dissents and votes to reverse the judgment of conviction and dismiss the indictment, with the following opinion: Appellant Franklin Rosa Moya, 24 years old at the time of the trial, appeals from a judgment of the Supreme Court, Kings County (Lombardo, J.), rendered June 16, 1981, convicting him, upon a jury verdict, of robbery in the first degree, robbery in the second degree, assault in the second degree (two counts) and grand larceny in the third degree, and imposing sentence.
By this appeal, there is once more placed before us a typical one-witness identification case charging cruel and vicious crimes against an individual never previously charged or convicted of any crime.
My confreres of the majority, in seeking to sustain the guilty verdict, arrive at several conclusions, i.e., (1) the hospital showup was, in light of its promptness, not unduly suggestive or conducive to mistaken identification; (2) the complainant had sufficient independent basis for her in-court identification; and (3) appellant has failed to preserve for appellate review: (a) an alibi charge which, in my opinion, is fatally flawed and cries for reversal, and (b) a prosecutor’s summation, labeled by the majority as harmless, which, in my opinion, is clearly unconstitutional. Except for the third conclusion, I find it difficult to follow, let alone agree with, the logic or the law that dictates these findings. Sad and sober experience alerts us to the deadly defects and inherent flaws that so frequently and tragically accompany eyewitness identification (e.g., Borchard, Convicting the Innocent [1932]; O’Connor, "That’s the man”: A Sobering Study of Eyewitness Identification and the Polygraph, 49 St. John’s L Rev 1; Sobel, Eye Witness Identification [1981]). The conclusion long since reached by most students of this ever-recurring problem is eloquently expressed in the following quote: "No amount of cross-examination can reveal the numerous suggestive influences that subtly bias the recollections of an honest but mistaken witness, and even the most eloquent of closing arguments will not persuade a jury to disregard sincere and apparently truthful eyewitness testimony” (Woocher, Note, Did Your Eyes Deceive You? Expert Psychological Testimony on the Unreliability of Eyewitness Identification, 29 Standford L Rev 969, 1029).
Although this is neither the time nor the forum for an in-*772depth review of the sad and sordid history of eyewitness identification, these fleeting references are enlightening and seem to me to be entirely relevant.
It requires no more than a cursory review to establish that wrong man cases have been a constant plague on our house of justice for many many years.
In one of England’s famous wrong man cases, in which one Adolph Beck was twice convicted of crimes he did not commit, a Royal Committee of Inquiry in 1904 concluded: "[E]vidence as to identity based on personal impressions, however bona fide, is perhaps of all classes of evidence the least to be relied upon, and therefore, unless supported by other facts, an unsafe basis for the verdict of a jury” (Borchard, Convicting the Innocent, at 271-272, n 6 [1932]).
The late lamented Justice Felix Frankfurter in 1927 exclaimed: "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” (Frankfurter, The Case of Sacco and Vanzetti, at 30 [1927]).
In 1945, the New York State Judicial Council, speaking of a legislative report concerning eyewitness identification, noted that in the nineteenth century a student of the problem was asked what remedy would prevent erroneous identification. The only reply the young man could make was: " 'It lies alone in caution and prudence. Observation and sad experience admonish courts and juries to the use of utmost care, caution and prudence’ ”. (NY Judicial Council, Fourteenth Annual Report and Studies, at 236.) The Judicial Council’s report then inquires: "Must the twentieth century render the same response? In view of the great advances made in the field of psychology it would seem that we should know more about the psychological factors which affect identification. In light of this knowledge, it might be possible to erect legislative safeguards which would prevent or minimize the recurrence of the tragic errors which have characterized the subject of identification testimony”. (Ibid., at 236-237.)
Now, as we stand on the threshold of the twenty-first century, the question remains unanswered and the problem looms as large as ever at a time when, it would appear, the strictures of the trilogy, United States v Wade (388 US 218), Gilbert v California (388 US 263), and Stovall v Denno (388 *773US 293), have been largely diluted and dismantled. Under these circumstances, for how long can we continue to reject, for example, scientific help in this sorely troubled field? That, however, is for another day. So let us proceed to a patient, cautious and prudent evaluation of the facts before us.
FACTS
It was about 6:00 p.m., on the evening of Saturday, November 10, 1979, when Carol Howard, a college-trained professional artist, with extensive experience in observing and noting features and details, was returning home from the movies. Outside it was dark and, as she let herself into the vestibule of her Brooklyn Heights home, there was no interior light. However, a street light, directly in front of the house, enabled her to see, as she moved towards the inside door, that a man, a total stranger, was standing in the outer doorway about four or five feet away. When she denied his demand for her jewelry a pushing match ensued, and, because the two were now "a few inches” apart, Mrs. Howard clearly saw the culprit’s face and the four- to five-inch knife in his right hand, with which he cut the straps of her shoulder bag and then stabbed her in the stomach. With the purse and its contents clutched in his hand, the perpetrator ran from the building, crossed to the other side of Willow Street, turned into Pierrepont and ran towards the Promenade.
Despite her serious wound, Mrs. Howard followed behind the culprit screaming and yelling at him. "I had screamed so loud that everybody had called 911”, she said. Within a minute or two the police arrived and, according to her testimony, she gave them a somewhat detailed description of her attacker.
Within minutes, the complainant was taken to the Long Island College Hospital in the first police car that responded to the alarm. Officers Ryan and Romano arrived in the second police car and later testified for the People. It is clear they never talked to Mrs. Howard at the scene and had to rely entirely on a description of the perpetrator received on their police radio. From the trial transcript, it appears that Romano’s memo book read: six feet male Hispanic, black jacket. Ryan’s memo book showed: male Hispanic, black jacket, light blue tee shirt and blue dungarees. No reference was made in Ryan’s memo book to height, nor to weight, nor to moustache, nor to beard (indeed no facial description of any kind), nor to sneakers or running shoes.
*774Immediately Ryan and Romano backed their radio patrol car out of Willow Street, and turned into Pierrepont in the direction taken by the perpetrator. As Romano testified "We were looking * * * [for] [o]ne male Hispanic, approximately six feet, black jacket; that’s all I had at the time”. A very short time later, an additional radio message advised that the perpetrator had a moustache, but no mention was made of a beard.
The two officers had proceeded less than a half a block to Pierrepont and Columbia—the very next corner—when they spotted a man walking in the direction of the crime scene accompanied by a 12-year-old child. The man appeared to fit the general description—male, Hispanic, approximately six foot, black jacket—and when questioned he promptly responded that he was coming from his uncle’s place, the Standish Arms Hotel, down the block. Immediately, he was taken into custody and rushed to the complainant’s bedside in the emergency room of the Long Island College Hospital. Mrs. Howard was told that the police were coming and as the curtains enclosing the cubicle where she lay were separated, she saw the appellant standing between two policemen who were holding him by the arms. His hands were held behind his back as though he were wearing handcuffs, and, indeed, complainant testified that she thought appellant was in cuffs. In substance, the complainant uttered the haunting words of Wade (United States v Wade, 388 US 218, supra), "That’s the man”, and appellant was promptly arrested.
No lineup was ever held.
We turn now to the events that transpired after the police left Pierrepont Street and we listen in amazement as a totally credible alibi surfaces almost instantaneously! Appellant’s 12-year-old companion, Robert Lopez, who turned out to be his cousin, was told to "beat it”. The boy ran up to the apartment house on Pierrepont Street where has father worked as a janitor and blurted out the story of appellant’s arrest. His father telephoned his wife, Carmen Lopez, but neither one could understand what it was all about.
Shortly thereafter, the police arrived at the Lopez’ apartment and searched it in vain and found nothing.
Carmen Lopez, aunt of appellant, immediately went to the 84th Police Precinct and there and then told the police the same story she told at the trial. In substance, she said that the defendant and his wife and two children had been at her *775home at the Standish Arms Hotel on Columbia Heights between 5:30 p.m., and 6:00 p.m., on that afternoon. She said she prepared food for her husband’s supper and put it in a bag for young Robert to take to him inasmuch as he worked in a nearby building on Pierrepont Street. At the trial this story, which sounds too true to be so quickly concocted, was recited under oath by Carmen, her son Robert, and the appellant. Extended cross-examination failed to impeach this testimony in any substantial way.
It is worthy of note that when appellant was stopped and questioned by Ryan and Romano, his young companion was carrying a paper bag. Apparently accepting the usual police version that they had the "right man”, the prosecution made no effort to disprove appellant’s alibi. What was in that bag? Was it, as appellant and his witness averred under oath, edible food? Surely this would tend to further corroborate the alibi. On the other hand, as bizarre as that might be, did the bag contain proceeds of the crime just committed? Would this not utterly destroy and disprove the alibi? Alas, no steps were taken to ascertain the facts which would go a long way to convict or acquit.
It is also interesting to note that a physical examination of appellant failed to disclose any blood on his body or clothing nor was the knife or any proceeds of the crime ever found. Moreover, complainant’s stolen credit cards were used by an unknown person three days later.
At the trial, the only real evidence linking appellant to the crime was the hospital showup and the in-court testimony of the complainant. The prosecution called Nancy Pearsall, complainant’s neighbor and long-time friend, but she could only say that she believed she recognized appellant as the man she saw in the hospital at complainant’s bedside.
She testified that the man she saw running down the street was a tall man with black hair, wearing a black jacket and blue jeans, but she noticed neither a moustache nor a beard. She said that he had dark hair and high cheek bones, but with commendable honesty she carefully refused to make a positive identification.
On cross-examination, the best this witness could say was:
"A. I believe that this is the man that I saw has [sic] the same general appearance, same height, same build, same hairdo, same outfit * * *
"Q. You’ve identified him as the person you saw in the *776hospital. What I’m asking, to be straight forward; are you saying you’re identifying him as the man you saw running down the street?
"A. I believe it is the same man” (emphasis supplied).
Not only was the witness not sure that appellant was the man she saw in the emergency room, she could say only that she believed he was the same man.
Worse than that, she was far from positive that appellant was the man she saw running from the scene of the crime. She only said that he had the same general appearance, same height, same build, same hairdo, same outfit. What deadly and totally insufficient testimony.
Despite it all, the jury found the accused guilty as charged, and my colleagues have now voted to affirm that judgment. These decisions cause me great concern because, in my opinion, there is here presented a classic affirmation of what students and forensic scientists have long since determined: there is no evidence presented in our courts which is so unreliable as eyewitness identification and, by the same token, there is no testimony, again generally speaking, which is more readily accepted by petit juries.
We now turn to examine and evaluate the accuracy of Mrs. Howard’s testimony and it is essential to note that the record strongly suggests that she is, by any standard, an outstanding woman of complete integrity and total honesty. A student at Skidmore College and Pratt Institute, she studied art both here and abroad for many years and she readily conceded that she was trained to observe features and details and remarked that she often painted from live models. From her emergency room bed, under grossly suggestive circumstances, she positively identified appellant as her assailant. No one could or would infer that she spoke less than the truth and thus we are brought face to face with the most frustrating and difficult problem that arises so often in pure eyewitness identification cases. In the main, such witnesses are thoroughly convinced, and honestly so, that they are telling the truth and most, if not all of them, as noted in People v Daniels (88 AD2d 392, 400), would pass any lie detector test with flying colors. At the crucial moment of the assault, it was complainant’s testimony that she and her assailant, both of about equal height, were facing each other in a well-lighted area only inches apart (eyeball to eyeball, to use a hackneyed expression), and that she clearly saw the culprit’s face and the four- to five-inch *777knife held in his right hand (minute details)! She described with considerable particularity her assailant’s height, weight, the color of his clothes, all possible aspects of the appearance of the hair on his head, and noted that he wore a moustache. Such a general description could be applied to many young, tall, Hispanic men.
A mug shot, taken on the day of the crime, shows appellant with a goatee beard. Although not in evidence and, hence, dehors the record, under the circumstances, it can and should be viewed. It shows that the goatee beard is black in color and clearly visible to the naked eye, especially in a lighted area. By her own testimony, Mrs. Howard affirms these conclusions. The record reads:
"Q. He has a beard?
"A. Yes.
"Q. It’s a noticeable type beard. I’m indicating the defendant has a goatee. It’s a rather noticeable beard, or goatee; is it not?
"A. Yes, it is.
"Q. And that’s the type of beard of goatee or facial hairs that a lady, woman of your experience, type of work you do, would recognize pretty much; isn’t that correct?
"A. Yes. I’m just amazed. I really didn’t register it carefully”.
A careful examination of that mug shot indicates that appellant had a moustache that curves down into the beard. That Mrs. Howard could see, as she noted, the moustache and not the beard is beyond comprehension. Could it be that she did not see the goatee because appellant was not the perpetrator? Could not Mrs. Howard make an honest mistake? Did the People sustain the heavy burden that, under these circumstances, the law demands?
THE LAW
THE SHOWUP
The majority concludes that the showup, with all its attendant horrors, constituted, at its worst, harmless error. I disagree. As already noted, appellant was displayed to complainant in her hospital room, flanked by two uniformed police officers holding his arms, with his hands behind his back as though in handcuffs. Under no view of the evidence do I think it can be said that the showup was not suggestive, and any *778reference to this suggestive showup should not have been allowed at trial (see, People v Adams, 53 NY2d 241; People v Cobenais, 39 NY2d 968).
AHEAD, A HEAVY BURDEN, INDEED!
The single most important error facing us, however, is the total failure of the People to carry the enormous burden, as required by law, of proving, by clear and convincing evidence, the existence of an independent basis for an in-court identification untainted by the prior suggestive showup (see, United States v Wade, 388 US 218, 240, supra; People v Ballott, 20 NY2d 600, 606-607). The factors to be considered in making this determination were clearly established by the Supreme Court in United States v Wade (supra, at p 241). Clearly the complainant had an excellent opportunity to observe the assault upon her. However, from the record it is clear, at least to me, that there exists a sharp and decisive discrepancy between the complainant’s description of her assailant compared to appellant and his real-life actual appearance. Complainant, as already noted, totally failed to see a black, clearly visible goatee beard. Admittedly, the findings of fact made by the hearing court are entitled to great weight (see, People v Duncan, 75 AD2d 823). Nevertheless, here the hearing court’s only rationale for finding an independent source—that it was "very favorably impressed with [complainant], and with her intelligence, and with her powers of observation”—is a total irrelevancy because here we are confronted not with deliberate error but a highly possible, and very likely, human error. I cannot agree that sufficient proof was adduced to support, by clear and convincing evidence, a finding of an independent source. Hence, no in-court identification should have been permitted (see, People v Boyce, 89 AD2d 623; People v Taylor, 68 AD2d 864).
IDENTIFICATION CHARGE
Although the charge on identification might be acceptable under some earlier Court of Appeals decisions (see, People v Whalen, 59 NY2d 273; People v Seppi, 221 NY 62), here, where the only real evidence linking appellant to the crime was the one highly questionable and inconclusive eyewitness identification and the appellant presented a highly credible alibi, justice can be heard crying from the rooftops that this is one of those appropriate cases, as the Court of Appeals noted in *779People v Whalen (supra, at p 279), that nisi prius should have "exercise[d its] discretion by giving a more detailed charge”. It bears noting that the identification charge recommended in People v Daniels (88 AD2d 392, 401-402, supra), unlike the one at bar, emphasizes the possibility of unreliable identification testimony and lays out precisely the factors to be considered by the jury in evaluating that testimony, including the mental and emotional state of the witness and any distinctive features which the eyewitness would be likely to remember. Here, the complainant, who had just been subject to a most harrowing and emotionally disturbing experience, was still able to observe and note a myriad of details: height, race, hair around the ears and the back of the head, and a moustache, and yet somehow totally failed to see a clearly visible black goatee beard. The court’s charge totally failed to even invite the jury to consider these salient and perhaps decisive facts. In my opinion, this constitutes reversible error.
ALIBI CHARGE
Moreover, as suggested earlier, the court’s charge on alibi was fatally flawed. With regard to that defense, the court instructed the jury:
"In all criminal cases, members of the jury, the defendant never has the burden of proof. The prosecution at all times has the burden of proof, and the prosecution must prove a defendant guilty beyond a reasonable doubt.
"If the evidence as to the alibi raises a reasonable doubt in your mind as to whether the defendant was present at the time and place, when and where the crimes alleged in the indictment were committed, or that he aided, abetted and participated in the commission of the crimes charged in the indictment, then the defendant is entitled to have that testimony fairly treated like any other testimony in the case, although, he is not obliged to establish that it was impossible for him to have committed, aided or participated in the commission of the crimes charged in the indictment. If under the evidence tending, if true, to prove the alibi, it may be possible for the defendant to have committed the crime. It is still for the jury to determine whether, if that evidence is true, he availed himself of the possibility afforded him to commit the crime.
"If the proof as to an alibi, when taken into consideration with all of the other evidence in the case, raises a reasonable *780doubt as to the defendant’s guilt of the crimes, the defendant is entitled to an acquittal of all of the crimes charged in the indictment”.
The charge, as delivered, is virtually identical in thrust and substance (in haec verba), if not precise verbiage (ipsissimis verbis), to the one that the Court of Appeals held to be inadequate in People v Victor (62 NY2d 374, 376-377). As the high court noted in that case (supra, p 377), such a charge tends to shift to the defendant the burden of proving the alibi defense through the use of such phrases as: " 'If under the evidence tending, if true, to prove that alibi it may be possible for the * * * defendant to have committed the crimes, it is still for the jury to determine whether, if that evidence is true, he availed himself of the possibility afforded [him to commit the crime]’ ” (emphasis supplied), and a Trial Judge must unequivocally state to the jury that the People bear the burden of disproving an alibi beyond a reasonable doubt. It is of little moment, as the majority contends, that the court instructed the jury that the prosecution bore the burden of proof. The charge in Victor also contained the identical instruction, which, when juxtaposed to the erroneous statements of law, was deemed insufficient to convey the proper burden to the jury and to "unequivocally state that burden in the jury charge” (People v Victor, supra, at p 378; see also, People v Chestnut, 99 AD2d 515; People v Landor, 92 AD2d 625; People v Daniels, 88 AD2d 392, supra).
The absolute necessity for a proper Victor charge is buttressed by the facts before us. As noted above, it is apparent that the prosecution made no effort to disprove appellant’s alibi. No attempt was made to examine the contents of the bag carried by young Robert at the time of appellant’s arrest. What was in that bag? Was it edible food? Or did the bag contain proceeds of the crime just committed? Had a proper charge been given, were these not questions that would intrigue a jury? And would not the jury be compelled to reply with resounding voice: no, the People did not disprove beyond a reasonable doubt appellant’s alibi? Nor did they. It is utterly inconceivable that the majority can decline to address this error previously denounced by the Court of Appeals merely because defense counsel failed to record an objection.
people’s summation
Finally, the prosecutor’s summation contained references *781highly prejudicial to defendant. On direct examination, the complainant testified that during the hospital showup she had asked defendant: "You got my purse. Why did you stab me?” On cross-examination, defense counsel, in an obvious attempt to elicit testimony regarding his client’s knowledge of English, inquired of complainant: "You had never spoke sic] to him at the hospital, did you?” In response, the witness volunteered: "Yes, but he said nothing”. Rather then let this fortuitous comment drop, the prosecutor took it up on summation and ran with it, telling the jury: "[The complainant] told you that when [defendant] was in the hospital standing by her hospital bed and she said to him, 'Why did you have to stab me? You already had my pocketbook,’ he just stood there, stared at his feet, kept his head down. I submit to you Ladies and Gentlemen, he was caught”. Whatever the source of defendant’s silence—there was testimony that defendant had difficulty with English and he may have simply not understood the question—the prosecutor’s only conceivable purpose in referring to defendant’s silence and indeed coloring it by supplying details was to invite the jury to infer from it a consciousness of guilt. To seriously suggest that this violent violation of due process is nothing more than harmless error defies all logic and reason and, in itself, does violence to any basic concept of due process and fair play. It has long been held that the use of such tactics impermissibly infringes upon a defendant’s constitutional right to remain silent and not incriminate himself (US Const 5th, 14th Amends; NY Const, art I, § 6; Griffin v California, 380 US 609, 615, reh denied 381 US 957; People v Von Werne, 41 NY2d 584, 587-588). Upon the facts before us, I simply cannot subscribe to the majority’s position that there was no reasonable possibility that this constitutional error contributed to the conviction in what I am convinced beyond cavil is but another unfortunate wrong man case. The haunting comments of Wade (supra) return again.
In conclusion, I find that here the identification evidence was inadequate as a matter of law, the judgment under review should be reversed and the indictment dismissed. Arguendo, even assuming an in-court identification by complainant was permissible, the court’s charge and the prosecutor’s summation (admittedly unobjected to), coupled with trial references to the suggestive showup, served to deprive defendant of his due process rights to a fair trial under the Federal and State Constitutions and warrant a reversal (US Const 14th Amend; *782NY Const, art I, § 6; People v Adams, 115 AD2d 746; People v Daniels, 88 AD2d 392, supra). Accordingly, I dissent.