APA Newsletters
Spring 2000
Volume 99, Number 2

Newsletter on Philosophy and Law

Articles

Previous | Next

Reasonable Doubt Jury Instructions

Jennifer Faust
Department of Philosophy
California State University at Los Angeles

Since the 16th century, when jurors became evaluators of testimony and evidence (rather than collectors of evidence or persons assumed to have firsthand knowledge of a crime), courts’ instructions to jurors have been a crucial factor in ensuring due process. A fair trial demands that jurors consider all and only the legitimate evidence at trial, that they refrain from discussing the case with parties outside the process, and that they apply the appropriate standards of proof. To this end, they are admonished to consider firsthand testimony but not hearsay, to consider the statements of sworn witnesses but not those of attorneys, and so forth. Surely one of the more important instructions to criminal trial juries concerns the correct standard of proof—to wit, that the prosecution must prove its case beyond a reasonable doubt. Embedded in this charge is both the claim that the burden of proof rests on the prosecution and the notice that the standard of proof is set extremely high. The importance of this jury instruction is underscored by the fact that every court in the land must inform jurors of the "beyond a reasonable doubt" standard of proof. While the charge is universal, however, courts differ both on whether and on how to define "proof beyond a reasonable doubt" for jurors. Most courts that do offer jurors an explanation of this concept attempt to define the embedded term "reasonable doubt." In this brief essay, I shall attempt to establish that some (quite popular) jury instructions concerning reasonable doubt are philosophically unsound.

In criminal trials, it is the burden of the prosecution to bring jurors to conviction—wherein the appropriate epistemic state of the juror is a vote for the legal state of guilt of the defendant. Thus, while conviction denotes a finding of guilt on the legal end, it also denotes an epistemological category. But, on the epistemic end, the word is ambiguous. Construed widely, "conviction" is often equated with a kind of belief as in "his religious conviction prevents him from considering some courses of action" and "it is my conviction that Clinton is guilty of something." Here, we may speak—loosely—of conviction (i.e., of being convinced) even where we lack conclusive evidence for our beliefs; in this sense, conviction is a measure of the tenacity with which one holds onto one’s belief. In another more straightforwardly epistemic sense conviction is a measure of the justification or evidence for a belief. In the second sense, to be convinced is to be persuaded of the truth of a claim by evidence sufficient to justify that claim. In an epistemically ideal world, the two senses of conviction would never come apart—i.e., we would cling tenaciously only to those beliefs that the evidence convinced us to endorse. But we often err by believing without sufficient evidentiary support and/or by refusing to believe where evidence is strong.

It is for this reason that jurors are admonished to decide cases solely on the basis of the evidence presented, for one may harbor beliefs about a defendant’s guilt or innocence without properly considering the evidence presented. Indeed—and this is crucial to the present investigation—one may believe that a defendant is guilty while at the same time recognizing that the prosecution has not met its burden of proof beyond a reasonable doubt. In other words, it is possible for a juror to both believe that a defendant is guilty and to fail to be convinced of this fact by the evidence presented at trial. In such cases, a juror may be in the strange position of believing a given claim while reasonably doubting that very claim. To show how this might arise, we must consider the situations in which reasonable doubt legitimately arises in a trial.

To put reasonable doubt in the proper context, however, we must first say something about doubt simpliciter. We can distinguish between the following two senses of doubt (where S is a person, p a statement).1

S doubts1 that p =df S believes that not p.

S doubts2 that p =df S does not believe that p.

In the first sense (doubt1), doubting is equivalent to disbelieving. In this sense, the atheist—but not the agnostic—doubts that God exists. In the second sense (doubt2), one’s doubting does not entail any positive belief state at all. This, clearly weaker, sense of doubting is compatible with taking no position at all with regard to the claim (p) in question. In this weak sense, we may say that the agnostic doubts that there is a god. Consider the case in which "p" stands in for "the defendant is guilty of the crime(s) charged." A juror may have doubts with regard to this claim in either the stronger sense (doubt1) or in the weaker sense (doubt2). In the first case, his doubt1 is equivalent to a belief in the defendant’s innocence; in the latter case, his doubt2 is merely the absence of belief with regard to the defendant’s guilt. Insofar as a juror believes that the defendant is guilty, the juror does not doubt (in either sense) that the defendant is guilty.

The relevance of this distinction to the philosophy of law lies in the fact that some reasonable doubt jury instructions equivocate between doubt1 and doubt2, with the potential result that jurors use incorrect standards when determining whether the prosecution has met its burden of proof. In order to establish this claim, I will first argue that the appropriate sense of doubt embedded in "reasonable doubt" is the weaker doubt2. Ultimately, I will show that some prevalent reasonable doubt instructions nonetheless employ doubt1.

In the context of a criminal trial, jurors are faced with two legitimate options with regard to the defendant’s status in the alleged crime: guilty or not guilty (where the latter is not equivalent to innocence). At the outset, I suggested that the prosecution must attain conviction in the epistemically technical sense, wherein jurors are provided sufficient evidence for belief in the defendant’s guilt. In such a case, a reasonable person would come to believe that the defendant is guilty; thus, on the assumption that jurors are reasonable persons, whenever the prosecution has met its burden of proof, the jury will convict. On the other hand, when the jurors have not been provided sufficient evidence for the defendant’s guilt, they are said to have reasonable doubt. It follows that reasonable doubt is itself equivalent to a lack of sufficient evidence for a juror to believe that the defendant is guilty. Now, compare this result to the "reasonable doubt" correlates of doubt1 and doubt2, where "p" stands in for "the defendant is guilty of the crime(s) charged."

S reasonably doubts1 that p =df S has sufficient reason to believe that not p.

S reasonably doubts2 that p =df S does not have sufficient reason to believe that p.

In the first option, a juror (S) has reasonable doubt only if he has sufficient reason to believe that the defendant is innocent. While jurors may be presented with sufficient reason to believe that a defendant is innocent in some (relatively rare) cases, the court certainly cannot insist that jurors attain such a state in order to acquit. Clearly, this requirement would fly in the face of the presumption of innocence, which the court allows the defendant. Reasonable doubt1 suggests that the relevant issue is whether or not the jurors have been convinced of the defendant’s innocence, thus incorrectly placing the burden of proof on the defense. The relevant issue, of course, is whether or not the jurors have been convinced of the defendant’s guilt. If not, they are said to have reasonable doubt. Thus, reasonable doubt is clearly defined in the weaker sense of doubt2. To see why this is so, we need only consider what is required for the prosecution to establish guilt beyond a reasonable doubt—the prosecution must present convincing evidence that the defendant is guilty of the crime charged, and there must be no plausible explanation of the facts established in court other than the explanation that incorporates the defendant’s guilt.

There are two quite different scenarios in which jurors will fail to be convinced of a defendant’s guilt. First and most obvious, the prosecution may fail to convince jurors of a defendant’s guilt beyond a reasonable doubt simply because their case is too weak and the evidence presented at trial is not sufficient to warrant the belief that the defendant is guilty. In such cases, the reasonable doubt need not be something that a juror can put her finger on, as it were, for there need not be any positive epistemic state with which the doubt could be equated. Instead, there is merely an absence of conviction in the juror’s mind. One would expect these "insufficient evidence" cases to be relatively rare in trial situations, for one would expect prosecutors to recognize such cases and attempt a plea bargain rather than bring an extremely weak case to trial. Nonetheless, such cases are often prosecuted, for a variety of reasons.

A second—and vastly more common—situation in which reasonable doubt may arise involves the underdetermination of theory by evidence. The thesis of underdetermination makes the claim that two (or more) theories are underdetermined by a given class of empirical evidence whenever each of the theories equally well incorporates that evidence and there is no way to determine which of the competing theories is true by appealing to that evidence. Where two competing theories are mutually consistent, underdetermination need not be a serious problem (for we could then simply decide to accept both theories unless or until further empirical evidence comes along and falsifies one of the theories). However, in those cases of underdetermined but mutually inconsistent theories, we are faced with a vexing epistemological situation: we know that at least one of the two theories must be false, but we cannot determine which, for our available evidence supports the theories equally well. It was just this situation that astronomers supposedly faced in the late 16th century when Copernicus’s heliocentric cosmology and Tycho Brahe’s modification of Ptolemy’s geocentric cosmology appeared to be equally well supported by the available astronomical data (prior to input from Kepler).

Just as scientists are presented with empirically underdetermined theories, jurors are sometimes presented with two theories—one advanced by the prosecution, one by the defense—that square equally well with the evidence presented in a trial. Technically speaking, of course, the statements of attorneys are not considered evidence in a trial. Courts make clear, via instructions to jurors, that the evidence that jurors are to weigh in deliberating is limited to the testimony of witnesses, the exhibits entered into evidence during the course of the trial, and the facts stipulated by the judge or attorneys. In reality though, counsel for each side aims to paint a coherent picture that incorporates the evidence and suggests the guilt or innocence of the defendant (depending on whether the prosecution or the defense is doing the painting). Where the defense proposes a scenario that incorporates the trial evidence and is at least plausible, and the jurors have no way of distinguishing which of the two stories is true by appealing to that evidence (and to "reason and common sense"), the jury is faced with a case of underdetermination of theory by evidence. Given the presumption of innocence, cases of underdetermination must be decided in the defendant’s favor. For where the prosecution cannot present evidence that falsifies a plausible scenario that points to innocence, they have failed to prove their case beyond a reasonable doubt. Indeed, Devitt and Blackmar explicitly acknowledge this in their proposed pattern jury instructions on reasonable doubt:

Unless the government proves, beyond a reasonable doubt, that the defendant has committed each and every element of the offense charged in the indictment, you must find the defendant not guilty of the offense. If the jury views the evidence in the case as reasonably permitting either of two conclusions—one of innocence, the other of guilt—the jury must, of course, adopt the conclusion of innocence.2

While the law makes it clear that when the evidence presented at trial underdetermines the mutually inconsistent conclusions of innocence and guilt, jurors must acquit, it is difficult to say just when a case "reasonably permits either of [these] two conclusions." Suppose a defense attorney suggests in both opening and closing arguments that the evidence entered at trial is consistent with his client’s innocence, but does not put on an affirmative defense; further, suppose that the scenario that the defense puts forth is possible (though not likely). Is a juror warranted in voting for acquittal on the basis that the prosecution did not show that the defense’s story was not in fact what happened? Many courts attempt to head off this sort of situation by instructing jurors on reasonable doubt, to the effect that "a reasonable doubt . . . is not based purely on speculation," "is not mere possible, imaginary, or fanciful doubt," and so forth.3 This suggests that defense counsel must put on a case in order for there to be reasonable doubt based on underdetermination.

Further questions present themselves as well; for instance, must the underdetermined "theories" presented at trial be equally plausible? This seems like too stringent a requirement for the defense, for the prosecution’s burden in a criminal trial goes beyond a preponderance of the evidence. That is, consider the situation in which there are two coherent scenarios on the table, but the defense scenario is less plausible than the prosecution’s, yet supported by some evidence. A juror cannot vote for guilt in this situation, it seems, for that would be to inappropriately use the weaker standard of proof used in civil cases to decide a criminal case. Though we cannot say precisely where lies the line between mere speculation and legitimate alternative, it is reasonably clear that if the defense goes beyond mere speculation and offers a plausible and coherent scenario that incorporates the evidence entered at trial, then the jury is faced with true underdetermination of theory and should acquit.

We have identified two quite different types of situations that legitimately give rise to reasonable doubt. In the first class of cases, the prosecution simply fails to present evidence sufficient to convince jurors of the defendant’s guilt. In the second class of cases, the juror is presented with two explanations for the evidence presented at trial: one offered by the prosecution points to the defendant’s guilt and another offered by the defense points to the defendant’s innocence. In this case, the evidence underdetermines the mutually inconsistent theories. In both kinds of cases, jurors must vote to acquit, as the prosecution has not proven its case beyond a reasonable doubt. In practice, however, courts often ignore or misconstrue these reasonable doubt scenarios in their pattern jury instructions, at the defendant’s expense.

One way in which courts can err in their reasonable doubt jury instructions is to fail to admonish jurors to acquit in cases of underdetermination. In fact, few state courts have adopted the language of Devitt and Blackmar’s instructions (above), leaving jurors to determine how they should vote when the evidence "permits two conclusions." In practice, this favors the prosecution in legitimate cases of underdetermination. Suppose that the prosecution has put on a strong case for the defendant’s guilt (such that in the absence of a defense case the prosecution’s evidence constitutes sufficient evidence for the defendant’s guilt), but that the defense has convincingly called into question some of the elements of that case. The totality of evidence in such a case would underdetermine the choice between guilt and innocence, yet in the absence of explicit instructions jurors may feel obliged to convict on the grounds that the prosecution’s evidence established guilt.

Worse yet, many courts define reasonable doubt so as to exclude the insufficient evidence cases from counting as legitimate cases of juror reasonable doubt. It is important to note that in insufficient evidence situations, jurors may not be able to state precisely why they are not convinced for they may have nothing "concrete" on which to hang their doubts. In practice, some pattern jury instructions ignore this fact by defining reasonable doubt as a positive epistemic state (more akin to doubt1 above than the more appropriate doubt2). The vast majority of state courts begin their pattern instructions on reasonable doubt with the claim that "a reasonable doubt is a doubt based on reason," suggesting that the relevant doubt is a positive epistemic state (i.e., one for which a person could produce a reason). Given the likelihood that the prosecution’s evidence in some cases that go to trial falls short of the beyond a reasonable doubt standard, such instructions are in error. Indeed, they may violate the defendant’s due process rights.

The claim that a reasonable doubt is a doubt based upon reason is so ubiquitous in American jury instructions that one commentator dubbed it "one of the law’s most famous tautologies."4 Similarly, the First Circuit Court judged the phrase to be "harmlessly circular."5 I contend, however, that such instructions are neither tautologies nor harmless, for the phrase could be interpreted in such a way as to shift the burden of proof onto the defense. That is, "a reasonable doubt is a doubt based on reason," admits of two very different readings. On one reading, it is akin to the reasonable person standard widely used in the law. In this sense, the statement is to be read as "a reasonable doubt is a doubt founded upon (or arising from) the faculty of reason" as opposed to doubts based upon whimsy, speculation, wishful thinking, or prejudice. Indeed, several courts endorse this reading in their jury instructions. For example, Maryland’s instruction reads (in relevant part):

A reasonable doubt is a doubt founded upon reason. It is not a fanciful doubt, a whimsical doubt or a capricious doubt.6

Similarly, Texas instructs jurors that:

A "reasonable doubt" is a doubt based upon reason and common sense after a careful and impartial consideration of all the evidence. It is the kind of doubt that would make a reasonable person hesitate to act in the most important of his or her own affairs.7

In such cases, the court telegraphs the correct reading to jurors—either by contrasting reasonable doubt with unreasonable (whimsical, etc.) doubts or by relating the "reason" embedded in the charge to another faculty, common sense.

However, instructions from other jurisdictions suggest another, incorrect reading of "a reasonable doubt is a doubt based upon reason." Consider the following example from Washington:

A reasonable doubt is one for which a reason exists.8

Another example of this reading comes from New York:

A doubt of the defendant’s guilt, to be a reasonable doubt, must be a doubt for which some reason can be given.9

These jury instructions clearly require that jurors must provide a reason for acquittal but that requirement seems to violate one of the basic components of due process. That is, such instructions seem to shift the burden of proof to the juror (or, perhaps, the defense). Recall that, properly understood, reasonable doubt is cashed out in terms of the weaker sense of doubt2 (wherein the lack of sufficient evidence itself constitutes reasonable doubt). In other words, the state of reasonable doubt is itself defined as the absence of evidence. Yet, jury instructions like those of Washington and New York suggest that one must provide a reason or evidence for being in this state.

One might argue in defense of such instructions in the following way. The reason that a juror is in the state of reasonable doubt is either that the evidence presented was insufficient to convince the juror of the defendant’s guilt or that the evidence underdetermines the choice of guilty or not guilty. In either case, the reasonable person should be able to specify which scenario has given rise to their doubt. Thus, insofar as the juror can identify which of these routes led her to doubt that the defendant is guilty, she has met the requirements of the instructions above. Her doubt is one "for which a reason exists."

In reply, I have only to note that a juror cannot be expected to reason along these lines. It is quite possible that a (reasonable) juror would understand these instructions to demand that she articulate some reason other than "the prosecution has not convinced me of the defendant’s guilt." This possibility could have a devastating impact on due process by effectively shifting the burden of proof, thus violating the defendant’s presumption of innocence. Given this possibility and the fact that there are alternative formulations of the reasonable doubt jury instruction, it seems incumbent on courts to avoid misleading definitions of reasonable doubt such as those found in Washington and New York’s pattern jury instructions.

Notes

This research was supported by grants from the National Endowment for the Humanities and California State University, Los Angeles. I am grateful to my colleagues Mark Balaguer, Sharon Bishop, and Ann Garry and an anonymous reviewer for the APA Newsletter on Philosophy and Law for helpful comments on earlier versions of this paper.

1. Nathan Salmon defines doubt as a disjunction of these two senses, in "Being of Two Minds: Belief with Doubt," Nous 29, no. 1 (1995): 1–20. Salmon credits Bertrand Russell with a similar usage in his Theory of Knowledge (London: Routledge, 1992).

2. Devitt and Blackmar, Instructions for Federal Criminal Cases, Section 12.10.

3. Both the 5th and 9th circuit courts have adopted such language, as have fully half of the state courts.

4. See Stephen J. Fortunato, Jr. (1996) "Instructing on Reasonable Doubt After Victor v. Nebraska: A Trial Judge’s Certain Thoughts on Certainty" 41 Vill. L. Rev. 365.

5. United States v. Romero, 32 F. 3d 641 (at 652).

6. Maryland Criminal Jury Instructions (MCJI) Section 1.04—Reasonable Doubt (1975).

7. Texas Criminal Jury Charges, Section C-1 "General Jury Instruction: Reasonable Doubt, Presumption of Innocence" (1997–98).

8. Washington Pattern Jury Instructions (W.P.J.I.)—Criminal, No. 4.01 (1977).

9. New York Criminal Jury Instruction No. 3.07 (1st edition 1983).

Previous | Next

APA Newsletters Spring 2000 Volume 99, Number 2

APA Newsletters
Spring 2000
Volume 99, Number 2