MR. JUSTICE REHNQUIST delivered the opinion of the Court.
This case presents the question whether the testimony of a witness in respondent's state court trial for rape must be excluded simply because police had learned the identity of the witness by questioning respondent at a time when he was in custody as a suspect, but had not been advised that counsel would be appointed for him if he was indigent. The questioning took place before this Court's decision in Miranda v. Arizona, 384 U.S. 436 (1966), but respondent's trial, at which he was convicted, took place afterwards. Under the holding of Johnson v. New Jersey, 384 U.S. 719 (1966), therefore, Miranda is applicable to this case. The United States District Court for the Eastern District of Michigan reviewed respondent's claim on a petition for habeas corpus and held that the testimony must be excluded.1 The Court of Appeals affirmed.2
On the morning of April 19, 1966, a 43-year-old woman in Pontiac, Michigan, was found in her home by a friend and co-worker, Luther White, in serious condition. At the time she was found the woman was tied, gagged, and partially disrobed, and had been both raped and severely beaten. She was unable to tell White anything about her assault at that time and still remains unable to recollect what happened.
While White was attempting to get medical help for the victim and to call for the police, he observed a dog inside the house. This apparently attracted White's attention for he knew that the woman did not own a dog
herself. Later, when talking with police officers, White observed the dog a second time, and police followed the dog to respondent's house. Neighbors further connected the dog with respondent.
The police then arrested respondent and brought him to the police station for questioning. Prior to the actual interrogation the police asked respondent whether he knew for what crime he had been arrested, whether he wanted an attorney, and whether he understood his constitutional rights.3 Respondent replied that he did understand the crime for which he was arrested, that he did not want an attorney, and that he understood his rights.4 The police further advised him that any statements he might make could be used against him at a later date in court.5 The police, however, did not advise respondent that he would be furnished counsel free of charge if he could not pay for such services himself.
The police then questioned respondent about his activities on the night of the rape and assault. Respondent replied that during the general time period at issue he had first been with one Robert Henderson and then later at home, alone, asleep. The police sought to confirm this story by contacting Henderson, but Henderson's story served to discredit rather than to bolster respondent's account. Henderson acknowledged that respondent had been with him on the night of the crime but said that he had left at a relatively early time. Furthermore, Henderson told police that he saw respondent the following day and asked him at that time about scratches on his face -- "asked him if he got hold of a wild one or something."6 Respondent answered: "Something like
that."7 Then, Henderson said, he asked respondent "who it was,"8 and respondent said: "Some woman lived the next block over,"9 adding: "She is a widow woman" or words to that effect.10
These events all occurred prior to the date on which this Court handed down its decision in Miranda v. Arizona, supra, but respondent's trial occurred afterwards. Prior to trial respondent's appointed counsel made a motion to exclude Henderson's expected testimony because respondent had revealed Henderson's identity without having received full Miranda warnings. Although respondent's own statements taken during interrogation were excluded, the trial judge denied the motion to exclude Henderson's testimony. Henderson therefore testified at trial, and respondent was convicted of rape and sentenced to 20 to 40 years' imprisonment. His conviction was affirmed by both the Michigan Court of Appeals11 and the Michigan Supreme Court.12
Respondent then sought habeas corpus relief in Federal District Court. That court, noting that respondent had not received the full Miranda warnings and that the police had stipulated Henderson's identity was learned only through respondent's answers, "reluctantly" concluded that Henderson's testimony could not be admitted.13 Application of such an exclusionary rule was necessary, the court reasoned, to protect respondent's Fifth Amendment right against compulsory self-incrimination. The court therefore granted respondent's petition for a writ of habeas corpus unless petitioner
retried respondent within 90 days. The Court of Appeals for the Sixth Circuit affirmed. We granted certiorari, 414 U.S. 1062 (1973), and now reverse.
Although respondent's sole complaint is that the police failed to advise him that he would be given free counsel if unable to afford counsel himself, he did not, and does not now, base his arguments for relief on a right to counsel under the Sixth and Fourteenth Amendments. Nor was the right to counsel, as such, considered to be persuasive by either federal court below. We do not have a situation such as that presented in Escobedo v. Illinois, 378 U.S. 478 (1964), where the policemen interrogating the suspect had refused his repeated requests to see his lawyer who was then present at the police station. As we have noted previously, Escobedo is not to be broadly extended beyond the facts of that particular case. See Johnson v. New Jersey, 384 U.S., at 733-734; Kirby v. Illinois, 406 U.S. 682, 689 (1972); Frazier v. Cupp, 394 U.S. 731, 739 (1969). This case also falls outside the rationale of United States v. Wade, 388 U.S. 218, 224 (1967), where the Court held that counsel was needed at a post-indictment lineup in order to protect the "right to a fair trial at which the witnesses against [the defendant] might be meaningfully cross-examined." Henderson was fully available for searching cross-examination at respondent's trial.
Respondent's argument, and the opinions of the District Court and Court of Appeals, instead rely upon the Fifth Amendment right against compulsory self-incrimination and the safeguards designed in Miranda to secure that right. In brief, the position urged upon this Court is that proper regard for the privilege against compulsory self-incrimination requires, with limited exceptions not
applicable here, that all evidence derived solely from statements made without full Miranda warnings be excluded at a subsequent criminal trial. For purposes of analysis in this case we believe that the question thus presented is best examined in two separate parts. We will therefore first consider whether the police conduct complained of directly infringed upon respondent's right against compulsory self-incrimination or whether it instead violated only the prophylactic rules developed to protect that right. We will then consider whether the evidence derived from this interrogation must be excluded.
The history of the Fifth Amendment right against compulsory self-incrimination, and the evils against which it was directed, have received considerable attention in the opinions of this Court. See, e. g., Kastigar v. United States, 406 U.S. 441 (1972); Miranda v. Arizona, supra; Murphy v. Waterfront Comm'n, 378 U.S. 52 (1964); Ullmann v. United States, 350 U.S. 422, 426 (1956); Counselman v. Hitchcock, 142 U.S. 547 (1892). At this point in our history virtually every schoolboy is familiar with the concept, if not the language, of the provision that reads: "No person . . . shall be compelled in any criminal case to be a witness against himself . . . ." This Court's decisions have referred to the right as "the mainstay of our adversary system of criminal justice," Johnson v. New Jersey, supra, at 729, and as "'one of the great landmarks in man's struggle to make himself civilized.'" Ullmann, supra, at 426. It is not surprising that the constitution of virtually every State has a comparable provision. 8 J. Wigmore, Evidence § 2252 (McNaughton rev. 1961) (hereinafter Wigmore).
The importance of a right does not, by itself, determine its scope, and therefore we must continue to hark back
to the historical origins of the privilege, particularly the evils at which it was to strike. The privilege against compulsory self-incrimination was developed by painful opposition to a course of ecclesiastical inquisitions and Star Chamber proceedings occurring several centuries ago. See L. Levy, Origins of the Fifth Amendment (1968); Morgan, The Privilege Against Self-Incrimination, 34 Minn. L. Rev. 1 (1949); 8 Wigmore § 2250. Certainly anyone who reads accounts of those investigations, which placed a premium on compelling subjects of the investigation to admit guilt from their own lips, cannot help but be sensitive to the Framers' desire to protect citizens against such compulsion. As this Court has noted, the privilege against self-incrimination "was aimed at a . . . far-reaching evil -- a recurrence of the Inquisition and the Star Chamber, even if not in their stark brutality." Ullmann, supra, at 428.
Where there has been genuine compulsion of testimony, the right has been given broad scope. Although the constitutional language in which the privilege is cast might be construed to apply only to situations in which the prosecution seeks to call a defendant to testify against himself at his criminal trial, its application has not been so limited. The right has been held applicable to proceedings before a grand jury, Counselman v. Hitchcock, supra ; to civil proceedings, McCarthy v. Arndstein, 266 U.S. 34 (1924); to congressional investigations, Watkins v. United States, 354 U.S. 178 (1957); to juvenile proceedings, In re Gault, 387 U.S. 1 (1967); and to other statutory inquiries, Malloy v. Hogan, 378 U.S. 1 (1964). The privilege has also been applied against the States by virtue of the Fourteenth Amendment. Ibid.
The natural concern which underlies many of these decisions is that an inability to protect the right at
one stage of a proceeding may make its invocation useless at a later stage. For example, a defendant's right not to be compelled to testify against himself at his own trial might be practically nullified if the prosecution could previously have required him to give evidence against himself before a grand jury. Testimony obtained in civil suits, or before administrative or legislative committees, could also prove so incriminating that a person compelled to give such testimony might readily be convicted on the basis of those disclosures in a subsequent criminal proceeding.14
In more recent years this concern -- that compelled disclosures might be used against a person at a later criminal trial -- has been extended to cases involving police interrogation. Before Miranda the principal issue in these cases was not whether a defendant had waived his privilege against compulsory self-incrimination but simply whether his statement was "voluntary." In state cases the Court applied the Due Process Clause of the Fourteenth Amendment, examining the circumstances of interrogation to determine whether the processes were so unfair or unreasonable as to render a subsequent confession involuntary. See, e. g., Brown v. Mississippi, 297 U.S. 278 (1936); Chambers v. Florida, 309 U.S. 227 (1940); White v. Texas, 310 U.S. 530 (1940); Payne v. Arkansas, 356 U.S. 560 (1958); Haynes v. Washington, 373 U.S. 503 (1963). See also 3 J. Wigmore, Evidence § 815 et seq. (Chadbourne rev. 1970). Where the State's actions offended the standards of fundamental fairness under the Due Process Clause, the State was then deprived of the right to use the resulting confessions in court.
Although federal cases concerning voluntary confessions often contained references to the privilege against compulsory self-incrimination,15 references which were strongly criticized by some commentators, see 8 Wigmore § 2266,16 it was not until this Court's decision in Miranda that the privilege against compulsory self-incrimination was seen as the principal protection for a person facing police interrogation. This privilege had been made applicable to the States in Malloy v. Hogan, supra, and was thought to offer a more comprehensive and
less subjective protection than the doctrine of previous cases. In Miranda the Court examined the facts of four separate cases and stated:
"In these cases, we might not find the defendants' statements to have been involuntary in traditional terms. Our concern for adequate safeguards to protect precious Fifth Amendment rights is, of course, not lessened in the slightest . . . . To be sure, the records do not evince overt physical coercion or patent psychological ploys. The fact remains that in none of these cases did the officers undertake to afford appropriate safeguards at the outset of the interrogation to insure that the statements were truly the product of free choice." 384 U.S., at 457.
Thus the Court in Miranda, for the first time, expressly declared that the Self-Incrimination Clause was applicable to state interrogations at a police station, and that a defendant's statements might be excluded at trial despite their voluntary character under traditional principles.
To supplement this new doctrine, and to help police officers conduct interrogations without facing a continued risk that valuable evidence would be lost, the Court in Miranda established a set of specific protective guidelines, now commonly known as the Miranda rules. The Court declared that "the prosecution may not use statements, whether exculpatory or inculpatory, stemming from custodial interrogation of the defendant unless it demonstrates the use of procedural safeguards effective to secure the privilege against self-incrimination." Id., at 444. A series of recommended "procedural safeguards" then followed. The Court in particular stated:
"Prior to any questioning, the person must be warned that he has a right to remain silent, that any statement he does make may be used as evidence
against him, and that he has a right to the presence of an attorney, either retained or appointed." Ibid.
The Court said that the defendant, of course, could waive these rights, but that any waiver must have been made "voluntarily, knowingly and intelligently." Ibid.
The Court recognized that these procedural safeguards were not themselves rights protected by the Constitution but were instead measures to insure that the right against compulsory self-incrimination was protected. As the Court remarked:
"We cannot say that the Constitution necessarily requires adherence to any particular solution for the inherent compulsions of the interrogation process as it is presently conducted." Id., at 467.
The suggested safeguards were not intended to "create a constitutional straitjacket," ibid., but rather to provide practical reinforcement for the right against compulsory self-incrimination.
A comparison of the facts in this case with the historical circumstances underlying the privilege against compulsory self-incrimination strongly indicates that the police conduct here did not deprive respondent of his privilege against compulsory self-incrimination as such, but rather failed to make available to him the full measure of procedural safeguards associated with that right since Miranda. Certainly no one could contend that the interrogation faced by respondent bore any resemblance to the historical practices at which the right against compulsory self-incrimination was aimed. The District Court in this case noted that the police had "warned [respondent] that he had the right to remain silent," 352 F.Supp. 266, 267 (1972), and the record in this case clearly shows that respondent was informed that any evidence taken could be used against him.17 The record is also clear that
respondent was asked whether he wanted an attorney and that he replied that he did not.18 Thus, his statements could hardly be termed involuntary as that term has been defined in the decisions of this Court. Additionally, there were no legal sanctions, such as the threat of contempt, which could have been applied to respondent had he chosen to remain silent. He was simply not exposed to "the cruel trilemma of self-accusation, perjury or contempt." Murphy v. Waterfront Comm'n, 378 U.S., at 55.
Our determination that the interrogation in this case involved no compulsion sufficient to breach the right against compulsory self-incrimination does not mean there was not a disregard, albeit an inadvertent disregard, of the procedural rules later established in Miranda. The question for decision is how sweeping the judicially imposed consequences of this disregard shall be. This Court said in Miranda that statements taken in violation of the Miranda principles must not be used to prove the prosecution's case at trial. That requirement was fully complied with by the state court here: respondent's statements, claiming that he was with Henderson and then asleep during the time period of the crime were not admitted against him at trial. This Court has also said, in Wong Sun v. United States, 371 U.S. 471 (1963), that the "fruits" of police conduct which actually infringed a defendant's Fourth Amendment rights must be suppressed.19 But we have already concluded that the
police conduct at issue here did not abridge respondent's constitutional privilege against compulsory self-incrimination, but departed only from the prophylactic standards later laid down by this Court in Miranda to safeguard that privilege. Thus, in deciding whether Henderson's testimony must be excluded, there is no controlling precedent of this Court to guide us. We must therefore examine the matter as a question of principle.
Just as the law does not require that a defendant receive a perfect trial, only a fair one, it cannot realistically require that policemen investigating serious crimes make no errors whatsoever. The pressures of law enforcement and the vagaries of human nature would make such an expectation unrealistic. Before we penalize police error, therefore, we must consider whether the sanction serves a valid and useful purpose.
We have recently said, in a search-and-seizure context, that the exclusionary rule's "prime purpose is to deter future unlawful police conduct and thereby effectuate the guarantee of the Fourth Amendment against unreasonable searches and seizures." United States v. Calandra, 414 U.S. 338, 347 (1974). We then continued:
"'The rule is calculated to prevent, not to repair. Its purpose is to deter -- to compel respect for the constitutional guaranty in the only effectively available way -- by removing the incentive to disregard it.' Elkins v. United States, 364 U.S. 206, 217 (1960)."20 Ibid.
In a proper case this rationale would seem applicable to the Fifth Amendment context as well.
The deterrent purpose of the exclusionary rule necessarily assumes that the police have engaged in willful, or at the very least negligent, conduct which has deprived the defendant of some right. By refusing to admit evidence gained as a result of such conduct, the courts hope to instill in those particular investigating officers, or in their future counterparts, a greater degree of care toward the rights of an accused. Where the official action was pursued in complete good faith, however, the deterrence rationale loses much of its force.
We consider it significant to our decision in this case that the officers' failure to advise respondent of his right to appointed counsel occurred prior to the decision in Miranda. Although we have been urged to resolve the broad question of whether evidence derived from statements taken in violation of the Miranda rules must be excluded regardless of when the interrogation took place,21 we instead place our holding on a narrower ground. For at the time respondent was questioned these police officers were guided, quite rightly, by the principles established in Escobedo v. Illinois, 378 U.S. 478 (1964), particularly focusing on the suspect's opportunity to have retained counsel with him during the interrogation if he chose to do so.22 Thus, the police asked respondent if he wanted counsel, and he answered that he did not. The
statements actually made by respondent to the police, as we have observed, were excluded at trial in accordance with Johnson v. New Jersey, 384 U.S. 719 (1966). Whatever deterrent effect on future police conduct the exclusion of those statements may have had, we do not believe it would be significantly augmented by excluding the testimony of the witness Henderson as well.
When involuntary statements or the right against compulsory self-incrimination are involved, a second justification for the exclusionary rule also has been asserted: protection of the courts from reliance on untrustworthy evidence.23 Cases which involve the Self-Incrimination Clause must, by definition, involve an element of coercion, since the Clause provides only that a person shall not be compelled to give evidence against himself. And cases involving statements often depict severe pressures which may override a particular suspect's insistence on innocence. Fact situations ranging from classical third-degree torture, Brown v. Mississippi, 297 U.S. 278 (1936), to prolonged isolation from family or friends in a hostile setting, Gallegos v. Colorado, 370 U.S. 49 (1962), or to a simple desire on the part of a physically or mentally exhausted
suspect to have a seemingly endless interrogation end, Watts v. Indiana, 338 U.S. 49 (1949), all might be sufficient to cause a defendant to accuse himself falsely.
But those situations are a far cry from that presented here. The pressures on respondent to accuse himself were hardly comparable even with the least prejudicial of those pressures which have been dealt with in our cases. More important, the respondent did not accuse himself. The evidence which the prosecution successfully sought to introduce was not a confession of guilt by respondent, or indeed even an exculpatory statement by respondent, but rather the testimony of a third party who was subjected to no custodial pressures. There is plainly no reason to believe that Henderson's testimony is untrustworthy simply because respondent was not advised of his right to appointed counsel. Henderson was both available at trial and subject to cross-examination by respondent's counsel, and counsel fully used this opportunity, suggesting in the course of his cross-examination that Henderson's character was less than exemplary and that he had been offered incentives by the police to testify against respondent.24 Thus the reliability of his testimony was subject to the normal testing process of an adversary trial.
Respondent contends that an additional reason for excluding Henderson's testimony is the notion that the adversary system requires "the government in its contest with the individual to shoulder the entire load." 8 Wigmore § 2251, p. 317; Murphy v. Waterfront Comm'n, 378 U.S., at 55; Miranda v. Arizona, 384 U.S., at 460. To the extent that this suggested basis for the exclusionary rule in Fifth Amendment cases may exist independently of the deterrence and trustworthiness rationales, we think it of no avail to respondent here. Subject
to applicable constitutional limitations, the Government is not forbidden all resort to the defendant to make out its case. It may require the defendant to give physical evidence against himself, see Schmerber v. California, 384 U.S. 757 (1966); United States v. Dionisio, 410 U.S. 1 (1973), and it may use statements which are voluntarily given by the defendant after he receives full disclosure of the rights offered by Miranda. Here we deal, not with the offer of respondent's own statements in evidence, but only with the testimony of a witness whom the police discovered as a result of respondent's statements. This recourse to respondent's voluntary statements does no violence to such elements of the adversary system as may be embodied in the Fifth, Sixth, and Fourteenth Amendments.
In summary, we do not think that any single reason supporting exclusion of this witness' testimony, or all of them together, are very persuasive.25 By contrast, we find the arguments in favor of admitting the testimony quite strong. For, when balancing the interests involved, we must weigh the strong interest under any system of justice of making available to the trier of fact all concededly relevant and trustworthy evidence which either party seeks to adduce. In this particular case we also "must consider society's interest in the effective prosecution of criminals in light of the protection our pre-Miranda standards afford criminal defendants." Jenkins
v. Delaware, 395 U.S. 213, 221 (1969). These interests may be outweighed by the need to provide an effective sanction to a constitutional right, Weeks v. United States, 232 U.S. 383 (1914), but they must in any event be valued. Here respondent's own statement, which might have helped the prosecution show respondent's guilty conscience at trial, had already been excised from the prosecution's case pursuant to this Court's Johnson decision. To extend the excision further under the circumstances of this case and exclude relevant testimony of a third-party witness would require far more persuasive arguments than those advanced by respondent.
This Court has already recognized that a failure to give interrogated suspects full Miranda warnings does not entitle the suspect to insist that statements made by him be excluded in every conceivable context. In Harris v. New York, 401 U.S. 222 (1971), the Court was faced with the question of whether the statements of the defendant himself, taken without informing him of his right of access to appointed counsel, could be used to impeach defendant's direct testimony at trial. The Court concluded that they could, saying:
"Some comments in the Miranda opinion can indeed be read as indicating a bar to use of an uncounseled statement for any purpose, but discussion of that issue was not at all necessary to the Court's holding and cannot be regarded as controlling. Miranda barred the prosecution from making its case with statements of an accused made while in custody prior to having or effectively waiving counsel. It does not follow from Miranda that evidence inadmissible against an accused in the prosecution's case in chief is barred for all purposes, provided of course that the trustworthiness of the evidence satisfies legal standards." Id., at 224.
We believe that this reasoning is equally applicable here. Although Johnson enabled respondent to block admission of his own statements, we do not believe that it requires the prosecution to refrain from all use of those statements, and we disagree with the courts below that Henderson's testimony should have been excluded in this case.26
480 F.2d 927, reversed.
MR. JUSTICE STEWART, concurring.
In joining the opinion of the Court, I add only that I could also join MR. JUSTICE BRENNAN's concurrence. For it seems to me that despite differences in phraseology, and despite the disclaimers of their respective authors, the Court opinion and that of MR. JUSTICE BRENNAN proceed along virtually parallel lines, give or take a couple of argumentative footnotes.
MR. JUSTICE BRENNAN, with whom MR. JUSTICE MARSHALL joins, concurring in the judgment.
The Court finds it unnecessary to decide "the broad question" of whether the fruits of "statements taken in violation of the Miranda rules must be excluded regardless of when the interrogation took place," ante, at 447, since respondent's interrogation occurred prior to our decision in Miranda v. Arizona, 384 U.S. 436 (1966). In my view, however, it is unnecessary, too, for the Court to address the narrower question of whether the principles of Miranda require that fruits be excluded when obtained as a result of a pre-Miranda interrogation without the requisite prior warnings. The Court, in answering this question, proceeds from the premise that Johnson v. New Jersey, 384 U.S. 719 (1966), makes Miranda applicable to all cases in which a criminal trial was commenced after the date of our decision in Miranda,
and that, since respondent's trial was post-Miranda, the effect of Miranda on this case must be resolved. I would not read Johnson as making Miranda applicable to this case.1
Frank acknowledgment that retroactive application of newly announced constitutional rules of criminal procedure may have a serious impact on the administration of criminal justice has led us, since Linkletter v. Walker, 381 U.S. 618 (1965), to determine retroactivity in terms of three criteria: (1) the purpose served by the new rules; (2) the extent of law enforcement officials' justifiable reliance on prior standards; and (3) the effect on the administration of justice of a retroactive application of the new rules. See, e. g., Michigan v. Payne, 412 U.S. 47, 51 (1973); Stovall v. Denno, 388 U.S. 293, 297 (1967); Tehan v. United States ex rel. Shott, 382 U.S. 406, 410-418 (1966). We have as a general matter limited our discussion of the relevant "purpose" of new rules to their functional value in enhancing the reliability of the factfinding process. See, e. g., Williams v. United States, 401 U.S. 646, 653 (1971); id., at 663 (concurring opinion); Desist v. United States, 394 U.S. 244, 249-250 (1969); Roberts v. Russell, 392 U.S. 293, 294 (1968); Tehan v. United States ex rel. Shott, supra; Linkletter v. Walker, supra, at 638-639. This limiting approach has been taken in recognition that "the basic purpose of a trial is the determination of truth," Tehan v. United States ex rel. Shott, supra, at 416; see Stovall v. Denno, supra, at 297-298, and that the principal legitimate interest of a convicted defendant is therefore assurance
that the factfinding process at his trial was not unduly impaired by adherence to the old standards.
In Johnson v. New Jersey, supra, the Court was called upon to determine whether the newly announced procedures in Miranda v. Arizona should be retroactively applied to upset final convictions based in part upon confessions obtained without the prior warnings required by Miranda. Aware that Miranda provided new safeguards against the possible use at trial of unreliable statements of the accused, we nonetheless concluded that the decision should not be retroactively applied.2 The probability
that the truth-determining process was distorted by, and individuals were convicted on the basis of, coerced confessions was minimized, we found, by the availability of strict pre-Miranda standards to test the voluntariness of confessions. 384 U.S., at 730. In addition, we recognized that law enforcement agencies had justifiably relied on our prior rulings and that retroactive application would necessitate the wholesale release and subsequent retrial of vast numbers of prisoners. Id., at 731. Then, in statements unnecessary to our decision -- since all of the convictions of the petitioners in Johnson had long since become final at the time of our decision in Miranda -- we went on to say that our newly announced Miranda rules should be applied to trials begun after the date that decision was announced. Id., at 732.
The conclusion that the Miranda rules should be applied to post-Miranda trials made good sense, where criminal defendants were seeking to exclude direct statements made without prior warning of their rights. Exclusion of possibly unreliable pre-Miranda statements made in the inherently coercive atmosphere of in-custody interrogation, see Miranda v. Arizona, 384 U.S., at 457-458, 467, 470, could be obtained at a relatively low cost. For, although the police might have relied in good faith on our prior rulings in interrogating defendants without first advising them of their rights, Miranda put the police on notice that pre-Miranda confessions obtained without prior warnings would be inadmissible at defendants' trials.
Since defendants who had made pre-Miranda confessions had not yet gone to trial, and the police investigations into those cases were still fresh, Johnson envisioned "no undue burden [being] imposed upon prosecuting authorities by requiring them to find evidentiary substitutes for statements obtained in violation of the constitutional protections afforded by Miranda." Jenkins v. Delaware, 395 U.S. 213, 219-220 (1969); see Johnson v. New Jersey, 384 U.S., at 732.
Application of the Miranda standards to the present case, however, presents entirely different problems. Unlike the situation contemplated in Johnson, the burden imposed upon law enforcement officials to obtain evidentiary substitutes for inadmissible "fruits" will likely be substantial. The lower courts, confronted with the question of the application of Miranda to fruits, have provided differing answers on the admissibility issue.3 The police, therefore, could not reasonably have been expected to know that substitute evidence would be necessary. As a result, in a case such as the present one, in which law enforcement officials have relied on trial and appellate court determinations that fruits are admissible, a contrary ruling by this Court, coming years after the commission of the crime, would severely handicap any attempt to retry the defendant. The burden on law enforcement officers, in that circumstance, would be comparable to that in Jenkins v. Delaware, supra, where we declined to apply the Miranda rules to post-Miranda retrials of persons whose original trials were commenced prior to Miranda. There, we said:
"Concern for the justifiable reliance of law enforcement
officials upon pre-Miranda standards militates against applying Miranda to retrials . . . . As we stated in Stovall [v. Denno, supra ], 'Inquiry would be handicapped by the unavailability of witnesses and dim memories.' 388 U.S., at 300. The burden would be particularly onerous where an investigation was closed years prior to a retrial because law enforcement officials relied in good faith upon a strongly incriminating statement, admissible at the first trial, to provide the cornerstone of the prosecution's case." 395 U.S., at 220 (footnote omitted).
Moreover, the element of unreliability -- a legitimate concern in Johnson because of the inherently coercive nature of in-custody interrogation -- is of less importance when the admissibility of "fruits" is at issue. There is no reason to believe that the coercive atmosphere of the station house will have any effect whatsoever on the trustworthiness of "fruits."
Since excluding the fruits of respondent's statements would not further the integrity of the factfinding process and would severely handicap law enforcement officials in obtaining evidentiary substitutes, I would confine the reach of Johnson v. New Jersey to those cases in which the direct statements of an accused made during a pre-Miranda interrogation were introduced at his post-Miranda trial. If Miranda is applicable at all to the fruits of statements made without proper warnings, I would limit its effect to those cases in which the fruits were obtained as a result of post-Miranda interrogations. Cf. Stovall v. Denno, 388 U.S. 293 (1967); Desist v. United States, 394 U.S. 244 (1969).4
Since I agree that the judgment of the Court of Appeals must be reversed, I concur in the judgment of the Court.5
MR. JUSTICE WHITE, concurring in the judgment.
For the reasons stated in my dissent in that case, I continue to think that Miranda v. Arizona, 384 U.S. 436 (1966), was ill-conceived and without warrant in the Constitution. However that may be, the Miranda opinion did not deal with the admissibility of evidence derived from in-custody admissions obtained without the specified warnings, and the matter has not been settled by subsequent cases.
In Orozco v. Texas, 394 U.S. 324 (1969), it appeared that petitioner, who was convicted of murder, had been arrested and interrogated in his home without the benefit of Miranda warnings. Among other things, petitioner admitted having a gun and told the police where it was hidden in the house. The gun was recovered and ballistic tests, which were admitted into evidence along with various oral admissions, showed that it was the gun involved in the murder. Petitioner's conviction was affirmed, the applicability of Miranda being rejected by the state courts. Petitioner brought the case here, urging in his petition for certiorari, which was granted, that the ballistic evidence was a fruit of an illegal interrogation -- "the direct product of interrogation" without indispensable constitutional safeguards. His brief on the merits suggested that it was error under Miranda to admit into evidence either his oral admissions or the evidence of ballistic tests performed on the pistol, which
was referred to as "an illegally seized object." This Court reversed the conviction but after referring to the ballistic evidence, went on to hold only that the admission into evidence of Orozco's statements made without benefit of Miranda warnings was fatal error. Although the issue was presented, the Court did not expressly deal with the admissibility of the ballistic tests and gave no intimation that the evidence was to be excluded at the anticipated retrial.
Miranda having been applied in this Court only to the exclusion of the defendant's own statements, I would not extend its prophylactic scope to bar the testimony of third persons even though they have been identified by means of admissions that are themselves inadmissible under Miranda. The arguable benefits from excluding such testimony by way of possibly deterring police conduct that might compel admissions are, in my view, far outweighed by the advantages of having relevant and probative testimony, not obtained by actual coercion, available at criminal trials to aid in the pursuit of truth. The same results would not necessarily obtain with respect to the fruits of involuntary confessions. I therefore concur in the judgment.
MR. JUSTICE DOUGLAS, dissenting.
In this case the respondent, incarcerated as a result of a conviction in a state court, was granted a writ of habeas corpus by the District Court. The basis for the writ was the introduction at respondent's trial of testimony from a witness whose identity was learned solely as a result of in-custody police interrogation of the respondent preceded by warnings which were deficient under the standards enunciated in Miranda v. Arizona, 384 U.S. 436 (1966). The District Court concluded that "the introduction by the prosecution in its case in chief of testimony of a third
person which is admittedly the fruit of an illegally obtained statement by the [accused violates the accused's] Fifth Amendment rights." 352 F.Supp. 266, 268 (ED Mich. 1972). The Court of Appeals affirmed. 480 F.2d 927 (CA6 1973).
Prior to interrogation, respondent was told of his right to the presence of counsel but he was not told of his right to have an attorney appointed should he be unable to afford one. Respondent is an indigent who has been represented at all times in both state and federal courts by court-appointed counsel. In Miranda, supra, we said:
"The need for counsel in order to protect the privilege [against self-incrimination] exists for the indigent as well as the affluent. . . . While authorities are not required to relieve the accused of his poverty, they have the obligation not to take advantage of indigence in the administration of justice. . . .
"In order to fully apprise a person interrogated of the extent of his rights under this system then, it is necessary to warn him not only that he has the right to consult with an attorney, but also that if he is indigent a lawyer will be appointed to represent him." 384 U.S., at 472-473.
I cannot agree when the Court says that the interrogation here "did not abridge respondent's constitutional privilege against compulsory self-incrimination, but departed only from the prophylactic standards later laid down by this Court in Miranda to safeguard that privilege." Ante, at 446. The Court is not free to prescribe preferred modes of interrogation absent a constitutional basis. We held the "requirement of warnings and waiver of rights [to be] fundamental with respect to the Fifth Amendment privilege," 384 U.S., at 476, and without
so holding we would have been powerless to reverse Miranda's conviction. While Miranda recognized that police need not mouth the precise words contained in the Court's opinion, such warnings were held necessary "unless other fully effective means are adopted to notify the person" of his rights. Id., at 479. There is no contention here that other means were adopted. The respondent's statements were thus obtained "under circumstances that did not meet constitutional standards for protection of the privilege [against self-incrimination]." Id., at 491 (emphasis added).
With the premise that respondent was subjected to an unconstitutional interrogation, there remains the question whether not only the testimony elicited in the interrogation but also the fruits thereof must be suppressed. Mr. Justice Holmes first articulated the "fruits" doctrine in Silverthorne Lumber Co. v. United States, 251 U.S. 385 (1920). In that case the Government had illegally seized the petitioner's corporate books and documents. The Government photographed the items before returning them and used the photographs as a basis to subpoena the petitioner to produce the originals before the grand jury. The petitioner refused to comply and was cited for contempt. In reversing, the Court noted that "the essence of a provision forbidding the acquisition of evidence in a certain way is that not merely evidence so acquired shall not be used before the Court but that it shall not be used at all." Id., at 392.
The principle received more recent recognition in Wong Sun v. United States, 371 U.S. 471 (1963). There one Toy had made statements to federal agents and the statements were held inadmissible against him. The statements led the agents to one Yee and at Yee's home
the agents found narcotics which were introduced at trial against Toy. In reversing Toy's conviction the Court held that the narcotics discovered at Yee's home must be excluded just as Toy's statements which led to that discovery.
The testimony of the witness in this case was no less a fruit of unconstitutional police action than the photographs in Silverthorne or the narcotics in Wong Sun. The petitioner has stipulated that the identity and the whereabouts of the witness and his connection with the case were learned about only through the unconstitutional interrogation of the respondent. His testimony must be excluded to comply with Miranda 's mandate that " no evidence obtained as a result of interrogation [not preceded by adequate warnings] can be used against" an accused. 384 U.S., at 479 (emphasis added).
In Johnson v. New Jersey, 384 U.S. 719 (1966), the Court held that statements obtained in violation of Miranda standards must be excluded from all trials occurring after the date of the Miranda decision. MR. JUSTICE BRENNAN suggests that Johnson be limited and that the fruits derived from unlawful pre-Miranda interrogations be admissible in trials subsequent to the Miranda decision. Though respondent's trial occurred subsequent to the Miranda decision, his interrogation preceded it. I disagree, as I disagreed in Johnson, that any defendant can be deprived of the full protection of the Fifth Amendment, as the Court has construed it in Miranda, based upon an arbitrary reference to the date of his interrogation or his trial.
In Linkletter v. Walker, 381 U.S. 618 (1965), the Court held the exclusionary rule of Mapp v. Ohio, 367 U.S. 643 (1961), inapplicable to convictions which had become "final" prior to the Mapp decision. As Mr.
Justice Black, joined by me, noted, the result was as follows:
"Linkletter, convicted in the state court by use of 'unconstitutional evidence,' is today denied relief by the judgment of this Court because his conviction became 'final' before Mapp was decided. Linkletter must stay in jail; Miss Mapp, whose offense was committed before Linkletter's, is free. This different treatment of Miss Mapp and Linkletter points up at once the arbitrary and discriminatory nature of the judicial contrivance utilized here to break the promise of Mapp by keeping all people in jail who are unfortunate enough to have had their unconstitutional convictions affirmed before June 19, 1961." 381 U.S., at 641 (dissenting opinion).
I find any such reference to the calendar in determining the beneficiaries of constitutional pronouncements to be a grossly invidious discrimination. Miranda was interrogated on March 13, 1963; Tucker was interrogated more than three years later in April 1966. I can conceive of no principled way to deprive Tucker of the constitutional guarantees afforded Miranda. The reason put forward for refusing to apply the strictures of Miranda to interrogations which preceded the decision is that the purpose of Miranda 's rules is the deterrence of unconstitutional interrogation. " The inference I gather from these repeated statements is that the rule is not a right or privilege accorded to defendants charged with crime but is a sort of punishment against officers in order to keep them from depriving people of their constitutional rights. In passing I would say that if that is the sole purpose, reason, object and effect of the rule, the Court's action in adopting it sounds more like lawmaking than construing the Constitution." 381 U.S., at 649 (Black, J., dissenting). Miranda 's purpose was
not promulgation of judicially preferred standards for police interrogation, a function we are quite powerless to perform; the decision enunciated " constitutional standards for protection of the privilege" against self-incrimination. 384 U.S., at 491. People who are in jail because of a State's use of unconstitutionally derived evidence are entitled to a new trial, with the safeguards the Constitution provides, without regard to when the constitutional violation occurred, when the trial occurred, or when the conviction became "final."
As Mr. Justice Black said in Linkletter : "It certainly offends my sense of justice to say that a State holding in jail people who were convicted by unconstitutional methods has a vested interest in keeping them there that outweighs the right of persons adjudged guilty of crime to challenge their unconstitutional convictions at any time." 381 U.S., at 653.
I would affirm the judgment below.
* Briefs of amici curiae urging reversal were filed by Evelle J. Younger, Attorney General, Jack R. Winkler, Chief Assistant Attorney General, and Doris H. Maier, Assistant Attorney General, for the State of California, and by Frank G. Carrington, Jr., Fred E. Inbau, William K. Lambie, Wayne W. Schmidt, Glen Murphy, Paul Keller, and James B. Zagel for Americans for Effective Law Enforcement et al.
The Civil Liberties Committee, State Bar of Michigan, filed a brief as amicus curiae urging affirmance.
1. 352 F.Supp. 266 (1972).
2. 480 F.2d 927 (1973).
3. Tr. of Prelim. Hearing 99.
5. Id., at 99-100.
6. Tr. of Trial 223.
8. Id., at 224.
11. 19 Mich. App. 320, 172 N. W. 2d 712 (1969).
12. 385 Mich. 594, 189 N. W. 2d 290 (1971).
13. 352 F.Supp., at 268.
14. The Court has also held that comment on a defendant's silence or refusal to take the witness stand may be an impermissible penalty on exercise of the privilege. See Griffin v. California, 380 U.S. 609 (1965).
15. For example in Bram v. United States, 168 U.S. 532, 542 (1897), the Court stated: "In criminal trials, in the courts of the United States, wherever a question arises whether a confession is incompetent because not voluntary, the issue is controlled by that portion of the Fifth Amendment to the Constitution of the United States, commanding that no person 'shall be compelled in any criminal case to be a witness against himself.'" As noted in the text, the privilege against compulsory self-incrimination was not held applicable against the States until Malloy v. Hogan, 378 U.S. 1 (1964).
16. Wigmore states his objection in the following terms: "Today in the United States confessions, and probably even lesser self-incriminating admissions, are excluded despite their trustworthiness if coerced. The policies leading to this recent extension of the confession rule are quite similar to those underlying the privilege against self-incrimination. It is thus not surprising that the privilege, with its unclear boundaries and apparently unending capacity for transmogrification and assimilation, is now sometimes invoked to effect exclusion even though the disclosure was not compelled from a person under legal compulsion. Distortion of the privilege to cover such situations is not necessary. If trustworthy confessions are to be excluded because coerced, it should be done frankly as an exception to the principle . . . that the illegality of source of evidence is immaterial. It should be done, as it usually is, on the ground that the combination of coercion and use of the evidence in the particular case violates the relevant constitutional due process clause." Id., at 402. (Citations omitted.)
17. See n. 5, supra.
18. See nn. 3 and 4, supra.
19. In Wong Sun the police discovered evidence through statements made by the accused after he had been placed under arrest. This Court, finding that the arrest had occurred without probable cause, held that the derivative evidence could not be introduced against the accused at trial. For the reasons stated in the text we do not believe that Wong Sun controls the case before us.
20. The opinion also relied upon Mapp v. Ohio, 367 U.S. 643, 656 (1961); Tehan v. United States ex rel. Shott, 382 U.S. 406, 416 (1966); and Terry v. Ohio, 392 U.S. 1, 29 (1968). See 414 U.S., at 348.
21. Brief for United States as Amicus Curiae 31 et seq.; Brief for Respondent 9 et seq.
22. As previously noted, the defendant in Escobedo had repeatedly asked to see his lawyer who was available at the police station. Those requests were denied, and the defendant ultimately confessed. Thus, in direct contrast to the situation here, the defendant in Escobedo was told he did not have a right to see his lawyer, although he had expressly stated his desire to do so.
23. The Court has made clear that the truth or falsity of a statement is not the determining factor in the decision whether or not to exclude it. Jackson v. Denno, 378 U.S. 368 (1964). Thus a State which has obtained a coerced or involuntary statement cannot argue for its admissibility on the ground that other evidence demonstrates its truthfulness. Ibid. But it also seems clear that coerced statements have been regarded with some mistrust. The Court in Escobedo, for example, stated that "a system of criminal law enforcement which comes to depend on the 'confession' will, in the long run, be less reliable and more subject to abuses" than a system relying on independent investigation, 378 U.S., at 488-489. The Court then cited several authorities concerned with false confessions. Id., at 489 n. 11. Although completely voluntary confessions may, in many cases, advance the cause of justice and rehabilitation, coerced confessions, by their nature, cannot serve the same ends.
24. Tr. of Trial 226-234.
25. It has been suggested that courts should exclude evidence derived from "lawless invasions of the constitutional rights of citizens," Terry v. Ohio, 392 U.S., at 13, in recognition of "the imperative of judicial integrity." Elkins v. United States, 364 U.S. 206, 222 (1960). This rationale, however, is really an assimilation of the more specific rationales discussed in the text of this opinion, and does not in their absence provide an independent basis for excluding challenged evidence.
26. Our Brother BRENNAN in his opinion concurring in the judgment treats the principal question here simply as a lineal descendant of the one decided in Linkletter v. Walker, 381 U.S. 618 (1965), to be analyzed only in terms of the retroactivity framework established in that and subsequent decisions. While his approach has a beguiling simplicity, we believe it marks a significant and unsettling departure from the past practice of the Court in this area. Our retroactivity cases, from Linkletter v. Walker, supra, to Gosa v. Mayden, 413 U.S. 665 (1973), all have in common a particular factual predicate: a previous constitutional decision of this Court governs the facts of an earlier decided case unless the constitutional decision is not to have retroactive effect. The doctrine of retroactivity does not modify the substantive scope of the constitutional decision but rather determines the point in time when it is held to apply. That common factual predicate is absent here. No defendant in Miranda sought to block evidence of the type challenged in this case, and the holding of Miranda, even if made fully retroactive, would not therefore resolve the question of whether Henderson's testimony must also be excluded at trial. Contrary, therefore, to the suggestion in our Brother's opinion that the question here is whether to "limit the effect of Johnson v. New Jersey," post, at 454 n. 1, Johnson has never been thought controlling on the question of fruits, for the simple reason that the parent Miranda case did not reach that issue. Our Brother BRENNAN's method of disposition is to determine in the present case the retroactivity of a holding which the Court has yet to make. He would say, in effect, that if the Court should later determine that Miranda requires exclusion of fruits such as the testimony of Henderson, nonetheless that determination shall not be applied retroactively. But this approach wholly subverts the heretofore established relationship between the parent case and the subsidiary case determining whether or not to apply the parent case retroactively. Under the framework of the analysis established in Linkletter, supra, and in subsequent cases, it would seem indispensable to understand the basis for a constitutional holding of the Court in order to later determine whether that holding should be retroactive. Yet ex hypothesi our Brother has no such analysis available, since the case has yet to be decided. Cases which subsequently determine the retroactivity of a constitutional holding have given the Court enough occasion for concern without substantially increasing the difficulty of that type of decision by making it before, rather than after, the constitutional holding.
1. Although the petition for certiorari did not urge us to limit the effect of Johnson v. New Jersey, this issue was raised in petitioner's brief as well as in the amicus curiae brief of the State of California, filed in support of petitioner. See Mapp v. Ohio, 367 U.S. 643, 646 n. 3 (1961); Stovall v. Denno, 388 U.S. 293, 294 n. 1 (1967).
2. In Johnson we commented -- as we have on a number of occasions in deciding to apply new constitutional rules of criminal procedure retroactively -- that "we do not disparage a constitutional guarantee in any manner by declining to apply it retroactively." 384 U.S., at 728; Michigan v. Payne, 412 U.S. 47, 55 n. 10 (1973). This is so, because a prospective application of new rules will often serve important purposes other than the correction of serious flaws in the truth-determining process. The Fifth Amendment privilege against compulsory self-incrimination -- guaranteed full effectuation by the Miranda rules -- serves a variety of significant purposes not relevant to the truth-determining process. See Tehan v. United States ex rel. Shott, 382 U.S. 406, 415-416 (1966). A number of these purposes were catalogued in Murphy v. Waterfront Comm'n, 378 U.S. 52, 55 (1964): "The privilege against self-incrimination 'registers an important advance in the development of our liberty -- "one of the great landmarks in man's struggle to make himself civilized."' Ullmann v. United States, 350 U.S. 422, 426. It reflects many of our fundamental values and most noble aspirations: our unwillingness to subject those suspected of crime to the cruel trilemma of self-accusation, perjury or contempt; our preference for an accusatorial rather than an inquisitorial system of criminal justice; our fear that self-incriminating statements will be elicited by inhumane treatment and abuses; our sense of fair play which dictates 'a fair state-individual balance by requiring the government to leave the individual alone until good cause is shown for disturbing him and by requiring the government in its contest with the individual to shoulder the entire load,' 8 Wigmore, Evidence (McNaughton rev., 1961), 317; our respect for the inviolability of the human personality and of the right of each individual 'to a private enclave where he may lead a private life,' United States v. Grunewald, 233 F.2d 556, 581-582 (Frank, J., dissenting), rev'd 353 U.S. 391; our distrust of self-deprecatory statements; and our realization that the privilege, while sometimes 'a shelter to the guilty,' is often 'a protection to the innocent.' Quinn v. United States, 349 U.S. 155, 162." (Footnotes omitted.)
3. Compare the decisions of the Michigan courts in the instant case, 19 Mich. App. 320, 172 N. W. 2d 712 (1969), and 385 Mich. 594, 189 N. W. 2d 290 (1971), with United States v. Cassell, 452 F.2d 533 (CA7 1971), and People v. Peacock, 29 App. Div. 2d 762, 287 N. Y. S. 2d 166 (1968).
4. Three approaches have been taken in deciding what cases should be affected by prospective application of new constitutional rules of criminal procedure. In Linkletter v. Walker, 381 U.S. 618 (1965), the Court held the exclusionary rule of Mapp v. Ohio, 367 U.S. 643 (1961), applicable to all cases in which direct review had not come to an end at the time Mapp was announced. See also Tehan v. United States ex rel. Shott, 382 U.S. 406 (1966). That approach, as we have observed, was abandoned in Johnson v. New Jersey, where we stated that the Miranda rules were applicable to all trials commenced after the date of that decision. In more recent decisions, we have regarded the cutoff point as that at which law enforcement officials could first begin to guide their conduct in accordance with our new rules. Thus, in Stovall v. Denno, 388 U.S. 293 (1967), the confrontation rulings of United States v. Wade, 388 U.S. 218 (1967), and Gilbert v. California, 388 U.S. 263 (1967), were made applicable to cases in which the confrontations took place after the date of those decisions, and in Desist v. United States, 394 U.S. 244 (1969), the exclusionary ruling of Katz v. United States, 389 U.S. 347 (1967), was made applicable only to cases in which the search and seizure took place after the announcement of Katz. See also Michigan v. Payne, 412 U.S. 47, 57 n. 15 (1973); Williams v. United States, 401 U.S. 646, 656-657 (1971). But cf. Fuller v. Alaska, 393 U.S. 80, 81 (1968) (holding that Lee v. Florida, 392 U.S. 378 (1968), which ruled evidence seized in violation of § 605 of the Federal Communications Act, 47 U. S. C. § 605, inadmissible in state trials, applicable to all cases in which the evidence was introduced after the date of decision in Lee). The trend of our decisions since Johnson has thus been toward placing increased emphasis upon the point at which law enforcement personnel initially relied upon the discarded constitutional standards. See Jenkins v. Delaware, 395 U.S. 213, 218 and n. 7 (1969). As has been noted by an eminent judicial authority, such an emphasis is wholly consistent with the underlying rationale for prospective application of new rules, i. e., justified reliance upon prior judicial standards. Schaefer, The Control of "Sunbursts": Techniques of Prospective Overruling, 42 N. Y. U. L. Rev. 631, 645-646 (1967).
5. My Brother REHNQUIST argues that this concurrence "marks a significant and unsettling departure from the past practice of the Court" in respect of retroactivity. Ante, at 452 n. 26. He argues that Miranda did not decide the question of the admissibility of fruits, and therefore that there is no "parent" decision for retroactive application. But the assumption upon which the concurrence rests, namely, that Miranda requires the exclusion of fruits, necessarily treats Miranda as a "parent" decision. For the assumption is that exclusion is necessary to give full effect to the purposes and policies underlying the Miranda rules and to its holding that "unless and until [the Miranda ] warnings and waiver are demonstrated by the prosecution at trial, no evidence obtained as a result of interrogation can be used against [the defendant]." 384 U.S., at 479 (emphasis added). It necessarily follows that Miranda itself is the "parent" decision.