Miranda v. Arizona

Mr. Justice White, with whom Mr. Justice Harlan and Mr. Justice Stewart join,

dissenting.

I.

The proposition that the privilege against self-incrimination forbids in-custody interrogation without the warnings specified in the majority opinion and without a clear waiver of counsel has no significant support in the history of the privilege or in the language of the Fifth Amendment. As for the English authorities and the common-law history, the privilege, firmly established in the second half of the seventeenth century, was never applied except to prohibit compelled judicial interrogations. The rule excluding coerced confessions matured about 100 years later, “[b]ut there is nothing in the reports to suggest that the theory has its roots in the privilege against self-incrimination. And so far as the cases reveal, the privilege, as such, seems to have been given effect only in judicial proceedings, including the preliminary examinations by authorized magistrates.” Morgan, The Privilege Against Self-Incrimination, 34 Minn. L. Rev. 1, 18 (1949).

Our own constitutional provision provides that no person “shall be compelled in any criminal case to be a witness against himself.” These words, when “[cjonsid-ered in the light to be shed by grammar and the dictionary . . . appear to signify simply that nobody shall be *527 compelled to give oral testimony against himself in a criminal proceeding under way in which he is defendant.” Corwin, The Supreme Court’s Construction of the Self-Incrimination Clause, 29 Mich. L. Rev. 1, 2. And there is very little in the surrounding circumstances of the adoption of the Fifth Amendment or in the provisions of the then existing state constitutions or in state practice which would give the constitutional provision any broader meaning. Mayers, The Federal Witness’ Privilege Against Self-Incrimination: Constitutional or Common-Law? 4 American Journal of Legal History 107 (1960). Such a construction, however, was considerably narrower than the privilege at common law, and when eventually faced with the issues, the Court extended the constitutional privilege to the compulsory production of books and papers, to the ordinary witness before the grand jury and to witnesses generally. Boyd v. United States, 116 U. S. 616, and Counselman v. Hitchcock, 142 U. S. 547. Both rules had solid support in common-law history, if not in the history of our own constitutional provision.

A few years later the Fifth Amendment privilege was similarly extended to encompass the then well-established rule against coerced confessions: “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.’ ” Bram v. United States, 168 U. S. 532, 542. Although this view has found approval in other cases, Burdeau v. McDowell, 256 U. S. 465, 475; Powers v. United States, 223 U. S. 303, 313; Shotwell v. United States, 371 U. S. 341, 347, it has also been questioned, see Brown v. Mississippi, 297 U. S. 278, 285; United States v. Carignan, *528 342 U. S. 36, 41; Stein v. New York, 346 U. S. 156, 191, n. 35, and finds scant support in either the English or American authorities, see generally Regina v. Scott, Dears. & Bell 47; 3 Wigmore, Evidence § 823 (3d ed. 1940), at 249 (“a confession is not rejected because of any connection with the privilege against self-crimination”), and 250, n. 5 (particularly criticizing Bram); 8 Wigmore, Evidence § 2266, at 400-40.1 (McNaughton rev. 1961). Whatever the source of the rule excluding coerced confessions, it is clear that prior to the application of the privilege itself to state courts, Malloy v. Hogan, 378 U. S. 1, the admissibility of a confession in a state criminal prosecution was tested by the same standards as were applied in federal prosecutions. Id., at 6-7, 10.

Bram, however, itself rejected the proposition which the Court now espouses. The question in Bram was whether a confession, obtained during custodial interrogation, had been compelled, and if such interrogation was to be deemed inherently vulnerable the Court’s inquiry could have ended there. After examining the English and American authorities, however, the Court declared that:

“In this court also it has been settled that the mere fact that the confession is made to a police officer, while the accused was under arrest in or out of .prison, or was drawn out by his questions, does not necessarily render the confession involuntary, but, as one of the circumstances, such imprisonment or interrogation may be taken into account in determining whether or not the statements of the prisoner were voluntary.” 168 U. S., at 558.

In this respect the Court was wholly consistent with prior and subsequent pronouncements in this Court.

Thus prior to Bram the Court, in Hopt v. Utah, 110 U. S. 574, 583-587, had upheld the admissibility of a *529 confession made to police officers following arrest, the record being silent concerning what conversation had occurred between the officers and the defendant in the short period preceding the confession. Relying on Hopt, the Court ruled squarely on the issue in Sparf and Hansen v. United States, 156 U. S. 51, 55:

“Counsel for the accused insist that there cannot be a voluntary statement, a free open confession, while a defendant is confined and in irons under an accusation of having committed a capital offence. We have not been referred to any authority in support of that position. It is true that the fact of a prisoner being in custody at the time he makes a confession is a circumstance not to be overlooked, because it bears upon the inquiry whether the confession was voluntarily made or was extorted by threats or violence or made under the influence of fear. But confinement or imprisonment is not in itself sufficient to justify the exclusion of a confession, if it appears to have been voluntary, and was not obtained by putting the prisoner in fear or by promises. Wharton’s Cr. Ev. 9th ed. §§ 661, 663, and authorities cited.”

Accord, Pierce v. United States, 160 U. S. 355, 357.

And in Wilson v. United States, 162 U. S. 613, 623, the Court had considered the significance of custodial interrogation without any antecedent warnings regarding the right to remain silent or the right to counsel. There the defendant had answered questions posed by a Commissioner, who had failed to advise him of his rights, and his answers were held admissible over his claim of involuntariness. “The fact that [a defendant] is in custody and manacled does not necessarily render his statement involuntary, nor is that necessarily the effect of popular excitement shortly preceding. . . . And it is laid down *530 that it is not essential to the admissibility of a confession that it should appear that the person was warned that what he said would be used against him, but on the contrary, if the confession was voluntary, it is sufficient though it appear that he was not so warned.”

Since Bram, the admissibility of statements made during custodial interrogation has been frequently reiterated. Powers v. United States, 223 U. S. 303, cited Wilson approvingly and held admissible as voluntary statements the accused’s testimony at a preliminary hearing even though he was not warned that what he said might be used against him. Without any discussion of the presence or absence of warnings, presumably because such discussion was deemed unnecessary, numerous other cases have declared that “[t]he mere fact that a confession was made while in the custody of the police does not render it inadmissible,” McNabb v. United States, 318 U. S. 332, 346; accord, United States v. Mitchell, 322 U. S. 65, despite its having been elicited by police examination, Wan v. United States, 266 U. S. 1, 14; United States v. Carignan, 342 U. S. 36, 39. Likewise, in Crooker v. California, 357 U. S. 433, 437, the Court said that “the bare fact of police ‘detention and police examination in private of one in official state custody’ does not render involuntary a confession by the one so detained.” And finally, in Cicenia v. Lagay, 357 U. S. 504, a confession obtained by police interrogation after arrest was held voluntary even though the authorities refused to permit the defendant to consult with his attorney. See generally Culombe v. Connecticut, 367 U. S. 568, 587-602 (opinion of Frankfurter, J.); 3 Wigmore, Evidence § 851, at 313 (3d ed. 1940); see also Joy, Admissibility of Confessions 38, 46 (1842).

Only a tiny minority of our judges who have dealt with the question, including today’s majority, have considered in-custody interrogation, without more, to be a violation of the Fifth Amendment. And this Court, as *531 every member knows, has left standing literally thousands of criminal convictions that rested at least in part on confessions taken in the course of interrogation by the police after arrest.

II.

That the Court’s holding today is neither compelled nor even strongly suggested by the language of the Fifth Amendment, is at odds with American and English legal history, and involves a departure from a long line of precedent does not prove either that the Court has exceeded its powers or that the Court is wrong or unwise in its present reinterpretation of the Fifth Amendment. It does, however, underscore the obvious — that the Court has not discovered or found the law in making today’s decision, nor has it derived it from some irrefutable sources; what it has done is to make new law and new public policy in much the same way that it has in the course of interpreting other great clauses of the Constitution.1 This is what the Court historically has done. Indeed, it is what it must do and will continue to do until and unless there is some fundamental change in the constitutional distribution of governmental powers.

But if the Court is here and now to announce new and fundamental policy to govern certain aspects of our affairs, it is wholly legitimate to examine the mode of this or any other constitutional decision in this Court and to inquire into the advisability of its end product in terms of the long-range interest of the country. At the very least the Court’s text and reasoning should withstand analysis and be a fair exposition of the constitutional provision which its opinion interprets. De *532 cisions like these cannot rest alone on syllogism, metaphysics or some ill-defined notions of natural justice, although each will perhaps play its part. In proceeding to such constructions as it now announces, the Court should also duly consider all the factors and interests bearing upon the cases, at least insofar as the relevant materials are available; and if the necessary considerations are not treated in the record or obtainable from some other reliable source, the Court should not proceed to formulate fundamental policies based on speculation alone.

III.

First, we may inquire what are the textual and factual bases of this new fundamental rule. To reach the result announced on the grounds it does, the Court must stay within the confines of the Fifth Amendment, which forbids self-incrimination only if compelled. Hence the core of the Court’s opinion is that because of the “compulsion inherent in custodial surroundings, no statement obtained from [a] defendant [in custody] can truly be the product of his free choice,” ante, at 458, absent the use of adequate protective devices as described by the Court. However, the Court does not point to any sudden inrush of new knowledge requiring the rejection of 70 years’ experience. Nor does it assert that its novel conclusion reflects a changing consensus among state courts, see Mapp v. Ohio, 367 U. S. 643, or that a succession of cases had steadily eroded the old rule and proved it unworkable, see Gideon v. Wainwright, 372 U. S. 335. Rather than asserting new knowledge, the Court concedes that it cannot truly know what occurs during custodial questioning, because of the innate secrecy of such proceedings. It extrapolates a picture of what it conceives to be the norm from police investigatorial manuals, published in 1959 and 1962 or earlier, without any attempt to allow for adjustments in police practices that may *533 have occurred in the wake of more recent decisions of state appellate tribunals or this Court. But even if the relentless application of the described procedures could lead to involuntary confessions, it most assuredly does not follow that each and every case will disclose this kind of interrogation or this kind of consequence.2 Insofar as appears from the Court’s opinion, it has not examined a single transcript of any police interrogation, let alone the interrogation that took place in any one of these cases which it decides today. Judged by any of the standards for empirical investigation utilized in the social sciences the factual basis for the Court’s premise is patently inadequate.

Although in the Court’s view in-custody interrogation is inherently coercive, the Court says that the spontaneous product of the coercion of arrest and detention is still to be deemed voluntary. An accused, arrested on probable cause, may blurt out a confession which will be admissible despite the fact that he is alone and in custody, without any showing that he had any notion of his right to remain silent or of the consequences of his admission. Yet, under the Court’s rule, if the police ask him a single question such as “Do you have anything to say?” or “Did you kill your wife?” his response, if there is one, has somehow been compelled, even if the accused has *534 been clearly warned of his right to remain silent. Common sense informs- us to the contrary. While one may say that the response was “involuntary” in the sense the question provoked or was the occasion for the response and thus the defendant was induced to speak out when he might have remained silent if not arrested and not questioned, it is patently unsound to say the response is compelled.

Today’s result would not follow even if it were agreed that to some extent custodial interrogation is inherently coercive. See Ashcraft v. Tennessee, 322 U. S. 143, 161 (Jackson, J., dissenting). The test has been whether the totality of circumstances deprived the defendant of a “free choice to admit, to deny, or to refuse to answer,” Lisenba v. California, 314 U. S. 219, 241, and whether physical or psychological coercion was of such a degree that “the defendant’s will was overborne at the time he confessed,” Haynes v. Washington, 373 U. S. 503, 513; Lynumn v. Illinois, 372 U. S. 528, 534. The duration and nature of incommunicado custody, the presence or absence of advice concerning the defendant’s constitutional rights, and the granting or refusal of requests to communicate with lawyers, relatives or friends have all been rightly regarded as important data bearing on the basic inquiry. See, e. g., Ashcraft v. Tennessee, 322 U. S. 143; Haynes v. Washington, 373 U. S. 503.3 *535 But it has never been suggested, until today, that such questioning was so coercive and accused persons so lacking in hardihood that the very first response to the very first question following the commencement of custody must be conclusively presumed to be the product of an overborne will.

If the rule announced today were truly based on a conclusion that all confessions resulting from custodial interrogation are coerced, then it would simply have no rational foundation. Compare Tot v. United States, 319 U. S. 463, 466; United States v. Romano, 382 U. S. 136. A fortiori that would be true of the extension of the rule to exculpatory statements, which the Court effects after a brief discussion of why, in the Court’s view, they must be deemed incriminatory but without any discussion of why they must be deemed coerced. See Wilson v. United States, 162 U. S. 613, 624. Even if one were to postulate that the Court’s concern is not that all confessions induced by police interrogation are coerced but rather that some such confessions are coerced and present judicial procedures are believed to be inadequate to identify the confessions that are coerced and those that are not, it would still not be essential to impose the rule that the Court has now fashioned. Transcripts or observers could be required, specific time limits, tailored to fit the cause, could be imposed, or other devices could be utilized to reduce the chances that otherwise indiscernible coercion will produce an inadmissible confession.

On the other hand, even if one assumed that there was an adequate factual basis for the conclusion that all confessions obtained during in-custody interrogation are the product of compulsion, the rule propounded by *536 the Court would still be irrational, for, apparently, it is only if the accused is also warned of his right to counsel and waives both that right and the right against self-incrimination that the inherent compulsiveness of interrogation disappears. But if the defendant may not answer without a warning a question such as “Where were you last night?” without having his answer be a compelled one, how can the Court ever accept his negative answer to the question of whether he wants to consult his retained counsel or counsel whom the court will appoint? And why if counsel is present and the accused nevertheless confesses, or counsel tells the accused to tell the truth, and that is what the accused does, is the situation any less coercive insofar as the accused is concerned? The Court apparently realizes its dilemma of foreclosing questioning without the necessary warnings but at the same time permitting the accused, sitting in the same chair in front of the same policemen, to waive his right to consult an attorney. It expects, however, that the accused will not often waive the right; and if it is claimed that he has, the State faces a severe, if not impossible burden of proof.

All of this makes very little sense in terms of the compulsion which the Fifth Amendment proscribes. That amendment deals with compelling the accused himself. It is his free will that is involved. Confessions and incriminating admissions, as such, are not forbidden evidence; only those which are compelled are banned. I doubt that the Court observes these distinctions today. By considering any answers to any interrogation to be compelled regardless of the content and course of examination and by escalating the requirements to prove waiver, the Court not only prevents the use of compelled confessions but for all practical purposes forbids interrogation except in the presence of counsel. That is, instead of confining itself to protection of the right against com *537 pelled self-incrimination the Court has created a limited Fifth Amendment right to counsel — or, as the Court expresses it, a “need for counsel to protect the Fifth Amendment privilege . . . .” Ante, at 470. The focus then is not on the will of the accused but on the will of counsel and how much influence he can have on the accused. Obviously there is no warrant in the Fifth Amendment for thus installing counsel as the arbiter of the privilege.

In sum, for all the Court’s expounding on the menacing atmosphere of police interrogation procedures, it has failed to supply any foundation for the conclusions it draws or the measures it adopts.

IV.

Criticism of the Court’s opinion, however, cannot stop with a demonstration that the factual and textual bases for the rule it propounds are, at best, less than compelling. Equally relevant is an assessment of the rule’s consequences measured against community values. The Court’s duty to assess the consequences of its action is not satisfied by the utterance of the truth that a value of our system of criminal justice is “to respect the inviolability of the human personality” and to require government to produce the evidence against the accused by its own independent labors. Ante, at 460. More than the human dignity of the accused is involved; the human personality of others in the society must also be preserved. Thus the values reflected by the privilege are not the sole desideratum; society’s interest in the general security is of equal weight.

The obvious underpinning of the Court’s decision is a deep-seated distrust of all confessions. As the Court declares that the accused may not be interrogated without counsel present, absent a waiver of the right to counsel, and as the Court all but admonishes the lawyer to *538 advise the accused to remain silent, the result adds up to a judicial judgment that evidence from the accused should not be used against him in any way, whether compelled or not. This is the not so subtle overtone of the opinion — that it is inherently wrong for the police to gather evidence from the accused himself. And this is precisely the nub of this dissent. I see nothing wrong or immoral, and certainly nothing unconstitutional, in the police’s asking a suspect whom they have reasonable cause to arrest whether or not he killed his wife or in confronting him with the evidence on which the arrest was based, at least where he has been plainly advised that he may remain completely silent, see Escobedo v. Illinois, 378 U. S. 478, 499 (dissenting opinion). Until today, “the admissions or confessions of the prisoner, when voluntarily and freely made, have always ranked high in the scale of incriminating evidence.” Brown v. Walker, 161 U. S. 591, 596; see also Hopt v. Utah, 110 U. S. 574, 584-585. Particularly when corroborated, as where the police have confirmed the accused’s disclosure of the hiding place of implements or fruits of the crime, such confessions have the highest reliability and significantly contribute to the certitude with which we may believe the accused is guilty. Moreover, it is by no means certain that the process of confessing is injurious to the accused. To the contrary it may provide psychological relief and enhance the prospects for rehabilitation.

This is not to say that the value of respect for the inviolability of the accused’s individual personality should be accorded no weight or that all confessions should be indiscriminately admitted. This Court has long read the Constitution to proscribe compelled confessions, a salutary rule from which there should be no retreat. But I see no sound basis, factual or otherwise, and the Court gives none, for concluding that the present rule against the receipt of coerced confessions is inadequate for the *539 task of sorting out inadmissible evidence and must be replaced by the per se rule which is now imposed. Even if the new concept can be said to have advantages of some sort over the present law, they are far outweighed by its likely undesirable impact on other very relevant and important interests.

The most basic function of any government is to provide for the security of the individual and of his property. Lanzetta v. New Jersey, 306 U. S, 451, 455. These ends of society are served by the criminal laws which for the most part are aimed at the prevention of crime. Without the reasonably effective performance of the task of preventing private violence and retaliation, it is idle to talk about human dignity and civilized values.

The modes by which the criminal laws serve the interest in general security are many. First the murderer who has taken the life of another is removed from the streets, deprived of his liberty and thereby prevented from repeating his offense. In view of the statistics on recidivism in this country4 and of the number of instances *540 in which apprehension occurs only after repeated offenses, no one can sensibly claim that this aspect of the criminal law does not prevent crime or contribute significantly to the personal security of the ordinary citizen.

Secondly, the swift and sure apprehension of those who refuse to respect the personal security and dignity of their neighbor unquestionably has its impact on others who might be similarly tempted. That the criminal law is wholly or partly ineffective with a segment of the population or with many of those who have been apprehended and convicted is a very faulty basis for concluding that it is not effective with respect to the great bulk of our citizens or for thinking that without the criminal laws, *541 or in the absence of their enforcement, there would be no increase in crime. Arguments of this nature are not borne out by any kind of reliable evidence that I have seen to this date.

Thirdly, the law concerns itself with those whom it has confined. The hope and aim of modern penology, fortunately, is as soon as possible to return the convict to society a better and more law-abiding man than when he left. Sometimes there is success, sometimes failure. But at least the effort is made, and it should be made to the very maximum extent of our present and future capabilities.

The rule announced today will measurably weaken the ability of the criminal law to perform these tasks. It is a deliberate calculus to prevent interrogations, to reduce the incidence of confessions and pleas of guilty and to increase the number of trials.5 Criminal trials, no *542 matter how efficient the police are, are not sure bets for the prosecution, nor should they be if the evidence is not forthcoming. Under the, present law, the prosecution fails to prove its case in about 30% of the criminal cases actually tried in the federal courts. See Federal Offenders: 1964, supra, note 4, at 6 (Table 4), 59 (Table 1); Federal Offenders: 1963, supra, note 4, at 5 (Table 3); District of Columbia Offenders: 1963, supra, note 4, at 2 (Table 1). But it is something else again to remove from the ordinary criminal case all those confessions which heretofore have been held to be free and voluntary acts of the accused and to thus establish a new constitutional barrier to the ascertainment of truth by the judicial process. There is, in my view, every reason to believe that a good many criminal defendants who otherwise would have been convicted on what this Court has previously thought to be the most satisfactory kind of evidence will now, under this new version of the Fifth Amendment, either not be tried at all or will be acquitted if the State’s evidence, minus the confession, is put to the test of litigation.

I have no desire whatsoever to share the responsibility for any such impact on the present criminal process.

In some unknown number of cases the Court’s rule will return a killer, a rapist or other criminal to the streets and to the environment which produced him, to repeat his crime whenever it pleases him. As a consequence, there will not be a gain, but a loss, in human dignity. The real concern is not the unfortunate consequences of this new decision on the criminal law as an abstract, disembodied series of authoritative proscriptions, but the impact on those who rely on the public authority for protection and who without it can only engage in violent self-help with guns, knives and the help of their neighbors similarly inclined. There is, of *543 course, a saving factor: the next victims are uncertain, unnamed and unrepresented in this case.

Nor can this decision do other than have a corrosive effect on the criminal law as an effective device to prevent crime. A major component in its effectiveness in this regard is its swift and sure enforcement. The easier it is to get away with rape and murder, the less the deterrent effect on those who are inclined to attempt it. This is still good common sense. If it were not, we should posthaste liquidate the whole law enforcement establishment as a useless, misguided effort to control human conduct.

And what about the accused who has confessed or would confess in response to simple, noncoercive questioning and whose guilt could not otherwise be proved? Is it so clear that release is the best thing for him in every case? Has it so unquestionably been resolved that in each and every case it would be better for him not to confess and to return to his environment with no attempt whatsoever to help him? I think not. It may well be that in many cases it will be no less than a callous disregard for his own welfare as well as for the interests of his next victim.

There is another aspect to the effect of the Court’s rule on the person whom the police have arrested on probable cause. The fact is that he may not be guilty at all and may be able to extricate himself quickly and simply if he were told the circumstances of his arrest and were asked to explain. This effort, and his release, must now await the hiring of a lawyer or his appointment by the court, consultation with counsel and then a session with the police or the prosecutor. Similarly, where probable cause exists to arrest several suspects, as where the body of the victim is discovered in a house having several residents, compare Johnson v. State, 238 Md. 140, 207 A. 2d 643 (1965), cert. denied, 382 U. S. 1013, it will often *544 be true that a suspect may be cleared only through the results of interrogation of other suspects. Here too the release of the innocent may be delayed by the Court’s rule.

Much of the trouble with the Court’s new rule is that it will operate indiscriminately in all criminal cases, regardless of the severity of the crime or the circumstances involved. It applies to every defendant, whether the professional criminal or one committing a crime of momentary passion who is not part and parcel of organized crime. It will slow down the investigation and the apprehension of confederates in those cases where time is of the essence, such as kidnapping, see Brinegar v. United States, 338 U. S. 160, 183 (Jackson, J., dissenting) ; People v. Modesto, 62 Cal. 2d 436, 446, 398 P. 2d 753, 759 (1965), those involving the national security, see United States v. Drummond, 354 F. 2d 132, 147 (C. A. 2d Cir. 1965) (en banc) (espionage case), pet. for cert. pending, No. 1203, Misc., O. T. 1965; cf. Gessner v. United States, 354 F. 2d 726, 730, n. 10 (C. A. 10th Cir. 1965) (upholding, in espionage case, trial ruling that Government need not submit classified portions of interrogation transcript), and some of those involving organized crime. In the latter context the lawyer who arrives may also be the lawyer for the defendant’s colleagues and can be relied upon to insure that no breach of the organization’s security takes place even though the accused may feel that the best thing he can do is to cooperate.

At the same time, the Court’s per se approach may not be justified on the ground that it provides a “bright line” permitting the authorities to judge in advance whether interrogation may safely be pursued without jeopardizing the admissibility of any information obtained as a consequence. Nor can it be claimed that judicial time and effort, assuming that is a relevant consideration, *545 will be conserved because of the ease of application of the new rule. Today’s decision leaves open such questions as whether the accused was in custody, whether his statements were spontaneous or the product of interrogation, whether the accused has effectively waived his rights, and whether nontestimonial evidence introduced at trial is the fruit of statements made during a prohibited interrogation, all of which are certain to prove productive of uncertainty during investigation and litigation during prosecution. For all these reasons, if further restrictions on police interrogation are desirable at this time, a more flexible approach makes much more sense than the Court’s constitutional straitjacket which forecloses more discriminating treatment by legislative or rule-making pronouncements.

Applying the traditional standards to the cases before the Court, I would hold these confessions voluntary. I would therefore affirm in Nos. 759, 760, and 761, and reverse in No. 584.

Of course the Court does not deny that it is departing from prior precedent; it expressly overrules Crooker and Cicenia, ante, at 479, n. 48, and it acknowledges that in the instant “cases we might not find the defendants’ statements to have been involuntary in traditional terms,” ante, at 457.

In fact, the type of sustained interrogation described by the Court appears to be the exception rather than the rule. A survey of 399 cases in one city found that in almost half of the eases the interrogation lasted less than 30 minutes. Barrett, Police Practices and the Law — From Arrest to Release or Charge, 50 Calif. L. Rev. 11, 41-45 (1962). Questioning tends to be confused and sporadic and is usually concentrated on confrontations with witnesses or new items of evidence, as these are obtained by officers conducting the investigation. See generally LaFave, Arrest: The Decision to Take a Suspect into Custody 386 (1965); ALI, A Model Code of Pre-Arraignment Procedure, Commentary § 5.01, at 170, n. 4 (Tent. Draft No. 1, 1966).

By contrast, the Court indicates that in applying this new rule it “will not pause to inquire in individual cases whether the defendant was aware of his rights without a warning being given.” Ante, at 468. The reason given is that assessment of the knowledge of the defendant based on information as to age, education, intelligence, or prior contact with authorities can never be more than speculation, while a warning is a clear-cut fact. But the officers’ claim that they gave the requisite warnings may be disputed, and facts respecting the defendant’s prior experience may be undisputed and be of such a'nature as to virtually preclude any doubt that the defendant knew of his rights. See United States v. Bolden, 355 F. 2d 453 *535 (C. A. 7th Cir. 1965), petition for cert. pending No. 1146, O. T. 1965 (Secret Service agent); People v. Du Bont, 235 Cal. App. 2d 844, 45 Cal. Rptr. 717, pet. for cert. pending No. 1053, Misc., O. T. 1965 (former police officer).

Precise statistics on the extent of recidivism are unavailable, in part because not all crimes are solved and in part because criminal records of convictions in different jurisdictions are not brought together by a central data collection agency. Beginning in 1963, however, the Federal Bureau of Investigation began collating data on “Careers in Crime,” which it publishes in its Uniform Crime Reports. Of 92,869 offenders processed in 1963 and 1964, 76% had a prior arrest record on some charge. Over a period of 10 years the group had accumulated 434,000 charges. FBI, Uniform Crime Reports — 1964, 27-28. In 1963 and 1964 between 23% and 25% of all offenders sentenced in 88 federal district courts (excluding the District Court for the District of Columbia) whose criminal records were reported had previously been sentenced to a term of imprisonment of 13 months or more. Approximately an additional 40% had a prior record less than prison (juvenile record, probation record, etc.). Administrative Office of the United States Courts, Federal Offenders in the United States District Courts: 1964, x, 36 (hereinafter cited as Federal Offenders: 1964); Administrative *540 Office of the United States Courts, Federal Offenders in the United States District Courts: 1963, 25-27 (hereinafter cited as Federal Offenders: 1963). During the same two years in the District Court for the District of Columbia between 28% and 35% of those sentenced had prior prison records and from 37% to 40% had a prior record less than prison. Federal Offenders: 1964, xii, 64, 66; Administrative Office of the United States Courts, Federal Offenders in the United States District Court for the District of Columbia: 1963, 8, 10 (hereinafter cited as District of Columbia Offenders: 1963).

A similar picture is obtained if one looks at the subsequent records of those released from confinement. In 1964, 12.3% of persons on federal probation had their probation revoked because of the commission of major violations (defined as one in which the probationer has been committed to imprisonment for a period of 90 days or more, been placed on probation for over one year on a new offense, or has absconded with felony charges outstanding). Twenty-three and two-tenths percent of parolees and 16.9% of those who had been mandatorily released after service of a portion of their sentence likewise committed major violations. Reports of the Proceedings of the Judicial Conference of the United States and Annual Report of the Director of the Administrative Office of the United States Courts: 1965, 138. See also Mandel et al., Recidivism Studied and Defined, 56 J. Crim. L., C. & P. S. 59 (1965) (within five years of release 62.33% of sample had committed offenses placing them in recidivist category).

Eighty-eight federal district courts (excluding the District Court for the District of Columbia) disposed of the eases of 33,381 criminal defendants in 1964. Only 12.5% of those cases were actually tried. Of the remaining cases, 89.9% were terminated by convictions upon pleas of guilty and 10.1% were dismissed. Stated differently, approximately 90% of all convictions resulted from guilty pleas. Federal Offenders: 1964, supra, note 4, 3-6. In the District Court for the District of Columbia a higher percentage, 27%, went to trial, and the defendant pleaded guilty in approximately 78% of the cases terminated prior to trial. Id., at 58-59. No reliable statistics are available concerning the percentage of cases in which guilty pleas are induced because of the existence of a confession or of physical evidence unearthed as a result of a confession. Undoubtedly the number of such cases is substantial.

Perhaps of equal significance is the number of instances of known crimes which are not solved. In 1964, only 388,946, or 23.9% of 1,626,574 serious known offenses were cleared. The clearance rate ranged from 89.8% for homicides to 18.7% for larceny. FBI, Uniform Crime Reports — 1964, 20-22, 101. Those who would replace interrogation as an investigatorial tool by modern scientific investigation techniques significantly overestimate the effectiveness of present procedures, even when interrogation is included.