The Griffin No-Comment Rule

The Griffin No-Comment Rule

The preceding sections followed a witness who faces incriminating questions and must decide, question by question, whether to answer. A criminal defendant on trial holds a larger form of the same protection: the defendant may stay off the witness stand altogether, the prosecutor may not call the defendant to testify, and the prosecutor may not force a claim of privilege to be made in front of the jury.¹ That protection leaves open what the state may do with the silence once the defendant has chosen it. Griffin v. California answers that the state may not turn the silence into evidence of guilt, and it reaches that answer by joining the penalty reasoning behind the privilege to the same-standards holding of Malloy v. Hogan.²

Griffin was tried by a jury in a California court for first-degree murder. The evidence placed him with the victim on the evening of her death and in the alley where her body was found, and he did not testify on the question of guilt, although he did testify at the separate proceeding on penalty.³ California's constitution then allowed the court and counsel to comment on a defendant's failure to explain or deny the evidence against him, and allowed the jury to consider that failure.⁴ The prosecutor used the license fully. He listed what the defendant would know, including how the victim got down the alley, how the blood reached the concrete steps, how long the two were together, how her wig came off, and whether he beat her, and then told the jury that these were things the defendant "has not seen fit to take the stand and deny or explain."⁵ The trial judge added an instruction that, as to evidence the defendant could reasonably be expected to deny or explain from facts within his knowledge, the jury could treat a failure to do so as tending to indicate the truth of that evidence and as making the inferences unfavorable to him the more probable. The judge also told the jury that the defendant had a constitutional right not to testify and that his silence did not create a presumption of guilt or relieve the prosecution of its burden.⁶

The Court reversed. It started from Malloy's statement that the same standards must determine whether an accused's silence in either a federal or state proceeding is justified, and it said it took that statement in its literal sense.⁷ On that footing the Court held that the Fifth Amendment, in its direct application to the Federal Government and in its bearing on the States through the Fourteenth Amendment, "forbids either comment by the prosecution on the accused's silence or instructions by the court that such silence is evidence of guilt."⁸ This is the Griffin no-comment rule, and it prohibits two acts: argument by the prosecutor and instruction by the judge. Federal trials already barred such comment under a statute, now 18 U.S.C. § 3481, as Wilson v. United States had held, so the question in Griffin was whether the Fifth Amendment itself imposes the bar, statute or no statute, and the Court answered that it does.⁹

Before Malloy the opposite position had prevailed. In Adamson v. California the Court held that the Due Process Clause did not carry the Fifth Amendment's protection against compelled testimony into state trials, and that a state rule permitting comment on an accused's failure to testify therefore did not make the trial unfair.¹⁰ That holding depended on the premise that the privilege was not a limit on the states. Malloy removed the premise by making the privilege itself binding on state governments.¹¹ Once the privilege itself governed, a state comment rule had to satisfy the federal standard, and Griffin applied that standard and overruled Adamson in doing so.¹²

The Court's reasoning calls comment on silence a penalty, and that label has a mechanism behind it. The privilege gives a defendant a choice between testifying and staying silent. If silence can be argued to the jury as a sign of guilt, the choice stops being free, because the defendant who stays silent pays for it in the jury room and the only way to avoid the payment is to take the stand. The Court described comment as a remnant of the inquisitorial system and as a penalty imposed by courts for exercising a constitutional privilege, one that "cuts down on the privilege by making its assertion costly."¹³ The same penalty logic later condemned a statute challenged in Brooks v. Tennessee, which required a defendant who wished to testify to do so before any other defense witness, because it punished silence by excluding later testimony.¹⁴

Silence also proves less than comment assumes. Not every innocent person can safely take the stand, since timidity and nervousness in explaining suspicious circumstances can make an honest witness look evasive.¹⁵ The California Supreme Court itself acknowledged that a defendant might refuse to testify out of fear that prior convictions would be used to impeach him, so a second explanation exists besides inability to deny the charge.¹⁶ An inference of guilt therefore treats an ambiguous act as if it were an admission. The third strand explains why the rule reaches the prosecutor and the judge. The Court distinguished what jurors may infer without help from what they may infer when the court "solemnizes the silence of the accused into evidence against him."¹⁷ The rule thus targets official endorsement of the inference and does not purport to control what jurors privately think. The Court has also refused to abandon it on the ground that jurors will draw the inference anyway, and in 1999 it reaffirmed the rule, explaining that it has proved its utility and teaches that the question at trial is whether the government has carried its burden while respecting the defendant's rights.¹⁸

The holding was contested, and the disagreement shows the two premises the rule rests on. Justice Stewart, joined by Justice White, dissented because the California procedure involved no compulsion of the kind that produced the Fifth Amendment, which grew out of commands to answer backed by imprisonment, banishment, or mutilation.¹⁹ He located whatever pressure existed in the defendant's own choice to stay silent rather than in the comment, and he described the comment rule as a way of bringing into rational discussion a fact already impressed on the jury's consciousness.²⁰ Justice Harlan, who had dissented in Malloy, concurred separately in Griffin on the footing that Malloy had already settled that the privilege binds the states, and he later counted both decisions among those incorporating the privilege over his protest.²¹ The dissent attacks the penalty premise and the concurrence attacks the incorporation premise. The majority answers the first with its account of how comment makes silence costly and the second with Malloy's holding that the same standards govern silence in state and federal proceedings, and the majority's rule is the law.

Griffin has a defined reach. It applies directly only to criminal cases, only to the defendant's own privilege, and only to the right to stay off the stand, as distinct from the right to refuse particular questions, and comment in other settings is left to other law.²² Students who apply Griffin to a party's silence in a civil case carry it past its holding, because comment on silence is usually permissible there, and the Court has allowed a criminal defendant's assertion of the privilege to be used against him in a related civil proceeding, as in Baxter v. Palmigiano.²³ The rule does reach beyond the verdict. After a guilty plea the privilege still applies at sentencing, so a sentencing judge may not infer the circumstances and details of the crime from the defendant's failure to testify, as the Court held in Mitchell v. United States, the 1999 decision that reaffirmed Griffin.²⁴

A violation of the rule does not automatically require reversal. In Chapman v. California the prosecutor filled his argument with references to the defendants' silence and the court told the jury it could draw adverse inferences, and the California Supreme Court admitted the federal violation but affirmed under its own harmless-error provision.²⁵ The Court held that whether a conviction may stand after a federal constitutional violation is itself a federal question, and that some constitutional errors are so insignificant in the setting of a particular case that they do not require reversal.²⁶ The error is presumed prejudicial unless the reviewing court can declare it harmless beyond a reasonable doubt, and overwhelming other evidence of guilt can supply that showing.²⁷ Griffin therefore forbids two distinct acts by the state's representatives at trial, the prosecutor's argument and the judge's instruction, and the next two sections take them in turn.

Prosecutorial Comment

The first act that the Griffin rule forbids is argument by the prosecutor, and the closing in Griffin v. California shows how such an argument is built. The prosecutor did not stop at observing that the defendant stayed off the stand. Instead, the prosecutor marked out a series of open questions, each introduced by the phrase "He would know," and each aimed at something only the defendant could say: how the victim got down the alley, how the blood reached the concrete steps, how her wig came off, and whether the defendant beat her.³⁰ The prosecutor then assigned every one of those questions to a single person and a single choice, telling the jury that the victim could not give "her side of the story" and that the defendant "won't."³¹

That structure explains why the argument works as evidence of guilt. Ordinarily a question left open by the state's proof is a weakness in the state's case, because the state carries the burden on every element. The Griffin closing reversed that effect: by announcing that the defendant would know each answer and had declined to give it, the prosecutor invited the jury to fill every gap with the answer least favorable to the defendant, as if the silence were a concession. The state thereby collected the benefit of an admission without an oath, without cross-examination, and without having to prove what the silence supposedly shows. The Court described comment of this kind as a penalty that "cuts down on the privilege by making its assertion costly."³² Allowing the inference would also push defendants toward the witness stand, since the only way to avoid paying for silence in the jury room would be to testify, and staying off the stand is the very choice the privilege protects.³³

Prosecutorial argument is an especially effective vehicle for that inference, because it comes just before the jurors deliberate and speaks with the authority of the state, and so the rule does not wait for an express request to infer guilt. The prosecution may not tell the jury that the defendant could have taken the stand to tell the defendant's own story and chose not to.³⁴ Any explicit invitation to treat the failure to testify as tending to prove guilt is barred, whether it comes from the prosecutor or from counsel for a codefendant.³⁵ The bar also reaches argument that does no more than call the failure to testify to the jury's attention. A prosecutor who told the jury that the defendant did not take the stand, and that the jury therefore lacked the defendant's thoughts, was held to have commented improperly, and the result is the same when the prosecutor adds that the defendant has a right not to testify.³⁶ A reminder, even a courteous one, points the jurors at the silence just as they weigh the evidence, and a disclaimer cannot remove the thought it has just planted. Students go wrong when they treat a bare mention of the defendant's failure to testify as safe because the prosecutor never asked for an inference.

Remarks that refer to the silence only indirectly are harder to classify. Courts look first to whom the remark is aimed. An argument that the defendant has not explained evidence that the defendant above all would know about is a clear violation, while an argument that the defense left a witness's account uncontradicted is usually permissible, unless the facts are such that the jury would reasonably conclude only the defendant could have contradicted it by taking the stand. For remarks that only possibly refer to the silence, lower courts read the remark in context and ask whether the prosecutor manifestly intended a comment on it or whether the jury would naturally and necessarily take it that way.³⁷ A lead prosecutor who stood in front of the defendant, gestured, and demanded in a loud voice that the defendant say where the defendant had been satisfied that test.³⁸ Describing the government's own proof as uncontradicted has produced conflicting decisions. Such a remark is more likely to be accepted when it answers an argument the defense made, and it can stand so long as an inference from the defendant's silence is not its natural and necessary result.³⁹ Even where a comment crosses the line, an appellate court may still hold the error harmless.⁴⁰

Step two does most of the work, because the Griffin closing shows that a reference can be built entirely from what the defendant would know, without the word silence ever being spoken. The prosecutor's argument is one of two ways the state can turn silence into evidence. The judge's instruction, which speaks with the court's own authority, is the other, and the next section takes it up.

Adverse Judicial Instructions

The second act that the Griffin rule forbids comes from the bench rather than from counsel table, and the instruction given in Griffin v. California shows its shape. The trial judge first told the jury that the defendant had a constitutional right not to testify. The judge then added that, as to evidence the defendant could reasonably be expected to deny or explain because of facts within his knowledge, the jury could treat his failure to do so "as tending to indicate the truth of such evidence" and could regard the inferences unfavorable to him as the more probable ones.⁴⁶ The same charge cautioned that such a failure created no presumption of guilt, did not by itself warrant an inference of guilt, and did not relieve the prosecution of its burden of proof.⁴⁷ Those cautions did not rescue the instruction, because the permission at its center was the defect. The Court set aside the conviction on the ground that the Fifth Amendment "forbids either comment by the prosecution on the accused's silence or instructions by the court that such silence is evidence of guilt."⁴⁸

The instruction earns its own treatment because a judge speaks with a different authority than a prosecutor. Jurors expect a lawyer in closing to press one side and can discount the argument accordingly, but they hear the judge as the neutral source of the law they must apply. When that source tells them that silence may be counted toward the truth of the state's evidence, a suspicion jurors might have formed privately becomes an approved method of reasoning. The Court drew exactly this line, contrasting what jurors may infer with no help from the court with what they may infer "when the court solemnizes the silence of the accused into evidence against him."⁴⁹ The cost to the defendant is the same cost that makes prosecutorial argument unconstitutional. A defendant who knows the judge will tell jurors to hold silence against him can avoid that instruction only by testifying, and declining to testify is the choice the privilege protects.⁵⁰ For that reason an explicit invitation from the trial judge to treat the failure to testify as tending to prove guilt is barred as firmly as one from the prosecutor or from counsel for a codefendant.⁵¹

A state's own law authorizing the instruction does not make it lawful. In Chapman v. California the trial court charged the jury that it could draw adverse inferences from the petitioners' failure to testify, which the California Constitution then permitted the court and counsel to do.⁵² After Griffin, the California Supreme Court conceded that the petitioners had been denied a federal constitutional right, and the Court's own account of Griffin was that it held California's constitutional provision and practice invalid because they penalized the exercise of the privilege.⁵³ The privilege binds the states through the Fourteenth Amendment, so a state constitution or statute cannot supply what the Fifth Amendment withholds.

The contrast case is the neutral instruction. In Lakeside v. Oregon, as the Court recounted it in Carter v. Kentucky, the trial judge gave a no-inference instruction even though defense counsel's strategy was to avoid any mention of the client's failure to testify, and the petitioner argued that any judicial reference to the silence, without the defendant's agreement, invaded the privilege.⁵⁴ The Court rejected that argument, reasoning that the Fifth and Fourteenth Amendments bar only adverse comment and that an instruction telling jurors to draw no adverse inference of any kind is "comment" of an entirely different order.⁵⁵ Its purpose is "to remove from the jury's deliberations any influence of unspoken adverse inferences," so it cannot supply the pressure that Griffin condemned.⁵⁶ The Court also declined to build constitutional law on the speculation that jurors would not notice the silence or would ignore the instruction.⁵⁷ Both kinds of instruction mention the silence, and what separates them is what jurors are told to do with it: one authorizes them to count it against the defendant, and the other forbids them to. Students who treat any judicial mention of silence as a Griffin violation misread the rule, because the violation lies in the adverse use the instruction permits.

Not every instruction that touches on unexplained facts is an adverse instruction on silence. Instructions with some other primary function are often held permissible even though they may also draw the jury's attention to the defendant's failure to testify. The Supreme Court, for example, upheld an instruction permitting the jury to infer knowledge that property was stolen from unexplained possession of recently stolen property, observing without elaboration that the instruction could not fairly be understood as a comment on the failure to testify.⁵⁸ The test looks to the work the instruction does, so a charge violates the rule when its function is to treat the defendant's silence as proof.

The neutral instruction also raises a practical question that the next section answers. Defense lawyers frequently decline a no-inference instruction because they believe it does more harm than good by calling the jury's attention to the very silence the defendant wants ignored.⁵⁹ Whether a defendant who wants the instruction may demand it, as opposed to merely tolerating it, is the question in Carter v. Kentucky.