banality, but highlighting the increasingly common questions about what is—and not just about what we or some government ought to do—is the appropriate entry into what this book is all about. More than two centuries ago the philosopher David Hume warned against the fallacy of deriving “ought” from “is”—of moving from the descriptive to the normative, or from what is to what ought to be—without recognizing that the move requires judgments that cannot be based on facts alone.5 But it is no less a fallacy to attempt to derive “is” from “ought.” It would be nice if there were world peace and nonfat bacon, but wishing won’t make it so. This book is based on the premise that controversies about facts are important in their own right, but even more so because they provide the foundations for questions of personal choice and public policy. Leaving to others questions about how we or government ought to act, this book is an attempt to provide some insight into how we do—and, yes, should—confront the factual questions and controversies that are all around us.
The Idea of Evidence
It is one thing to say that there are facts; it is quite another to determine what those facts are. And once we shift from the idea of a fact to questions about figuring out what those facts are—to deciding which claims about the facts are true and which are false—we enter the realm of evidence, the subject of this book. Evidence is what provides the justification, or warrant, as philosophers are prone to put it, for believing that something is true—or false. Pieces of evidence are facts, but they are the facts that lead us to the conclusion that other facts do or do not exist. The fact that the skin tone and whites of someone’s eyes appear yellowish is evidence for the further fact that the person has hepatitis.6 The fact that the engine on my car is making a pinging sound is evidence for the fact that the gasoline in the tank is of too-low octane. The fact that Herman’s fingerprints are on the door handle of Susan’s stolen car is evidence for the fact that Herman was the thief. And the fact that then-nominee and now Supreme Court Justice Amy Coney Barrett signed an anti-abortion advertisement some years ago is evidence for what she then believed, which is in turn evidence for what she now believes, which is in turn evidence for how she might now decide.7
This book deals only indirectly with what makes a fact a fact, and with what makes a true statement true and a false one false. But it deals directly with how we know that statements or conclusions are true or false. In philosophical terminology, the focus of this book is on epistemology and not ontology (or metaphysics). In everyday language, the book is not about what is or what is not. It is about how we determine what is, and how we know what is not. It is about the assessment of factual truth and falsity in public policy, in public deliberation, and in personal decision making.
The Demand for Truth
Evidence is the prerequisite for judgments of truth (and falsity). But as we will see in Chapter 13, what psychologists call “motivated reasoning” is a large part of the evidentiary terrain. Unfortunately, how people perceive the facts of the world is often substantially influenced by their normative preferences about how they would like the world to be. It is a question of fact—in theory verifiable—whether the soccer ball crossed the line into the goal, but supporters of the team that kicked the ball will almost invariably believe in close cases that the ball crossed the line, just as supporters of the defending team will equally invariably believe that it did not. Although the location of President Barack Obama’s birth is a question of fact, it comes as no surprise that his supporters believed (correctly) that he was born in Hawaii, just as many of his political opponents believed (falsely) that he was born in Kenya. And when asked who “won” a debate between candidates for president, a question whose answer is admittedly far less factual and far less verifiable, it is still noteworthy that both prospective voters and pundits reach conclusions that are expectedly and tediously consistent with their political preferences.8 So too, as in recent events, with a closely contested election, where again many people’s assessments of the facts align all too conveniently with their political and outcome preferences.9
Even more fundamentally, however, evidence matters only to those for whom truth matters. And it is not clear that truth matters to everyone, in the same contexts, and to the same degree. If we think of truth as something that can be preferred (or not), we can then understand a preference for truth as competing with preferences for happiness, affection, friendship, ambition, wealth, health, lack of stress, and a myriad of other emotions and conditions that may at times conflict with and be more important to some people than truth. Although Henry Clay was talking more about policy than fact when he observed in 1839 that he would rather be right than president, his observation has endured precisely because we recognize that many (most?) politicians would rather be president than right.10 Being right is a preference for the truth, but not everyone has that preference, or has it in the same amount, or has it all the time.
If truth is a preference, then we should recognize that there can be a market for truth. Suppliers of truth, and thus suppliers of evidence, will understand that not everyone wants evidence, or wants it to the same degree. And if this conclusion seems uncomfortable, it is not uncomfortable for the publishers of supermarket tabloids. Many people enjoy reading about the travails of the rich and the famous, producing a demand for information—the more detailed the better—about these travails. And the suppliers of information about such travails, the supermarket tabloids prominently among them, seek to satisfy the demand, whether the travails actually exist or not. Obviously at some point stories about the misfortunes and missteps of the rich and famous will be so untethered from reality—from the evidence—that the demand will contract. And equally obviously, at some point the reality will be sufficiently uninteresting that the demand will again contract. The successful publishers of sensationalist supermarket tabloids are those who can best assess this relationship between the demand for sensationalism and the demand for truth. These financially successful tabloid publishers have located the optimal equilibrium—the point beyond which more truth (and thus more evidence) will decrease demand for the information, and the point beyond which less sensationalism will also decrease the demand for the information. Success in the tabloid business depends on discovering whatever balance between sensation and truth maximizes readership (and thus sales and advertising revenue).
This is not a book about the publishing industry, whether tabloid publishing or academic publishing or anything in between. But this detour into tabloid publishing illustrates that it is hardly obvious—as if recent events were not enough to make the point—that more truth (or more knowledge) is equally important for all people (or institutions) at all times and for all subjects. And it is equally not obvious that evidence, the basis for our knowledge and the basis for our judgments of truth and falsity, is equally important for all people at all times and for all subjects. This book is about evidence, and is premised on the idea that evidence is often important because truth is often important. But understanding the place of evidence in the world, in public policy, and in personal decision making requires understanding that evidence and truth are not the only values there are.
Of Facts and Opinions
It is a fact that Adolph Hitler had a mustache, and it is a fact that he was born in Braunau am Inn, Austria, on April 20, 1889. It is also a fact that he was evil, but this latter fact, unlike the others, is evaluative. It involves a judgment, or what some would call an “opinion.” In this case the judgment is easy, and few people, or at least few with whom I associate, would dissent from characterizing Hitler as evil. Frequently, however, the evaluation is less obvious and more contested. And often, and even more of a problem, the evaluation hides within a seemingly factual statement. The philosopher Philippa Foot gave us the idea of a “thick ethical concept,” labeled as such by the philosopher Bernard Williams.11 A thick ethical concept is a description combining the factual with the evaluative, as we see in adjectives such as “rude” and “generous,” nouns like “coward” and “hero,” adverbs like “carelessly” and “carefully,” and even verbs such as “hurry” and “loiter.”
Offering thick descriptions, as with stating bald facts, requires evidence. And in that sense this book applies as much to evidence for (or against) the conclusion that Francesco Schettino, the captain of the ill-fated Costa Concordia who in 2012 abandoned ship, leaving thirty-two crew and passengers to die, was a “coward” as it does to the conclusion that the Costa Concordia sank.12 But what counts as evidence for describing some behavior as “cowardly” requires reference to criteria for cowardice that are neither as obvious nor as uncontested as the criteria for describing an event as a “sinking.”
Looking at some recent political events and controversies, for example, we can say that to accuse someone of being a member of the Communist Party USA, an actual organization with actual members, is to make an accusation that is, in most cases, either true or false. But accusing someone of being a small-c communist, or, more frequently these days, a socialist, is to make an accusation that presupposes a contested definition.13 Much the same applies to asking someone if she believes that there is systemic racism in the United States, as Justice Barrett was asked in her confirmation hearings. The question presupposes a definition of the value-laden and contested idea of systemic racism, and as such is a question about a contested political issue masquerading as a purely factual question with a nonevaluative answer.14
This book is not about rhetorical or political strategy, nor is it a book of normative political or moral argument. It is a book about evidence and thus about how we know whether something or someone qualifies as a something under specified criteria or a specified definition. Was that a bird or a bat that was flying around the house at dusk? Is that 1930s car a Dodge or an Oldsmobile, and how do we know? But the focus of this book is not on what the definition of (or criteria for) some evaluative attribute—rudeness, cowardice, bravery, systemic racism—should be. Or, to put it another way, not all questions are questions about evidence, despite how common it is to try to make questions about values look as if they are questions only about bald facts and thus only about evidence.
A Few Words about the Law
This book is not about the law of evidence that governs trials in courts of law. Nor is the book about the law of evidence as it is portrayed on television and in the movies, which is emphatically not the same thing. Still, it would be a mistake to ignore the law in examining the way in which evidence is and should be treated in public policy and in everyday life. Although the law of evidence occasionally affects our nonlegal evidentiary determinations, that is rare. The greater importance for the law of evidence in nonlegal contexts comes from the fact that the legal system has been thinking about evidence somewhat systematically and often carefully for almost three centuries.15 This is especially true of legal systems in common-law countries.16 Sometimes the result of this process—the legal system being what it is—is that bad ideas become entrenched over time because of the legal system’s occasional (or frequent) preference for long-standing poor ideas over novel good ones. Indeed, Justice Oliver Wendell Holmes once observed that it was “revolting” to be required to follow earlier conclusions—precedents—that “persist … for no better reasons than … so it was laid down in the time of Henry IV.”17 Nevertheless, sometimes a few hundred years of thinking by mostly intelligent and mostly thoughtful judges can generate sound ideas and valuable perspectives, especially when that thinking has taken place in the context of concrete cases involving real people with real problems. It would be a mistake to ignore this repository of accumulated wisdom, and I have no intention of doing that here, any more than I wish to ignore the accumulated, even if less systematic, wisdom about the nature and use of evidence that has been produced by natural and social scientists, historians, and philosophers, among others, as they wrestle with the problems of evidence.
A Preview
The organization of what follows is topical rather than strictly logical or serial. Each of the chapters deals with a particular topic within the broad subject of evidence, but the arrangement of the chapters does not follow a systematic progression of argument from beginning to end. Instead, the arguments within the chapters are somewhat independent of each other, making each chapter at least slightly self-standing. Occasional references back to earlier chapters—or announcements of what is to come in subsequent ones—will help tie things together, but the reader seeking an unwavering march from premises to conclusion will be disappointed. When Justice Holmes observed that he “care[d] nothing for the systems—only the insights,” he made clear that, for him at least, interstitial and incremental observations and analyses may often be more valuable than grand comprehensive theories.18 The topic of evidence is especially amenable to Holmes’s approach, and the organization of this book is premised on the belief that there are interesting and important things to say about evidence, even if not everything that is interesting and important about evidence will fit together neatly, systematically, or logically.
That said, the chapters that follow can be divided into four main groups. Chapters 2, 3, and 4 are largely about the concept of evidence itself. Here we focus on questions of inference and relevance, examining evidence through the lens of the idea of probability, with particular attention to the way in which the burden of proof is an idea of pervasive importance and common confusion. Specifically, Chapter 2 deals with the basics of inference and relevance, looking at how evidence bears (or does not bear) on the conclusions of fact that interest us. Chapter 3 focuses on the idea of the burden of proof, exploring how, in a world of uncertainty, we determine whether and when the inconclusive indications of the evidence are or are not good enough for one or another purposes. And Chapter 4 aims to connect these notions of inference and burden of proof to ideas of probability and statistics, recognizing that questions of “how much,” even if not expressed in numbers, lie at the heart of many important questions about evidence.
Much that we know, we learn from what others tell us. The image of Sherlock Holmes in his deerstalker hat searching with an outsized magnifying class for physical clues is endearing and enduring, but much of the evidence we use comes not from what we see or find but from the statements—the testimony—of others. Chapters 5, 6, 7, and 8 focus on testimony in this broad sense. Chapter 5, making some use of philosophers’ recent interest in testimony, will explore the very idea of testimony, and look at when and why we should reach a conclusion just because of what someone has told us. But not everything that people say is accurate. Chapter 6 focuses on the devices lawyers and others have long used to attempt to assess the accuracy of what others say. Does it help that someone has sworn to something, whether in court or elsewhere? Does cross-examination help in testing testimony, whether eyewitness or otherwise, for truthfulness or accuracy? And how can we calibrate testimony once we recognize that knowing something about the testifier will help to assess the evidentiary value of the testifier’s testimony? Indeed, as we increasingly confront a world in which political figures and others say things—often intentionally—that are untrue, we often wish we had a good way of separating liars from truth-tellers. Chapter 7 looks specifically at lie detection, tracing both the development of lie-detection technology and the legal system’s persistent skepticism about the value of such technology—and its surprising lack of skepticism about the ability of judges and jurors to determine who is telling the truth and who is not. Not all sources of error, however, are the product of lies. Honest people make mistakes, even mistakes about things they believe they have seen with their own eyes and heard with their own ears. And they even make mistakes about things they themselves have experienced. Chapter 8 takes up the problem of honest mistakes in perception and recall, a problem that lies at the heart of our ability to treat the statements of others as evidence for what they have reported.
As recent controversies have prominently and often tragically demonstrated, expertise matters. It is sometimes said that we live in an “age of experts.”19 We live also, however, in an age of dueling experts, in which the conclusions of experts are marshaled on multiple sides, and in which the notion of “expert opinion” as a collective consensus judgment of multiple experts is elusive. Chapters 9, 10, and 11 explore the world of experts and expertise, concentrating on the ways in which expertise can be used, evaluated, and, at times, challenged. Chapter 9 examines the conclusions of science and scientists as evidence, using recent controversies about vaccination, climate change, Covid-19, and genetically modified organisms (GMOs) to probe the role of scientific experts in the evidentiary determinations that are increasingly central to public policy. Chapter 10, the chapter that is connected most explicitly with law and the courts, deals with current debates over the forensic techniques that have long dominated police-focused television as much as they have the activities of real courts and real detectives. But are these techniques—fingerprints, voiceprints, tire marks, ballistic tests, handwriting identification, and much more—as reliable as they are purported to be? And how do we know? Chapter 11 then leaves the world of science but not the world of expertise, looking at many of the factual conclusions—in realms as diverse as wine-tasting, art authentication, and the writing of history—that rely heavily, often for better but sometimes for worse, on the conclusions of experts, even though the methods that produce those conclusions are distant from anything that looks like science.
Chapters 12 and 13 deal with topics not so neatly grouped together but connected with what we can learn from contemporary research in cognitive and social psychology. Chapter 12 takes up the question of character and the related question of the relevance of past acts to current determinations. Does character matter? Does what people are like help us to determine what they have done? Can knowing what people have done in the past tell us what they might do in the future or what they might have done in the present? And Chapter 13 addresses the phenomenon of motivated reasoning. In evaluating evidence, do evaluators see what they want to see? How much of what we want to be the case influences—or distorts—our judgments of what is the case?
As the foregoing should make clear, the topic of evidence implicates a wide variety of only loosely connected subtopics. But there is nothing wrong with that. Sometimes, even if not always, the whole really is greater than the sum of its parts, and perhaps the vast subject of evidence is good evidence of that phenomenon. Still, there are a few themes that run erratically through what is to follow. First is that probabilities matter. A lot. The very idea of evidence is about inductive reasoning and thus about probability, even if not necessarily with numbers attached; and the probabilistic nature of evidence will be a recurring focus. Second, evidence comes in degrees. Sometimes, and seemingly more often recently than in the past, people reach conclusions and make statements for which there is, literally, no evidence. But although “no evidence” really is no evidence, weak evidence is still evidence. And weak evidence often has its uses. Repeatedly, the question of how much evidence there is will be as important as the question whether there is any evidence at all. The third running theme follows from the second. Whether what evidence we have, or how much evidence we have, is good enough depends on what we are to do with it, and on the consequences of the evidence we have being sufficient or insufficient to support some conclusion. In other words, questions of “Compared to what?” and “For what?” will be a recurring motif. All of this, I hope, will become much clearer as the book proceeds.