APPEAL COURT, HIGH COURT OF JUSTICIARY
Lord Justice Clerk
Appeal Nos: XC956/03, XC959/03, XC958/03
OPINION OF THE COURT
LORD JUSTICE CLERK
in the references by
THE SCOTTISH CRIMINAL CASES REVIEW COMMISSION
in the cases of
(First) THOMAS CAMPBELL; (Second) JOSEPH STEELE and
(Third) THOMAS GRAY
For Campbell: Bell, QC, Shead; Beltrami Berlow, Glasgow
For Steele: Gebbie, Miss McColl, McLaughlin; McClure Collins
For Gray: Wheatley, solicitor advocate, Macdonald, solicitor advocate; McClure Collins
For the Crown: Mulholland AD, solicitor advocate, Balfour, advocate; Crown Agent
17 March 2004
 The Scottish Criminal Cases Review Commission (the Commission) has referred to us the convictions of Thomas Campbell, Joseph Steele and Thomas Gray at Glasgow High Court on 10 October 1984. In each case the Commission considers that a miscarriage of justice may have occurred and that it is in the interests of justice that the reference should be made.
The ice cream wars
 These cases arise from the Glasgow ice cream wars of the 1980s. At that time rival sellers used violence and intimidation to win control of the routes of ice cream vans in the outskirts of the city. In the east end of Glasgow the principal van operators were Marchetti and Fifti Ices. Campbell and members of his family were associated with Fifti Ices. Andrew Doyle, known as Fat Boy, then aged 18, drove a Marchetti van. It seems that he had resisted attempts to intimidate him.
The shooting of Andrew Doyle's van
 On 29 February 1984 Andrew Doyle was working in his van, with his assistant Anne Wilson, in the Garthamlock area. At about 8 pm he stopped the van in Balveny Street. A red Volvo drew up beside the van. A masked man with a shotgun got out of the car and fired two shots through the windscreen.
The murder of the Doyles
 On 16 April 1984 Doyle's home at Bankend Street, Ruchazie, was set on fire. He and five members of his family, one of them a baby, were burned to death.
 Campbell, Steele and Gray were tried with Thomas Lafferty, John Campbell and Gary Moore on an indictment containing sixteen charges, all relating to the ice cream wars. These references concern only charges (9) and (15).
 Charge (9) was a charge of attempted murder relating to the shooting of Andrew Doyle's van. On this charge Campbell was convicted of assault to danger of life and sentenced to 10 years imprisonment. Gray was convicted of attempted murder and sentenced to 14 years imprisonment.
 Charge (15) was a charge of murder relating to the deaths of the Doyles. It charged Campbell, Steele, Gray and Moore with having committed the murder while acting along with Joseph Granger (otherwise Grainger). The charge against Moore was withdrawn at the end of the trial. Gray was acquitted by direction of the judge on the ground of insufficiency of evidence. Campbell was convicted as libelled and sentenced to life imprisonment, with a recommended minimum period of 20 years imprisonment. Steele was convicted as libelled and sentenced to life imprisonment.
II The cases against Campbell, Steele and Gray
The case against Campbell - charges (9) and (15)
 The Crown case was that Campbell was guilty art and part of the shooting and of the murders because he had planned them.
 The crucial witness for the Crown on both charges was William Love. Like most of the principal personalities in these cases, Love had a criminal record. On his own admission, he had been directly involved in the shooting. On 25 March 1984 he was arrested and charged with an armed robbery at a filling station. He was remanded in custody. He was in custody on the date of the murders. He was later interviewed by the police. On 9 May 1984 he was precognosced on oath. Later that day, he was allowed bail on the armed robbery charge. On 12 July 1984 he appeared on petition on other charges and was remanded in custody. Later in July he was acquitted on the robbery charge. He was still in custody on the other charges when he gave evidence at the trial.
 There were two sources of evidence against Campbell on this charge. The first was the evidence of Love. Love gave evidence as a socius. He said that he drove the Volvo to the locus, but that Gray had fired the shots. He said that some time later, in the Barge public house, Campbell had thanked him for the "message" that he had done in driving the car and said that he would square up with him later. The second source of evidence was the police evidence of a statement allegedly made by Campbell at the time of his arrest, which incriminated him in both the shooting of the van and the murders. We shall discuss this later.
 There was no evidence that Campbell was at the locus when the fire was started. The Crown relied on three sources of evidence. The first was the evidence of Love. Love said that in March 1984 in the Netherfield public house he overheard a conversation between Campbell, Steele, Gray, Granger and Duncan Moore. Campbell and Gray spoke about setting fire to Fat Boy's door to give him a fright. Steele was listening and was saying "aye" as if agreeing with the proposal. There were several others standing round at the time, all of them either nodding or saying "Aye, aye." The second source was the statement allegedly made by Campbell to which we have referred. The third was the finding in Campbell's house of a briefcase allegedly containing a map on which Bankend Street was marked with a circle with a cross inside it, apparently at the location of Andrew Doyle's house. The advocate depute accepts that this element of the evidence was not crucial.
 Campbell's defence was that he was at home with his wife on the night of the shooting and on the night of the fire. He alleged that Love had fired the shots at Doyle's van. Love had lied to avoid prosecution. The police had fabricated the evidence of his alleged statement and had planted the map in his briefcase. On the day of his arrest, Detective Chief Inspector Dunwoodie had told him that he was to be fitted up for the murders.
 Campbell's alleged statement constituted corroboration of Love on both charges (9) and (15). Counsel were agreed that but for the alleged statement, there was no case against Campbell on either charge. That takes us to the police evidence. It is the central issue.
The police evidence against Campbell
 On 12 May 1984 Campbell was arrested at his home by Detective Inspector William McCafferty, now deceased, Detective Sergeant Andrew Hyslop, Detective Constable Alexander Geddes and Detective Constable Ian Cargill and taken to Easterhouse Police Station. He was arrested on a petition warrant that related only to the shooting incident.
 DI McCafferty said that he cautioned Campbell and that Campbell then said "I only wanted the van windaes shot up. The fire at Fat Boy's was only meant to be a frightener which went too far." He noted this statement in his notebook at the time. When they arrived at the police station, he did not confer with DC Geddes regarding the statement. He did not report the statement to the officer in charge of the enquiry, Detective Superintendent Norman Walker, until the following day.
 DS Hyslop said that he had noted the reply as soon as it was made. DC Geddes said that Campbell had made the reply. He gave evidence about it without reference to his notebook. Both officers spoke to the reply in identical terms, except that they referred to "windows" rather than "windaes"; but in our view nothing should be made of that.
 DC Cargill spoke to the same statement in identical terms, except that he spoke of the windows being "shot out." He said that the officers had not compared their notes on their return to the police station.
 The officers' notebooks have all been destroyed; but the Commission has recovered the officers' police statements. We may reasonably assume that these statements correctly transcribed the entries in the notebooks.
 DI McCafferty's statement notes Campbell's alleged words as follows: "I only wanted the van windaes shot up, the fire at fat boys was only ment (sic) to be frightener which went too far." DS Hyslop's statement notes Campbell's words in almost identical terms, including the omission of the apostrophe in "boy's" and including the misspelling of the word "meant" but with the insertion of "the" before "fat boys". DC Geddes' statement says that Campbell said "I only wanted the van windaes shot up, the fire at 'fat boys' was only meant to be a frightener which went too far." DC Cargill's statement says that Campbell said "I only wanted the van windaes shot up, the fire at the 'fat boys' was only meant to be a frightener which went too far."
 DS Hyslop and DC Geddes were not asked whether the arresting officers had compared their notes. The Commission interviewed Mr Hyslop and Mr Geddes. Mr Hyslop, now retired, said that to the best of his knowledge the police officers had not compared notebooks and that that would have been inappropriate. Mr Geddes, now a Detective Inspector, could not recall if the arresting officers had compared notes, but he said that normal practice was not to do so and that it would have been inappropriate to copy a statement from another officer's notebook. The Crown has re-precognosced the three surviving officers. They confirm that each officer recorded the statement independently and that they did not compare their notebooks.
 Campbell denied that he made the statement. His denial was corroborated by his partner, Mrs Elizabeth Donaldson, who said that she was present when he was arrested.
The case against Steele - charge (15)
 There were three sources of evidence against Steele. The first was the evidence of Love about the conversation in the Netherfield, to which we have referred. The Crown proposition was that in that conversation Steele was assenting to a plan to start a fire at Andrew Doyle's house.
 The second source was the evidence of William Ferguson who said that a few months after the fire he overheard a conversation in the Barge public house between Steele and Moore in which Moore said to Steele that it was an easy £300 he got for torching a cellar and that Campbell would see him alright. Steele had said something that he had been unable to hear. Ferguson's evidence was confused and unsatisfactory. He was unwell. He suffered from a heart condition and was brought to court from hospital. He kept changing the pronouns in his account of the conversation. He referred to Moore as having said that Campbell would see "them" alright, see "you" alright and see "us" alright. He spoke only to what Moore had said. In re-examination, he said that his impression was that both Steele and Moore were to get the £300.
 The third source was the police evidence of an incriminating statement allegedly made by Steele while he was being taken in a police car to Easterhouse Police Station. In this case, too, the police evidence is the central issue.
The police evidence against Steele
 Steele, like Campbell, was an experienced criminal. On 1 June 1984 Detective Inspector McKillop, now deceased, and Detective Sergeant Stewart Clark arrested him at his home. Detective Constable Geddes and Detective Constable Andrew Grainger remained outside. Steele was cautioned and replied, according to DI McKillop and DS Clark, "I thought you would have been here before this." DS Clark said that he noted this reply in his notebook at once.
 Steele was taken by car to Easterhouse Police Station to be interviewed. DC Geddes drove the car. DI McKillop sat in the front passenger seat. Steele sat in the rear between DS Clark and DC Grainger. According to DS Clark, DC Grainger and DC Geddes, just as the car was drawing away, Steele said "I'm no' the one that lit the match." According to DI McKillop, he said "I am no' the one that lit the match." When Steele said this, DC Geddes stopped the car and noted Steele's remark in his notebook. DI McKillop did not make a note of it at the time. Neither of the officers in the rear of the car made a note because, they said, they were squashed. Why DI McKillop did not note Steele's remark at the time has never been properly explained. DI McKillop, DS Clark and DC Grainger said that they noted the remark in their notebooks when they reached the police station. After Steele made the remark in the car, DI McKillop either cautioned Steele or reminded him that he was still under caution. According to DI McKillop, he replied "I'm saying fuck all else." DI McKillop said that he wrote these statements in his notebook from memory on arrival at the police station.
 Steele was then interviewed at the police station by DI McKillop and DS Clark. DS Clark took notes during the course of the interview. DI McKillop did not. He said that he made notes of the interview in his notebook later. DS Clark said that in the course of the interview DI McKillop told Steele that he knew that Steele had made a statement to the police that Campbell had asked him to set the Doyles' house on fire. DC Clark said that Steele replied "I cannae admit that. If I said that in court I would get fucking killed." DI McKillop was not asked about this.
 All four officers insisted that they did not discuss the first of the remarks made by Steele in the car. They denied that they had colluded in their evidence about it.
 Steele said that on the night of the fire he was in bed with 'flu at his mother's house. He said that the evidence of the police and of Ferguson was false.
The case against Gray - charge (9)
 On this charge Love accused Gray of having fired the shots. The Crown relied for corroboration on the finding of Gray's fingerprints on a plastic bag in the Volvo and on a statement allegedly made by Gray to Detective Inspector Forsyth and Detective Inspector Wylie at Baird Street Police Office on 12 May 1984, namely "I don't give a fuck what the charges are. The fire's no' doon tae me. The shooting, aye, but I'm no' saying another word until I see my lawyer."
 Gray's defence was that he was at home at the time of the shooting. He had been in the habit of giving plastic bags to a friend, David Coyle, for use when shopping. He had heard that Coyle had been seen in the Volvo. That might be how one of the bags found its way into the Volvo; or it could have been planted by the police. He had not made the alleged statement. Love was lying. Love had fired the shots.
The evidence of Joseph Granger and Alexander Reynolds - charge (15)
 Joseph Granger, with whom those accused of charge (15) were alleged to have acted, was led as a Crown witness. He did not speak to the precognition that he had given to the Crown. He gave no evidence of any value to the Crown case. The advocate depute put to him that on 23 and 25 May he had given statements to the police that were inconsistent with his evidence on oath. Granger admitted that he had signed a police statement dated 25 May, but he said that the police had made it up and that he had been forced to sign it by means of bullying and violence. He denied that the contents were true. The trial judge warned Granger of the consequences of not telling the truth on oath. The advocate depute took him through the substance of the statement dated 25 May, as he was entitled to do (cf. Criminal Procedure (Scotland) Act 1975, s. 147). He put it to Granger that in his statement he had told the police that before the fire he and others had gone with Campbell in a car driven by Gray to the side of the motorway opposite Andrew Doyle's house. Campbell had then pointed out the house. A week later, he was picked up by Campbell, Gray, Steele and Moore in the same car. Campbell said that they were going to Fat Boy's house to burn his door. Gray parked the car near a wall in the street next to where Andrew Doyle lived. They all got out of the car. Campbell and Gray stood at the wall. Steele, Granger and Moore walked to the Doyles' close. Moore was carrying a petrol container. Moore and Steele went up the close while Granger stayed at the close mouth and kept a lookout. He could see Campbell and Gray standing at the wall. A few minutes later, Moore and Steele came downstairs and all of them got back into the car.
 If Granger had given evidence along these lines, his evidence would have been highly incriminating against all four accused. The Crown called two police officers who said that Granger had made the statement dated 25 May and that they had not used violence or threats to make him give it. They did not speak to the contents of the statement.
 The questions that the advocate depute put to Granger about the terms of the 25 May statement were not of course evidence against any of the accused; but the mere asking of the questions was highly prejudicial to them.
 Another Crown witness, Alexander Reynolds, failed to speak to a police statement dated 26 May that he was alleged to have made. He too said that the police had forced him to sign it by means of violence and threats. If Reynolds had given evidence along the lines of that statement, he would have incriminated Campbell, Gray and Moore.
III The trial judge's charge
 The trial judge dealt with the credibility and reliability of the police officers in the context of Campbell's line of defence. He said the following.
"The credibility and the reliability of the witnesses whom you heard are matters for you. You have to decide who you believe, where there is a conflict on the evidence. Now, in this case an attack has been made by counsel on the credibility of Love, and Ness, and many of the detective officers involved in investigating these crimes. I have already given you a direction upon the evidence of Love; you have to consider whether you are to accept Love's evidence which incriminates some of the accused. So far as the detectives are concerned, Mr Macaulay delivered a vehement and sustained attack upon the integrity of a number of detective officers involved in this case, some of considerable experience and in superior positions, some with less experience and in lower positions. He used such words as 'rotten', 'Strathclyde Police rotten', and you will remember he used such expressions as 'There are good policemen, bad policemen' and then reference was made to 'the ugly'; and they have been submitted to be liars and bullies. Well, of course, you appreciate that this attack is made on behalf of Thomas Campbell: Mr Macaulay is acting upon the instructions of his client, either express or implied, because counsel do not hold any views on these matters; Mr Macaulay said this to you himself: so what this attack amounted to was Mr Macaulay on behalf of Thomas Campbell alleging that the police were liars and bullies. Now, the force or the validity of any attack of this kind must be judged on the evidence in the case, not on evidence in other cases, what other policemen may have done in other cases, or on anything else. You have to ask yourselves however 'What is the evidence on which this attack is based that the police are liars and bullies? What is it based on?' That is for you to say, but so far as I can gather from the evidence which you have heard it is based upon the evidence of the accused, Mrs Campbell, and on the young man Hamilton, who says he was bullied by the police into making a statement which he says is not true, Joseph Granger, who said also that he was bullied, or definitely he said he had his hair pulled in order to be forced to make this statement or to sign a space on a plan, a dot on a plan, and the witness Reynolds, who said a 13-page statement was put before him, and he was told to sign it, and it was made up by the police: there may be others. Now, against that body of testimony you have the evidence of the detectives themselves to whom these allegations were put, and who denied them. You have to choose. It is only if you accept the evidence of the accused and the others to whom I have referred that you could agree with Mr Macaulay's submission. If you do, you must consider what follows. What follows is that you are saying that not one or two or four but a large number of detectives have deliberately come here to perjure themselves, to build up a false case against an accused person, and they have carried this through right to the end; a conspiracy of the most sinister and serious kind. They have formed this conspiracy to saddle the accused wrongly with the crimes of murder and attempted murder, and murder of a horrendous nature. If so, it involves their making up and persisting in a concocted story, concocted statements attributed wrongly, falsely, to the accused. Now, what do you prefer, ladies and gentlemen? It is up to you: you saw the witnesses in the witness box; you heard how the questions were answered. You have to make up your mind what to believe (at pp. 26-29)."
The trial judge returned to this theme when he came to the question of the credibility and reliability of Love. He said:
"That is the background against which you have to consider the evidence of Love. If he is untruthful, if this is a concocted story, a false story, then it follows, doesn't it, that he has played a gigantic confidence trick on the police, senior police officers, the procurator fiscal and indeed the Crown authorities, because they accepted his story, allowed him out on bail and have proffered him as a witness before you. They according to the submission of the defence have swallowed hook, line and sinker Love's story, which they say is false. Well, you have go to consider this submission, ladies and gentlemen, because it is the foundation of this charge; because if he is lying he is playing a colossal deceit on you: but if he is telling the truth, whatever may have been his motive for not remaining silent, then his evidence is powerful and damning against the accused. It is for you to decide. You saw Mr Love, you heard him giving his evidence, and to some extent you are in a position to assess his character and his intelligence and his intellect. Ask yourselves, was he capable of hoodwinking the police and the Crown and you with an utterly false and concocted story? Could he have conceived of such a story in the first place with no foundation in fact to work on? Could he have persisted in it, successfully, so far as the Crown is concerned, and now before you, without being tripped up, without being shown up as a liar or as a complete inventor of the accounts which he gave? He was subjected to the most skilled and exhaustive cross-examination, designed to test his credibility: did he come through it unscathed, or did you think he was as he is branded by the accused an out and out liar? These are questions for you, not for me" (pp. 61-62).
The trial judge gave the following general direction on the subject of statements made by one accused incriminating another.
"Statements made by an accused before trial which refer to other accused are not admissible evidence against that other accused if the statements were made in the absence of that other accused, because not having heard the accusation or whatever it was, the person who is referred to in the statement is in no position immediately to deny or comment on the statement made. Now, in this case there are a number of items of evidence consisting of statements made by an accused concerning another, and as I say, they are only to be used when considering the evidence against the accused who made the statement, not against the accused referred to in the statement" (p. 24).
The trial judge dealt with the evidence of Granger and Reynolds. By that stage Gray had been acquitted on a submission. The trial judge said this.
"Now, ladies and gentlemen, they are charged with Joseph Granger, and Joseph Granger was adduced as a witness. His evidence does not assist you at all, ladies and gentlemen, whatever you may think of his credibility: he did not give any evidence which implicated Steele or Thomas Campbell. The same applies to Reynolds. Whatever you may think of Reynolds' evidence, none was given by him upon which you could found in support of the Crown allegation against Thomas Campbell and Joseph Steele" (pp. 98-99).
On the question of Ferguson's evidence, the trial judge said this.
"Much will depend again what you make of the evidence of Mr Ferguson: he was the gentleman who was not very well, who gave evidence some time ago, late on in the afternoon. He referred to overhearing a conversation after the fire in a public house between Joseph Steele and Gary Moore. Now, you heard and you saw Mr Ferguson, and you know that literally construed his evidence did not consist, because it altered from time to time according to the particular pronoun which he used. It is however your decision as to what you take from Mr Ferguson's evidence. You have heard the submissions on it, and you have to take these into account. You have to decide what Mr Ferguson was trying to tell you that he overheard being said. It is said that he has been listening to gossip. Well, you must consider that submission as well. Is Mr Ferguson likely to have made up that kind of story? You have to think about that too. If you think that Mr Ferguson was telling the truth, you still have to decide what he was telling you, and if what he was telling you necessarily involved Joseph Steele as one of the persons responsible for the fire" (pp. 101-102).
IV Subsequent proceedings
(a) The trial of Granger and Reynolds
 In consequence of their evidence at this trial, Granger and Reynolds were tried at Glasgow High Court in March 1985 on charges of perjury. Granger was convicted of perjury in relation to the statement of 23 May, but acquitted in relation to the statement of 25 May. The sequel to that conviction need not concern us (Granger, Petr, 2001 JC 183). Reynolds was convicted in relation to his statement of 26 May.
(b) The appeals by Campbell, Steele and Gray against conviction (1985)
 All three appealed against conviction. On 4 June 1985 the appeals were refused, for reasons set out in each case in an Opinion of the Court delivered by Lord Justice Clerk Wheatley. In his grounds of appeal and in a written submission to the court, Campbell argued that there had been a misdirection in the passage of the charge at page 61 that we have quoted. Unfortunately, in rejecting that argument, the court mistakenly referred to the wrong passage in the charge; but that is no longer a live issue since we have considered the whole question afresh.
(c) The Secretary of State's reference (1996) - Campbell and Steele
 In 1996, the Secretary of State for Scotland referred the cases of Campbell and Steele to this court under the then section 124(2) of the Criminal Procedure (Scotland) Act 1995 (Campbell v HM Adv, 1998 JC 130).
 The reference was made on the basis of an affidavit sworn by Love to the effect that he had lied at the trial at the instigation of D Supt Walker and DI McKillop under various inducements and threats and that he had carried out the shooting. Love's affidavit was supported by an affidavit of his sister Agnes, known as Mrs Carlton, who had not been called at the trial.
 On 29 April and 4 July 1984 Mrs Carlton gave statements to the police that she saw the shooting but did not know who fired the shots. On 7 July she gave a statement to the police in which she said that on the day of the shooting Love had appeared at her house, from which she could see the locus of the shooting; that Love had a sawn-off shotgun with him; that he left her house with the stated intention of firing it; that after he left, she looked out of the window and saw a masked man, who was wearing the clothes that Love had been wearing, shooting at the van; and that soon after, Love had come back to her house with the shotgun and concealed it. Later, she gave a precognition on similar lines to Steele's solicitors.
 The proposed evidence of Mrs Carlton was crucial to the reference (Campbell v HM Adv, supra, at pp. 145ff). Although it related to charge (9), it had implications for the Crown case on charge (15).
 This court accepted that Love's affidavit contained new evidence that was capable of meeting the tests set out in section 106 of the Criminal Procedure (Scotland) Act 1995 (the 1995 Act); but, by a majority, the court held that the proposed evidence of Mrs Carlton was not capable of supporting the requirement that there should be a reasonable explanation why Love had not given evidence at the trial on the lines of his affidavit (ibid, at pp. 158-159; 180-181).
V The present references
The Commission's approach
 These references arise, in the cases of Campbell and Steele, from an entirely new matter, namely the reports obtained by the Commission from Professor Brian Clifford and Dr Peter French on the police evidence of the statements allegedly made by Campbell and Steele. Professor Clifford has been Professor of Cognitive Psychology at the University of East London since 1996. He is an expert of international renown in the field of psycho-linguistics as applied to memory. He has written extensively on the subject. He advises on matters of forensic psychology. Dr French is an expert in forensic linguistics. In general, these experts conclude that it is unlikely in either case that all of the officers concerned could have recalled the relevant statement, and noted it in their notebooks, in virtually identical words. This
conclusion is based on their own experiments, on their professional experience and on the scientific literature.
 The Commission's reference in Campbell's case is confined to charge (15). The Commission considers that the terms of the reports by Professor Clifford and Dr French are sufficiently strong to cast doubt on the evidence of the police officers who arrested Campbell. Its decision is "based upon its view that had the jury rejected the evidence of Campbell's arresting officers, it appears that there would have been insufficient evidence to convict him of murder" (Statement of Reasons, p. 95). It is surprising that the reference should be confined to charge (15), because if doubt is cast on the police evidence of Campbell's alleged statement, that doubt affects the conviction on charge (9) also.
 The Commission has concerns about all three sources of evidence against Steele. It considers it arguable that Ferguson's evidence was not evidence against Steele at all. It is also of the view that the manner in which Love's evidence against Steele was presented to the jury may have led them erroneously to believe that they could convict even if they did not accept this evidence as corroboration of Steele' s alleged admission to the police (Statement of Reasons, p. 106).
 In the case of Gray, the reference is concerned only with the possibility that the professional representation of Gray may have been defective by reason of the failure of Gray's defence team to lead the evidence of Mrs Carlton incriminating Love (Statement of Reasons, pp. 63-66).
 The Commission is to be commended for the thoroughness of its investigation and for the excellence of its analysis of the issues. In the light of these Reports and
the submissions of counsel, we accept that it was in the interests of justice that each of these references should have been made.
The additional grounds of appeal
 Proceedings of this kind fall to be treated as an appeal (Higgins v HM Adv, 1956 JC 69, at p. 74; Kilpatrick v HM Adv, 1992 JC 120, at p. 123). The court is not confined to the scope of the reference itself (Beattie v HM Adv, 1995 SCCR 93). The court can entertain any ground of appeal that may be tabled, even one that has been decided already (Campbell v HM Adv, supra; Boncza-Tomaszewski v HM Adv, 2000 SCCR 657, at pp. 664-665). Counsel for Campbell has widened his submissions to include an appeal against conviction on both charges.
 In addition to relying on the new evidence, counsel for Campbell and Steele submitted that the trial judge misdirected the jury in his presentation of the evidence; that he misdirected them on the evidence of Granger; and that the court should hear new evidence from Love and Mrs Carlton, being the evidence offered at the 1996 reference. Counsel for Steele also submitted that there was insufficient evidence to entitle the jury to convict him, and that the trial judge misdirected the jury on the evidence of Ferguson. The solicitor advocate for Gray adopted the points made by the Commission in his case, which are in essence an Anderson ground, and submitted that the trial judge misdirected the jury on the evidence of Love and of Granger. He did not rely on the new evidence, which, as the Commission observes, is of limited significance to Gray's case; nor did he dispute that there was a sufficiency of evidence to warrant Gray's conviction.
VI The evidence of Professor Clifford and Dr French
 On 20 February 2003, after a hearing on the point, we rejected a submission for the Crown that the proposed evidence of Professor Clifford and Dr French was inadmissible on the ground that it was opinion evidence on the credibility and reliability of the police witnesses. We held that the proposed evidence was evidence of the ability of witnesses, in the circumstances spoken to at the trial, to recall a statement accurately and of the likelihood that several witnesses to the same statement could recall it in almost identical terms. That was relevant and admissible evidence of fact that could have a significant bearing on the credibility of the police witnesses. We gave leave to the Crown to lead evidence on the subject.
 We have now heard the evidence of Professor Clifford. He was adduced as a witness on behalf of Campbell, but in reality on behalf of Steele also. Professor Clifford spoke to the findings of his experimental studies on which he reported to the Commission. The terms of Dr French's reports were agreed. Dr French was not called on behalf of Campbell or Steele, but counsel for Campbell and Steele relied on his reports to the extent that they generally supported Professor Clifford's conclusions. The advocate depute did not in the event call Dr French, but he referred to his reports and submitted that they did not cast doubt on the convictions. The advocate depute put to Professor Clifford certain criticisms of his methodology and conclusions. In his submissions, he criticised Dr French's Report on similar points. He led no evidence in rebuttal of either of them.
 We have set out the almost identical versions of the 24-word statement imputed to Campbell by the four officers who arrested him. The Commission asked Professor Clifford the following questions:
"(1) How likely is it that all four officers were able to note the remark in such similar terms?
(2) What is the likelihood that all four officers were able to note the remark in such similar terms in the absence of any comparison or collaboration whatsoever between them?"
In the case of Steele, the Commission asked Professor Clifford to consider two of the statements allegedly made to the police by Steele. The first was the statement allegedly made in the house, "I thought you would have been here before this". The second was the statement allegedly made in the police car, "I'm no' the one that lit the match" or, as DI McKillop noted it, "I am no' the one that lit the match". The Commission asked Professor Clifford the following questions:
"(1) How likely is it that all four officers noted Steele's comment in the car in such similar terms - or would the memories of those officers who did not note the statement until their arrival at the police station be affected by the time lapse stipulated?
 We have disregarded the questions put to Professor Clifford in both cases relating to the reliability of the police officers' evidence in relation to these statements. In our view, that is not a matter for expert opinion.
 Professor Clifford said that most cognitive psychologists agreed that there was a need to postulate some capacity limitation on immediate short-term memory and that the upper limit was set at about seven to nine discrete items. Beyond that, the meaning of statements was retained but not the actual words, unless the individual was an actor. The short sentences imputed to Steele might be capable of immediate recall verbatim, but the longer statement imputed to Campbell would not. It would be subject to the phenomenon of semantic processing of sentences that had been demonstrated by psycho-linguistic research.
 Professor Clifford carried out four experimental studies. The first was designed to test the capacity of individual subjects to retain and retrieve immediately heard utterances of different lengths. The second study was carried out for the same purpose, but with a sample of participants who were Scots. The third study tested the effect of delay upon memory of a heard utterance. The fourth study tested the participants' ability to recall an eight-word sentence after a lapse of five, ten or fifteen minutes. In each study, the participants had to listen to taped statements.
 Since the Crown relied at the trial upon two statements in particular, Campbell's alleged statement on arrest and Steele's first alleged statement in the police car, we shall concentrate on Professor Clifford's studies in relation to them.
The first and second studies
 In the first study there were 57 participants, both male and female, with a mean age of 31. They were drawn from a wide range of occupations both private and public. All of them were English and English was their first language. They were divided into two groups. One was given relevant information. The other was not. The purpose of having an informed group was to simulate the knowledge that the police officers had when they arrested Campbell and Steele. The participants in the informed group were told that arson and discharge of a shotgun were involved, and that the experiment related to a real life criminal case at the heart of which was the possibility of a miscarriage of justice centred on the reliability of memory for sentences. The participants in the uninformed group had no idea what the topic would be. All the participants were presented with the statements one at a time and in each case were requested to write the statement down immediately word for word. They were told that the utterances were to be spoken by a Scotsman, whose accent might present a difficulty for them, and that they would have to pay close attention to what he said. The source voice spoke the key sentences after extensive practice so that he had them word perfect. He had a Glasgow accent. The mood in which he spoke them was intended to mimic surprise, indignation and resignation, which was thought to be the way in which they had been uttered.
 None of the participants could recall all 24 words of Campbell's alleged statement. The majority of the informed group could recall only 40% of the words. The majority of the uninformed group could recall only 30%. The majority of the participants could recall only about 70% to 80% of the eight words of Steele's statement. It made no significant difference whether they were in the informed or the uninformed group.
 These results suggested that verbatim recall of the complete Campbell statement of 24 words was impossible. While some individuals could recall verbatim the short sentences that they had heard up to three seconds before, the majority could not. Verbatim recall of the first of Steele's statements in the car was therefore possible; but there remained the question whether the short sentence could be recalled by several hearers in identical terms.
 Having considered the results of the first study in the light of psycho-linguistic knowledge, and having compared them with the police evidence, Professor Clifford concluded that it was "very improbable" that all four police officers who noted Campbell's statement would have identical, non-conferenced verbatim recall of it.
 Since it was to be assumed that all of the police officers who noted the statements were Scottish, Professor Clifford decided to repeat the study with a sample of Scottish participants. All of these participants were informed in broad terms of the purpose of the study. According to Professor Clifford, this second study stacked the cards in favour of results more comparable with the police officers' recall ability.
 In this study there were 74 participants. They included firefighters, nurses, factory workers, a karate group and persons in other occupations chosen at random. Among them there were 14 police officers. All the participants were from Perthshire. The results were closely similar to those found in the first study. There were no significant differences in the data. The low verbatim recalls found in both studies were considered not to be explicable by the linguistic difference between the two sets of participants. Professor Clifford concluded that the low verbatim recalls resulted from the difficulty of the task.
 In these two studies, the range of recall for long sentences was wide, but no participant even came close to the apparent recall ability of the arresting police officers. Professor Clifford thought that the difficulties were caused by the structural constraint of the sheer length of the sentence rather than by individual differences in the size of memory span. Both of the long sentences used, one of which was Campbell's 24-word statement, exceeded the most capacious estimates of an individual's short-term memory ability.
 The fact that none of the participants equalled or even came close to the recall ability seemingly achieved by the arresting police officers in these cases raised serious doubts in Professor Clifford's mind about the officers' evidence.
 In the second study, the police participants performed numerically better than some non-police groups, but the numerical differences were not statistically reliable. There was therefore evidence within the data, according to Professor Clifford, that policemen do not perform any better in terms of accuracy of recollection than other members of society. That was supported by other published data. He concluded:
"Given the extant publications and the current results, the memory performance that apparently was exhibited by the police officers in the case under review must be seen as truly remarkable. So remarkable in fact as to be doubtful."
 As an extension of these two studies, Professor Clifford also carried out a semantic analysis of the statement imputed to Campbell. He said that when sentences exceed short term memory, it is the semantic units constituting the statement that are encoded, stored and retrieved. The Campbell statement comprised four idea units; namely, shooting at van windows; fire at Fat Boy's; a frightener; and going too far. It was found that not all semantic units were equally well recalled. The most poorly recalled were those with the greatest criminal significance, namely the shooting and the fire. On a semantic analysis, whether the studies were taken singly or in combination, the same pattern appeared. The semantic study was based on a more lenient criterion of recall, yet even in this study no participant in the combined sample of 131 could recall all four semantic units.
 Professor Clifford concluded that in the face of these data the probability that any random four or random two officers could recall a sentence of 24 words in identical terms was "infinitesimal." In answer to the questions put to him by the Commission, he concluded that the findings of the first and second studies, taken together or separately, strongly suggested the unlikelihood that in Campbell's case all four officers could note the utterance in such similar terms, and indicated clearly that to obtain such similar recalls, each officer could not have been acting alone and independently. The degree of similarity in the statements and in the evidence of the four officers could have resulted only from comparison or collaboration.
The third study
 The third study related to the alleged statement of Steele, "I'm no' the one that lit the match." It was designed to assess the probability of its being accurately recalled five, ten or fifteen minutes after it was said. There were 20 participants from England and 20 from Scotland. There were 21 males and 19 females, with a mean age of 28 years. The same set of tapes was used in this study as in the previous two. All of the participants were informed about the background to the study and the key elements of arson and the discharge of a firearm, and about the "current plight" of Campbell and Steele. Professor Clifford described the results of this study as "clear and somewhat stark." Recall of a fairly short sentence was found to be very poor after a delay of only five minutes. Only two participants succeeded in recalling it verbatim. 12 recalled it correctly in the semantic sense, but not verbatim. 26 failed to recall anything.
 The word for word recalls of the Steele statement by three of the arresting officers after a delay of at least ten minutes, and possibly as long as fifteen, gave Professor Clifford cause for concern. On the basis of his third study, he could say that identical verbatim delayed recall was "not readily available to humans".
 In the light of the findings of the third study, Professor Clifford concluded that the verbatim similarity in the police officers' noting of Steele's first statement in the police car was "an extremely unlikely occurrence." The fact that all three of the officers who noted it at the police station managed to recall it verbatim was "quite remarkable." He said that it was hard to escape the conclusion that these similarities were "due to other than independent retrieval of a stored memory of a heard utterance."
The fourth study
 Professor Clifford undertook a fourth study, to test recall after specified intervals of time. This study, like the third, concerned only Steele's first statement in the car. Its purpose was to examine whether verbatim memory for an eight-word utterance was possible after delays of five, ten or fifteen minutes under the conditions spoken to by the police officers. For this study there were 50 female and 26 male participants, with an average age of 291/2 years, who were undergraduates and professional people.
 Two tapes were prepared. One had the eight word statement, followed by a long caution, followed by the words "I'm saying fuck all else." The other had the same statements, but with a short caution in between. The participants were tested in groups of four to simulate conditions in the police car. They were told that they would be presented with a number of utterances on tape as part of an important experiment. Their task was to remember only the first utterance that they heard and to be prepared to write it down later word for word. They were told that when the tape was stopped they should sit quietly and await further instructions. The first instruction was repeated. The tape was then played and the participants sat in silence for either five, ten or fifteen minutes. At the end of whichever period they were given, they were instructed to write down, if they could, word for word, the first utterance that they had heard.
 Only one participant recalled the critical sentence verbatim. That participant was in the ten-minute delay group who heard the short caution. The results showed that participants could recall verbatim only about half of the words of the critical utterance and that recall was delay-dependant. The participants could retain the semantics of the sentence fairly well, irrespective of the time delay and the length of the intervening utterances; but as delay increased, so did error in recall.
 Professor Clifford considered that it was extremely unlikely that three police officers could independently recall the alleged utterance in identical or near identical terms. That opinion, he said, was now based on the experience of a total of 244 participants who had heard the critical sentence and had recalled it either immediately, in the case of 151 of them, or after some delay, in the case of the others. The data showed that the individual does not normally process a sentence by storing it verbatim even when he is specifically instructed to do so, as in the first three studies, or when, because of its importance, he intends to remember it for later noting, as in the case of the arresting officers.
Whether this was new evidence
 Professor Clifford also dealt with the question whether evidence of this kind could have been obtained at the time of the trial. It is our impression that if it had been available then, it is unlikely that the court would have admitted it; but since the advocate depute did not dispute that this was new evidence, we need not consider the question further.
The Crown criticisms of the new evidence
 The advocate depute compared Professor Clifford's studies unfavourably with some of those of Dr. French. His principal criticism was that the information provided to the participants fell far short of the information that the police had when they arrested Campbell and Steele. The experiments could not replicate the significance which the statements had for the police officers. This was one of the most important cases that they would have dealt with. They knew that they would have to remember the statements. It was wrong to assume that the arresting officers started to note Campbell's statement only after he completed it. There was no warrant for the assumption that it was made as one statement of 24 words. It could be regarded as two sentences. Moreover, there was no basis on which to make any assumption about the speed at which it was spoken. In the second experiment, the participants came from Perthshire. They would not have a Glaswegian accent. The evidence of Professor Clifford and Dr. French was not so significant that the convictions of Campbell and Steele should be regarded as miscarriages of justice. Their evidence was not sufficiently relevant and cogent in relation to the evidence at the trial to satisfy the test laid down in Cameron v HM Adv (1991 JC 251) and Kidd v HM Adv (2000 SCCR 513).
(1) The evidence of Professor Clifford and Dr French (the Campbell and Steele references)
 This evidence raises the principal issue on which these references have been made. Before we even consider it, we are surprised by certain aspects of Campbell's alleged statement, and the circumstances in which it was made. It seems remarkable that an experienced criminal like Campbell should have said anything at all when cautioned, let alone have made so incriminating a statement, and that in answer to a caution relating to the shooting, he should have volunteered an incriminating admission about the murders. If Campbell made such a statement, it is remarkable that he was not asked to authenticate it, by signing or initialling the note of it, in one or more of the police notebooks. Perhaps the most remarkable feature of all is that this crucial statement, which was a breakthrough in the police enquiry, was not reported by DI McCafferty to D Supt Walker until the following day.
 In Steele's case, too, the police evidence is surprising. It is remarkable that Steele too should have spontaneously volunteered such an incriminating statement, that he should have done so in a police car on his way to a police interview, and that only the driver of the car should have noted it, while the senior officer, whose hands were free, did not.
 We are sceptical about this evidence even without the benefit of Professor Clifford's conclusions; but that is merely our own reaction. We acknowledge that all of these points were put to the jury, who nevertheless convicted Campbell and Steele unanimously. We need not pursue the point further, because we consider that we can decide the Campbell and Steele references on the basis of the new evidence and of certain directions of the trial judge.
The reports of Dr French
 Dr French did not give evidence. His analysis was therefore not tested. For that reason alone, we consider that his study carries much less weight than that of Professor Clifford. Moreover, Professor Clifford has expressed reservations about Dr French's methodology and about the small size and unreliability of the samples that he used. Nevertheless we think that Dr French's work is of value to the limited extent that his findings and conclusions point in the same direction as those of Professor Clifford.
The evidence of Professor Clifford
 The Crown has had since November 2001 to consider the substance of the new evidence. It has had a year since we held that the new evidence was admissible, to canvass expert advice on it. It has not produced any expert evidence to the contrary.
 We have to consider whether the new evidence is capable of being regarded by a reasonable jury as both credible and reliable (Kidd v HM Adv, supra, at para ). In the making of that judgment, the quality of the new evidence and its cogency are of critical importance. The evidence is based on established theory and on experimental proof. It is uncontradicted, both as to theory and method.
 Professor Clifford's reasoning appears to be logical and his conclusions appear to be cogent. Nothing put to him on behalf of the Crown persuades us that we should not accept his key conclusions. We accept that there is substance in the advocate depute's comments on the limitations of his attempt to replicate the conditions in which the alleged statements were heard. It was impossible fully to replicate those conditions, but in our view Professor Clifford achieved a reasonable proxy for them. Moreover, we are impressed by the robustness of his results.
 On that view, we consider that the evidence is highly relevant to the crucial issue in the trials of Campbell and Steele.
 We then have to consider what effect the evidence would have had on the course of the trial. We have to decide that question in the knowledge that without the evidence of the statements, the Crown had no case against either Campbell or Steele.
 In our opinion, if the new evidence had been given at the trial, the trial judge could not have directed the jury as he did. The contention that the statements were fabricated was crucial to the defence case; but the defence had no independent evidence to support it. The trial judge emphasised that the defences for Campbell and Steele involved an attack on the integrity of the police. He asked the jury to consider what evidence there was to support it. He emphasised that the conflict lay between the police officers and the defence witnesses, including the accused; and that if the jury were to accept the defence case, it would follow that a large number of detectives had engaged in a conspiracy. In our view, it is hardly surprising that, on the evidence before them, the jury resolved the conflict in favour of the Crown.
 The alleged statements were recorded in the arresting officers' notebooks, in their police statements, and no doubt in the Crown precognitions. For those officers, therefore, there could be no retreat from the statements. If the evidence that we have now examined had been available to the defence, there would have been independent evidence from a reputable source that undermined the crucial evidence on which the convictions of Campbell and Steele depended.
 In our view, any jury hearing Professor Clifford's evidence would have assessed the police evidence in an entirely different light. Moreover, it is obvious that while that evidence would have been more convincing in relation to Campbell's alleged 24-word statement, any doubt that it cast on that statement would have carried over to the alleged statement of Steele.
 In our view, the new evidence is of such significance that the verdicts of the jury, having been returned in ignorance of it, must be regarded as miscarriages of justice (cf Kidd v HM Adv, supra).
 In his charge the trial judge made no reference to Love's evidence in his analysis of the case against Steele; but in his report to the appeal court in 1985, he relied on Love's evidence as a material part of the case against Steele.
 In giving the Opinion of the Court in that appeal, Lord Justice Clerk Wheatley dealt with this point by saying that Love had given evidence "about a conversation among a party of people, of whom [Steele] was one, which establishes that [Steele] was a member of a group discussing plans for the commission of the offence and was accepting his involvement in what was being discussed (Opinion of the Court, p 2)." Later, the Lord Justice Clerk said that the evidence of Steele's listening to the conversation between Campbell and Gray and saying "aye, aye" was not just evidence of a confession, but was "independent evidence of the appellant's participation in a plan to commit the offence which subsequently was carried out." That, we think, was an overstatement. That evidence was consistent with an innocent interpretation, so far as Steele was concerned. Nevertheless, it was also consistent with his being involved in the plan. So long as the latter was a possible interpretation, as in our view it was, the jury were entitled to take the evidence into account (HM Adv v Fox, 1998 JC 94). That evidence and the evidence of the police constituted a sufficiency of evidence. We therefore reject this ground of appeal.
(a) The trial judge's presentation of the evidence
 Counsel for Campbell and Steele submitted that the directions of the trial judge inverted the onus of proof; trespassed on the province of the jury; showed a bias in favour of the credibility of the police witnesses; and undermined the defence cases.
 The criticisms now made of the trial judge's charge in these respects were made at the hearing on the Secretary of State's reference. On that occasion Lord Justice Clerk Cullen, as he then was, was the only member of the court expressly to deal with the matter. He rejected the submission (Campbell v HM Adv, supra, at p. 143A-E).
 Counsel for Campbell submitted that since a Commission reference was excepted from the scope of section 124(2) of the 1995 Act, the court's previous decision on the question was not conclusive. He also submitted that since, in the Secretary of State's reference, the court held that it could entertain an issue that had previously been determined by the court, so long as it was not a matter of interpretation of the criminal law (Campbell v HM Adv, supra, at p. 135C-G), we were not precluded from considering the question. In any event, the alleged misdirections in these respects had now to be seen against the background of the new evidence and the new ground of appeal relating to misdirection on the question of onus. The advocate depute did not dispute these submissions. In our view, it is right that we should consider these questions again, since we have extensively examined the trial evidence with the assistance of the new evidence that we have discussed and in the light of more focused grounds of appeal.
 In our opinion, the trial judge did not materially misdirect the jury in his comments that the police, the procurator fiscal and the Crown authorities had "accepted" Love's story (at p. 61D-E). In the circumstances in which Love came to give evidence for the Crown, it was in a sense true to say that the police, the procurator fiscal and the Crown had accepted Love's story. His was the account on which the prosecution was based. The defence approached his evidence on the same footing. If one reads the charge as a whole, it becomes clear that the trial judge repeatedly emphasised that Love's credibility was crucial and that the assessment of it was a matter for the jury. In our view, the trial judge did not pre-empt the jury's assessment of Love's credibility.
 We do not accept that the trial judge trespassed on the province of the jury in his treatment of charge (9) when he described Love's evidence, if true, as "powerful and damning" (at p. 62). If Love's evidence were true, "powerful and damning" would be a fair description of it. Neither the Crown nor the defence can have doubted that. We cannot see how that can be considered to be a misdirection.
 Comments such as the trial judge made about the defence criticisms of the police (at p. 28C) were not uncommon twenty years ago; but we have to decide this case by the judicial standards of today. The essence of a judge's charge is the giving of directions in law. In strict theory, the judge need say nothing about the facts. But in modern practice, in all but the simplest cases, he should refer to the main points of the evidence on which the Crown and the defence rely. In that way, he can give some context to his directions on legal concepts such as corroboration. But whenever the judge refers to the evidence, he should do so with restraint. In this case, the judge not only referred to the defence in forthright terms, but went on to warn the jury that if they accepted the defence cases they would be accepting that there had been a police conspiracy "of the most sinister and serious kind." That sort of comment was appropriate for the advocate depute's speech; but it was perhaps inopportune in the charge. Nevertheless, it did not misrepresent the defence. On the contrary, as we can see from the judge's references to the speech of counsel for Campbell, that was a central point in the defence. In our view, the judge's comments were not a misdirection.
 Counsel for Campbell and Steele submitted that there was a misdirection on onus (at pp. 28C-29B) when the trial judge said that it was only if the jury accepted the evidence of the accused that they could accept the defence submission, and that it was for the jury to decide which account they preferred. We do not agree. Those directions must be seen in their context. At the outset of his charge the trial judge gave the standard directions on onus and on reasonable doubt. Later, in the passage complained of, he dealt with the specific issue of the conflict between the evidence of the police and the evidence of the accused about the statements. In that context, he was quite right to say that it was only if they believed the accused that they could accept the submission of counsel for Campbell that the police were liars and bullies. It is obvious that on that issue it was for the jury to decide which account they preferred. We reject this submission.
(b) The direction on Granger's evidence
 The trial judge directed the jury that Granger's evidence did not assist them at all. He said that the same applied to Reynolds. In his Report to the appeal court, he said that this direction "was intended as a direction to the jury to disregard the whole of the evidence of the witnesses so far as charge (15) was concerned" (at p. 22). In our opinion, that direction was inadequate.
 Campbell and Steele took similar points in the 1985 appeals. Campbell argued then that in relation to both Granger and Reynolds the trial judge misdirected the jury by failing to warn them that if they did not believe a witness' denial of a point of evidence put to him, they were not entitled to infer that the opposite was true. Lord Justice Clerk Wheatley referred to the passage that we have quoted and said
"To suggest, in the face of such clear directions that nothing could be taken out of the evidence of these witnesses to implicate the appellant, that a further direction in the terms desiderated was necessary to make the point clear to the jury, reeks more of an insult to their intelligence than a requirement for their proper guidance. We reject this complaint out of hand" (at p. 7).
Counsel for Steele argued then that the trial judge should have directed the jury that the contents of the statements of Granger and Reynolds were not evidence against him, and that the jury should put the contents out of their minds. Counsel for Steele founded on Paterson v HM Adv (1974 JC 35). In that case, the mother of the accused denied that she had given a statement to the police that incriminated him. The Crown led two police officers who spoke to the terms of the statement. It was held that the trial judge had misdirected the jury in failing to tell them that the evidence of the police officers was not evidence against the accused.
 The appeal court, referring to the same passage in the charge, distinguished Paterson v HM Adv on the ground that in this case the Crown had not led evidence of Granger's statement. We are not convinced that that was a distinction of any great significance since after the advocate depute's questions to Granger about his statement, the jury were as well informed of its contents as they would have been if a police witness had read it out. Lord Justice Clerk Wheatley continued as follows.
"We not consider the case of Paterson to be apposite here, nor do we consider that the direction called for was necessary in the circumstances, having regard to the manner with which the judge dealt with them. For good measure, the judge had already pointed out to the jury that any statement made by one accused inculpating a co-accused made outwith the presence of that co-accused could not be evidence against that co-accused. Counsel for the appellant submitted that the judge had confined his direction to the situation where the statement was made by one accused involving another, and here these witnesses were not co-accused. Even on the basis that the principle in law therein involved is confined to co-accused, Granger was really in the position of a co-accused. The accused libelled in charge 15 are charged with acting along with Granger in its commission, and when he was being interviewed by the police who took his statement he was very much in the position of a co-accused and equally liable to attempt to exonerate himself at some one else's expense. It follows from this, of course, that nothing of evidential or corroborative value can flow from this chapter of the evidence, nor was anything of such a nature sought by the Crown."
In our opinion, these comments missed the point. While the contents of the statement were not evidence against Campbell or Steele, the putting of its contents to Granger created a serious risk that if the jury disbelieved his denial that he had made the statement, as they well might, they would conclude that the account of events put to him by the advocate depute represented the truth.
 In the present case, counsel for Campbell and Steele have confined the submission to the question of Granger's evidence. In our opinion, their submissions are well founded. On a matter of such sensitivity the trial judge was obliged, in our view, to direct the jury that Granger had said nothing to incriminate Campbell or Steele, and that they must not take into account the advocate depute's questions to Granger as to the contents of the statement, or the police evidence as to the making of it, when they were considering the case against Campbell or Steele. The failure of the trial judge to direct them in such terms was, in our opinion, a material misdirection. In the circumstances of this trial, that misdirection constituted a miscarriage of justice in relation to Campbell and Steele. It did not constitute a miscarriage of justice in relation to Gray, because by then he had been acquitted of this charge.
 The trial judge left the jury to make what they could of Ferguson's evidence. Counsel for Steele submitted that it was not clear what Ferguson's evidence was and that, in any event, it related only to what Moore had said. This point was raised by Steele in the 1985 appeal. On that occasion Lord Justice Clerk Wheatley said the following.
"Counsel for [Steele] submitted that while it is for the jury to assess the value of a witness' evidence, where a witness cannot make up his own mind about what he heard it is for the judge to do something to clear it up. That seems to be exactly what happened here. Between them the advocate depute and the judge seem to have extracted from the witness his difficulty in using the right personal pronoun in the right place, but, with that explanation, it was for the jury to make up their minds on what the witness was in effect saying and what value should be attached to it. That being so, there is no reason why the jury should not have been allowed to consider Mr Ferguson's evidence and, in doing so, been entitled, in the light of all the other evidence in the case, to put on it the evidential significance against [Steele] which the Crown desired" (at p. 7).
We do not agree. In our opinion, the trial judge should have directed the jury that Ferguson's evidence was not evidence against Steele. Whatever interpretation could be put on his evidence, the key point was that Ferguson spoke only to what he heard Moore say. His evidence in re-examination that he thought that both were to get the £300 was merely his interpretation of Moore's words. It was not evidence against Steele. The trial judge's approach was therefore a material misdirection. The case against Steele was at best so thin that we cannot exclude the possibility that the jury considered Ferguson's evidence, and not that of Love, to be corroboration of the alleged statement to the police. In our view, this misdirection caused a miscarriage of justice.
(5) Proposed new evidence of Love and Mrs Carlton (the Campbell and Steele references)
 This ground of appeal is to the effect that this court in its judgment on the Secretary of State's reference (Campbell v HM Adv, supra) erred in its refusal to admit the new evidence of Love. Counsel for Campbell accepted that this ground of appeal raises nothing new. It merely asks us to reach a different conclusion on exactly the same point. Counsel for Campbell suggested that if we were to take a different view from the court which heard the 1996 reference, we should convene a larger court in order to consider the point.
 In our view, the opinions of the majority of the court in the 1996 reference were correct. We refer to and adopt their reasons. Since nothing has changed in the interim, we reject this ground of appeal.
(6) The Anderson appeals (the Campbell and Gray references)
 The Commission considers that of all of the possible tactical reasons for not calling Mrs Carlton as a witness for Gray, none would appear to be immediately convincing. In its view, the inconsistencies in her account, to which we have referred, did not necessarily indicate her lack of credibility. Even if her evidence had been heard and believed, the jury would still have been entitled to convict Gray of assault to severe injury on an art and part basis, but in that event the conviction would have resulted in a shorter period of imprisonment.
 Campbell and Gray were separately represented by counsel at the trial, but were represented by the same solicitor, Mr Robert T McCormack. Mr McCormack was cited as a witness for Gray and, in consequence of that, could take little part in the conduct of the defence. He was represented in court by another agent.
 Mrs Carlton was precognosced and cited by Steele's solicitor. She was also cited on behalf of John Campbell and was on the list of witnesses for Lafferty. Although not cited on behalf of Campbell or Gray, counsel for either of them could have called her. Mrs Carlton's precognition for Steele identified Love as having fired the shots. She was the only known source of that evidence. If she had given it, it would have undermined Love's credibility completely.
 It is not disputed that there was co-operation between the defence teams. Whether or not Mr McCormack knew it, counsel for Campbell and Gray had seen the precognition that Mrs Carlton gave to Steele's agents. Counsel for Gray put to Love the substance of Mrs Carlton's account of the shooting.
 Mr McCormack, when interviewed by the Commission, could not remember why Mrs Carlton was not called. He suggested that perhaps it was because she was changing her story. In his statement to the Commission, he confirmed that Mrs Carlton was not precognosced. He said that he was not aware of her existence at the time of the trial, and that her evidence incriminating Love did not exist at the time of trial, so far as he was concerned. Having seen the terms of the precognition that she gave to Steele's agents, Mr McCormack agreed that her evidence was important to Gray's defence. He also said that the notice of incrimination lodged on Gray's behalf was probably based on information given by Gray himself or on "general street wisdom."
 Counsel for Campbell submitted that those representing Campbell had a duty to precognosce Mrs Carlton and to cite her as a witness, rather than to put the substance of her statement to Love in cross-examination. No lawyer acting reasonably and responsibly on Campbell's behalf could have failed in that duty. He accepted that on the information before the court, he could not exclude the possibility that the calling of Mrs Carlton was carefully considered by those representing Campbell and, for sound reasons, rejected.
 The solicitor advocate for Gray submitted that any solicitor professing normal competence would have had Mrs Carlton precognosced, even if he was unaware of her proposed evidence; and that once it was known that she had given a precognition to Steele's solicitor on the lines that we have described, it was important that she should be precognosced specifically from the standpoint of Gray. Once counsel for Gray had cross-examined Love, it became vital, in view of Love's denials of the allegations put to him in cross, that Mrs Carlton should be led. Without her evidence, there was nothing to support the special defence. After Love gave evidence there was ample time for Mrs Carlton to be precognosced and cited. It appeared that there had been a rumour at the trial that she might change her evidence. If that was the case, there was an even more urgent reason to find out whether or not she would. At worst for Gray, even if she had changed her evidence, she could not have significantly harmed the defence case. In the result, as the trial judge reminded them (Charge, pp. 66, 83), the jury never heard any evidence in support of Gray's incrimination of Love.
 In our opinion there is no substance in these submissions. A finding by an appeal court that an advocate or solicitor has failed in his professional responsibilities is a serious judgment that should be made only on clear and cogent evidence. While certain of Mrs Carlton's statements underlined her potential significance to the defence, it does not follow that failure to precognosce, cite or call her constituted an Anderson error. For this ground of appeal even to be arguable, counsel for Campbell and Gray would have to establish not only that those acting for them were aware of Mrs Carlton's importance as a witness, but that they made a considered decision not to have her precognosced and not to call her as a witness. Even if that were to be established, counsel for Campbell and Gray would have to show that those decisions constituted a fundamental error of judgment going beyond the scope of counsel's professional discretion.
 In our view, counsel for Campbell and Gray have failed to establish any of these points. In the ground of appeal and in the submissions of counsel, there has been no specification of the person or persons whom Campbell and Gray blame for the alleged failure in representation. In this complex trial, important decisions such as those relating to Mrs Carlton's evidence would be made by counsel. These events took place nearly 20 years ago. Counsel who defended Campbell have no recollection of the circumstances. Leading counsel for Gray is dead. His junior counsel has no material information within his recollection. While Mr McCormack appears to have been unaware even of Mrs. Carlton's existence, we must bear in mind that, as a witness in the trial, he took only a limited part in the defence. We cannot assume, therefore, that he has any useful insight into the decisions that counsel made or the reasons for which they made them.
 In our view, we should be slow to pass judgment on the conduct of counsel on the meagre information before us. We are not prepared to draw any adverse conclusion against counsel from the mere fact that Mrs Carlton was neither precognosced nor cited on behalf of Campbell and Steele. The failure to precognosce Mrs Carlton is immaterial, in our view. The crucial evidence on which the present complaint is based was her evidence incriminating Love. That was all in the precognition given to Steele's defence. So the real question is whether counsel for Campbell and Gray were culpable in failing to have Mrs Carlton cited and led as a witness While it is certainly a possibility that counsel were culpable in this respect, it is in our view more likely that they carefully considered the whole question and, for good reasons, decided not to call her. Counsel would have had every reason to think that she might be an unreliable witness. We now know that she gave conflicting pre-trial statements on 29 April, 4 July and 7 July. In all likelihood, counsel for Campbell and Gray were aware of that. But despite her incrimination of Love in her statement of 7 July, there was always the distinct possibility that when confronted with having to incriminate her brother in court, she would fail to do so.
 The solicitor advocate for Gray argued that in that event, the defence would be no worse off. We disagree. If there was every indication that Mrs Carlton might be an unreliable witness, there was no knowing what she might say once she was in the witness box. An experienced and sensible counsel would have had that in mind.
 This case is distinguishable from Hemphill v HM Adv (2001 SCCR 361) on which counsel for Campbell relied. In that case there was an agreed failure by the defence to secure expert evidence on a point that had a vital bearing on the line of defence and which, if given, would have been "authoritative" (at para ). It was therefore a straightforward matter for the court to conclude that the failure to obtain such evidence had led to a miscarriage of justice. None of those circumstances can be said to apply in this case.
 On the whole matter, we refuse to find fault with the experienced and able counsel who defended Campbell and Gray. We therefore refuse to disturb the convictions on this ground.
 We shall quash the convictions of Campbell on charges (9) and (15) on the
grounds of (i) the new evidence and (ii) the misdirection relating to the evidence of Granger.
 We shall quash the conviction of Steele on charge (15) on the grounds of (i) the new evidence, (ii) the misdirection relating to the evidence of Granger and (iii) the misdirection relating to the evidence of Ferguson.
 We shall not disturb Gray's conviction on charge (9) on the ground referred to in the reference and we shall refuse his grounds of appeal.