England and Wales Court of Appeal (Criminal Division) Decisions

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CRIMINAL DIVISION)
ON APPEAL FROM THE CROWN COURT AT BRISTOL
THE HON. MR. JUSTICE SWINTON THOMAS

		Royal Courts of Justice Strand, London, WC2A 2LL
		17th June 2003

Lord Justice Auld :

1. This is an appeal, on a reference from the Criminal Cases Review Commission, by Anthony Keith Poole and Gary Mills against their conviction, on 26^th January 1990 before Swinton Thomas J. and a jury in the Crown Court at Bristol, of the murder of Hensley Wiltshire. The prosecution case was that at about midnight on 5^th/6^th January 1989, in the course of a fight, both appellants attacked Wiltshire, Mills with a crowbar and Poole with a knife, causing him severe injuries from which he died the following day. Mills' defence was that he had struck Wiltshire with a crowbar, but only in self-defence. Poole's defence was that he had not used any weapon or otherwise taken any part in the assault. They also denied that the injuries inflicted on Wiltshire in the fight had caused his death.
2. Since the convictions this case has had a considerable forensic history. On 16^th April 1996 the full Court, constituted by Otton LJ, Ian Kennedy and Keene JJ., granted an extension of time of four years and leave to appeal, but dismissed the appeals. The Court certified a point of law of general public importance, but refused leave to appeal to the House of Lords. The House of Lords granted leave to appeal, but on 24^th July 1997 dismissed the appeals. In October 1998 there was a trial in which Detective Inspector Gladding, the second most senior officer involved in the investigation, sued Channel 4, the makers of a television programme, for libellous allegations about his conduct in the investigation and at the trial. The jury upheld Channel 4's defence of justification. The appellants then sought a reference from the Criminal Cases Review Commission ["the CCRC"] to this Court, which, in November 2000, the Commission, after a thorough investigation, declined to make. The appellants sought judicial review of that decision, which, in December 2001, the Divisional Court, constituted by Lord Woolf, CJ and Ouseley J., dismissed. The Court held that the Commission had been entitled not to make a referral, but Lord Woolf, giving the judgment of the Court, expressed the view that "the Court of Appeal could now have a doubt about the safety of the convictions". Following a further application to, and careful consideration by, the Commission, it has made this referral.

The evidence at trial

3. Only two prosecution witnesses gave direct evidence of the fatal assault on Wiltshire, which took place in Poole's ground floor front room at 34 Conduit Street in Gloucester. The first was a 21 years old woman named Kimberley Stadden, without whose evidence the prosecution had no viable case against either appellant. The second was a young man called Paul White, who claimed to have been looking through the window from outside for a short time while the fight was under way.
4. At about midnight on the night in question Stadden had gone to Poole's room to obtain and take amphetamine. In the room were Poole, Mills, Wiltshire and a man called Juke. It is plain from all the evidence that Wiltshire had had a lot to drink – and also some amphetamine – and was in a very aggressive mood. Stadden knew Wiltshire; she had been passing stolen cheques for him over the previous few days. In her evidence she gave the following account of what happened shortly after her arrival. Mills and Wiltshire were arguing. Wiltshire was behaving aggressively, trying to pick a fight with anyone, but at that stage with Mills in particular. The two of them began to fight, punching and kicking each other, Mills with a crowbar in his hand and Wiltshire unarmed. Poole intervened and, she thought, kicked Wiltshire twice. Mills then hit Wiltshire with the crowbar several times on the legs and head, causing him to slump into a sitting position on the floor. Mills continued to hit him on the legs with the crowbar. At that point Poole pushed the blade of a knife into Wiltshire's arm. The effect of this combined assault led Wiltshire to attempt escape by the window, but he was unsuccessful and fell back onto a settee. Mills continued to attack him, punching him with his fists, causing him to fall to the floor. While he was on the floor Poole kicked him and then stabbed him four or five times in the buttocks with the knife. Much shouting followed, in the course of which Wiltshire threatened his two assailants and to smash Poole's flat. At that point the fourth man in the room, Juke, who had played no part in the violence, grabbed at Poole, saying that Wiltshire had had enough. However, the violence continued, and Juke tried to shield Stadden from sight of it by putting her head on his shoulder and covering her eyes with his hands. However, she heard something of it, which she described as a "thud" and a "squelching" noise. That was the end of the violence.
5. Thus, according to Stadden, all three men were aggressive and using violence; Mills and Poole had each used a weapon on Wiltshire; he was unarmed; and he, not they, suffered serious injuries in the course of the fight.
6. Continuing with Stadden's evidence, the incident ended with Juke saying that they should all leave. Wiltshire said that he needed an ambulance. Juke dragged him out onto the pavement and he, Mills, Poole and Stadden left him there. Stadden walked home alone. Shortly afterwards Mills and Poole also went there, where they stayed for a minute or two. Both of them were blood-stained and wanted to call for a taxi. She told them where to find a telephone kiosk and they left.
7. Stadden did not immediately contact the police. But a day or two later, as a result of a television news broadcast about the matter in a programme called "Crime Line", she telephoned the programme makers and told them what had happened. It emerged as a result of cross-examination of her by Mr. Ian MacDonald, QC, counsel for both appellants, that she read out to Crime Line over the telephone a written statement that she had dictated to and discussed with a friend, Kathryn Halliday (who was later to gain notoriety by selling to the Daily Mirror an account of her sexual liaisons with Fred West and as a witness in the trial of Rosemary West). That account corresponded broadly with what she, Stadden, was to say in evidence, including the allegation that Poole had attacked Wiltshire with a knife. But it also contained an allegation, not made in her evidence, namely that Mills too had stabbed him, more than once, with a knife, in addition to attacking him with the crowbar. In cross-examination, she held to her account that Poole had used a knife, but in re-examination said that her allegation in her Crime Line statement that Mills had also done so was wrong.
8. It also emerged in her cross-examination that her first witness statement to the police, given to WPC Moore and DC Churchill on 9^th January, though referring to Mills' attack on Wiltshire with the crowbar, had not implicated Poole in the fight at all. But, in a second witness statement given to DC Churchill on the following day, she had implicated him in the fight, saying that she had seen him kicking and jabbing Wiltshire with a knife several times in the buttocks. Her explanation in cross-examination for the discrepancy in this and other respects between her two witness statements was that she had told WDC Moore and DC Churchill about Poole's part in the attack when making her first statement, but that they had mistakenly not recorded it; hence her second witness statement in which it was mentioned. This account differed from her evidence in the committal proceedings, in which she had been less certain about what Poole had done with the knife and said that she had forgotten, when making her first witness statement, to mention his involvement in the attack on Wiltshire.
9. Looking forward in the prosecution evidence for a moment, three police officers were called to deal with this important discrepancy between Stadden's first witness statement and her evidence. WDC Moore and DC Churchill said that when WDC Moore had taken the first witness statement Stadden had said that she had seen Poole stab Wiltshire with a knife. WDC Moore said that she had inadvertently omitted to record it in the written statement. She had reported the omission to her superior officers and had been instructed to take a further witness statement. Superintendent Bennett, the officer in overall charge of the investigation, confirmed her evidence to the extent that, when he had learned of the omission, he had instructed her to take a second witness statement. All the officers denied expressly or implicitly the suggestion put in cross-examination by Mr. Macdonald that they had done some sort of deal with Stadden to persuade her falsely to implicate Poole in the attack on Wiltshire.
10. Cross-examination of Stadden also revealed an inconsistency between her evidence in chief and the account in her first witness statement of what had happened after she and the others left Poole's room. In the witness statement she had said that she had met Mills and Poole on the way back to her home, an assertion that she said in evidence had been a lie because she had been frightened about her involvement in passing stolen cheques for Wiltshire and her drug taking.
11. The general thrust of Mr. MacDonald's cross-examination was that she was lying about Mills' violence, other than in self-defence against a very aggressive Wiltshire, and about Poole's involvement in any assault of Wiltshire or use of any weapon. He suggested that the police had put improper pressure on her to lie by threatening to prosecute her for her involvement with Wiltshire in passing the stolen cheques and that she had told lies about Mills and Poole to save her own skin. She denied those suggestions and held to her story, though she acknowledged the lie we have mentioned about meeting Mills and Poole on the way home from the violence in Poole's room. She also acknowledged that the police had cautioned her, but told her that they would not prosecute her, for the cheque offences.
12. As we have said, the only other person to give direct evidence of what had taken place in Poole's room was Paul White. His account was at best limited and of only a short part of the fight. He said that he had visited the house twice that night, first at about 9 p.m., when he had seen Poole, and later just before midnight. On the second visit, he went there alone in a car he had borrowed from a man called Andrew Neale. On arrival outside the house, he left the car and was about to enter the front door, when he heard shouting and screaming from Poole's room. He returned to stand by the car, where he remained for about five to ten minutes, and looked up at the window of Poole's lighted room. Through a gap of about one and a half to two inches between the curtains, he saw Poole in silhouette, striking downwards and heard someone, as if on the floor, shout in a pleading way, "no, Tony no". He saw, also as a silhouette through the window, the arm of a man pointing upwards as if to protect himself. He then drove away and returned about half an hour later with Neale, by which time the police had arrived. In cross-examination, Mr. MacDonald suggested to him that he too was lying because the police had offered him an inducement to implicate Poole, which he denied.
13. In the course of the trial the defence produced photographs to suggest that the line of visibility from the path outside 34 Conduit Street into Poole's ground floor front room window was such that White could not have seen what he described in evidence. The prosecution relied upon the evidence of a police plan drawer indicating that a partial view into the room was possible. Mr. Alun Jenkins, QC, in his closing speech for the prosecution, invited the jury to rely on White's evidence as well as that of Stadden. And the Judge, in his summing-up and without further comment, told them that White's credibility and reliability were matters for them.
14. Juke, the only other potential eyewitness of what happened in Poole's room, was not called to give evidence by the prosecution or the defence. DI Gladding gave evidence of having interviewed him as a potential witness and of his belief that a witness summons had been served on him to attend to give evidence at the committal proceedings. He said that Juke had not attended the committal proceedings. In cross-examination, he denied that he had told Juke not to attend those proceedings.
15. Returning to what happened to Wiltshire after he had been dragged outside the house on the night of the fight, ambulancemen soon arrived and took him to hospital. On examination at the hospital, he was undressed and found to have a kitchen knife tucked down one of his socks. He had a large number of wounds on his arms, legs and face, including a large and deep cut on the top of his face. However, he was so uncooperative and abusive that the doctors could not examine him properly or stitch all of his wounds. He refused to remain at the hospital and was discharged into police custody at about 4.30 a.m., as a warrant for his arrest was outstanding. A police surgeon examined him shortly after his arrival at the police station and, on the surgeon's advice, he was returned to the hospital. This time he was more co-operative and, after further treatment, was returned again to police custody at about 6.15 a.m. Later, after about midday, he collapsed and was taken to hospital for the third time where, shortly after arrival, he suffered two cardiac arrests and died at about 2.30 p.m.
16. The post-mortem showed the following physical injuries. There were injuries to the head and upper and lower body at the front and the back, each of the head injuries having been caused by a separate blow. There were ovoid stab wounds to the shins, which could not have been caused by a knife or a crowbar, but by a drill-shaped object. There were 17 stab wounds, 13 of them in the pelvic area, which must have been caused by a thin, sharp knife. There were fractures to the left ribs and the right fibula. The injuries to the head, body and legs could have been caused by a crowbar. A cut to the face and nostril was caused by a sharp instrument, such as a knife. It was difficult to say what instrument could have caused the ovoid leg wounds, but they had destroyed much more muscle tissue than any of the other individual injuries – the main muscle damage overall was to the left leg and the arms.
17. The post-mortem also revealed that Wiltshire had died from damage to his heart function caused by the wounds. One of the issues at trial was whether his death was caused by the injuries sustained in the fight or as a result of intervening acts of assault and/or negligence. The jury clearly found against the appellants on this issue; the first Court of Appeal rejected their challenge of that finding; and it is not an issue on this appeal.
18. Examination of Poole and Mills showed only superficial injuries.
19. DI Gladding gave evidence of having arrested the appellants on 9^th January. They both put their hands up and Mills said, "Fair enough".
20. In interview by DI Gladding, Mills gave the following account of having acted in self-defence. There was a scuffle and a "second fight" in which Wiltshire drew a knife and slashed at him with it, causing him superficial wounds. He, Mills, took a crowbar from the wall and swung it at Wiltshire, striking him on the head, he thought, and a few times on the leg, in order to get him to drop the knife. Wiltshire continued to brandish the knife; Mills snatched it from him and Wiltshire seized the crowbar, striking him with it. They wrestled down onto the floor, with Wiltshire ending up on top of him and still striking him with the crowbar. Mills, who was then in a panic because he feared for his life, stabbed Wiltshire in the back of the legs and the backside. He, Mills, had not started the fight, and he did not see Poole or Juke fighting with Wiltshire.
21. DI Gladding also interviewed Poole, who gave the following account. He had taken no part in the fight with Wiltshire. Wiltshire started the fight, which then continued, he and Mills, at different stages of it, striking each other with the knife and the crowbar. At one stage they were on the floor and Wiltshire had the crowbar and Mills was stabbing him on the backside and in the back of the legs. He, Poole, grabbed hold of Mills, and Juke pushed Wiltshire, who was bleeding, out of the room and into the street.
22. Mills and Poole gave much the same account in evidence as they had each given in interview.
23. The Judge, in his summing-up to the jury, gave them the conventional direction on joint enterprise but, in addition, directed them that, on the facts, joint enterprise only arose if they were sure that both of the appellants had assaulted Wiltshire, each with a weapon. In response to a question from the jury towards the end of his summing-up, he repeated that direction in the following terms:

. "In this case the question of joint enterprise can only arise if you find – which is of course quite contrary to the case which these two defendants put forward – that the two accused men were both engaged in attacking Wiltshire, one with the knife and the other with crowbar. It cannot arise on the facts of this case unless you make that finding, that is to say, that both of the two men were attacking Wiltshire with weapons."

24. Before turning to the forensic history of the case following the appellants' conviction and to the grounds of appeal, we should mention post-trial discoveries about the police handling of White, who gave evidence, and of Juke, who did not.

White

25. After the trial, information came to light suggesting that he had lied when he said that he had driven alone to Poole's address in Neale's car just before midnight. After the trial Neale made a written statement indicating that White had been with him from early in the evening and that they had both gone in his car to Poole's address at about 11.30 p.m. and found the police already there. On that account, White could not have borrowed Neale's car to drive alone to the address and could not have seen what he claimed in his evidence at the trial.
26. In addition, two police officers, DCs Paine and Cheminais, later made statements that, shortly after the fight and when interviewing White as a potential witness, he had told them that Neale had driven him to the house and had remained there while he, White, had watched the incident. DC Paine, in a written statement provided in 1992 to a Police Complaints Authority Inquiry, said that when he and DC Cheminais had first spoken to White he had been reluctant to make a witness statement. DC Paine said that when he and DC Cheminais had interviewed him again a few days later he gave them the same account and agreed to make a witness statement about what he had seen, but said that he would not include in it any mention of Neale having been with him at the time. DC Paine said that he had then taken a witness statement from White with that omission, an omission that White carried through into his evidence.
27. DC Paine, in his 1992 statement, thus acknowledged that he had prepared and White had signed a witness statement, that White had indicated was in part untrue and which untruth he knew White was going to and did continue in his evidence. However, DC Paine later gave a different account in evidence to the first Court of Appeal in 1996, an account of which the Court expressed itself to be sceptical.
28. DC Paine and DC Cheminais were clearly alive to the importance of the difference between White's oral account to them on the one hand and his witness statement and evidence on the other. Despite an internal police report by DC Paine at the time that White had refused to mention Neale's presence in his witness statement, no police action was taken to trace and check the position with Neale or to inform the Crown Prosecution Service about it. Thus, the prosecution served as part of its case a witness statement, in accordance with which White gave evidence that the police knew to be false in a material particular. In addition, and as a result, the defence were deprived of the opportunity to use this fact to discredit White's evidence at trial and of the possibility of calling Neale as a defence witness.

Juke

29. We have already mentioned that neither the prosecution nor the defence called Juke to give evidence. The police, as part of their investigation and preparation for the trial, had taken two witness statements from him.
30. In his first statement, provided to APS Griffin on 6^th January 1989, he gave the following account, one which, it should be noted, largely supported Mills' case of self-defence and did not suggest that Poole had had any part in the violence. Wiltshire, who was drunk and aggressive, challenged each of them to a fight. He then attacked Mills with a knife, prompting Mills to take hold of a crowbar in order to defend himself. In the course of the ensuing fight, Wiltshire seized the crowbar and struck Mills with it, and Mills, in his turn, seized the knife and jabbed Wiltshire in the legs with it. At that point he, Juke, put his hands over Stadden's eyes to shield them from sight of the blood. He concluded the statement by saying "The trouble happened because of Willie's [Wiltshire's] attitude. Willie was definitely looking for a fight, even after he got one beating, he still went back".
31. In his second statement, provided this time to DI Gladding and DS Jeynes, on 10^th January 1989, Juke gave a different account, indicating that Mills' violence towards Wiltshire at the end of the fight went beyond self-defence and also implicating Poole in the violence towards Wiltshire. He stated that Poole had stabbed Wiltshire with the knife and that at the end of the fight Mills stood over a helpless Wiltshire, giving him "a severe beating" with the crowbar. He also described how he, Juke, dragged Wiltshire out of the room into the corridor and out into the street, and then left the scene once he knew that an ambulance was on its way.
32. Juke later said on two occasions that his second witness statement was untrue and made as a result of pressure from the police to implicate Poole. He said it to Mr. Gadd, the appellants' solicitor, in interview before the committal proceedings. And he said it three years later in the Police Complaints Authority Inquiry.
33. Before the "old style" committal proceedings, Mr. Jenkins decided that the prosecution would not rely on Juke as a witness because he considered that he would not tell the truth and would seek to depart from or contrive an explanation for his second witness statement if called. He informed defence counsel of the existence of Juke's witness statements and gave him his name and address, but, in accordance with the practice of the time, did not disclose their contents. However, at the request of Mr. Gadd, who wanted to call him to "flush out" the two undisclosed witness statements, the prosecution warned Juke to attend the committal proceedings.
34. Shortly before the committal proceedings took place DI Gladding telephoned Juke. In the conversation, which, unknown to DI Gladding, Juke recorded and which was subsequently transcribed, the officer told Juke not to attend the proceedings and warned him that he would be arrested if he did so. DI Gladding did not tell the CPS or prosecuting counsel of this telephone conversation and, in consequence the prosecution did not disclose it to the defence. However, Mr. Gadd, the defence solicitor, learned about it from Juke whom he interviewed again before trial. Juke told him that he had not seen Poole take any part in the attack on Wiltshire and that he had turned his head away at about the time he had shielded Stadden's eyes from what was happening.
35. At trial Mr. Jenkins maintained his refusal to disclose Juke's witness statements, and, as we have said, neither side called him to give evidence. In the light of Mr. Gadd's interview of Juke and the latter's provision of the tape recording of the telephone conversation with DI Gladding, Mr. MacDonald asked DI Gladding in general terms about Juke's non-attendance at the committal proceedings. In his replies, DI Gladding incorrectly gave the impression that the prosecution had required him to attend those proceedings as a prosecution witness. He also said that officers had tried to find him in order to secure his attendance and that they, the police, wanted him to attend. In the light of the telephone conversation DI Gladding had had with Juke, this was dishonestly misleading evidence and Mr. MacDonald knew it. However, Mr. MacDonald apparently did not consider he could put the contents of the taped conversation to DI Gladding without calling Juke to prove it, a view that Lord Woolf in the Divisional Court considered surprising.
36. Despite the absence of Juke from the witness box, the substance of the account in his second witness statement was put before the jury - with the agreement of defence counsel - as part of the transcripts of interviews by DI Gladding and DS Jeynes of Mills and Poole. DS Jeynes' questioning, particularly of Mills, about Juke's account became of importance in different ways in the later forensic part of the story: first as a possible answer to a complaint made on behalf of the appellants that they were prejudiced by lack of knowledge of what Juke could have said if called to give evidence; second as an indication of possible misconduct on the part of DS Jeynes because he misrepresented in questioning what, by implication, Juke had said; and third, only late in the course of this second appeal, that the appellants, particularly Poole, were prejudiced by the admission of such hearsay evidence in the form of questions which Mills, in his answers, did not accept. We set out here one of DS Jeynes' questions to Mills, clearly drawing in large part on Juke's second witness statement undermining Mills' defence of self-defence and implicating Poole. As can be seen, Mills did not accept the main part of the assertions, and he and Poole continued to deny them whenever put to them.

"Q. I think it only fair to tell you Mr. Mills and see if you wish to make any comment upon it. But it is alleged by witnesses that were in the room and have made statements to us that, when Willie bounced off the window that they gained the impression that you thought he was attacking you again. That you put him to the floor and started to have a rough and tumble with him. It was at this point that Mr. Poole came up behind him and stabbed him in the buttocks on a number of occasions and that you then got the better of Willie, became hot as they call it or lost your cool for the first in reality because they are basically saying you have been quite restrained with the antics of Mr. Wiltshire up to now. But that you then get the crowbar and in their view you go over the top and decide to teach him a lesson and that you rain a vast number of blows aimed on the calves and thighs of Wiltshire and that it is quite clear to the one person in particular that you meant to make him learn a lesson and that he gained the impression that you were intending to break his legs in order to settle the matter once and for all. And that your intention was to teach a lesson for all the trouble he had caused that night I mean that is couched in with the fact that up until then, you had held back in reality, but at that stage you finally lost your cool and you gave him a lesson he wouldn't forget. That is what the people are saying they saw in the room, again you don't have to make comment if you don't want to but if you wish to deny that or say anything to it you can.

A. No, as I said, that's not how I remember the incident. I'm not denying that I hit his legs several times with that bar but that was with a knife in his hand."

The references by DS Jeynes in that long question to Mills intending to break Wiltshire's legs and to teach him a lesson were not to be found in anything that Juke or Stadden had said in their witness statements. DS Jeynes, in his evidence in the libel trial suggested that they were his interpretation of what Juke had indicated when he took his statement and that he had not recorded in it everything that Juke had said.

37. As a result of the evidence of DI Gladding about Juke's non-attendance at the committal proceedings, the jury were left with the impression that Juke was to have been a prosecution witness. Given that indication and the obvious assistance that Juke could have given to the jury as an eye-witness of the fight causing Wiltshire's fatal injuries, coupled with DS Jeynes' references in his interviews of Mills and Poole to what Juke had said in his second witness statement, it is not surprising that the jury, towards the end of the Judge's summing-up, sent him a note asking: "Can you please tell us why Neville Juke has not appeared? He seems to [have] been in the room most of the time". The Judge answered the question in the only way he could:

"We have not heard from Juke. As I told you in answer to the note you passed to me yesterday; no doubt there is a good reason why neither the Crown nor the defence have called Juke as a witness. As I said to you, please do not speculate as to why a witness has not been called in the case; you must decide the case on the basis of the evidence you have heard."

38. When reviewing in his summing-up the evidence of the interviews of Mills, the Judge did not mention the passages in which DS Jeynes had put to him and Poole the content of Juke's second witness statement. Nor did he direct them on the subject of such hearsay, nor to ignore DS Jeynes' rehearsal of Juke's allegations. He simply told them to "refer to any passages … [they] thought were relevant". The jury had been provided with transcripts of the interviews and would have had an opportunity to consider those questions in them when considering their verdicts.
39. As we have mentioned, Juke, in a third statement, made for the Police Complaints Authority Inquiry in 1992, said that his allegations in his second witness statement of Poole's involvement in the violence had been untrue. He said that he had made them as a result of police pressure to implicate Poole, the police having taken a statement from Stadden implicating him that they wanted him to confirm. He said that they had suggested that if he did not co-operate in that way they might have to conclude that, as he was the last to leave the house that night he, Juke, might have had something to do with Wiltshire's injuries.

The first appeal to the Court of Appeal

40. In their first appeals against conviction to this Court in 1996 the appellants relied on a number of grounds, including the following: 1) there was fresh medical evidence casting doubt on the cause of death; 2) Stadden's evidence was unreliable; 3) the fresh evidence of Neale and DCs Paine and Cheminais, and/or previously undisclosed prosecution material, cast doubt on the evidence of White; 4) there had been an abuse of process in that the prosecution had failed to disclose the pre-trial witness statements of Juke, and that, although, at the request of the defence, it had required his attendance at the committal proceedings, DI Gladding had warned him not to attend and had threatened him with arrest if he did so; and 5) by reason of DI Gladding's senior position and close involvement with the investigation and preparation of the case, his misconduct towards Juke and his lies in the witness box about it had undermined the integrity of the prosecution so as to amount to an abuse of process of the court.
41. The Court dismissed both appeals.

1) As to causation - it held that, although the precise medical effect of the wounds leading to Wiltshire's heart failure and death was not as had been advanced by the prosecution medical experts at trial, the evidence at trial and the fresh evidence that the Court had heard showed that the wounds had nevertheless caused or had, at least, contributed significantly to his death.

2) As to Stadden's evidence – the weaknesses in her evidence had been thoroughly explored at trial, and the jury had been entitled to act on it. Otton LJ, giving the judgment of the Court, said at page 76 of its judgment:

"Many of these matters were explored and brought out in cross-examination at trial. The jury clearly believed Stadden and could not have convicted on the basis of disbelieving her and believing White."

.

3) As to the evidence of White - the Court observed that if the fresh evidence it had heard from DC Paine and the post-trial statement of Neale and the other associated information, suggesting that White had lied in his evidence at the trial, had been made available to the defence at the time, it would have provided them strong material for cross-examination to discredit his evidence and/or of the possibility of calling one or more of them to give evidence to contradict his account. The Court, regarded with some scepticism DCs Paine's explanation to it that he had believed White's account in his witness statement of having gone to the scene on his own. And it described the non-disclosure to the defence of what White had said to him and DC Cheminais as a material irregularity. For those and other reasons, Otton LJ said, at page 52 of the judgment, that the Court was "in a state of considerable unease about White's reliability as a witness", but that it regarded his evidence as so seriously flawed in any event that it "could have carried little weight with the jury":

"It is clear to us that by this stage White's evidence was not seen to be critical. The jury must have been sceptical about what he saw and what he claimed to have seen. The photographs clearly indicate that it would have been impossible to see from the pavement the hand and arm of a man on the floor of the front room. There was evidence that the hi fi equipment was playing and thus there must have been doubt in the jury's mind about his claim to have heard the deceased shout 'no Tony no, no'. This remark was not heard by Stadden. She did not describe the deceased lying on the ground with his hand in the air as Poole struck down at him. Thus it is clear that White's evidence could have carried little or no weight with the jury. Whatever explanation White might have given concerning the discrepancies in what he told the police, and even if Neale had given evidence, this additional material would have done little to undermine further the credibility of an already unreliable witness. Moreover in spite of the material irregularity and our scepticism we are satisfied that they are not sufficient to render the conviction unsafe."

4) As to Juke - the Court of Appeal, unlike the jury, heard his evidence; it also heard the evidence of DI Gladding on the subject of their telephone conversation warning him not to attend the committal proceedings, together with the threat that accompanied it. The Court regarded that conversation and DI Gladding's denial of it in evidence at trial as a serious matter that caused it "great concern", but did not consider that it rendered the convictions unsafe. As to the non-disclosure of Juke's witness statements, the Court held that it did not amount to a material irregularity and that, in any event, any prejudice caused by such non-disclosure had been "largely eliminated" by references to the account he had given in his second statement in the questions put by the police to Mills and Poole in interview. Further, if the defence had been able to call him, his evidence would not have raised doubts in the jury's minds as to the appellants' guilt, principally because of the discrepancy between his account of the injuries inflicted on Wiltshire and those that the doctors found. This is how Otton LJ put it at pages 72b-73F of the judgment:

"If our conclusion had been that that there was a material irregularity in the non-disclosure of Juke's statement, we would have had to consider whether for that reason either conviction was unsafe. Having heard Juke in the witness box we are sure that his evidence would not have raised doubts in the jury's minds as to the guilt of either defendant. We have already referred to the conflict between the injuries spoken of by Juke and the casualty doctor's findings. It was Juke who took the deceased from the front room in which the incident happened and left him on the street outside. He was thus well placed to know in general terms what injuries the deceased had sustained,

Miss Stadden said that Juke had for a period covered her eyes so that she would not see what was happening. Juke agreed that he had done so, and that in doing so he had turned his head away from the incident and towards Miss Stadden. But, he insisted, he had covered her eyes only for a moment, and only turned his head for that moment. Thus it was not possible to explain his not having seen the infliction of the injuries to the deceased's buttocks and below his kneecaps on that account.

In our judgment, Juke's evidence would only have served to underscore the disproportion between the extensive injuries to Wiltshire seen at casualty and the minor injury to Mills. Any question of who began the violence aside, the distribution of those injuries to the deceased that were seen pre arrest told most strongly against primarily Mills, but Poole also.

Moreover, it is likely the jury would have connected these two pieces of evidence and drawn the inference (as we unhesitatingly do) that Juke covered her eyes to spare her witnessing the savagery of the attack on the helpless Wiltshire – a visitor from the London drug scene who was to be taught a lesson for attempting to muscle in on the Gloucester drug scene on which Poole and Mills were obvious key players.

We are fully satisfied that had Mr. McDonald called Juke at the trial his evidence would not have assisted either appellant in any way."

Otton LJ obliquely criticised the admission into evidence of DS Jeynes' question to Mills in interview rehearsing what Juke had told him of the violence, despite Mills' denial of it, expressing surprise that the defence had made no attempt to exclude it. But, he described it as "an accurate summary of the substance of his second statement" and observed that it "was inconsistent with Mills' defence of self-defence and Poole's denial of any involvement".

5) On the question whether DI Gladding's misconduct towards Juke may have so infected the whole police investigation as to amount to an abuse of process and/or to render the conviction unsafe on that account, the Court, as we have said, clearly accepted that he had been guilty of serious misconduct, but was of the view that its effect was not such as to render the convictions unsafe. This is how Otton LJ dealt with the matter at pages 54-58 of the judgment:

"This is a serious matter and has caused us great concern. We have read the transcript of the tape, and the transcript of Gladding's cross-examination on this point at trial. The decision is undoubtedly unequivocal. ….

We have come to the conclusion that Mr. Gladding behaved in an exceedingly unwise manner. … He should never have gone so far as to threaten Juke that he would be arrested if he were to attend court. It is most unfortunate that he answered proper questions in cross-examination so readily in an improper and misleading manner. We apprehend that this conduct will have been dealt with by disciplinary procedures. …

We do not accept that what occurred at committal or at trial was so unfair and wrong that the judge seized of all the facts would not have allowed the case to proceed with what was in all other respects a regular proceeding. We do not consider that the conduct as revealed before us, reprehensible as it was, amounts to an abuse of process to the extent that we should set aside the convictions."

The appeal to the House of Lords

42. The appellants, with the leave of the House of Lords, appealed to their Lordships on a point certified by the Court of Appeal concerning the prosecution's failure to disclose to the defence before or at trial the contents of the two witness statements of Juke. The House of Lords dismissed the appeal (R v. Mills & Poole [1998] AC 382), holding that, although prosecuting counsel's failure to provide copies of Juke's witness statements to the defence was a material irregularity, the Court of Appeal had been entitled to conclude that if Juke had given evidence for the defence the jury would still have convicted and, accordingly, the convictions were not unsafe. In so holding, their Lordships took account of the information Juke had given to Mr. Gadd, the defence solicitor, when the latter interviewed him before the trial and the fact that the police had put Juke's second statement to Mills in the course of interview. Lord Hutton, in a speech with which all their Lordships agreed, said, as the Court of Appeal had done, that, on the facts, there was little or no prejudice to the defence from non-disclosure and that if the defence had been able to call Juke it would not have assisted either defendant. As to non-disclosure, he said, at pp 405-406:

"Therefore the final question for decision is whether this material irregularity made either of the convictions unsafe. In my opinion it did not for two reasons. First, although the non-disclosure of Juke's two statements meant that the defence did not have precise knowledge of what Juke had told the police in those statements about the actions of the two defendants, nevertheless the information which Juke gave to Mr. Gadd, the solicitor for the defendants, in the detailed interview which the latter conducted with him, together with the substance of Juke's second statement which was put to Mills by the police when they questioned him, meant that the defence were aware of the general nature of the information which Juke had given to the police and were alerted to the risks which would be involved in calling him as a defence witness. Accordingly I am in agreement with the view of the Court of Appeal that any prejudice from the non-disclosure was thereby largely eliminated."

As to the deprivation of the opportunity for the defence to call Juke, he rehearsed and agreed with the Court of Appeal's view that if Juke had given evidence, the jury would still have convicted. As we have said, the sole issue before the House of Lords was as to the non-disclosure of what Juke had told the police. Understandably, the House did not, therefore, consider the propriety or otherwise of the admission into evidence of DS Jeynes' account, when interviewing Mills, of what Juke had told him. Lord Hutton merely observed, at page 392, that the form of DS Jeynes' question must have made it obvious to Mills that Juke was the prime source of the version that he was putting to him.

"Because the police referred to allegations of 'witnesses that were in the room and have made statements to us' it must have been clear to Mills that the police were putting to him what Stadden and Juke had said to them in their statements, and the reference to 'the one person' followed by 'he gained the impression' must have made it clear to Mills that Juke had told them that in his view Mills was intending to break Wiltshire's legs in order to settle the matter once and for all."

The libel action

43. In January 1994 Channel 4 broadcast a television programme in its "Trial and Error" series, which it later included in a book, in which it was alleged that DI Gladding had lied at the criminal trial about his dealings with Juke. In the libel action brought against it by DI Gladding, Channel 4, in its pleas of justification, maintained that he had perverted the course of justice and/or had perjured himself at the criminal trial. At the trial in 1998 before Morland J. and a jury, the jury found in favour of Channel 4. The clear implication of the jury's verdict was that DI Gladding, in his dealings with Juke and/or in his lies about them to the criminal jury, had perverted the course of justice and had committed perjury.
44. In the course of the libel trial, both DI Gladding and DS Jeynes were closely questioned about the former's treatment of Juke and the latter's misleading description of Juke's account in his questioning of Mills. DI Gladding accepted that it was a misleading summary. But he denied that he had knowingly allowed DS Jeynes to mislead Mills in that way or that he had realised afterwards, when reading the transcript of the interview, the discrepancy between it and what was in Juke's written statements. And DS Jeynes said that his summary in the interview accurately reflected what witnesses, or possibly just Juke, had told him, but that he had not recorded everything they had said to him and that, if he had misled Mills, it had been unintentional.

The CCRC's decision in November 1998 not to refer

45. In the meantime, the appellants had applied, in January 1998, to the CCRC to refer their convictions back to the Court of Appeal on the strength of new evidence going to the issue of causation. They later also drew the CCRC's attention to the outcome of the libel trial maintaining that it confirmed the allegations in the Channel 4 broadcast that DI Gladding had committed perjury at the criminal trial about his dealings with Juke and had thereby perverted the course of justice. They invited the CCRC to consider against the back cloth of that development a number of other complaints, including those aired in their appeals to the Court of Appeal and the House of Lords about the conduct of the police with regard to Stadden's and White's witness statements and about the quality of their evidence. They also complained that the Court of Appeal had wrongly accepted that DS Jeynes' summary, in his interview of Mills, of Juke's second statement was correct. They maintained that the summary was inaccurate and misleading in its references to "breaking" someone's legs and to an intention to "teach him [Wiltshire] a lesson", neither of which had appeared in the statements of Juke or, for that matter, of Stadden.
46. In November 2000 the CCRC decided that the libel trial had not produced any significant material that the Court of Appeal had not already considered and rejected as a basis for finding the convictions unsafe. In doing so, it noted the Court of Appeal's view that the criminal jury had been entitled to rely on the evidence of Stadden and its, the Court's, poor view of White as a witness at trial and of Juke in the appeal.
47. We should note that the Commission was not troubled by the point taken on behalf of the appellants that DS Jeynes' references to Juke's account in his interview of Mills were inaccurate and misleading. It noted that the references had not deterred Mills from insisting that he had been acting in self-defence. And, on its reading of the transcript of the libel trial, it appears to have accepted that DS Jeynes had not intended to mislead Mills, and to have rejected the suggestion that DI Gladding had knowingly allowed him to do so. It concluded on this point at paragraph 4.3. of its decision:

"… if the Court of Appeal was now to be made aware of the inaccurate nature of the question posed in Mr Mills' interview, it would be likely to consider that had an application (informed by documents disclosed prior to the appeal) been made to exclude the question and any relevant answers, it would undoubtedly have been successful. The question then to be considered would be whether or not the mere existence of the question, being a matter which had gone before the jury, would have been so prejudicial as to upset the safety of the conviction. In the Commission's view there is no real possibility that the Court would come to such a view. The Commission considers that the Court would consider the evidence in the round and would take account of both the weight of the other evidence available, as well as the fact that the inaccurate summary provided did not in fact give rise to any admissions by Mr. Mills."

It is plain that the Commission, like everyone else up to that stage, did not go beyond considering the effect, if any, of the inaccuracy of DS Jeynes' representation to Mills of what Juke had said. That is, it did not go on to consider the possible prejudice to the appellants, Poole in particular, of the indirect admission in that way of challenged hearsay evidence, whether accurately represented or not, on matters central to the issues that the jury had to decide.

The Divisional Court

48. The appellants sought judicial review of the Commission's refusal to refer the matter to this Court. In December 2001 the Court, consisting of Lord Woolf CJ and Ouseley J, dismissed the application, holding that the Commission had been entitled to conclude, for the reasons it gave, that the matters relied on in the application for the reference had "in their essence" already been considered by the Court of Appeal. However, Lord Woolf, in giving the judgment of the Court, took what he described as an exceptional course of expressing unease about the safety of the convictions. He said:

"114. … the Court of Appeal now could have a doubt about the safety of these convictions. In expressing that view we note the recent statement of Lord Bingham of Cornhill in R v. Pendleton [2001] UKHL 66, at para. 19 that in considering whether to allow an appeal the Court of Appeal should 'bear very clearly in mind the question for consideration is whether the conviction is safe not whether the accused is guilty'.

115. First of all in relation to Poole, there is only one witness who gave evidence that might be reliable in directly making him a party to the offence though we have not forgotten the telling nature of the deceased's injuries. The reliability of Miss Stadden's evidence and of her second statement might have been viewed differently in light of the full knowledge of the array of misconduct. Secondly, in relation to Mills, though again we bear in mind the telling nature of the injuries suffered by the deceased, his argument as to self-defence might have been viewed differently in the light of all the misconduct.

116. Although the Court of Appeal had the material to form their own view of D.I. Gladding's behaviour, the verdict of the jury in the libel action is damning condemnation of his conduct and that verdict was reached after a more detailed examination than would have been possible in the Court of Appeal when his conduct was only one of many issues before that Court. He was found guilty of perjury and perverting the course of justice in relation to this case. Any jury would be and, in our view, would rightly be, deeply influenced by this finding if they could have known of it. It just might have affected their view of Miss Stadden who was such a critical witness. The trouble with a senior officer behaving as D.I. Gladding has been found to have behaved is that a jury would find it difficult to be sure that there was not other misconduct for which he could have been responsible.

117. Then there is the problem with the interview. It could all have been an innocent mistake but was certainly unfortunate and worrying because it was done under D.I. Gladding's supervision.

118. To this has to be added what happened in relation to Mr. White, which was again at least unfortunate. Finally there is the non-disclosure as to Mr. Armstrong. This relates to an entirely separate issue but it again reflects badly on how this case was being handled by the police. There was at least a chapter of errors. Every aspect of the case including Miss Stadden's two statements is not straightforward. Almost every aspect of the prosecution is tarnished.

119. In expressing the views we have we do not suggest the Court of Appeal's previous decision can be faulted. … But although the new material may not be that significant it can still be sufficient to tip the balance, from upholding the conviction to allowing an appeal.

120. It is entirely a matter for the Commission to decide what if any weight to attach to the view we have expressed. If the matter is to be reconsidered by them it is desirable this is done while the facts are still fresh in the Commission's mind."

The referral

49. The Commission duly reconsidered the matter and on 22^nd May 2002 referred it to this Court, expressing the view that there was now "a real possibility that the Court of Appeal would find these convictions to be unsafe". It gave five reasons for that view.
50. First, the Commission saw no evidence to suggest that Stadden's second witness statement incriminating both appellants was a result of any attempt by DI Gladding to corrupt her, as Lord Woolf had suggested in the Divisional Court. However, it concluded, in the light of DS Jeynes's explanation at the libel trial about his use of the words "punishment beating" when questioning Mills, that it could not be certain that everything said by Stadden in evidence "represented a full and accurate recollection" of what she had said at the time.
51. Second, the Commission stated that the Court of Appeal's conclusion that the jury must have dismissed Mr. White's evidence as incredible should be re-visited in the light of R v. Pendleton [2002] 1 Cr App R 34, HL.
52. Third, the Commission expressed concern about the failure of counsel on both sides to remove from the transcript of DS Jeynes' interview of Mills his misleading summary of Juke's second witness statement before it was put before the jury. It said that, having regard to Pendleton, the misleading questioning was no longer relevant only to whether it had prompted Mills to make an incriminating admission (which it had not), but to whether the prosecution had proposed to call evidence to support the assertion:

"4.4. … the Crown, the Commission concludes, should not have been allowed to place before the jury a question, whether denied by the witness [sic] or not, in the form that the suggestion of a punishment beating came from a witness whom the Crown did not intend to call. Counsel for Messrs, Mills and Poole rightly points out that there is danger that such a question might have had some prejudicial effect upon the jury.

4.5 The Commission notes further that defence counsel did not invite the trial judge to draw the sting of the inference left by the question, by means of an appropriate direction."

53. Fourth, the Commission expressed the view, notwithstanding the Court of Appeal's holding that Juke's evidence before it, if given at trial, would not have helped either appellant, that there could be no certainty as to what evidence Juke might have given at the committal or trial or as to the value he might have had as a defence witness. Having considered various possibilities, it concluded, in paragraph 4.18 of its Statement of Reasons that it could not say:

"with certainty, that the defence suffered no detriment from its inability to explore Mr. Juke's evidence at the committal proceedings, caused by D.I. Gladding's warning to Mr. Juke. The Commission does not consider it likely that such prejudice accrued, however, it cannot go further than that."

54. Fifth and in reaching its final decision, the Commission, whilst rejecting the suggestion that there had been an abuse of process, observed that the appellants' case was that they had not received a fair trial because of police misconduct and inefficiency of which the prosecution had been unaware. It concluded, at para 6.1 of its Statement of Reasons that, in the light of all the information before it "regarding the police handling of the case …, in particular, with regard to the judgment in …Pendleton," that there was a real possibility that this Court would find both convictions unsafe.

The issues in the appeal

55. The appellants maintain, largely for the reasons given by Lord Woolf CJ, that their convictions are unsafe on the following common grounds:

1) the close examination in the libel trial of DI Gladding's conduct in relation to Juke revealed a level of impropriety that tainted the whole police investigation more profoundly than had been recognised in the previous appeal, so as to cast fresh doubt on all the prosecution evidence, and in particular, that of Stadden;

2) in the light of Pendleton, the first Court of Appeal was wrong to hold the conviction safe on the basis that, although there had been serious improprieties in the handling of White's evidence, the jury would not, in any event, have relied on it to convict;

3) in the light of Pendleton, the first Court of Appeal was wrong to hold that, even if the defence had called Juke to give evidence, it would not have raised doubts in the jury's minds as to the appellants' guilt;

4) the inaccuracy of DS Jeynes' rehearsal of Juke's second witness statement in questioning Mills, only identified after the first appeal, introduces a new and separate argument from that considered by the House of Lords in its ruling that any prejudice to the defence arising from non-disclosure of the statement was largely overcome by the substance of it having been put to Mills in interview; and

5) the admission into evidence of that hearsay evidence of DS Jeynes, whether accurate or inaccurate, was unfairly prejudicial to the appellants.

The law

56. The Commission may only refer a conviction to the Court of Appeal if it considers that there is a real possibility that the Court would not uphold the conviction because of new argument or evidence, or because there are exceptional circumstances for making it; section 13 of the Criminal Appeal Act 1995. However, once the matter is referred, the appeal is not confined to the Commission's reasons for the referral. It may be on any ground; section 14(5) of the 1995 Act. That may consist of or include a ground that has already been aired in a previous appellate hearing in the matter. But normally the proper exercise of the Court's discretion under section 14(5) involving departure from its previous reasoning should equally be confined to exceptional circumstances; R v. Ian Thomas [2002] EWCA Crim 941; and R v. Wallace Duncan Smith [2002] EWCA Crim 941.
57. Exceptional circumstances may exist where, for example, there was some cogent argument advanced, but not properly developed at the previous appellate hearing, but which as now developed could persuade the Court that the conviction was unsafe (cf. R v. Chard (1984) 78 Cr. App. R. per Lord Diplock at 113, under the former procedure of reference by the Home Secretary, which, though referring to the development of an argument not previously advanced, had much the same principle in mind). Other examples, as the Court observed in Thomas, at paragraph 74, may be where there has been a development of the law requiring the adoption of a different approach by the Court to the issue before it, or where there has arisen some tension between overlapping principles such as that between the statutory criterion for safety of a conviction and the ECHR concept of a fair trial. As the Court said:

"…the exceptional circumstances, whatever they are, would have to be such as would convince the Court that if the matter had been arguable and argued in that way before the previous Court, it would – not might – have quashed the conviction. The Court should in any such cases be very slow to differ from its previous judgment."

58. This approach, it seems to us, is consistent with the reasoning in Pendleton of Lord Bingham, at paragraph 19, and Lord Hope at paragraph 33, of the need of the Court itself to be sure of the safety of the conviction, as distinct from sureness of guilt or that, if such circumstances had been before the jury, it would still have convicted. See also per Lord Hobhouse in Pendleton at paragraph 36 and R v. Maloney [2003] EWCA Crim 1373, at para. 45.
59. Provided that the Court, on a second appeal as well as the first, keeps in mind its own need to be sure of the safety of a conviction, there is nothing in the judgment of this Court in R v Wallace Duncan Smith, to which Miss Baird referred the Court, to suggest that a restraint of exceptionality on the exercise of the Court's discretion under section 14(5) is inappropriate where it is considering on a reference its earlier decision on the same argument or evidence. Smith was a directions application in which the Court decided that the combined effect of a reference and section 14(5) enabled an appellant following a reference to argue without leave any ground he chose even if it had not been considered or had been rejected by the Commission. It is no authority for the proposition, in particular, in Buxton LJ's observations in paragraph 30 of the judgment of the Court, that, it would be a proper exercise of the Court's discretion under section 14(5) in the absence of exceptional circumstances to revisit its previous decisions when there was no new argument, whether of law or fact, or new evidence.
60. This restraint of exceptionality on the exercise by the Court, in the absence of new argument or evidence, of its discretion under section 14(5) to consider a ground of appeal considered by it before, does not, in our view, govern its consideration of grounds rejected by the Commission. Whilst the Court will, of course, treat the Commission's reasoning on making a reference with considerable respect, its discretion to consider a ground not included in the Commission's reasons for the reference or to reject or follow the Commission's reasoning one way or another on any matter considered by it is unfettered; see R v. Garner [2002] EWCA 1166.
61. Accordingly, where there is no new argument or evidence, there is no reason why the Court should only observe the restraint on the exercise of its power to revisit its former decision where the Commission has supported the former decision, as suggested contingently by Miss Baird. And neither the full citation from the relevant passage in Lord Diplock's speech in Chard, at p. 113, nor Thomas nor Smith is an authority for such proposition. The same applies to observations by a superior court exercising another jurisdiction, such as the Divisional Court, presided over by Lord Woolf CJ in this case. His observations, though of high persuasive authority, should not deter us from keeping in mind, alongside the safety of convictions, the public and private interests in an orderly, as well as just, system that secures finality of decisions.
62. It follows that the task for the Court on this appeal is to determine: 1) whether there is some new argument or evidence not previously before, or properly developed before, the Court; 2) if not, whether there are any exceptional circumstances which, in our view, make either conviction unsafe; and 3) the effect of Pendleton on how we approach our task.
63. The first point to note about Pendleton is that it was a fresh evidence case. The House had to consider how the Court should assess the effect of such evidence on the safety of the conviction. Lord Bingham, with whom the majority of their Lordships agreed, stressed that the Court, when considering such evidence, should focus on its effect on the safety of the verdict rather than on the guilt of the accused. In doing so, he said that it was not the role of the Court to usurp the role of the jury. However, he acknowledged that the Court cannot, save in clear cases either way, entirely exclude from its thought processes questions of how evidence at trial, if put alongside the fresh evidence before it, might or would have affected the jury's decision. It is clear from his reasoning that in those cases where the answer to the question, safety or unsafety, is not immediately clear, the Court may have to ask itself, for example, whether and to what extent the jury relied on a particular piece of evidence in convicting an accused and how their decision might reasonably have been affected if they had known what the Court now knows. This is how Lord Bingham, at paragraph 18, spoke of the varying degrees to which the Court may intrude on the jury's domain according to the nature of the issue in play and the obviousness of the impact of the fresh evidence on it:

"18. Where the Court of Appeal has heard oral evidence under section 23(1) (c) … the evidence will almost always have appeared, on paper, to be capable of belief and to afford a possible ground for allowing the appeal. By the time the Court comes to decide whether the appeal should be allowed or dismissed, it will have heard the evidence, including cross-examination and any submissions made on its effect. It may then conclude, without doubt, that the evidence cannot be accepted or cannot afford a ground for allowing the appeal. … The Court may, on the other hand, judge the fresh evidence to be clearly conclusive in favour of allowing the appeal. … The more difficult cases are of course those which fall between these extreme ends of the spectrum. …The Court of Appeal can make its assessment of the fresh evidence it has heard, but save in a clear case it is at a disadvantage in seeking to relate that evidence to the rest of the evidence which the jury heard. For these reasons it will usually be wise for the Court of Appeal, in a case of any difficulty, to test their own provisional view by asking whether the evidence, if given at the trial, might reasonably have affected the decision of the trial jury to convict. If it might, the conviction must be thought to be unsafe."

64. In our view, the Pendleton jury impact test, looked at as a range of permissible intrusion into the jury's thought processes for confirmatory purposes, is equally applicable where the new matter is one of argument, either of law or of interpretation of, or of inference from, the evidence at trial. The Court may also have to ask itself similar questions as to the effect on the jury of evidence improperly given or of other irregularities at trial.

The submissions

65. Miss Vera Baird, QC, for Mills, and Mr. Edward Fitzgerald, QC, for Poole (neither of whom appeared below), placed at the forefront of their submissions the point that so exercised Lord Woolf, the proven misconduct and/or incompetence of a number of police officers. First, there was the improper conduct of DI Gladding toward Juke and his lies about it in evidence at the criminal trial. They maintained, with illustrative references to R v. Maxine Edwards [1996] 2 Cr App R 345 and R v. Zomparelli 23^rd March 2000, the potential for an officer in his position to infect the whole investigative process and evidence so as to render the conviction unsafe. They acknowledged that his misconduct had all been before the Court of Appeal. However, they maintained that the more detailed examination of that conduct in the libel trial, coupled with the libel jury's conclusion that he had perverted the course of justice and perjured himself at the criminal trial, added something new which, if the criminal jury had been aware of it, might have led them to acquit. Second, there was the misconduct of DCs Paine and Cheminais in their dealings with White, adding to the impression of a dishonest and incompetent investigation from top to bottom. Third, there was the behaviour of DS Jeynes in his misrepresentation to Mills in interview of what Juke had said about the attack on Wiltshire. In summary, they submitted that the cumulative effect of what had happened since the first appeal and the evidence and arguments in that appeal is seriously to discredit an already tainted investigation and to render the convictions unsafe.
66. Mr. Jenkins rejected the notion that DI Gladding, by reason of his seniority and established misconduct in his behaviour towards Juke, with which the libel trial had been concerned, had somehow infected the whole police investigation of this matter. He said that there was no evidence to support such a wide-ranging suggestion and that, in any event, no new evidence or argument bearing on it had emerged in the libel trial which had not already been before and considered by the first Court of Appeal.
67. Miss Baird and Mr. Fitzgerald also relied on the unreliability of the evidence of Stadden and White.
68. As to Stadden, they maintained that she had dubious credibility, given: 1) her drug habit and involvement with Wiltshire in cheque fraud; 2) her contradictory accounts as to whether Mills had stabbed Wiltshire and as to Poole's involvement in the attack on him; 3) her and WPC Moore's and DC Churchill's conflicting explanations for her initial failure to mention Poole's involvement; 4) her uncertainty as to when she had called the "Crime Line" programme; 5) a police action note not disclosed to the defence at trial, but which appears to have led nowhere, recording a suggestion that a friend of Wiltshire named Williams had given Stadden a prepared statement with regard to the fatal attack; and 6) the verdict of the jury in the libel trial. Stadden was just the sort of person, Miss Baird and Mr. Fitzgerald suggested, who might have been vulnerable to police pressure or persuasion to give untrue evidence. For that reason, she had required honest and careful handling by the police, the sort of handling of which a fully informed jury might not have been confident. At the very least, they added, given DS Jeynes's evidence of his manner of recording Juke's witness statements, it may be that observations by her favourable to the defence were not recorded in her witness statements.
69. Mr. Jenkins, on the other hand, submitted, in agreement with the Commission, that there is no evidence to support the suggestion that Stadden's second witness statement incriminating Poole resulted from any attempt by DI Gladding to corrupt her. In so submitting, he made the valid point that she had made her statement to Crime Line implicating Poole before she made either of her witness statements and before DI Gladding had any contact with her. He added that the broad consistency of her account as to Poole's involvement, save for the lack of reference to it in her first witness statement, was supported by the evidence of WDC Moore and DC Churchill that she had in fact mentioned it when making that statement, but that WDC Moore had inadvertently omitted to record it. In summary, Mr. Jenkins submitted that there is no new evidence or argument in respect of Stadden's evidence; all the conflicts and uncertainties in it upon which the appellants rely were aired before the criminal jury and reconsidered in the light of the considerable body of fresh evidence heard by the first Court of Appeal.
70. As to White, Miss Baird's and Mr. Fitzgerald's main complaint was of the non-disclosure to the defence that the police had allowed him to make a witness statement claiming to have been on his own at the relevant time, when he had told them that he was with Neale. They challenged, by reference to Stadden's evidence at the committal proceedings and that of White at the trial, the first Court of Appeal's view that his evidence could have had little impact on the jury because of the Court's doubt as to what he could have seen and heard of the fight inside Poole's room. They also drew attention to the prosecution's strong reliance on his evidence throughout the trial as an honest and reliable witness. They submitted that, had the Court been armed with the much fuller examination of DI Gladding's role in the libel trial, including a reference to him in the undisclosed internal police note mentioning White's account of having been with Neale at the relevant time, and the libel jury's verdict, it might have reached a different view. They maintained that the Court, in concluding that White's evidence would not have influenced the jury, undertook an exercise, namely substituting its view for that of the jury, that was inconsistent with and could not survive Pendleton.
71. Mr. Jenkins relied on the first Court of Appeal's view that White's evidence to the jury would have carried little or no weight and that the additional material with regard to Neale would have done little to undermine further his already damaged credibility. He maintained that, even looking at the matter with Pendleton hindsight, the first Court was entitled to conclude, in holding the convictions to be safe, that the jury must have reached them in reliance on Stadden's evidence.
72. As to Juke, Miss Baird and Mr. Fitzgerald had four complaints: 1) that, but for DI Gladding's misconduct in preventing him from attending and from his possible deployment as a defence witness at the committal proceedings, the defence might have been able to expose that misconduct and the misrepresentation by DS Jeynes to Mills of what Juke had said, and to raise questions as to the propriety of the police conduct generally in their treatment of prosecution witnesses; 2) that the first Court of Appeal's rejection of him as a credible witness, particularly because of his lack of explanation for the injuries suffered by Wiltshire, was not, in the light of Pendleton, a safe basis on which to dismiss him as a witness who could not have swayed the jury; 3) that the first Court of Appeal and the House of Lords had mistakenly proceeded on the basis that the account put by DS Jeynes in his interview of Mills accurately reflected what Juke had said; and 4) that neither tribunal had considered the question of prejudice to the appellants in the admission into evidence in this way of the hearsay account of Juke, whether accurate or inaccurate.
73. Mr. Jenkins submitted that, on the issue of non-disclosure, there was no new argument or evidence and no exceptional circumstances entitling this Court to go behind the decision of the first Court of Appeal, however much we might disagree with the basis of its reasoning for rejection of Juke as a credible witness. As to the first Court of Appeal's and the House of Lords' unawareness that Jeynes had put a misleading version to Mills of what Juke had said, he said that, though it was a new argument, it could not in itself provide a rational basis for holding the convictions to be unsafe. And, as to the argument that, misleading or not, the inclusion in DS Jeynes' evidence of Juke's hearsay account was prejudicial to both appellants, he conceded that it is a new argument. But, he asked "rhetorically", if it is of such significance to the safety of the verdicts, why has it not been relied on before now as a ground of appeal?

Conclusions

General

74. Although the appellants were separately represented on this appeal, they had, as we have indicated, the same counsel at the trial. Their defences were different, but consistent one with another. Mills relied on self-defence; Poole supported Mills' account and denied that he had been a party to any of the violence. As Miss Baird noted in her submission, they were witnesses for each other so that any unfairness to one in the trial would have been likely to diminish the defence of the other. In addition, Mills' defence of self-defence would have been substantially undermined by any evidence, admissible or inadmissible, suggesting that both of them were engaging Wiltshire in the fight.
75. Many of the complaints of Miss Baird and Mr. Fitzgerald of the police treatment of Stadden, White and Juke, and about their vulnerability and/or unreliability, are not new arguments nor do they arise from any new evidence. They were fully considered and rejected by the first Court of Appeal after hearing the evidence of, amongst others, DI Gladding and Juke. We have considered whether any new matter or significance has emerged as a result of the more detailed examination of DI Gladding's and DS Jeynes' conduct in the libel trial, such as to raise a new argument of the sort considered by Lord Woolf of overall infection of the police investigation so as to cast new doubt on the reliability of, in particular, Stadden, White and/or Juke. We have also considered whether Pendleton gives rise to a new argument based on the first Court of Appeal's examination of the likely credibility of those witnesses and the effect of that authority on our approach.

DI Gladding and the integrity of the police investigation

76. As to the significance of DI Gladding's misconduct towards Juke as a pointer to infection, or undermining the integrity, of the police investigation as a whole, nothing new by way of argument or evidence has emerged since the first Court of Appeal's decision that could render the decision unsafe. Nor are there any exceptional circumstances pointing to unsafety that would justify us re-visiting the first Court's decision on that account. The first Court had the advantage of hearing Juke as well as DI Gladding; it heard the evidence of the offending telephone conversation between the two men and as, a result, could not have avoided the conclusion that DI Gladding, in that conversation and in his lies about it to the jury, had at least attempted to pervert the course of justice and had perjured himself. As Lord Woolf acknowledged, at paragraph 116 of his judgment in the Divisional Court, the first Court, as a result, "had all the material to form their own view of DI Gladding's behaviour". The Court of Appeal's view of the reprehensible nature of this conduct is plain from its observations about it in its judgment. The fact that Otton LJ, in giving the judgment of the Court, did not understandably feel able to give it the labels "perversion of the course of justice" and "perjury" in a proceeding to which he was not a party, does not detract from the Court's understanding of what DI Gladding had done and the seriousness of it. And the fact that, following a later and more detailed examination by a libel jury, those labels have now been applied, does not affect the underlying conduct of which the first Court of Appeal was fully aware or its decision that it had not been such as to render the criminal trial an abuse of process or the convictions unsafe.
77. If, as the Commission, in its Statement of Reasons for Referral, and we consider, there was and is no evidence to suggest that DI Gladding attempted to corrupt Stadden or of any "systemic" dishonesty in the investigation such as to undermine the integrity of the prosecution as a whole, it is difficult to see on what basis this argument could succeed. The most important witness for the prosecution was Stadden; the prosecution could not succeed unless the jury at least found her to be credible and her account reliable; but that is a different issue.

Stadden – credibility and reliability

78. As we have indicated, the first Court of Appeal held that the inconsistencies and weaknesses of her evidence had been fully explored at the trial and that the jury had been entitled to act on it. There was, as Mr. Jenkins put at the forefront of his argument on this point, also the broad consistency between her Crime Line statement, her second witness statement and her evidence, coupled with an explanation for the inconsistency of her first statement vouched for by police officers whose evidence may suggest error, but not dishonesty. She had a critical role in the prosecution case, the only eye-witness to give evidence for the prosecution of nearly all that occurred and potentially the only such witness of any of it. In those circumstances, the Court was entitled, even post Pendleton, to express the only common sense view that the jury clearly believed her and could not have convicted on the basis of disbelieving her and believing White. The only new matters on which the appellants can rely are the undisclosed police action note suggesting that a friend of Wiltshire had prepared her statement for Crime Line, a suggestion that led nowhere, and the argument that DS Jeynes' lax method of recording witness statements may have been symptomatic of the general practice of the officers in the investigation, including those who took her witness statements. Both of these are highly speculative; there is no evidential support for either of them. As to the second, the highest that the Commission felt able to put it in favour of the appellants was that it could not be certain that everything said by Stadden represented a full and accurate recollection of what she had said when making her witness statements. Having ventured that modest caution - one that must apply to many, if not most witness statements, it concluded on the point, as we do, that the essential question is the reliability of her evidence. In our view, there is nothing new, by way of argument or evidence or any exceptional circumstances for departing from the response of the first Court of Appeal to this challenge to the safety of the convictions.

White

79. The Judge, in his summing-up of White's evidence to the jury, described him as an "important witness" and expressly told them that it was for them "to decide whether he [was] a truthful and reliable witness". In our view, the first Court's view that the jury, even without the Neale evidence, would have regarded his evidence as to what he saw and heard as unreliable offends the Pendleton principle of substituting the Court's for the jury's view of the evidence. There is no knowing what the jury thought of the credibility or reliability of White's evidence of what he could and did see and hear of the fight from his vantage point outside on the street. The prosecution had not conceded that he was unreliable. And, as we have mentioned, the Judge, in his summing-up, described him as "an important witness" and concluded his references to his evidence by directing that it was for them "to decide whether he [was] a truthful and reliable witness". Since, as the House of Lords held, the failure to disclose the Neale information was a material irregularity, this Court must consider the effect of the non-compliance of the first Court of Appeal with Pendleton in, as it seems to us, intruding on the jury's role in assessing the credibility and reliability of an important witness in the trial. He was obviously important, as the trial Judge acknowledged, because he was the only potential prosecution eye-witness apart from Stadden. His evidence, if accepted by the jury, might reasonably have been regarded by them as reassuring confirmation of Stadden's account – enough perhaps to clinch any tentative doubts that they may have felt when considering whether they believed and could rely on her. In saying this, we are conscious that we too are in danger of crossing the Pendleton line, but we do so only because we are sufficiently uneasy about the effect of the inability of the defence to exploit the Neale material on the safety of the verdict. Put another way, we do not agree with Mr. Jenkins' submission that, even looking at the matter with Pendleton hindsight, the jury must - or would - have convicted on reliance on Stadden's evidence, if she had been the only eye-witness.

Juke

80. It is an irony that the original focus of complaint about the treatment of Juke and of the non-disclosure of his witness statements to the defence may have distracted attention from the prejudicial hearsay use of his second witness statement, regardless of its inaccurate representation by DS Jeynes. As it was, the Court of Appeal, though it did not regard the non-disclosure as a material irregularity, considered that the officer's hearsay references to it in the interview would have largely eliminated any prejudice that might have resulted from the non-disclosure. And it considered that, even if he had been called by the defence, his evidence was likely to have been so unreliable as not to have assisted the appellants. Of course, the Court of Appeal did not know, and nor did the House of Lords, which was focusing on the single issue of non-disclosure before it, that DS Jeynes had misrepresented to Mills in interview some of what Juke had alleged in his second witness statement. This point seems to have been identified only by the appellants' advisers on the first application to the Commission for a reference, and to have been rejected by it as immaterial because the inaccuracy had not deterred Mills from saying that he had acted in self-defence. Lord Woolf's concern about this point, briefly expressed in paragraph 113 of his judgment, does not seem to have gone beyond the possibility that the misrepresentation had resulted from DI Gladding's evil influence. The Commission, when it reconsidered referral, after and in the light of Lord Woolf's observations, seems to have come a little closer to considering the possibility of prejudice from the hearsay itself (see paragraph 52 above). However, its focus still appears to have been on the inaccuracy of the hearsay rather than its main thrust, an allegation by Juke, who saw what happened but did not give evidence, inculpating both Mills and Poole.
81. In our view, the important point, on which Miss Baird and Mr. Fitzgerald have in this appeal for the first time given due emphasis, is that DS Jeynes's rehearsal in his interview of Mills of Juke's account in his second witness statement, whether accurate or inaccurate, became hearsay evidence unfairly prejudicial to both appellants on issues central to their respective defences. It was strongly supportive of Stadden's evidence. It contradicted Mills' account of self-defence and it implicated Poole in the attack. It is surprising that defence counsel did not seek exclusion of the offending question and others like it eliciting denials from the copies of the transcripts put before the jury. Whatever their reason for not doing so, the primary concern of the Court now is the effect on the safety of the verdicts of admitting and leaving such critical hearsay material before the jury.
82. In our view, the hearsay was capable of having had a seriously and unfairly prejudicial effect on the jury's deliberations. First, there was the unfortunate impression left with them by Mr. MacDonald in his cross-examination of DI Gladding that Juke was to have been a prosecution witness. Second, the significance of what he might have said if called was clearly a matter of interest to the jury, as their note to the Judge at the end of the trial indicated. Third, DS Jeynes' question effectively introduced a hearsay account of the only other person present in the room throughout the fight with Wiltshire, apart from Stadden and the appellants, an account that they, the appellants, denied when it was put to them. Fourth, quite apart from the partly misleading nature of the question, it was also, as Miss Baird said, disingenuous in its lack of reference to what Juke had said in his first statement. Fifth, although the Judge made no reference to the question is his summing-up, the jury had it before them in writing in the transcript of the interviews when considering their verdicts, in a form and at a time when it could have left a disproportionate impression on them. The Judge could have alerted them to the dangers of such hearsay allegations, though for him to have done so in any specific way would probably have only made matters worse. And, as the jury had the transcripts, it was unfortunate that, instead of possibly directing them in general terms as to what was and was not relevant, his only observation on the subject was to invite them to refer to any passages that they thought were relevant.
83. The only other complaint about Juke's role, or lack of it, in the trial, upon which the Court could properly act on this second appeal is that the first Court's rejection of him as a credible witness, largely because of his lack of explanation for the injuries suffered by Wiltshire, does not accord with Pendleton. However, we acknowledge the force of the reasoning of the first Court on this point (see paragraph 41(4) above). And the likely effect on the jury of his evidence, if he had given it, in our view, falls between the extreme ends of the spectrum of clarity to which Lord Bingham referred in Pendleton (see paragraph 63 above). Nevertheless it is probably not an analysis that the Court could undertake today, though it might, without expressing it in that way, reach the same conclusion as to the effect of this aspect of the trial on the safety of the verdicts.
84. In the result, we are of the view that both convictions are unsafe for two main reasons. The first related to the evidence of an important eye-witness, White, and the second related to the possibly more influential hearsay use of a witness statement of one who was not a witness, Juke. Both could have had an important influence on the jury's conclusion as to the credibility and reliability of the key eye-witness in the case, Stadden. The non-disclosure of the Neale material deprived the jury of the opportunity to test the truth of White's evidence, the credibility and reliability of which were essentially matters for them. As to the hearsay use of Juke's witness statement, we agree with Miss Baird's and Mr. Fitzgerald's submission that, to put before the jury a document containing a graphically phrased, inaccurate, damning and inadmissible account central to the case was improper, and, in the way in which the Judge left it with them, was unfairly prejudicial to the defence of both appellants.
85. Accordingly, we allow both appeals against conviction.
______________