Van Dongen & Anor, R v

The appellants, Anthony and Mitchell van Dongen, are brothers. Anthony van Dongen is 31. Mitchell van Dongen is 24. They each appeal against their conviction of murder by a unanimous jury at the Central Criminal Court on 14th May 2004 before HHJ Forrester. They were each sentenced to life imprisonment with a recommended minimum term of 12 years 11 months. Contingently upon the outcome of their appeals against conviction, there are appeals against the length of their recommended minimum terms of imprisonment. This judgment deals with the conviction appeals.

The trial was a retrial, necessary because the judge at the first trial became indisposed.

The person who died was a 26 year old engineering student, Myo Thaung. Mitchell van Dongen admitted causing his death by kicking his head. His defences were self-defence, a lack of the necessary intent for murder and accident. Anthony van Dongen’s case was that he was unconscious for most of the incident. He regained consciousness towards the end of the incident and may in self-defence have struck Alan Knapp, a friend of the man who died. But he had nothing to do with the injuries from which the victim died.

A defence of provocation was not advanced in evidence or forensically for either defendant. Whether or not to rely on a defence of provocation is a dilemma which often faces those representing defendants in murder trials whose main defence is self-defence or lack of intent. A defence of provocation may be intrinsically inconsistent with, or may otherwise weaken, other defences. But it is well established that the judge should direct the jury to consider a defence of provocation, even if it is not the defendant’s overt case, if there is some evidence from whatever source from which the jury could find that there was provoking conduct which resulted in the defendant losing his self-control – see for instance R v Acott [1997] 2 Cr. App. R 94.

In the present case, Mr Dein QC, on behalf of Mitchell van Dongen, invited the judge to direct the jury as to provocation. Mr Waters QC for the Crown resisted this. The judge decided not to leave provocation to the jury. Mitchell van Dongen’s main ground of appeal is that the judge was wrong here and that the conviction for murder is in consequence unsafe.

Mr Hadley, on behalf of Anthony van Dongen, adopts this ground of appeal, although he recognises that the nature of Anthony van Dongen’s case, that is that he did not participate in the fatal attack, weakens the ground of appeal in his case. He submits, however, that, if Mitchell van Dongen’s appeal succeeds on this ground, justice requires that Anthony van Dongen’s appeal should also succeed. We are inclined to agree.

There is a second insubstantial ground of appeal which Mr Dein advanced without enthusiasm.

At about 1.30 am on 6th April 2003, police were called to Westferry Road, Isle of Dogs. They there found Myo Thaung unconscious in the road with severe head and internal injuries. After being treated at the scene, he was taken to hospital where he received surgery. But he died two days later.

The prosecution case was that the appellants jointly and deliberately attacked and beat the victim intending either to kill him or to cause him really serious bodily injury. The appellants denied that they had participated in a joint enterprise to attack the victim.

Mitchell van Dongen’s case was that, although he admitted that he had inflicted with his foot the head injuries which caused the victim’s death, he did so accidentally in reasonable defence of himself and his brother. He did not intend to kill the victim, nor to cause him really serious injury. Anthony van Dongen’s case was that he was unconscious for most of the time. He had not intended to kill the victim nor cause him really serious injury. Indeed he had not inflicted any injury to him, although he may in self-defence have struck Alan Knapp.

Alan Knapp gave evidence that he and the victim were flat mates. They had been out that night celebrating his own birthday. They had been for a meal and then went to a bar with two other friends. He admitted that he became drunk. His memory was fragmented and he did not remember earlier parts of what had happened. The victim had drunk less than he had. He appeared in a relaxed and good mood. He became a larger and livelier person in drink, but was not aggressive. They walked home along Westferry Road. He was on the right hand side of the road and the victim was on the left. He became aware that the victim was in a scuffle with two white men. By the time he himself got to the other side of the road, the victim was on the ground and the two white men were both kicking him. The kicks were hard and aggressive, more than one kick from each man. One of the men stopped kicking the victim and came towards him aggressively. He put out his hand in peaceful gesture to try to diffuse the situation. The advancing man hit him on the back of his head with his fist causing him to drop the telephone he had in his right hand. He had already called the emergency services. Both men ran off down the street. He did not recall anyone holding or throwing anything.

The two other people, with whom the victim and Alan Knapp had been out that night, each gave evidence of the victim’s condition and of how in general drink would affect him.

There was further eye witness evidence from five people who were driving past the incident. Four of these were in one car and the fifth in another car which was following the first car. They each gave detailed evidence and were cross-examined. These witnesses were Sarah Darwood, the driver of the first car, a BMW; Louise Everingham, her front seat passenger; Richard King, who was in the back seat behind the front passenger; and Tina Stephens, who was in the back seat behind the driver. Samantha Taylor was in a following BMW.

A fairly short summary of their combined evidence will suffice. There were some differences between their accounts, and some inconsistencies between their evidence in this trial and in the first trial.

The combined effect of their evidence, if the jury were to accept it, was that they saw two white men attacking a Chinese man who was on the ground by repeatedly kicking or stamping on him. Both the white men were attacking him. One was at the head of the man on the ground, the other towards his feet. The man at the head end stamped on or kicked the Chinese man’s head. Some of these witnesses saw one of the white men with or wielding a road sign. Sarah Darwood said that the Chinese man had not been holding the road sign. On the ground, the Chinese man had his arms up trying to defend himself and offering no violence. The force used by the white men was hard and violent. Another man, who must have been Alan Knapp, shouted at the white men to stop. Samantha Taylor said that both white men were involved in the fighting. The Chinese man was on the floor scrunched up in a foetal position. One of the white men was kicking him. The other was just standing there about a foot away doing nothing to stop it.

One reason why it is not necessary to give a fuller account of this evidence lies in the post mortem and biological evidence, to which we will come in a moment. But this body of eye witness evidence was in our judgment amply sufficient, if the jury accepted it, for them to reject Anthony van Dongen’s evidence and case that he was unconscious and did not take any significant part in the attack. It was also broadly inconsistent with a defence of self-defence.

Post mortem examination of the victim found a total of about 30 injuries. There were fifteen separate head injuries which, in the opinion of Dr Heath, the pathologist, would have needed a minimum of fifteen separate blows. The head injuries included bruising and abrasions. Two bruising patterns were likely to have been caused by stamping rather than kicking. The brain was soft, congested and swollen with haemorrhage between the brain and skull. The head injuries had caused the brain damage which had caused the victim’s death. Injuries to his hands taken as a whole were defensive injuries, although it was possible that some of the injury resulted from him delivering a punch.

Generally one or two of the injuries could have been caused by falling, but not most of them. There was bruising to his chest and abdomen which could not have resulted from him falling to the ground. Bruises below the knee could have been so caused, but not one higher up on the knee. There were however 10 such injuries which, in Dr Heath’s opinion, were consistent with the victim being curled up in a defensive position. The injuries below the waist were not the cause of death.

Dr Heath agreed in cross-examination that one injury to Mitchell van Dongen’s back shown on photographs, which he had not seen before, could have been caused by a blow from a quite heavy road sign, which he was shown. The photographs showed another back injury which was not, in his opinion, an impact injury.

Bridget March, a biologist, had examined the scene and found bloodstains from the victim. There was blood on a drainpipe 46 cm above the ground. This, in her opinion, was blood spatter resulting from impact at low level into wet blood. There was blood on the middle panel of a window 102 cm above the ground. She again considered that this was airborne blood resulting from impact into wet blood while the bleeding person was upright at the time. These findings required at least three blows – one to cause the bleeding and the other two to cause the spatter. The victim’s blood was found on both sides of the road sign, and in her opinion he had been struck with both sides of the sign. There were no recoverable fingerprints from the sign.

Both appellants were arrested on 12th April 2003. They were cautioned. There was police evidence that Anthony van Dongen said: “I don’t believe this. I am a member of the constabulary.” He was indeed a Special Constable. Mitchell van Dongen said: “I know what this is about. He came at me with a sign and so I knocked him to the floor”. He then became emotional and said: “I knocked him to the floor and then I kicked him in the head to stop him getting up.” He was warned to say nothing further until he had spoken to his solicitor. He then said: “I didn’t know he had died”. When he was later asked to read and confirm a written account of this conversation, he said that he had meant that his brother had fallen to the ground when the man tried to hit him with the sign, because they had their arms around each other.

Mitchell van Dongen was medically examined for injuries to his back, which he said had been caused during the incident when the victim had hit him with a metal sign, causing him to fall to the ground. The doctor reported scabs which were old and dry and consistent with the incident, rather than being fresh.

The appellants each gave accounts in police interviews which were consistent with their evidence at their trial. The day after the interviews, Mitchell van Dongen made a unsolicited remark that he prayed that it was all recorded on CCTV because that, he said, would clear him and his brother.

Both appellants gave evidence. Mitchell van Dongen said that he had had four drinks that night. He was merry, but not drunk. The drink had not affected his temperament or his memory. His brother was quite drunk, unsteady and slurring his words. But he was not aggressive when he had been drinking.

After leaving the bar, they walked along the street with arms around each other. He became aware of someone behind them. He heard a scraping noise, looked behind and saw a man with a road sign in both hands raised above his head. He asked the man whether he was coming at him with it. The man did not reply, but struck him with the sign on his left shoulder. His brother went to the floor and the man struck him with the sign directly onto his back. He himself fell causing his knee to make contact with the ground. He turned, his only thought being for the well being of his brother, who was vulnerable on the ground. He stood up. The man with the sign came at him again. He managed to get the sign off him and pulled it away. In the struggle, he knocked the man in the face with his elbow and he fell flat onto his back. The man then got up and came charging at him with his head down and clenched fists. He himself moved to the left of the man and clipped him on the back of his head to stop him. The man fell down in the alleyway. He went to get up again and so he kicked him to keep him on the floor so that he would not attack any more. At that stage his foot had perhaps made contact with the man’s ribs and arm and the right side of his face banged on the cobbles. The man tried to get up again, so he kicked his arm to stop him. He tried to stamp on his hands to stop him and his foot glanced off his hand and caught the left side of his face. The man grabbed his right leg. He shook it to try to release the grasp and perhaps in doing so made contact with the man’s body. He may have kicked the man as he lay on the ground and stamped on him, but he did not intend to hurt him. He was just trying to get away. He was not aiming at anywhere in particular. He just wanted him to let go of his leg, which he had in a tight grip and he was pulling it towards his head. He kicked the man in the side and pressed his leg back to pin him down.

He then saw the man’s mate coming over. He was on the telephone. He himself swore at the friend and told him to put the phone down. He thought he was calling other people as reinforcements. The friend dropped the phone and approached his brother. He did not see what happened between them. He and his brother then left. He thought he had just been involved in a fight. He did not believe that the man was seriously hurt. If he had known afterwards that the man had died, he would have gone to the police. He had no reason to attack anyone or to hurt the man. His only intention was to make sure his brother was all right. His brother was not involved in the incident at any stage. He had not seen him make any physical contact with the man. He said in cross-examination that he did not know if the man had hit his brother with the road sign. He did not say to the officers who later arrested him that he had deliberately kicked the man in the head to stop him from getting up.

He accepted in cross-examination that he had caused a minimum of fifteen injuries to the victim’s head, although he could not say how many times his foot had been in contact with his head. He was shaking his leg to get away and the contact was accidental. The prosecution witnesses were mistaken. The broken ribs must have been caused when he stood on the man’s back: the leg injuries when the man fell to the floor.

Anthony van Dongen said that, when he and his brother left the bar that night, he was probably unstable, but not falling about. He remembered walking down Westferry Road arm in arm with his brother. The fresh air made him feel worse. He suddenly became unconscious in a way he had not experienced before. This was before he became aware of any part of the incident. The next thing he remembered was coming round on the pavement and seeing his brother struggling with someone he had never seen before. He thought his brother was being attacked, so he got up to help him. The man was lying by a gate and his brother was standing at his top end. Another man then appeared, so he went towards him as he thought he was going to attack him. It all happened quickly. He did not deny that he may have punched Alan Knapp on the back of his neck because he thought he was about to be attacked. His brother then called him and they left. He did not pay any attention to the man on the ground and did not see his condition. He had no idea at the time that he was injured. He did not know what had caused his injuries, but it was certainly not he. He himself was not injured. He agreed that his loss of consciousness was not caused by any injury.

Thus it was that Mitchell van Dongen’s main defences were accident and self-defence. He also said that he did not have the necessary intent for murder. These were unpromising defences in the light of the prosecution evidence, not least the medical evidence, which securely established the savagery of what must have been a sustained attack. There was no countervailing medical or scientific evidence.

Anthony van Dongen’s main defence was that he did not participate, being unconscious at the time – again an unpromising defence in the light of the clear evidence of the eye witnesses.

Subject to consideration of the two matters raised in the grounds of appeal, the jury were properly and fully directed by the judge in his summing up. They unanimously convicted the appellants of murder. They must therefore have accepted the main lines of the evidence of the eye witnesses and surely rejected the defences on which the appellants relied. In particular, they must have rejected the main parts of Mitchell van Dongen’s exculpatory evidence. They must have been sure that the injuries he inflicted were not accidental; sure that he did not inflict those injuries in reasonable self-defence; and sure that he intended to cause the victim at least really serious injury. As to Anthony van Dongen, they must have rejected his account that he was unconscious and played no part in the attack at least to the extent of accepting the evidence of the eye witnesses that both white men, of whom he must have been one, participated in the kicking and stamping attack on the man on the ground.

The jury’s verdicts do not imply whether or not they surely rejected Mitchell van Dongen’s evidence that the victim came at them wielding the road sign. As we understand it, the eye witnesses’ evidence did not exclude that possibility. Although Sarah Darwood said that the Chinese man had not been holding the road sign, and Louise Everingham said that a white man picked up a road sign and hit the Chinese man over the head with it, the Chinese man may have had the road sign before the cars drove by. Louise Everingham had said that she saw the Chinese man and one of the white men struggling with the road sign. Richard King first saw a man on the alleyway side of the pavement holding a road sign above his head. He could not say which man this was and agreed that the man on the ground could have had the road sign first. Tina Stephens saw a white man holding what looked like a road sign in both hands. She saw no scuffle holding the road sign. The jury’s verdict is consistent with a decision that, even if the victim did come at the appellants with a road sign, their responding violence quickly moved from defence to unnecessary and utterly disproportionate intentional attack, and thus was not reasonable self-defence.

Before he summed the case up, the judge was asked by defence counsel to include a direction to the jury on provocation. The prosecution opposed this request. The judge declined to give a provocation direction. The first ground of appeal is that he was wrong not to do so.

The judge’s attention was drawn to R v Miao [2003] EWCA Crim 3486, an appeal to this court in which the trial judge had also been invited to leave provocation to the jury, but had declined to do so. In the present case, the judge accepted Mr Waters’ submission that, if it were necessary in this case to give a provocation direction, such a direction would be necessary in every case in which the real issue was self-defence. He noted (at page 23B of the transcript) that the two appellants and the victim had never come across each other before. It was an incident which had happened in the street. The appellant’s account was that they were acting in self-defence. It was not an appropriate case for provocation to be left. They were “in the realms of what has been called a speculative possibility”.

Section 3 of the Homicide Act 1957 provides:

If there is evidence on which a jury could find that the accused was provoked to lose his self-control, the issue of provocation must be left to the jury even if, in the opinion of the trial judge, no reasonable jury could possibly conclude on the evidence that a reasonable person would have done as the defendant did and thus that on the evidence a verdict of manslaughter would be perverse – see R v Gilbert [1978] 66 Crim. App. R. 237. This is the plain meaning of the statutory provision

In R v Miao, the Vice-President, Rose L.J., quoted from the opinion of Lord Steyn in Acott at page 100E as follows:

The Vice-President then said in paragraph 19 of Miao:

In Miao, the Court said that there was evidence of potentially provoking conduct sufficient to be left to the jury, but that there was only a speculative possibility that the defendant may have lost his self-control. There was no evidence of a frenzied attack. Applying Lord Steyn’s yardstick in Acott, the Court held that the trial judge’s decision not to leave provocation was courageous, but also correct.

In the present case there was evidence of a frenzied attack, if only by deduction from the post mortem and biological evidence. There was thus evidence of loss of self-control by Mitchell van Dongen fit to go to the jury. Mr Waters accepted this. He submits that the judge was entitled to conclude that there was no sufficient evidence of provoking conduct. He accepts that Mitchell van Dongen’s evidence was that the victim came at him with a road sign. But he points to the evidence of Alan Knapp and the five other eye witnesses whose effect the judge summarised as being that, with one exception, who was neutral, they all said that the man with the road sign was not the victim. Mr Dein had agreed with that summary. Mr Waters submits in effect that the judge was entitled to make an evaluative judgment of the evidence as a whole and thereby to reach the conclusion that he did. Mr Dein says that, in the context of the possible defence of provocation, it was for the jury to determine whether Mitchell van Dongen’s account of the victim coming at him with the road sign was or may have been true. We think that, in the present case, Mr Dein is correct. The terms of section 3 of the 1957 Act, Acott and Miao all predicate a degree of judicial evaluation of the evidence bearing on the first two of the elements of a possible defence of provocation. But Mitchell van Dongen’s evidence that the victim came at him with a road sign was neither speculative nor minimal – nor indeed, we think, of entirely doubtful credibility. A speculative possibility will normally arise when there is no, or very little, direct evidence. Here there was direct evidence.

This constitution of this Court is less troubled than earlier constitutions appear to have been by the possibility that a judicial direction on provocation which the defence does not overtly rely on may appear to undermine other defences on which the defendant does rely. We acknowledge the forensic dilemma, but consider that a skilfully constructed summing up should be able to avoid the problem in most, if not all, cases. However that may be, in the present case the defence asked for a provocation direction, and Mr Waters acknowledges before us that, if a provocation direction was required, it should have been given.

We are more troubled by the difficulty for judges in deciding in borderline cases whether particular conduct is to be regarded as provoking conduct. The difficulty derives from the fact that section 3 of the 1957 Act requires the objective element of provocation to be left to the jury. Yet provocation or provocative conduct should, it is to be supposed, have some attribute which makes the conduct provocative. We are concerned with provocative conduct, not mere causative conduct. Yet a judgment, that particular specific conduct was causative but cannot properly be called provocation, risks straying into an evaluation of the objective element which statute has left to the jury. To decline to make the judgment would mean, as the judge in the present case thought, that judges would be obliged to give a provocation direction in all, or nearly all, murder cases in which there is a defence of self-defence and more than speculative evidence that the defendant may have lost his self-control because of things done or said or both.

It looks as if this difficulty, or a variant of it, arose for debate in Acott, where Lord Steyn said at page 102F:

In the present appeal, Mr Waters asks us to say that this was a borderline case and that the good sense of the trial judge should not be disturbed. With some hesitation, we disagree. There was evidence that Mitchell van Dongen lost his self-control. There was specific evidence – that of Mitchell van Dongen himself – that the victim came at him with the road sign. That was capable of being causative conduct. We are not confident that it was not capable, apart from the objective element, of being provocative conduct. We conclude that the judge should have directed the jury on provocation. In borderline cases such as this, it is the prudent course for judges to take, especially if the defence ask for a provocation direction to be given.

There was, therefore, a misdirection. Mr Dein submits that in these circumstances the appeal should be allowed and a manslaughter verdict substituted. Mr Waters accepted initially that this should follow – the prosecution would not seek a second retrial. At one stage in argument we had the quixotic forensic situation where Mr Dein conceded that it was open to the Court to conclude that the verdicts of murder were safe: but Mr Waters conceded with reference to his then understanding of the Privy Council decision of The Queen v Franco (P.C. Appeal No. 70 of 2000, 14th August 2001) that it was not open to this Court so to conclude. His then understanding was based on the Times Law Report summary version of the Privy Council decision. We adjourned the hearing of this appeal to enable counsel to make further written submissions about Franco, when they had considered the full report. We shall consider Franco later in this judgment.

The burden is on the prosecution to disprove provocation to the criminal standard. In many cases where provocation has not been left to the jury when it should have been, it would be no more than speculation to ask what the jury’s decision would have been, if had been left to them. Logically, if there is, as we have concluded in this case, evidence on which the jury could have concluded that Mitchell van Dongen was provoked to lose his self-control, we could not safely conclude that the jury would surely have decided that he was not so provoked, unless this can be safely deduced from the verdicts which they gave on the issues which they were directed to decide. There was clear evidence that Mitchell van Dongen may have lost his self-control. The jury would have had to consider whether he was provoked. That would depend on whether they accepted that the victim initially came at the brothers with the road sign or that he may have done. Given that the jury convicted both appellants and that they must have surely rejected the main significant parts of their evidence, they may well also have rejected Mitchell van Dongen’s account of the victim having the road sign. But we cannot be sure about this, since the defence of self-defence at least was scarcely viable on the medical and scientific evidence alone even if the victim initially had and used the road sign.

It remains to consider whether the jury could conceivably have concluded that any provocation which they might conceivably have found was enough to make a reasonable man do as Mitchell van Dongen did. Since this is the element of the defence of provocation which statute reserves exclusively for the jury, the first issue is whether it is open to this Court to conclude that the appellants’ convictions for murder are safe, even though the jury were never given the opportunity of performing this element of their statutory function. Mr Waters has remained coy about this. But we have a statutory duty to dismiss the appeals, if we do not think that the murder convictions are unsafe – section 2(1)(b) of the Criminal Appeal Act 1968, subject of course to relevant matters of law or statute.

In R v Cox [1995] 2 Cr. App. R. 513, this Court was clearly of the opinion that, as a matter of law, the court in an appropriate case might apply the proviso to the then section 2(1) of the Criminal Appeal Act 1968, where there had been a misdirection by a failure of the judge to leave the issue of provocation to the jury. [Section 2(1)] has of course since been amended, but not so as to alter the intrinsic nature of the question now under consideration.] The court considered R v Whitfield (1976) 63 Crim. App. R. 39 at 43, in which Lord Widgery CJ, giving the judgment of the court, had said that, if the court applied the proviso, they would be determining the issue otherwise than by verdict of the jury – see also R v Burgess and McLean [1995] Crim. LR 425 cited in Cox at 520E. Glidewell LJ, giving the judgment of the court said that, with all due respect of how the court in Whitfield dealt with the matter, section 3 of the 1957 Act spells out the function of the judge and jury at the trial. The section is dealing with the trial, not the appeal.

We were also referred in this context to R v Rossiter (1992) 95 Cr. App. R. 326 at 333 and to the passage there cited in the speech of Lord Tucker in Bullard v R. [1957] A.C. 635 at 642 where he said that every man on trial for murder has the right to have the issue of manslaughter left to the jury if there is any evidence upon which such a verdict can be given: and that to deprive him of this right must of necessity constitute a grave miscarriage of justice and it is idle to speculate what verdict of the jury would have reached.

In R v Dhillon [1997] 2 Cr. App. R. 104, the trial judge had thought that there should be a provocation direction, but neither prosecuting nor defence counsel thought it was necessary. So the judge did not give the direction. This court on appeal acknowledged that the judge was anxious to be fair, and had done what both experienced counsel had agreed that he should do. Nevertheless, the law was clear and on the facts the judge should have given a provocation direction. The question then was whether the omission made the conviction unsafe. Ward LJ, giving the judgment of the Court, said at page 114F:

The Queen v Franco was an appeal to the Privy Council from the Court of Appeal of Antigua and Barbuda. The appellant was convicted of murder and sentenced to death. His defence was self-defence. No defence of provocation was advanced at trial and the trial judge gave no direction to the jury on provocation. On appeal to the Eastern Caribbean Court of Appeal, the prosecution accepted that, on the evidence of the appellant, an issue of provocation had been raised and ought to have been left for the jury’s consideration. But the Court of Appeal concluded that, if the jury had been properly directed on provocation, they would inevitably have convicted the appellant. They accordingly applied the proviso to section 39(1) of the West Indies Associated States Supreme Court Act 1969 and dismissed the appeal because they considered that no miscarriage of justice had actually occurred. [The proviso in section 39(1) is equivalent to that in the unamended version of section 2(1) of the 1968 Act.] The central issue in the appeal to the Privy Council was whether the proviso was properly applied in the circumstances of the case. Section 9C of the Offences against the Person Act (Cap. 58) of Antigua, as amended in 1986, follows section 3 of the 1957 Act. This issue before the Privy Council in Franco was essentially the same as the issue we are now considering.

Lord Bingham of Cornhill delivered the judgment of the Judicial Committee. He said at paragraph 14 that the clear effect of section 9C was to deny to trial judges the power previously exercisable to withdraw the issue of provocation from the jury where there is evidence potentially capable of satisfying the subjective condition, even if the judge considers that there is no evidence which could lead a reasonable jury to conclude that the provocation was enough to make a reasonable man do as the particular defendant did.

The submission on behalf of the appellant was, not that section 9C of the Antigua statute absolutely precluded the Court of Appeal from applying the proviso when the trial judge had wrongly failed to leave the objective issue for determination by the jury, but that it would rarely be proper to do so. By applying the proviso, the appellate court was doing exactly what the statute forbade, by substituting the decision of judges for the decision of the jury on the question which the statute required to be decided by the jury. Counsel particularly relied on Whitfield. Lord Bingham cited at length from the judgment of the court, given by Lord Widgery CJ, at page 42, to which we have referred earlier in this judgment. Counsel accepted that there could be cases in which the proviso could be applied even where section 3 or its equivalent had not been observed, but the case of Franco was not such a case.

Counsel for the Crown submitted that it was appropriate to apply the proviso, if the court could be satisfied that the jury would inevitably have come to the same conclusion and returned the same verdict, even if the error or irregularity complained of had not occurred. Counsel relied in particular on Cox, and Lord Bingham cited from the judgment of Cox, to which again we have referred earlier in this judgment.

Lord Bingham then said at paragraph 17:

Lord Bingham then said:

The question for this court, as Mr Waters emphasises, is whether, notwithstanding the misdirection, the appellants’ convictions are safe. It is not suggested that for present purposes there is any material difference of substance between this question and the question arising under the former proviso. The unamended version of section 2(1) of the 1968 Act had as a ground for allowing an appeal, subject to the proviso, that the conviction was unsafe.

The judgment in Franco leaves open the possibility of this court deciding that a murder conviction is safe, where provocation, including the objective issue, should have been left to the jury but was not. The Board did not go to the length of saying that this would never be appropriate. We have to be very cautious in drawing inferences or making findings about how the jury would have resolved issues which were never before them. There is a danger of delving into the minds of the jury. We do not propose in this appeal to do this.

We take the evidence in support of a possible defence of provocation at the highest that it could reasonably be put. We suppose, therefore, for these purposes, that the victim quite suddenly came at Mitchell van Dongen from behind in an inebriated state at night in the street, aggressively wielding a quite heavy road sign, and that he struck Mitchell van Dongen in the way that he described. We proceed on the basis that Mitchell van Dongen’s evidence was or may have been true up to the point when the victim went to the ground. Thereafter, the logic of the jury’s verdict rejects the main lines of the appellants’ evidence and accepts that of the eye-witnesses. This is the logic of their verdicts, not speculation. We discount entirely as militating against the appellants the fact that the victim’s blood was found on both sides of the road sign. It would be somewhat speculative to ask how it got there. There is clear evidence that Mitchell van Dongen lost his self-control. In such a state, he at least intended to cause the victim really serious harm – so the jury must have found and so a defence of provocation has to suppose. The question then is, we think, are we sure that at least 10 members of a jury, not being perverse, would inevitably and surely have concluded that the provocative conduct was not enough to make a reasonable man do as Mitchell van Dongen did, taking into account everything done or said according to the effect which it would have on a reasonable man?

In addressing this question, we suppose that the jury would have been directed in accordance with the majority opinion of the Privy Council in Attorney General for Jersey v Holley [2005] UKPC 23 (15^th June 2005) as to the objective standard of self-control to be applied. We take this shortly since no submissions were addressed to us on this part of the topic and the facts do not in this respect raise issues of detail which might be controversial. There is no evidence that anything was said which might have been provocative. Mitchell van Dongen may have been to an extent intoxicated, but self-induced intoxication is not a matter to be taken into account when considering whether he exercised ordinary self-control – see Holley at paragraphs 24, 68 and 76, these representing, we think, virtual unaniminity for the purposes of the present appeal. We note that Lord Steyn said in Acott at page 102 that:

The Privy Council decision in Holley was published after the conclusion of oral submissions in this appeal and after we had received the parties’ written submissions in relation to Franco. An earlier draft of paragraph 60 of this judgment, written before Holley was published, referred to the decision of the majority of the House of Lords in R v Smith (Morgan) [2001] 1 AC 146, which the majority of the 9 member Board in Holley decided was erroneous. We assume, but do not decide, because it is not necessary to do so, that Holley, a decision of the Privy Council, would be taken as binding in England and Wales. Insofar as a Holley direction might academically be thought to be less favourable to Mitchell van Dongen’s case than a Smith (Morgan) direction would have been, we should say, first, that in our view a Holley problem does not arise on the facts of the present appeal; and, second, that, subject only to minor adjustments of detail only, succeeding paragraphs of the earlier draft of this judgment are unchanged in substance following the publication of Holley. Thus, the publication of Holley has had no effect on the outcome of this appeal.

Mr Dein has drawn attention to a number of matters which he submits should lead the court to be less than sure that the jury would have convicted the appellants of murder, if they had been directed as to provocation. He acknowledges that some of these have more bearing on the subjective elements, which we are assuming in the appellants’ favour. He identifies the glowing testimonials of Mitchell van Dongen’s good character; the fact that the victim would lose his inhibitions when he was drunk; the fact, as we accept, that no prosecution witness saw the beginning of the incident; the fact, as Mr Dein submits, that Mitchell van Dongen only kicked or stamped on the victim 8 times and that he was very angry; and the fact that he himself received injury to his back. Mr Dein submits that this court should follow the course taken in Dhillon and conclude that it is not possible to say that the outcome before the jury would still have been guilty of murder.

We have given these submissions the most anxious consideration, having the extreme caution required by Franco at all times in mind. We also bear well in mind Mr Waters’ less than whole-hearted support on this point for a conclusion that the murder convictions are safe. We think, however, that his standpoint derives from agnosticism on the law rather than from factual considerations.

As to the facts, the provoking conduct which we assume in the appellants’ favour is not in doubt, nor a matter of speculative inference. Neither is the fact that in consequence Mitchell van Dongen lost his self control. The assumed attack was by a drunken man with a heavy implement capable, we assume, of causing serious injury and in fact causing Mitchell van Dongen some injury. There is little room for a difference of view as to the gravity of this provocative conduct not least because things said are not part of it. The victim was soon disarmed and shortly afterwards on the ground. At about this stage, the evidence of the eye-witnesses and the logic of the jury’s verdict take over. Any submissions as to the number of kickings or stampings to the head has to accommodate the pathologist’s evidence and Mitchell van Dongen’s admission that he was responsible for no less than 15 head injuries, the seriousness of which was described by the pathologist and which resulted in the victim’s death. We are sure that the jury must have concluded that the injuries, or most of them, were inflicted while the victim was defenceless on the ground. You cannot otherwise kick or stamp on a person’s head. Provocative conduct, yes for present purposes. Loss of self control, yes. But no reasonable man could possible have been provoked by the assumed acts to do what Mitchell van Dongen surely did. Of that we are certain. Any other jury decision would have been perverse. The jury’s rejection of the defence of self-defence predicates as a minimum a decision that Mitchell van Dongen’s retaliation was disproportionate. The issue for provocation is not the same, but it is related.

Justice in a criminal trial rightly and necessarily concentrates on justice for the appellant. But the court must not overlook the matter of justice for those concerned with the victim also, nor the requirements of a proportionate criminal appellate system, which include that those who are surely and fairly shown to be guilty of murder, and have been so found by a jury, should not escape that consequence on gossamer grounds. In our judgment, the unavoidable facts of this case and the necessary logic of the jury’s verdict rule out any possibility of a miscarriage of justice – see Franco, paragraph 19.

We conclude, therefore, that, even though a provocation direction should have been given, the murder convictions are in this respect safe. It is unnecessary to distinguish between the two appellants, although quixotically a defence of provocation was theoretically stronger for Mitchell than for Anthony, even though Mitchell was on the evidence and his own admission the lead assailant.

The second ground of appeal is insubstantial. No complaint is made, except as to provocation, of the directions of law given by the judge in his main summing up. After the jury had retired, they asked a question as follows:

For these reasons, we conclude that the appellants’ convictions of murder are safe. Their appeals against conviction are dismissed. It remains to consider questions of sentence.