The Queen v Crawford

The respondent was convicted by the trial judge sitting without a jury of the offence of possession of an unlicensed firearm. The Court of Appeal quashed the conviction. The Crown appeals against that decision.

At about 03.00 on 18 November 2011, PC Gordon, on duty in a car park outside a club at closing time, received a complaint from two people leaving the club to the effect that the driver of a Nissan car (the respondent) had a gun. When the officer went to the car, the accusation was repeated in front of the respondent and one of the complaining men threw a punch at him through his car window. The respondent thereupon drove off at speed, evading the officer’s police car which partially blocked his way. PCs Bradley and Rabess were also in the vicinity in a patrol car driven by Rabess. Alerted by PC Gordon to the complaint about a gun, they gave chase. Not far on, the respondent crashed his car into the barrier by a roundabout and ran off across rough bushy land to the side of the road. PC Bradley chased him and eventually tackled him to the ground. There had been a passenger in the car, who was recognised by PC Rabess but who also ran away and was not detained.

PC Bradley’s evidence was that during the foot chase the respondent looked over his shoulder at his pursuer, and then took a silver handgun from his waistband and threw it into the bush away to his left hand side. Shortly after that, PC Bradley caught him. The respondent thereupon said to him “That guy had a gun”.

A search was mounted by other officers in the area of bushes indicated by PC Bradley. Additional lights were fetched to supplement the street lighting in the search, as was a metal detector. PC Soto gave evidence that at about 04.30, using the metal detector, he found a silver luger pistol in the undergrowth in the area indicated. The respondent was charged with unlawful possession of that gun.

Beyond his initial assertion to PC Bradley, the respondent said nothing else. He declined to answer questions put by the police and in due course did not give evidence at his trial. On his behalf, and no doubt carefully consistently with his instructions (if any), the case put to PC Bradley was that he was mistaken about what he saw thrown away. It was suggested that he had assumed it was a gun because of the complaint which had been made in the club car park, and/or because a gun was found in the bushes, and/or because he knew that the respondent was subject to a “long arm authority”, as a result of having been under suspicion of previously being in possession of a firearm. On the basis of the witness statement and notes of PC Rabess, it was suggested that he had told that officer only that he believed that what had been thrown had been a gun and/or that he had said that the object seen resembled a firearm, but he disagreed. His evidence was that he had been about five to six yards behind the respondent, that the latter had held the gun in his hand before throwing it, as a relay runner holds a baton, and that he had seen it travel something like 30 feet through the air. It was not suggested to PC Bradley that he had not seen something thrown, nor was it suggested what it was that he had seen, if it were not the gun. At no stage during the trial was it suggested on behalf of the respondent that the gun had been brought to the scene by police officers and planted.

Sometime after the gun was found in the bushes, PC Rabess took some photographs, one of which was of it in situ. Sometime later, probably about 40 minutes afterwards, a scenes of crime officer, PC Taylor, arrived and amongst his other duties also took photographs of the gun in situ. Both photographs were before the court. That taken by PC Taylor showed the breech of the pistol open (demonstrated by the elevation of the arm and confirmed by PC Taylor) and a live round lying immediately above it next to the barrel. That taken by PC Rabess also showed the breech open, but did not show the round. PCs Soto and Rabess were both questioned about the discrepancy. Both said that no-one had interfered with the gun between the taking of the two photographs. PC Soto said that although the bullet was not visible in the photograph of PC Rabess, it had been there from the beginning. PC Rabess, however, said that it did not appear to be visible in the photograph and that he did not recall any bullet. The photographs also showed a limited difference in the number of strands of foliage lying over the gun; there were perhaps up to a dozen fewer in the photographs taken second by PC Taylor than in the earlier one taken by PC Rabess, but this was not adverted to in the evidence and no-one was asked whether the foliage had been disturbed to give a clearer view to the camera or for any other reason. PC Bradley said that he remained by the road and went to look at the gun in situ when it was found. He was not asked about the photographs or the presence or absence of a round.

The gun was examined for fingerprints and DNA. There were no fingerprints. There was a mixed DNA profile, but the respondent could not have contributed to it. Later, one component was identified as belonging to another known man.

The trial judge directed himself with some care as to his function when trying a criminal case on his own. There has been no complaint about his self-direction. His judgment was some 344 paragraphs long and thorough; it incorporated his summary note of the evidence, witness by witness, together with his review of the arguments and his conclusions. His decision that the respondent was proved to be guilty was based substantially on his assessment of the evidence of PC Bradley as accurate and reliable.

There has been no dispute before the Board as to the proper role of an appellate court when reviewing a decision of a trial judge which amounts to a finding of primary fact based upon his assessment of the credibility and reliability of witnesses whom he has seen and heard. It is well established that an appellate court should recognise the very real disadvantage under which it necessarily operates when considering such a finding only on paper. There are many statements of this principle. It is enough to set out the formulation of it by Lord Sumner in The Hontestroom [1927] AC 37 at 47-48:

The present case is a good example of the difficulty necessarily facing an appellate court. The trial had occupied something over six working days. The judge had been immersed in the evidence in a way which could not be replicated in the Court of Appeal. He did not have merely the written words of the witnesses. He had seen the way in which the words were spoken and challenges were met, and he had been able to read the faces and body language as well as what could be put on a page. That sometimes exaggerated general claims may be made for the ability of experienced judges to determine the truth solely by assessing the demeanour of witnesses does not alter the fact that part of the judicial function is to read the witness as a whole, nor that the demeanour may sometimes contribute very significantly to the correct conclusion. A transcript cannot provide the same opportunity. In the present case, the Court of Appeal did not even have a full transcript, but only the judge’s note of evidence, as set out fairly extensively in his judgment.

The Court of Appeal did not remind itself explicitly of this well understood role of an appellate court, but there is no reason why this should be rehearsed in every case, and it would add only formalism if it were required. It based its decision upon its conclusion that the judge had erred in principle. The question in this further appeal is whether it was right so to conclude.

The Court of Appeal held that the judge had erred in his treatment and analysis of the evidence in two respects:

he failed to address the impact of the evidence of the photographs and the discrepancy between them; and

he misdirected himself as to the significance of the absence of the respondent’s DNA on the pistol.

As the Court of Appeal held, the discrepancy between the two photographs raised the real possibility that the gun had been moved or disturbed in some way, and that the round had been positioned by it, notwithstanding the evidence of both PC Rabess and PC Soto that it had not been touched. That is not the only possible explanation for the discrepancy, since no one asked PC Taylor whether the foliage had been parted or moved to get a better view, and the open breech might or might not have spilled a round, but it was certainly a possibility which was raised. If that is what had happened, it might or might not be sinister, but it went to the reliability of the testimony of those two officers. Indeed, the court might have added that PC Soto asserted that the round was there all along, whilst PC Rabess did not remember it. The Court of Appeal adverted to the possibility that the officers might have “improved” the evidence by moving the gun so that it appeared to lie “on top of” the foliage rather than “beneath” it, this being, it was suggested, more plausibly consistent with the gun having been thrown into the bushes. The court held that the judge had not properly addressed this possibility. It said this:

The judge, however, did address the evidence of the photographs. He had himself asked questions of PC Soto about them. In his judgment he referred to the differences both as to the presence of the round and the “grass coverage”. He also recorded the submission of counsel for the defence that this scene evidence was the most unfortunate part of the Crown’s case, that somebody must have moved the gun, and that what had occurred was “a set up”. He recorded the further defence submission that whilst PC Bradley may have been honest, it was necessary to ask whether his evidence was undermined by this aspect of the case. He reminded himself that there had, additionally, been errors of labelling by the scenes of crime officer. He went on to analyse the extent of the differences between the two photographs, pointing out that the breech arm was up in both of them, and concluding that there was no evidence of soil or vegetation attaching to the gun, such as might suggest that it had been in the bush for some time, whilst the evidence was that it was in full working order, if old and with an ill-fitting magazine. Further, he did directly address the possibility described by the Court of Appeal as “improvement” of the evidence, saying that it was his conclusion that the presence of the round by the barrel did not in fact strengthen the Crown case at all.

The judge directed himself that what mattered in the case was the evidence of PC Bradley. If it was truthful and reliable, the respondent was guilty; if there was any doubt about Bradley having seen a gun, the respondent was to be acquitted. The judge gave himself a self-direction that the evidence of seeing a gun should be subjected to scrutiny analogous to evidence of facial recognition and went through the process required in such cases by R v Turnbull [1977] QB 224. In that, he was certainly not unfair to the respondent; the peculiar risks of mistaken facial identification do not apply to the same extent to evidence of sightings of other objects, such as a motor car (R v Browning (1991) 94 Cr App R 109), or its numberplate (R v Hampton [2004] EWCA Crim 2139; [2006] Crim LR 60). He rejected the contention that PC Bradley had seen the respondent pulling up his trousers, rather than extracting what turned out to be the gun from his belt, observing that there was no evidence to support any looseness of trouser and positive contrary evidence that the respondent was wearing a belt. He accepted PC Bradley’s evidence that although not “breathing down [the respondent’s] neck” he was close enough to see the gun in his hand and its silver colour. He concluded that PC Bradley was accurate and reliable in his evidence of the gun in the respondent’s possession, and in consequence reached a verdict of guilty.

There can be no doubt that the judge was right to direct himself that the case depended on the evidence of PC Bradley. It is plain that his conclusions about that witness depended not only on what he had said, as reproduced to a limited extent in his note, but also on his judgment about him as a whole. Unless there is some error of principle, or material to show that he was plainly wrong, this is exactly the kind of assessment which ought not to be disturbed by an appellate court without clear reason. The criticism of the Court of Appeal was that disturbance of his conclusion was necessary because he did not make any finding as to the explanation for the photographic discrepancy, and nor did he set out how that evidence impacted on his assessment of that crucial witness.

As a general proposition, it is axiomatic that a trial court is not required to make findings upon every question raised in the course of the trial. It is required to make findings in relation to those matters which it is necessary to resolve in order to reach a conclusion on the issue before it. This general proposition is as true of criminal cases as of civil. It is illustrated by the direction which is invariably given to juries in criminal cases, that there may well be loose ends or unresolved questions, but that they do not have to be resolved unless necessary to decide whether the defendant is guilty or not shown to be so. A judge sitting alone is of course expected to articulate his reasoning issue by issue, in a way in which a jury is not, but the general proposition here set out applies equally to him, for it sets the parameters of which issues need to be resolved. There is no reason to think that the Court of Appeal in the present case ignored this general proposition. Its conclusion was that it was necessary to resolve the explanation for the photographs in order to determine whether PC Bradley’s evidence was reliable.

This, however, does not follow. This was not a case in which the issue was whether the gun might have been brought to the scene by one or other of the officers and planted in the position photographed. That was not at any stage even ventilated as a possibility; it would have involved (at least) a significant conspiracy going beyond the officers at the scene, access to a suitable gun in the middle of the night and a pantomime of sending for lights and a metal detector. The Court of Appeal rightly made it clear that the risk which troubled it was that PC Bradley may have seen what he expected to see and might have been “putting two and two together and making five”. What mattered in this case was therefore not what had or had not happened to the gun after it was found, but whether PC Bradley had seen it in the hands of the respondent a little earlier. Even if there was reason to doubt the truthfulness of one or more of PCs Soto, Rabess or Taylor, there was no basis for supposing that any untruthfulness affected, one way or another, the accuracy of PC Bradley’s observation. That fell to be evaluated independently, and that is what the judge did. Even if the difference of something under a dozen strands of vegetation crossing the gun justified the inference of its position having been altered so that it lay on top of rather than beneath the foliage, and even if, perhaps more plausibly, its position had been “improved” by the bullet being placed beside it, that simply did not bear on whether what PC Bradley had seen thrown had been this gun or some other unidentified object. The first was a question of truthfulness; the second was a matter of accuracy of observation.

It may be that it might have been clearer if the judge had spelled out the possibilities of what might have happened to the gun in the 40-minute period and then had stated explicitly his conclusion that, even if it had, it made no difference to his evaluation of PC Bradley. That, however, is a counsel of perfection. The judge made it absolutely clear that he was considering the submission that the discrepancy between the photographs undermined the evidence not only of PCs Soto, Rabess or Taylor, but of PC Bradley. He amply clearly rejected this contention. He considered and rejected the possibility that the gun had been there for some time. Moreover, he said expressly that in his view the presence of the round did not improve the Crown case. He was entitled to conclude that the presence or absence of the round by the barrel was of no assistance in asking whether the gun had been thrown by the escaping respondent, and that it was not easy to see why any police officer should think that, if placed neatly alongside the barrel, such a round would be supportive of the evidence that the gun got there by being thrown 30 feet or so through the air.

The Court of Appeal rightly identified the risk that PC Bradley’s evidence that what he saw was a gun might have been the product of assumption. He might have been expecting to see a gun, because he had had relayed to him the complaint about a gun which other people had made to PC Gordon in the club car park. But the judge did not ignore this. At para 313 of his judgment he addressed this possibility, reminding himself also that the complaint had not been proved in the trial and remained mere hearsay. Indeed he reminded himself also that PC Bradley knew that the respondent had previously been suspected of possessing a gun and was as a result on a long arm authority, which was a further reason why he might have made an assumption. Having thus addressed the risk of assumption, the judge went on to examine seriatim the strengths and weaknesses of PC Bradley’s sighting of the object thrown, in the light of this risk. That was the correct approach. There was no basis for an appellate court to depart from his conclusion.

There was in this case no requirement in law for corroboration. Nor was it a Turnbull case of facial identification in which it is wise, although not mandatory, to look for support. The judge would have been quite entitled to accept the evidence of PC Bradley even if there were nothing to support it. But it was of course relevant to look to see whether there was any such support. Before the Board, Mr Perry QC for the Crown suggested that such support could be found in (1) the fact that PC Bradley arrested the respondent immediately on apprehension for possession of a gun, (2) the absence of any evidence from the respondent, (3) the respondent’s reckless flight in the face of an accusation in the club car park of having a gun, persisted in even after crashing the car and (4) his plea to the arresting officer that somebody else had a gun.

The first of these constitutes no relevant support. It shows that PC Bradley was consistent, and that he believed there was a gun before one was found, but not that he was not mistaken. The second might have justified, according to the law of the Cayman Islands, an adverse inference against the respondent under section 18(b) of the Evidence Act, but the judge does not appear to have relied on such an inference, understandably limiting himself to the observation that the Crown evidence went without anything from the respondent to set against it. But the third and fourth were matters which were plainly capable of supporting the accuracy of PC Bradley’s observation. The judge relied on the fourth. As to the third, true it was that someone had aimed a punch at the respondent in the club car park, but it might be thought that that would hardly explain headlong flight when police officers were present in some numbers, still less reckless attempts to evade not the club patrons but the police. As to the fourth, the defendant’s solitary utterance after arrest did indeed suggest that what was in the forefront of his mind at the moment of detention was a gun, and that he was attempting to cast responsibility for it onto someone else, apparently his passenger. Moreover, in addition to the matters discussed in Mr Perry’s submissions, once it be established, as it was, that PC Bradley had seen something thrown into the undergrowth, the very fact that a gun was found in the area where he said the object had landed was support for what he said he had seen, and had accused the respondent of having before it was found. The judge perfectly properly relied on this also.

Accordingly, although the Court of Appeal held that there was no support, the Board takes the clear view that in fact there was, and that the matters relied upon by the judge were part of it.

The judge reminded himself of the DNA findings, including the mixed profile to which others but not the respondent had contributed. He directed himself that the absence of the respondent’s DNA (and fingerprints) was evidence which supported his case and which he took into account as such. He further directed himself that a person who handles a gun may not always leave either recoverable fingerprints or DNA upon it. Thirdly the judge directed himself that it followed that the presence or absence of DNA or fingerprints was not determinative of the guilt or innocence of any defendant. The Court of Appeal’s judgment contains a simple typographical error inverting the second self-direction, to suggest that the judge held that DNA was necessarily left by someone who handles a weapon, but this can be ignored; that court undoubtedly correctly appreciated the basis on which the judge had proceeded.

The Court of Appeal then held the judge had fallen into error of principle in thus directing himself. It said this:

Firstly, as has been common ground before the Board, the judge’s self-direction that DNA is not necessarily left by a person who handles an object was entirely correct. Such a warning is necessarily given to juries in every case in which the absence of such scientific evidence is asserted, as it often is, to be evidence demonstrating that the defendant did not handle the object or was not present on the relevant occasion. And since DNA once deposited may sometimes persist and sometimes may not, it is equally true that the presence of someone else’s on the object in question is evidence suggesting that that person has at some point handled the object, but does not tell anyone that the defendant has not done so also. The Court of Appeal referred to a criminal appeal in the English Court of Appeal, R v Mitchell [2004] EWCA Crim 1928: 8 July 2004. In that case, there was not only the absence of the defendant’s DNA in the mouth of a woman who had suffered oral rape, but the presence of the DNA of a male who could not have been him, in a place (her mouth) where it was hardly likely to have been present unless the donor was responsible for the rape. That is quite different from the real possibility of successive people handling an object such as the gun in this case.

Secondly, the judge did not treat the absence of DNA, still less the presence of someone else’s, as “neutral”, as had the judge in Mitchell. On the contrary, he directed himself that it went to give some support to the respondent’s case. What he told himself was that it was not determinative. He was right.

Thirdly, as has also been common ground, there was no evidential basis at all for the proposition that DNA was more likely to be retained on the gun either because the respondent’s hand was hot (if it was) or because he had a firm grasp of it. This kind of question, which is notoriously difficult, calls for, and is occasionally given, detailed scientific consideration by forensic scientists with extensive experience of the probabilities of DNA deposit and retention, who consider the limits of knowledge as to the point, together with a number of variables, such as the tendency of an individual to shed DNA-containing material, the tendency of the object to retain it, the probabilities of loss and so on. In the absence of this kind of evidence, the proposition is simply speculation.

Lastly, the Court of Appeal criticised the judge for not reverting to the DNA point when stating his conclusion in the penultimate paragraph of his judgment. But he had more than fully stated how he approached the point a little earlier in his judgment. In the penultimate paragraph he was simply summarising the positive factors leading him to conviction. A judge is entitled to structure his judgment in the manner of his choosing. The judge had set out the relevant factors impeccably. His conclusion was not vitiated by not repeating it at any particular stage of his judgment.

Neither in respect of the photographic evidence nor in respect of the DNA evidence were the criticisms of the Court of Appeal justified. The trial judge approached the case in respect of both matters correctly. There was no basis for departing from his verdict, which was based on his assessment of PC Bradley’s evidence and the elements of support available for it. The Board must, in those circumstances, humbly advise Her Majesty that the Crown’s appeal should be allowed and the respondent’s conviction restored.