1. DORAB PATEL, J.-Rambail Khan was shot dead in the house of one Saadullah Khan in village Pir Bai on 9-8-1973. The appellant, his brother and his cousins were tried for this murder by the learned Sessions Judge, who acquitted them by his judgment dated 10-7-1974. The State filed an appeal against this judgment in the Peshawar High Court, and by their judgment dated 27-1-1979, the learned Judges of the Peshawar High Court convicted the appellant for murder and sentenced him to death, but they held that the prosecution had failed to prove the guilt of the other four accused, therefore, they acquitted them. This judgment is The State v. Lal Pasand and 4 others PLD 1979 Pesh. 48As the High Court had set aside the appellant's acquittal, he has filed this appeal under Article 185 (2)(a) of the Constitution.
2. According to the prosecution, the accused including the appellant had chased Rambail Khan from his field into the house of Saadullah Khan. There the appellant's cousin Bashirullah field Rambail Khan in his arms whilst the appellant fired at him and killed him. In order to prove its case against the appellant, the prosecution examined four eye-witnesses, two ofwhom were the brother and the son of the deceased. As both the Courts have held that these two witnesses were thoroughly dishonest, it will not be necessary to refer to their evidence. The other two witnesses were Saadullah Khan and his grand son, Asif Jan, and as we pointed out, the actual murder was committed in Saadullah Khan's house in the presence of these two witnesses. Both Saadullah Khan and Asif Jan said that they had, on the hearing of the sound of firing, come out of their house into the compound of the house. They saw one man (namely the decreased) run into the compound followed by two armed men, one of whom held the deceased whilst the other shot him dead.
3. As the accused were arrested on or about the 21st of December, 1973, the first identification parade was held on the 28th of December, 1973. We depreciate this unexplained delay of one week in holding the parade. But, what was much more unfortunate was that only Saadullah Khan was invited to identify the accused at this parade. His grandson was asked to attend an identification parade after another delay of three months. This is absolutely shocking, and such delay in holding identification parades cannot but lead to a miscarriage of justice. However, reverting to the facts of the case, Saadullah Khan said in his evidence that he had identified the appellant, because he had green eyes and a thin neck. On the other hand, although the appellant had green eyes and a thin neck, Asif Jan was not able to identify him, and further neither of the witnesses were able to identify the appellant's cousin Bashirullah.
4. It will thus be seen that the prosecution case against the appellant rests solely on his identification by Saadullah Khan and as the High . Court sentenced the appellant to death solely on Saadullah Khan's opinion that the person whom he had seen committing the murder on 9-8-1973 was the person whom he had identified on the 28th of December, 1973, the main question in the appeal is whether the opinion of this solitary eye-witness was sufficient to prove the appellant's guilt. The learned Sessions Judge decided this question against the prosecution for two cogent reasons. The first was that the number of other persons intermingled with the accused in the identification parade of 28-12-1973 was not in the proportion a proved by the West Pakistan High Court in Kaim and others v. The State (If, in Wahid Bux and others v. The State (1969 P Cr. L J 1317), and in Shabbir Ahmad v. The State (1972 P Cr. L J 310The second was that Saadullah Khan had first identified the appellant at an identification parade held more than four months' after the murder. Therefore, as he was a complete stranger to the appellant, the learned Session; Judge did not think it would be safe to base a conviction on the witnesses capacity to identify a stranger whom he had seen for the first time in his lift for half a minute or a minute four or five months earlier.
5. Now, the learned Judges of the Peshawar High Court, who reversed the finding of the learned Sessions Judge, have given three reasons for sentencing the appellant to death solely on Saadullah Khan's opinion of his identification of the appellant. The first of these reasons is that Saadullah Khar was a witness of integrity. The second was that although Saadullah Khar did not know the appellant, he must have seen him sometime or the other, because they lived ,in the same village and the learned Chief Justice, who wrote the judgment of the Court appears to have been of the view that the possibility of such chance encounters between Saadullah Khan and the appellant made it safer to rely on Saadullah Khan's evidence. The third was that the witness "had a close up of the gunman," and because he was a retired Superintendent of Police, the learned Chief Justice had no doubt that the witness "had memorised his features" and, therefore, this ruled out the possibility of an erroneous identification.
6. Mr. Ibne A.I placed great stress on the finding of the High Court that Saadullah Khan was an honest witness. However, even if we assume for the sake of argument that the witness was a paragon of truth, this will not help the prosecution case, because it is not the appellant's case that Saadullah Khan had identified him dishonestly and falsely out of spite. On the contrary, the appellant's case is that Saadullah Khan was an old man, who had seen the murderer only for half a minute, or a minute, therefore, this old man could not possibly have identified the murderer after a delay of four or five months, the more so, as the old man had not given a description of the murderer in his statement to the Police, which was given very shortly after the murder. This is the real point which requires examination. Can the possibility of an honest but erroneous identification be ruled out?
7. And, even if Saadullah Khan had not been an old man, we do not see how, in the circumstances of the case, the possibility of an erroneous identification by the witness could be ruled out. It is true that the witness said that he had identified the appellant, because he had green eyes and a thin neck, but in this part of the world there are so many people with green eyes and thin necks. And, further, we do not know whether this explanation of the witness was a form of rationalisation after the event, because he had not given any descrip--petition of the murderer in his statement to the Police. But, if the murder had any physical features which distinguished him from other people whether it be green eyes or a thin neck, it was the witness---s duty as a retire Police Officer to give the description of the assailant whatever it be in hi statement to the Police and as he did not this cannot but react against the entire prosecution case.
8. This is so obvious that the learned Chief Justice has fallen back on the possibility that the witness was able to identify the appellant, because he must have had the opportunity to see him in the village. With all respect to the learned Chief Justice as the witness was emphatic that he had never seen the murderer except at the murder, the view taken by the learned Chief Justice involves the substitution of evidence by conjecture and w~ cannot agree with it. However as far as we can see, the main ground on which the learned Chief Justice based his finding that Saadullah Khan's opinion was sufficient to prove the appellant's guilt, was the assumption that Saadullah Khan had been able to memorise the appellant's features even though he had seen him for a fleeting moment on 9-8-1973, because he was a trained Police Officer.
9. With all respect to the learned Chief Justice, powers of observation vary from individual to individual, and though a trained Police Officer is likely to be more observant than the ordinary man, sometimes, if we may permitted to say so, country bumpkins are extremely observant, whilst & geniuses are proverbially absent-minded. Therefore, we are unable to subscribe to the proposition that a Police Officer has an infallible memory at least in the matter of remembering and identifying the faces of strangers. However, as we do not wish to lay down a wider proposition than is necessary for the decision of this case, we would observe that the position might have been different if the prosecution had led evidence to show that Saadullah Khan was a man with a perfect memory, who had never been wrong in his identification of strangers. But, as there is no such evidence, we do not see how the possibility of an erroneous identification by this witness can be ruled out.
10. It is unfortunate that the attention 'of the learned Chief Justice was not drawn to what was the real danger in the case. But, perhaps this was because of the total lack of research in the country into legal questions, by which we are all handicapped. However, the problem of illegal convictions based on erroneous identification has haunted the conscience of other Judges, therefore, we may, with advantage, point out that according to the report of the Criminal Law Revision Committee (1972) prepared by Lord Edmund-- Davies mistaken identification was "by far the greatest cause of actual or possible wrong convictions". Then a few years later, Lord Devlin was appointed to examine this specific problem of illegal convictions based on erroneous identifications and he made recommendations which we will presently consider. But, before doing so, we would quote from the observa--tions on this problem by Lord Devlin from his famous treatise "The Judge".
11. Lord Devlin begins by referring to some very celebrated cases of totally illegal convictions in the last seventy or eighty years. Thus about eighty years ago a man called Adolf Beck was convicted and Lord Devlin writes :- "Beck was twice wrongly convicted, having been identified in 1896, by eleven witnesses and in 1904, by four. This miscarriage of justice was the goad which finally pricked Parliament into setting up the Court of Criminal Appeal . . . . . In 1912, a man on a charge of murder was identified by no less than seventeen witnesses, but fortunately, was able to establish an irrefutable alibi. In 1928, Oscar Slater, after be had spent nineteen years in prison and after a public agitation in which many distinguished people joined, he had his conviction for murder quashed; he had been identified by fourteen witnesses."
12. Then the learned author refers to the two cases which led the Home Secretary to request him to hold a fresh enquiry. The learned author writes:- "In 1974, two shattering cases of mistaken identify came to light within four weeks of each other. In the first of them Mr. Dougherty was convicted of shop-lifting, having been identified by two witnesses, at a time when be was on an excursion with some twenty other persons. The accidents and blunders which led to his conviction, and to his appeal from it being dismissed, are not relevant here . . . . . After he had served most of his sentence in prison and on a reference back to the Court of Appeal by the Home Secretary, Mr. Dougherty got his alibi evidence before the Court and the prosecution threw up the sponge . . . . . . This was on 14th March, 1974. On 5th April, the Home Secretary discharged with the grant of a free pardon a Mr. Virag from the prison in which he had been for five years. As was conclusively proved in the subsequent inquiry, he had been wrongly identified by eight witnesses, four of them Police Officers, on six different occasions."
13. This brief list of illegal convictions is all the more depressing, because it is clear that errors in identification cannot be prevented by increasing the number of identifying witnesses or by having resort to the evidence of Police Officers, and indeed Virag's care proves beyond doubt the fallacy of assuming that Police Officers have extraordinary powers of observation. What then is the solution of this borrowing problem? In our opinion, the only safeguard against illegal convictions is that the Judge or Judges hearing the case should be conscious of the dangers inherent in the identification of accused by witnesses who are total strangers to them.
14. Reverting however to the learned author, he made two recommenda--tions to the British Government. In this connection, the learned author writes:- "The first recommendation was for an absolute and unconditional rule that the jury should be directed or warned about the dangers of identification evidence. The second was for a general rule that the jury should not be allowed to convict on eye-witness evidence alone. This rule had to be general and not absolute because admittedly there would be exceptions, e. g. When the witness was identifying someone he knew well or who had been under frequent or prolonged observation."
15. As we do not have the jury system, the first recommendation would translated into our procedure, mean that the Judges hearing a case should be aware of the dangers inherent in the identification of strangers by witnesses, who have seen them very briefly and this coincides with our own views on the matter. The second recommendation, however, is very wide, and it is not, necessary for us to decide so wide a proposition, as the question before us is whether it would be safe to base a conviction on the evidence of one identifying witness only, because he was an honest witness. Now, although! The witness was an honest witness, the dangers of errors in identification are so great that in our humble opinion, safe in exceptional circumstances, it would not be safe to base a conviction on the evidence of a solitary eye--witness if the witness has only had a fleeting glips of the assailant. And, as in the instant case, there are no special circumstances, we hold that the learned Chief Justice erred in convicting the respondent solely on Saadullah Khan's evidence.
16. We now turn to the question of the evidentiary value of the identification parade conducted by the police and as we explained, the Sessions Judge had rejected it, because the number of other persons intermingled with the accused in the identification parade was not in the proportion of nine or ten to one as laid down in a series of judgments of the West Pakistan High Court. The attention of the learned Judges of the Peshawar High Court was drawn to these judgments, but the learned Chief Justice overruled the view of the learned Sessions Judge, because he was of the view that the judgments of the West Pakistan High Court did not lay down an inflexible rule as held or assumed by the Sessions Judge.
17. The question is which of these two views is to be preferred. And, whilst we respectfully agree with the observation of the learned Chief Justice, that the judgments of the West Pakistan High Court cited earlier, do not lay down an inflexible rule about the number of persons to be intermingled with the accused, the prosecution has always to explain its failure to conform to a ratio accepted by a long line of Judges. And, we may, further point out that the ratio of ten other under trials to one accused had been prescribed by the Government of the United Provinces in its Government Orders for the Police (See paragraph 23 of the judgment in Asharff and another v. The State AIR 1961 All.
153. Therefore, it would be monstrous for the Courts to permit a departure from a rule accepted by Government in the absence of some explanation by the prosecution. Now, in the instant case, the explanation of the State is that there had been a joint identification parade with the result that it was not possible to find a sufficient number of other persons to be intermingled with the accused.
18. And, the learned Magistrate who conducted the identification parade appears to have assumed that all the five accused had to be identified. That assumption was wholly erroneous, because Saadullah Khan had seen only two of the five assailants.
19. However, on the footing that five persons had to be identified, it would have been unreasonable to mix them. With fifty other persons, because such a large number of persons could only have confused the identifying witnesses. Therefore, the proper course in such cases is to have separate identification parades for each accused. And, no explanation has been given in the instant case for not holding separate identification parades.
20. Mr. Ibne A.I, however, pointed out that as only two accused had to be identified, the real ratio between the accused and the other persons was eight to one,, therefore, the departure from the normal rule was not such as to react against the evidentiary value of the identification parade. A departure from the ratio recommended by the Courts can be justified by the prosecution because we would repeat that the ratio of nine or ten persons to one accused is not an inflexible ratio.
21. However, in the instant case, there was absolutely no explanation for departing from the normal rule, and, on the other hand, in the circumstance of the instant case, a departure from the rule was hazardous and fraught with danger to the accused, because the. Prosecution case rested , on the evidence of a solitary witness, who had seen the assailant for a' brief moment. , Therefore, on the facts of the case, we agree with the view -of the learned- Sessions Judge that it would not be safe to rely. On the corroboratory evidence of the identification parade on the further ground that the number of strangers mixed with the accused was no in accordance with the -rule prescribed by the -West Pakistan High Court in the cases cited earlier. Accordingly, on this ground also, the conviction of the respondent is illegal. We, therefore, allow his appeal, set aside his conviction and- direct that he shall be released forthwith unless required on any other charges..
…and 87 more citing cases