This appeal raises questions concerning the guidelines formulated by this court in United States v. Barber, 442 F.2d 517 (3d Cir. 1971), for jury instructions on identification. Appellant, Claude Lamott Wilford, Jr., was tried to a jury and convicted of one count of unlawful distribution of heroin in violation of 21 U.S.C. 841(a)(1). He was sentenced to a term of eight years plus a special parole term of three years. On appeal, he contends that the failure of the trial judge to specifically instruct the jury on eyewitness identification constitutes reversible error. We affirm.

On the afternoon of December 18, 1972, Agents Britt and Malloy of the Federal Bureau of Narcotics and Dangerous Drugs and James Vaughn, a paid Government informer, met and formulated plans to purchase narcotics from an individual known to Vaughn as 'Lamott' living at 2606 Bowers Street, Wilmington, Delaware. About 5:15 that afternoon, after searching Vaughn to assure that he had no narcotics or money on his person, the agents provided Vaughn with $40 and drove him to a housing project located near the intersection of 24th and Bowers Streets. The two agents remained in the car and observed Vaughan as he walked to 2606 Bowers Street, knocked on the front door, and talked to a man at the doorway. The agents then observed Vaughan leave the residence and return to the car where he turned over four glassine packages containing white powder. Field testing and chemical analysis showed that the substance was heroin. Wilford was arrested on January 4, 1973, and tried on May 4.

On direct examination Vaughn testified that on two or three prior occasions he had purchased narcotics from Wilford. He positively identified Wilford in court as the person from whom he purchased narcotics on December 18. Vaughn further testified that he purchased the narcotics as he and Wilford stood in the open doorway of the Bowers Street residence, after Wilford came downstairs in response to his knock.

During the transaction, Agents Britt and Malloy remained in the parked car, approximately 150-200 feet from the Bowers Street reisdence. Although both testified that they observed two men at the doorway and could identify one as Vaughn, neither was able to positively identify the individual with whom Vaughn was talking.1 Malloy, however, testified that just prior to his observations of the two men at the doorway he saw a Negro male leaning out of a second floor window of the residence. Malloy definitely identified the individual as Wilford2 and he adhered to this positive identification on cross-examination.3

Wilford took the stand in defense. He admitted that he, together with his mother, two brothers and a sister, lived at 2606 Bowers Street. He denied, however, ever having met or seen James Vaughn prior to his arrest on January 4. In addition, he denied that he sold Vaughn anything on December 18 or at any time. Although he did not deny that he could have been at his residence on December 18, he testified that he could not recall where he was at that time. Concerning Malloy's positive identification at the second story window, Wilford asserted that screens on the window prevented him from looking out and that even had he leaned out, poor lighting at the housing project would have made identification impossible.

Relying on our recent decision in United States v. Barber, Wilford's sole assignment of error is the trial court's failure to sua sponte instruct the jury that identification testimony should be 'received with caution and scrutinizied with care.'4 He maintains that the evidence introduced on that issue, the testimony of Vaughn and Malloy, failed to meet the four criteria outlined in Barber. Under such circumstances, he contends that under Barber, the trial court was compelled, on its own motion and absent a request, to give an appropriate cautionary instruction.

In Barber we dealt with a special set of facts. Seven defendants were charged in four counts of an indictment with assaulting FBI agents, conspiring to commit such assaults, and assisting a federal prisoner to escape from custody. Fifteen men had approached the two FBI agents, Snyder and Grant, and three to four persons had assaulted each agent. Thus, the confusion surrounding the short-lived episode could easily result in misidentification. Many of the eyewitnesses were unclear as to what had occurred and were often self-contradictory. As Judge Aldisert noted:

An affray of such short duration, involving so many participants, affords but limited opportunities for witnesses to observe and to make positive identifications.

442 F.2d 525. Our concern in Barber was with the crucial problem of mistaken identification-- a problem which is frequently presented when the prosecution relies on eyewitness testimony to establish the identity of the defendant.5 We noted the often 'skeletal' instructions given on this question and sought to insure that where appropriate, cautionary instructions should focus the jury's attention to the possibility of mistaken identification with sufficient particularity. To achieve this goal, we took occasion, under our supervisory power, to provide trial courts with guidelines that were 'require(d)' to be satisfied when cautionary instructions were given.

We first address the contention that Barber requires a cautionary jury charge even absent a request from counsel.6 We do not read Barber to require this result. Barber involved a case where a specific charge was requested. In that context we stated that instructions with sufficient particularity were required.7 We did not direct, as Wilford here contends, that in every case, even absent a request, such instructions are required.8 Our inquiry, therefore, is limited to the question of whether the trial court's failure to sua sponte instruct the jury that the testimony concerning identification should be 'received with caution and scrutinized with care' constitutes plain error. F.R. Crim. P. 52(b).

In the case sub judice, the issue of good faith mistake in identification did not play such a central role9 that the failure to give cautionary instructions requires reversal. The identification of Wilford by the Government's central witness Vaughn clearly met the four criteria outlined in Barber, and accordingly, required no cautionary instruction.10 The thrust of Wilford's cross-examination of Vaughn, moreover, was directed not to Vaughn's inability to observe the seller or remember the transaction but rather to Vaughn's basic credibility. The cross-examination developed that at the time of the alleged transaction Vaughn was a paid informer compensated by Government agents for 'useful information,' that he was a heroin addict, and that he was obtaining funds from his activities with the Government as a means of feeding his drug habit. The court accordingly cautioned the jury that the testimony of the informer must be examined and weighed with greater care than the testimony of an ordinary witness.11 The suggestion that Vaughn was not worthy of belief and had strong motivation to lie, reinforced by Wilford's denial of the sale, was the central and crucial issue in the case.

The contrast to that of Vaughn, Malloy's testimony concerning his identification of Wilford at the second story window and uncertain identification at the door12 did raise an issue of the possibility of mistaken identification. Both in his cross-examination of Malloy and in his testimony on direct examination, Wilford raised questions concerning Malloy's opportunity to observe the figure at the second floor. Malloy's testimony, however, was not persuasive standing alone. At best, it merely placed Wilford on the premises near the time of the sale and provided some slight support for the credibility of Vaughn on the crucial point of the identity of the man at the door. The issue of mistaken identification, therefore, was at most tangential to the central question in the case, i.e., Vaughn's credibility. We do not believe that under such circumstances the failure of the trial judge on his own motion to give special cautionary instructions on the issue of mistaken identification was reversible error.13

We share the concern expressed by Judge Adams in his dissent on the shortcomings of eyewitness identification testimony. We do not minimize 'the vagaries of eyewitness identification' or the hazards such testimony often raises. United States v. Wade, 388 U.S. 218, 228, 87 S. Ct. 1926, 18 L. Ed. 2d 1149 (1967). When convictions obviously turn on the testimony of eyewitnesses who are uncertain, unclear, or inconsistent, the difficulties raised by such evidence are manifest. In such circumstances, a cautionary instruction will help to obviate the danger of erroneous conviction.

The dissent apparently believes that an identification charge should have been given in the case sub judice because Malloy's testimony may have strengthened the Government's case in the minds of the jury. As we have indicated, we believe that when the Government's case rests chiefly on the credibility of an actual party to the alleged transaction, and at best only minimally on the identification testimony of another witness, absent a request by counsel, the failure to give a cautionary identification instruction does not constitute plain error.

The facts of every case vary, and the spectrum of the identification issue may range from the obvious to the obscure. In view of the many variations of fact and circumstance which must be considered in determining whether a special cautionary instruction should be given, we are reluctant to impose the responsibility for making this determination upon the trial judge. This is and should remain the responsibility of defense counsel. Counsel knows his case, his witnesses, and his strategy. He should have the duty to alert the court to the identification issue, especially when the issue is less than obvious. To hold that it is plain error to fail to give a special charge in the circumstances of this or similar cases would only encourage lethargy on the part of counsel, and would permit counsel to sow the seeds of error by remaining mute until after the verdict is returned.

The judgment of the district court will be affirmed.

Many of my own views coincide with those expressed in Judge Rosenn's thoughtful opinion. However, in the circumstances of this case, I regard the failure to give an identification instruction, even though defense counsel did not request one, as prejudicial to the defendant's ability to secure an adequately informed determination of his innocence or guilt. Further, such omission, in my estimation, requires reversal of the judgment of the district court. Accordingly, I respectfully dissent.

It seems settled that the potential for erroneous eyewitness identification at or prior to trial is substantial. The Supreme Court, when dealing with possibly suggestive pretrial identification practices, has emphasized the demonstrably significant likelihood of misidentification in some situations and has sought to provide a degree of protection from attribution of wrongdoing to innocent parties, in part, by augmenting the procedural safeguards available to suspected malefactors.

In United States v. Wade,1 the Supreme Court, in extending the Sixth Amendment right to counsel to post-indictment lineups, made the following observations:

. . . The vagaries of eyewitness identification are well-known; the annals of criminal law are rife with instances of mistaken identification. Mr. Justice Frankfurter once said: 'What is the worth of identification testimony even when uncontradicted? The identification of strangers is proverbially untrustworthy. The hazards of such testimony are established by a formidable number of instances in the records of English and American trials. These instances are recent-- not due to the brutalities of ancient criminal procedure.' The Case of Sacco and Vanzetti 30 (1927) . . . A commentator has observed that 'the influence of improper suggestion upon identifying witnesses probably accounts for more miscarriages of justice than any other single factor-- perhaps it is responsible for more such errors than all other factors combined.' Wall, Eye-Witness Identification in Criminal Cases 26. Suggestion can be created intentionally or unintentionally in many subtle ways. And the dangers for the suspect are particularly grave when the witness' opportunity for observation was insubstantial, and thus his susceptibility to suggestion the greatest.2

The Supreme Court, however, has not yet squarely faced the difficulties presented when a jury, undoubtedly unaware of the possible shortcomings of eyewitness identification testimony, is asked to render a decision of innocence or guilt when such testimony constitutes an important part of the evidence against the defendant. The Court of Appeals for the District of Columbia however, has abandoned the general rule that a trial judge need not submit a particular charge to the jury absent a request3 for an instruction relating to eyewitness identification testimony. In Macklin v. United States,4 the Court of Appeals for the District of Columbia stated that:

After the Supreme Court has focused on identification problems in Wade-Gilbert-Stovall trilogy, it is even more imperative that trial courts include, as a matter of routine, an identification instruction. In cases where identification is a major issue the judge should not rely on defense counsel to request so important a charge.5

It might be contended that the rule fashioned in Macklin by the Court of Appeals, pursuant to its supervisory power, required a new trial in every case where identification was a major issue and the trial court failed to give an identification instruction. However, United States v. Telfaire6 effectively eliminates any implication that Macklin leads automatically to such a result.

In Telfaire, the Court affirmed a conviction even though an identification instruction had not been given. It concluded, in effect, that the failure so to instruct the jury was harmless error because that case exhibited 'none of the special difficulties often presented by identification testimony that would require additional information be given to the jury in order for us to repose confidence in their ability to evaluate the reliability of the identification.'7 The Court added that the eyewitness had 'an adequate opportunity to observe' and had made an earlier spontaneous identification. It did, however, set forth a model identification instruction and stated '(that) failure to use this model, with appropriate adaptations, would constitute a risk in future cases that should not be ignored unless there is strong reason in the particular case.'8

Thus, it appears that the Court of Appeals for the District of Columbia has concluded that in any case where identification is a major issue, the failure to instruct the jury with respect to the possible unreliability of eyewitness identification testimony must 'be noticed although . . . not brought to the attention of the (trial) court' by way of a request for a specific instruction.9 The question left open for review by the Court of Appeals for the District of Columbia is whether the error in failing so to instruct 'affect(ed) substantial rights.'10

This Court treated extensively the problems associated with eyewitness identification testimony in United States v. Barber.11 Although holding that there was no error in the identification charge given by the trial court in that case, this Court, exercising its supervisory power, sought to provide guidance for the district courts in cases where the identification of the defendant is at issue. It stated that:

. . . where the circumstances surrounding the criminal act gave limited opportunity for observation or utilization of other sensory perception, or where uncertainty is expressed by the witness himself, or exposed by a past history of the witness' statements or demonstrated by cross-examination, the statement of identity should be considered as only an expression of opinion and should be accompanied by appropriate instructions as to its sufficiency and weight.12

The Barber court required that instructions satisfy the guidelines followed by Pennsylvania state courts, which arguably establish reversible error if the trial court fails to instruct, even in the absence of a request, when any of the conditions of uncertainty just enumerated are present.13 However, the Barber panel stated that when none of those conditions are present the trial judge need not, at least absent a request, give an identification instruction.

A full appreciation of the uncertainties associated with eyewitness identification testimony and of the judiciary's close attention to the impact of these uncertainties on the fact finding process should temper any decision to designate, or not to designate as plain error a failure to give identification instructions.

I agree with Judge Rosenn that Vaughn's confusion as to the correct name of the defendant did not raise the type of doubts concerning the reliability of Vaughn's eyewitness identification of the defendant that calls for a specific charge. I also agree that Barber, unlike Macklin, does not appear to preclude the applicability of the plain error rule in situations like this. I concede, however, that in light of the concerns expressed by the Supreme Court and this Court, it is difficult, in the absence of Telfaire-type instructions, to contemplate cases in which identification is a major issue where a serious question whether the correct identification has been made. In such situations, plain error would, in all probability, exist.

Although Judge Rosenn's characterization that the 'central and crucial issue in the case' is the credibility of Vaughn, the informer, may not be inaccurate, such characterization should not blind us to the important role identification testimony may have played in Wilford's conviction. Vaughn's testimony placing Wilford at the location where the narcotics were sold and implicating Wilford as the seller is corroborated by Agent Malloy's identification of Wilford as the person leaning from a second floor window. If the jury did credit Malloy's testimony-- and there is no way that we can ascertain this-- the government's case was substantially strengthened. If the jury discounted Malloy's testimony, the government's case was immeasurably weakened. This seems particularly so because the testimony of informants is frequently suspect since they may well believe that convictions will result in better treatment for themselves.14 Because of these considerations, I would be inclined to view this case as one in which identification is a major issue.

Evidence was introduced by Wilford, chiefly by way of his own testimony, which indicated that Agent Malloy's opportunity to observe may have been impaired by the dim lighting at the housing project. It would appear, therefore, that one of the conditions is present that, under Barber, transforms identification testimony from statements of fact into opinions.

Since this Court in Barber has recently and thoroughly grappled with the difficulties posed by eyewitness identification testimony, and in my view, such testimony may well have been a compelling factor in the jury's decision in this case, I would not hold that Wilford's failure to request an identification instruction prevents this Court from recognizing the trial court's error in not so charging the jury after Wilford had produced evidence suggesting that Agent Malloy's observations occurred in dim light.

In practice, it seems that a conclusion that plain error exists invariably leads to a conclusion that the error is not harmless; for quite frequently the considerations that prompt the former conclusion preclude a decision contrary to the latter.15 Contrasting this case with Telfaire strongly suggests that it is not an aberration. Judge Rosenn apparently acknowledges that the general instructions given here were not as likely to focus the jury's attention on the identification issue as those in Telfaire.16 Further, the witness in Telfaire did have an adequate opportunity to observe, and a spontaneous pretrial identification in Telfaire buttressed the confidence of the Court of Appeals for the District of Columbia in the reliability of the identification testimony.17

The procedure followed in Telfaire, suggesting model identification instructions in this type of case, merits further attention. If the trial courts are constantly aware that certain instructions, varied to meet specific cases, are to be given whenever identification is an important issue, it is more likely that the defendant's right to an informed consideration of his case will not be diluted and the number of appeals such as this may well be substantially reduced.18 On balance, it may be wiser to reverse the defendant's conviction, based in large part on the informant's possession, after allegedly meeting with the defendant, of four glassine packages of heroin, and, at the same time, take the opportunity to provide district courts with more definite guidance with respect to the necessity for and content of identification charges.

Accordingly, I would reverse the judgment of the district court and remand for a new trial.