dissenting in part and concurring in part.
The Court has again propounded a broad constitutional rule barring use of a wide spectrum of relevant and probative evidence, solely because a step in its ascertainment or discovery occurs outside the presence of defense counsel. This was the approach of the Court in Miranda v. Arizona, 384 U. S. 436. I objected then to what I thought was an uncritical and doctrinaire approach without satisfactory factual foundation. I have much the same view of the present ruling and therefore dissent from the judgment and from Parts II, IV, and V of the Court’s opinion.
The Court’s opinion is far-reaching. It proceeds first by creating a new per se rule of constitutional law: a criminal suspect cannot be subjected to a pretrial identification process in the absence of his counsel without violating the Sixth Amendment. If he is, the State may not buttress a later courtroom identification of the witness by any reference to the previous identification. Furthermore, the courtroom identification is not admis*251sible at all unless the State can establish by clear and convincing proof that the testimony is not the fruit of the earlier identification made in the absence of defendant’s counsel — admittedly a heavy burden for the State and probably an impossible one. To all intents and purposes, courtroom identifications are barred if pretrial identifications have occurred without counsel being present.
The rule applies to any lineup, to any other techniques employed to produce an identification and a fortiori to a face-to-face encounter between the witness and the suspect alone, regardless of when the identification occurs, in time or place, and whether before or after indictment or information. It matters not how well the witness knows the suspect, whether the witness is the suspect’s mother, brother, or long-time associate, and no matter how long or well the witness observed the perpetrator at the scene of the crime. The kidnap victim who has lived for days with his abductor is in the same category as the witness who has had only a fleeting glimpse of the criminal. Neither may identify the suspect without defendant’s counsel being present. The same strictures apply regardless of the number of other witnesses who positively identify the defendant and regardless of the corroborative evidence showing that it was the defendant who had committed the crime.
The premise for the Court’s rule is not the general unreliability of eyewitness identifications nor the difficulties inherent in observation, recall, and recognition. The Court assumes a narrower evil as the basis for its rule — improper police suggestion which contributes to erroneous identifications. The Court apparently believes that improper police procedures are so widespread that a broad prophylactic rule must be laid down, requiring the presence of counsel at all pretrial identifications, in *252order to detect recurring instances of police misconduct.1 I do not share this pervasive distrust of all official investigations. None of the materials the Court relies upon supports it.2 Certainly, I would bow to solid fact, but the Court quite obviously does not have before it any reliable, comprehensive survey of current police practices on which to base its new rule. Until it does, the Court should avoid excluding relevant evidence from state criminal trials. Cf. Washington v. Texas, ante, p. 14.
The Court goes beyond assuming that a great majority of the country’s police departments are following improper practices at pretrial identifications. To find the lineup a “critical” stage of the proceeding and to exclude identifications made in the absence of counsel, the Court must also assume that police “suggestion,” if it occurs at all, leads to erroneous rather than accurate identifications and that reprehensible police conduct will have an unavoidable and largely undiscoverable impact on the trial. This in turn assumes that there is now no adequate source from which defense counsel can learn about the circumstances of the pretrial identification in order to place before the jury all of the considerations which should enter into an appraisal of courtroom identification *253evidence. But these are treacherous and unsupported assumptions,3 resting as they do on the notion that the defendant will not be aware, that the police and the witnesses will forget or prevaricate, that defense counsel will be unable to bring out the truth and that neither jury, judge, nor appellate court is a sufficient safeguard against unacceptable police conduct occurring at a pretrial identification procedure. I am unable to share the Court’s view of the willingness of the police and the ordinary citizen-witness to dissemble, either with respect to the identification of the defendant or with respect to the circumstances surrounding a pretrial identification.
There are several striking aspects to the Court’s holding. First, the rule does not bar courtroom identifications where there have been no previous identifications in the presence of the police, although when identified in the courtroom, the defendant is known to be in custody and charged with the commission of a crime. Second, the Court seems to say that if suitable legislative standards were adopted for the conduct of pretrial identifications, thereby lessening the hazards in such con*254frontations, it would not insist on the presence of counsel. But if this is true, why does not the Court simply fashion what it deems to be constitutionally acceptable procedures for the authorities to follow? Certainly the Court is correct in suggesting that the new rule will be wholly inapplicable where police departments themselves have established suitable safeguards.
Third, courtroom identification may be barred, absent counsel at a prior identification, regardless of the extent of counsel’s information concerning the circumstances of the previous confrontation between witness and defendant — apparently even if there were recordings or sound-movies of the events as they occurred. But if the rule is premised on the defendant’s right to have his counsel know, there seems little basis for not accepting other means to inform. A disinterested observer, recordings, photographs — any one of them would seem adequate to furnish the basis for a meaningful cross-examination of the eyewitness who identifies the defendant in the courtroom.
I share the Court’s view that the criminal trial, at the very least, should aim at truthful factfinding, including accurate eyewitness identifications. I doubt, however, on the basis of our present information, that the tragic mistakes which have occurred in criminal trials are as much the product of improper police conduct as they are the consequence of. the difficulties inherent in eyewitness testimony and in resolving evidentiary conflicts by court or jury. I doubt that the Court’s new rule will obviate these difficulties, or that the situation will be measurably improved by inserting defense counsel into the investigative processes of police departments everywhere.
But, it may be asked, what possible state interest militates against requiring the presence of defense counsel at lineups? After all, the argument goes, he may do some good, he may upgrade the quality of identification evidence in state courts and he can scarcely do any *255harm. Even if true, this is a feeble foundation for fastening an ironclad constitutional rule upon state criminal procedures. Absent some reliably established constitutional violation, the processes by which the States enforce their criminal laws are their own prerogative. The States do have an interest in conducting their own affairs, an interest which cannot be displaced simply by saying that there are no valid arguments with respect to the merits of a federal rule emanating from this Court.
Beyond this, however, requiring counsel at pretrial identifications as an invariable rule trenches on other valid state interests. One of them is its concern with the prompt and efficient enforcement of its criminal laws. Identifications frequently take place after arrest but before an indictment is returned or an information is filed. The police may have arrested a suspect on probable cause but may still have the wrong man. Both the suspect and the State have every interest in a prompt identification at that stage, the suspect in order to secure his immediate release and the State because prompt and early identification enhances accurate identification and because it must know whether it is on the right investigative track. Unavoidably, however, the absolute rule requiring the presence of counsel will cause significant delay and it may very well result in no pretrial identification at all. Counsel must be appointed and a time arranged convenient for him and the witnesses. Meanwhile, it may be necessary to file charges against the suspect who may then be released on bail, in the federal system very often on his own recognizance, with neither the State nor the defendant having the benefit of a properly conducted identification procedure.
Nor do I think the witnesses themselves can be ignored. They will now be required to be present at the convenience of counsel rather than their own. Many may be much less willing to participate if the identifica*256tion stage is transformed into an adversary proceeding not under the control of a judge. Others may fear for their own safety if their identity is known at an early date, especially when there is no way of knowing until the lineup occurs whether or not' the police really have the right man.4
Finally, I think the Court’s new rule is vulnerable in terms of its own unimpeachable purpose of increasing the reliability of identification testimony.
Law enforcement officers have the obligation to convict the guilty and to make sure they do not convict the innocent. They must be dedicated to making the criminal trial a procedure for the ascertainment of the true facts surrounding the commission of the crime.5 To this extent, our so-called adversary system is not adversary at all; nor should it be. But defense counsel has no comparable obligation to ascertain or present the truth. Our system assigns him a different mission. He must *257be and is interested in preventing the conviction of the innocent, but, absent a voluntary plea of guilty, we also insist that he defend his client whether he is innocent or guilty. The State has the obligation to present- the evidence. Defense counsel need present nothing, even if he knows what the truth is. He need not furnish any witnesses to the police, or reveal any confidences of his client, or furnish any other information to help the prosecution's case. If he can confuse a witness, even a truthful one, or make him appear at a disadvantage, unsure or indecisive, that will be his normal course.6 Our interest in not con*258victing the innocent permits counsel to put the State to its proof, to put the State’s -case in the worst possible light, regardless of what he thinks or knows to be the truth. Undoubtedly there are some limits which defense counsel must observe7 but more often than not, defense counsel will cross-examine a prosecution witness, and impeach him if he can, even if he thinks the witness is telling the truth, just as he will attempt to destroy a witness who he thinks is lying. In this respect, as part of our modified adversary system and as part of the duty imposed on the most honorable defense counsel, we countenance or require conduct which in many instances has little, if any, relation to the search for truth.
I would not extend this system, at least as it presently operates, to police investigations and would not require counsel’s presence at pretrial identification procedures. Counsel’s interest is in not having his client placed at the scene of the crime, regardless of his whereabouts. Some counsel may advise their clients to refuse to make any *259movements or to speak any words in a lineup or even to appear in one. To that extent the impact on truthful factfinding is quite obvious. Others will not only observe what occurs and develop possibilities for later cross-examination but will hover over witnesses and begin their cross-examination then, menacing truthful fact-finding as thoroughly as the Court fears the police now do. Certainly there is an implicit invitation to counsel to suggest rules for the lineup and to manage and produce it as best he can. I therefore doubt that the Court’s new rule, at least absent some clearly defined limits on counsel’s role, will measurably contribute to more reliable pretrial identifications. My fears are that it will have precisely the opposite result. It may well produce fewer convictions, but that is hardly a proper measure of its long-run acceptability. In my view, the State is entitled to investigate and develop its case outside the presence of defense counsel. This includes the right to have private conversations with identification witnesses, just as defense counsel may have his own consultations with these and other witnesses without having the prosecutor present.
Whether today’s judgment would be an acceptable exercise of supervisory power over federal courts is another question. But as a constitutional matter, the judgment in this case is erroneous and although I concur in Parts I and III of the Court’s opinion I respectfully register this dissent.
Yet in Stovall v. Denno, post, p. 293, the Court recognizes that improper police conduct in the identification process has not been so widespread as to justify full retroactivity for its new rule.
In Miranda v. Arizona, 384 U. S. 436, 449, the Court noted that O’Hara, Fundamentals of Criminal Investigation (1956) is a text that has enjoyed extensive use among law enforcement agencies and among students of police science. The quality of the work was said to rest on the author’s long service as observer, lecturer in police science, and work as a federal crime investigator. O’Hara does not suggest that the police should or do use identification machinery improperly; instead he argues for techniques that would increase the reliability of eyewitness identifications, and there is no reason to suggest that O’Hara’s views are not shared and practiced by the majority of police departments throughout the land.
The instant case and its companions, Gilbert v. California, post, p. 263, and Stovall v. Denno, post, p. 293, certainly lend no support to the Court’s assumptions. The police conduct deemed improper by the Court in the three cases seems to have come to light at trial in the ordinary course of events. One can ask what more counsel would have learned at the pretrial identifications that would have been relevant for truth determination at trial. When the Court premises its constitutional rule on police conduct so subtle as to defy description and subsequent disclosure it deals in pure speculation. If police conduct is intentionally veiled, the police will know about it, and I am unwilling to speculate that defense counsel at trial will be unable to reconstruct the known circumstances of the pretrial identification. And if the “unknown” influence on identifications is “innocent,” the Court’s general premise evaporates and the problem is simply that of the inherent shortcomings of eyewitness testimony.
I would not have thought that the State’s interest regarding its sources of identification is any less than its interest in protecting informants, especially those who may aid in identification but who will not be used as witnesses. See McCray v. Illinois, 386 U. S. 300.
“The United States Attorney is the representative not of an ordinary party to a controversy, but of a sovereignty whose obligation to govern impartially is as compelling as its obligation to govern at all; and whose interest, therefore, in a criminal prosecution is not that it shall win a case, but that justice shall be done. As such, he is in a peculiar and very definite sense the servant of the law, the twofold aim of which is that guilt shall not escape or innocence suffer. He may prosecute with earnestness and vigor— indeed, he should do so. But, while he may strike hard blows, he is not at liberty to strike foul ones. It is as much his duty to refrain from improper methods calculated to produce a wrongful conviction as it is to use every legitimate means to bring about a just one.” Berger v. United States, 295 U. S. 78, 88. See also Mooney v. Holohan, 294 U. S. 103; Pyle v. Kansas, 317 U. S. 213; Alcorta v. Texas, 355 U. S. 28; Napue v. Illinois, 360 U. S. 264; Brady v. Maryland, 373 U. S. 83; Giles v. Maryland, 386 U. S. 66; Miller v. Pate, 386 U. S. 1.
One point of view about the role of the courtroom lawyer appears in Frank, Courts on Trial 82-83. “What is the role of the lawyers in bringing the evidence before the trial court? As you may learn by reading any one of a dozen or more handbooks on how to try a law-suit, an experienced lawyer uses all sorts of stratagems to minimize the effect on the judge or jury of testimony disadvantageous to his client, even when the lawyer has no doubt of the accuracy and honesty of that testimony. ... If such a witness happens to be timid, frightened by the unfamiliarity of court-room ways, the lawyer, in his cross-examination, plays on that weakness, in order to confuse the witness and make it appear that he is concealing significant facts. Longenecker, in his book Hints On The Trial oj a Law Suit (a book endorsed by the great Wigmore), in writing of the ‘truthful, honest, over-cautious’ witness, tells how ‘a skilful advocate by a rapid cross-examination may ruin the testimony of such a witness.’ The author does not even hint any disapproval of that accomplishment. Longeneeker’s and other similar books recommend that a lawyer try to prod an irritable but honest ‘adverse’ witness into displaying his undesirable characteristics in their most unpleasant form, in order to discredit him with the judge or jury. ‘You may,’ writes Harris, ‘sometimes destroy the effect of an adverse witness by making him appear more hostile than he really is. You may make him exaggerate or unsay something and say it again.’ Taft says that a clever cross-examiner, dealing with an honest but egotistic witness, will ‘deftly tempt the witness to indulge in his propensity for exaggeration, so as to make him “hang himself.” ‘And thus,’ adds Taft, ‘it may happen that not only is the value of his testimony lost, but the side which produces him *258suffers for seeking aid from such a source’ — although, I would add, that may be the only source of evidence of a fact on which the decision will turn.
“ ‘An intimidating manner in putting questions,’ writes Wigmore, 'may so coerce or disconcert the witness that his answers do not represent his actual knowledge on the subject. So also, questions which in form or subject cause embarrassment, shame or anger in the witness may unfairly lead him to such demeanor or utterances that the impression produced by his statements does not do justice to its real testimonial value.’ ”
See the materials collected in c. 3 of Countryman & Finman, The Lawyer in Modem Society; Joint Committee on Continuing Legal Education of American Law Institute and the American Bar Association, The Problem of a Criminal Defense 1-46 (1961); Stovall, Aspects of the Advocate’s Dual Responsibility, 22 The Alabama Lawyer 66; Gold, Split Loyalty: An Ethical Problem for the Criminal Defense Lawyer, 14 Clev.-Mar. L. Rev. 65; Symposium on Professional Ethics, 64 Mich. L. Rev. 1469-1498.