Steel, R v

This is the appeal against conviction of Anthony Steel, brought upon a reference by the Criminal Cases Review Committee (the “CCRC”) under section 9 of the Criminal Appeal Act 1995.

The appellant was convicted (by a majority of 11:1) of the murder of Carol Wilkinson and sentenced to life imprisonment. His trial, in the Crown Court at Leeds before Boreham J and a jury, concluded on 13 December 1979, now more than 23 years ago. He applied for leave to appeal on a number of grounds, of which the principal one related to an issue of causation of death, an issue which is of no relevance to the current appeal. That application was heard together with an appeal in the case of one Richard Malcherek, which raised an identical or very similar issue of causation, and came before this court on 17 March 1981 (Lord Lane CJ, Ormrod LJ and Smith J, nos 5284/B2/79 and 5490/A2/79). Both Malcherek’s appeal and the appellant’s application failed. The appellant’s application on that occasion also raised a number of other grounds, most of which have no relevance to the current appeal, but some of which bear indirectly on it and will be mentioned below. The appellant’s then application was unsuccessful in respect of all those grounds.

The CCRC’s reference was precipitated by fresh evidence in the form of a report from a consultant psychologist, Ms Olive Tunstall, dated 11 April 1996, which concluded that the appellant was educationally retarded and abnormally suggestible and thought to have been potentially vulnerable in the context of being interviewed by the police. Those interviews ended in a confession recorded in a statement made on 25 April 1979. The truth of the appellant’s confessions, oral and written, was the central issue of the trial. The CCRC’s Statement of Reasons drew attention to the similarity between the appellant’s case and that of R v. Ashley King [2000] 2 Cr App R 391, in which a report by Ms Tunstall had also played a major role.

The appellant was released on licence in 1998. Ms Tunstall interviewed and assessed him a second time in May 2001, repeating the tests administered in 1996. She produced a second report dated 30 May 2001, which concluded that he had a full scale IQ of 65 (falling within what had previously been termed the mentally retarded, or mentally handicapped, range), and that he functioned near the borderline of abnormality in the case of both suggestibility and compliance, terms which we will elucidate further below. The Crown in turn instructed Mr Charles Burdett, a consultant clinical psychologist, to interview, assess and report on the appellant. Mr Burdett’s report, dated 19 July 2001, confirmed the findings of Ms Tunstall and concluded that “any confession obtained from Mr Steel would be unsafe”.

The appellant’s appeal was presented by Mr Michael Mansfield QC. On the hearing of that appeal, Mr John Kelsey-Fry QC, who appeared for the Crown, did not seek to dissuade the court from receiving the three reports as new evidence in the appeal. While recognising that the question of the safety of a conviction is a matter for the court and not the Crown, he expressed the Crown’s view by saying that he did not feel able to argue that, had the jury heard the evidence of the two psychologists, they would have been bound to arrive at the same verdict. He considered that there was a compelling argument to the effect that the conviction could no longer be said to be safe.

We agreed, received the evidence of the three reports under section 23 of the Criminal Appeal Act 1968, and allowed the appeal. At the time we gave a short judgment, as follows:

“Mr Steel was released on licence in 1998. We are allowing this appeal and quashing his conviction in the light of new evidence from both defence and Crown consultant psychologists, which we accept, indicating that Mr Steel is and was mentally handicapped and at the borderline of abnormal suggestibility and compliability; and, therefore, a significantly more vulnerable interviewee than could have been appreciated at the time of trial.

“The essential question left to the jury in the summing-up of Boreham J was whether Mr Steel’s confession was voluntary and true; or whether, as he said, it had been produced as a result of the circumstances of his detention and interviews and had been put in his mouth by police insistence and suggestions. On that issue, the jury did not have the benefit of the new psychological evidence as to Mr Steel’s vulnerable personality.

“We are satisfied, as is the Crown, that if the jury had had the benefit of that evidence it could have made a significant difference to their thinking.

“In the circumstances, we cannot regard this conviction as safe and the appeal must be allowed. Other matters in support of this appeal have also been argued, which we will consider further. Our full reasons, of which this is only an indication, will be handed down in due course.”

These are those full reasons.

On 10 October 1977, at some time between 9 and 9.30 am, the victim, 20 year-old Carol Wilkinson, was savagely attacked in Woodhall Road, Bradford, as she walked to her place of work, a bakery in Gain Lane. She was partially stripped, sexually assaulted and battered about the head with a large stone. Shortly afterwards, she was found unconscious. She was taken to hospital, was found to have multiple skull fractures and severe brain damage, and was put immediately onto a life support machine. On 12 October, her doctors concluded that her brain had ceased to function and the life support machine was switched off. The case for the Crown depended almost entirely upon admissions both orally and in writing said to have been made voluntarily by the appellant under police questioning. The Crown contended that only the murderer could have known the details that he mentioned in interview. The Crown also relied on the fact that the appellant had given a key ring in the shape of a fish, said to have belonged to the victim and to have been taken from her handbag, to his future wife, Michelle Smith, at about the time of the murder.

The appellant lived at 45 Langdale Road, on the same estate as the victim. At that time he was living there with his partner, Pamela Ward, but he also had a relationship with Michelle Smith and subsequently married her. On 13 October 1977 police called at his home in the course of a systematic enquiry and asked him to complete a questionnaire about his whereabouts on the day of the murder. He said that he had arrived at work, as a gardener, at 8 am, but had left for home again at about 8.20, arriving there at 9.00, together with his workmate Eddie Hannon, and staying until 9.30. He said that Pamela Ward was at home. He was not arrested until 23 April 1979, some 18 months later. It appears that his mother-in-law had gone to the police with the fish key ring. Over that and the next two days he was interviewed seven times, although the seventh interview was said by the Crown to be only a purely friendly visit to the cells. On the fifth occasion, he made a written statement, confessing to the murder. During the first three interviews, he denied having had anything to do with the murder. He said that he and Eddie Hannon had gone to his home for a pot of tea, arriving there at about 8.45 and returning to work about 9.45. When told that Eddie Hannon was saying that he had stayed at work and not gone home with him, he said that was a lie. He could not remember where he obtained the fish key ring. On the fourth occasion, he began to make certain admissions of presence at the murder scene. On the fifth occasion he made a written statement, confessing to the murder.

Following his statement, he was interviewed a sixth time by Detective Chief Superintendent Hobson, who was in overall charge of the investigation and sought confirmation of the confession. There were, he told the appellant, one or two matters he wanted to clarify. At trial, there was an application to the judge to exclude this and the seventh interview (see below) on the grounds that under paragraph (d) of the preliminary principles set out as the prologue to the Judges’ Rules (1964, revised) he should have been charged and not interviewed further following his confession, and that what followed was oppressive. The judge was prepared to assume, while stating that he was far from holding that there had been a breach of the Judges’ Rules, but he nevertheless rejected the application. He said that he had no doubt that what happened thereafter was in no way oppressive to the defendant. The judge also ruled, in the course of the application just referred to, that there was no evidence that the appellant was at any time refused the services of a solicitor.

That sixth interview ended at 7.40 pm on 25 April 1979. At 8.01 the appellant saw a solicitor, Mr David Taylor, of the firm of T I Clough & Co. He had been at the station, on his evidence, for perhaps half an hour prior to seeing the appellant. The latter had had no previous access to a solicitor. He was with Mr Taylor until 8.35 pm. Mr Taylor said that the appellant was extremely agitated. On the conclusion of that conference the appellant retracted his confession. The solicitor told Superintendent Hobson that there were to be no further interviews.

So far the interviews had been recorded by the police in writing. These were the days before tape recordings.

The seventh interview was later that night and was unrecorded. DS Falconer went to the appellant’s cell, he said for a “purely friendly” visit, a “social call”. DS Falconer’s evidence was that the appellant said to him: “I’m sorry, Mr Falconer, I’m going to have to go back on my statement”, and that he replied: “That’s up to you lad, it’s a free country”. He said that the appellant then said: “It was my solicitor you see, he said I was a pillock for making it and that I should have kept my mouth shut…I just want you to know that I’m sorry if I’ve shown you up. It wasn’t my idea.” The appellant’s version of this event was that he was taken by DS Falconer to an interview room; the sergeant, red in the face and rubbing his hands, said “You’re a pillock for going back on your statement. If you had stuck to it I could have got you a visit and got you bail”; to which he replied: “I’m not admitting to something I hadn’t done”.

At the trial, Mr Taylor was called by the defence to give evidence as to what the appellant had said to him at their first conference, so as to rebut the inference to be drawn from DS Falconer’s evidence that the retraction of the confession was forced on the appellant by his solicitor. The evidence which the defence sought to adduce through Mr Taylor is set out in the judgment of this court on the appellant’s application for leave to appeal (at 22G): “It seems, according to the solicitor, that the applicant told him that the statement which he had made to the police was not true and that the reason why he had made it was because the police had “kept on at him”.” The judge would not permit that evidence to be given, on the ground that it was hearsay, and the appellant’s application to this court to give leave for a ground of appeal that the judge had erred in this respect was refused.

The appellant’s case at trial was that he had nothing to do with the murder. That morning he had arrived at work at about 8 am. After a while he had asked his workmate, Eddie Hannon, if he would like to come back to his house for a pot of tea, but Hannon had declined as they had previously got into trouble for skiving on the job. So he had returned alone, arriving at about 8.50. Pamela Ward was there, and they had talked. He left between 9.30 and 9.40 to go back to work. He had worked until 3.30 pm. Subsequently he had read about the attack on Carol Wilkinson in the newspapers.

When interviewed for the purposes of the questionnaire on 13 October 1977 he had not said that Hannon had returned home with him, but that Hannon could confirm that he had gone home for a pot of tea. He agreed that he had confirmed in his first three interviews in April 1979 that Hannon had gone home with him, but that was much later and he had been confused. (Hannon gave evidence and said that he had stayed at work and not returned home with the appellant.)

As for his interviews, he said that he had asked for a solicitor from the first and repeatedly, but had been refused. He had been shouted at and abused. He had been pushed to the floor and hit about the head. As for his fourth interview, he said that the admissions had not been made by him, at most he had accepted them under pressure as suggestions to him. He had been repeatedly shouted at, and had said: “You won’t take no for an answer”. He had again asked for a solicitor. He was shaking. As for his statement, the critical part of it was, with minor exceptions, made up of suggestions put to him which he had either completely denied or accepted under pressure. He had not marked crosses on a map of the area designed to show his and the victim’s routes. It was not a voluntary statement, nor dictated by him. DS Falconer had told him to add some words of his own, confirming the accuracy of his statement, and he had signed, but without appreciating the meaning of the words. He had not read the statement before signing, although he had been given the opportunity of doing so. As for the sixth interview with the superintendent, he had been frightened by DS Falconer, who had also been present, into demonstrating how he had attacked the victim. We have already referred to his account of the disputed seventh interview with DS Falconer later that evening.

As for the fish key ring, he accepted that in his first interview he had said that he could not recall it, but he said that he knew that he had given it to Michelle Smith but could not remember where he had got it from. He had, however, realised by the time of the second interview that he had got it from Pamela Ward’s drawer, where he had seen it before. He had not then mentioned it to the police, because he knew that they would not believe him. He also said, however, that he had realised this very recently, alternatively a week or two after his arrest.

In sum, the police witnesses denied the appellant’s allegations of violence, threats of violence, browbeating, fabrication of parts of the confession, and the repeated denial of access to a solicitor. The Crown’s case was that the confessions emanated from the appellant, were voluntary and therefore reliable. The appellant’s case, however, was that his response was one of despair in the face of pressure which he was unable or too frightened to withstand and that whatever he said or agreed to was not to be regarded as a voluntary, truthful confession. The judge summed up to the jury in those terms.

We now turn to the interviews in greater detail.

The first was on 23 April 1979, lasted two hours between 9.20 and 11.20 pm and was conducted by DI Mould with DS Falconer in attendance taking the note. The appellant had been arrested at about 6.30 that evening. He denied having anything to do with the murder. He said that he was at work that day. He was then presented with the questionnaire from 13 October 1977 and asked to confirm that it was true, which he did. The fish key ring was produced, and he was asked if he recognised it: he said “I don’t know, I don’t think so”.

The appellant did not dispute the content of this interview, but he nevertheless gave a very different account of its style from that which appears from the written page. He said that the police officers were speaking in loud voices, and he was shaking and very upset. “I had never before experienced it and I’m not much good with words”. He asked for a solicitor, but was refused and was told that he would get one when they had “finished with him”. DS Falconer called him an animal. He also took him by the collar, pushed him to the floor and struck him about the back of his head and said: “We’ll start again and while you are here we want ‘yes’ and not ‘no’.” The appellant once again asked for a solicitor and was again refused. All these allegations were denied by the police witnesses.

The custody record shows that the appellant received no refreshment on the day of his arrest, although he was interviewed until 11.20 pm.

The second interview was almost 24 hours later, and was much shorter, lasting from 8.32 to 9 pm. It was conducted by DI Mould with DCI Mitchell. The appellant was again asked about his whereabouts that morning, and confirmed that Hannon had returned home with him for a pot of tea. He was again asked about the key ring: “Can you remember where you got it from now?” He said he did not know, but agreed that he had given it to his wife, Michelle. He was told that it was identical to one owned by the victim. Again, the appellant accepted the essence of the content of the interview, but gave his own account of it. He said that the police officers took it in turns to shout at him, and that he was again refused his request for a solicitor. The police officers denied this account. But the appellant had continued to deny the allegations of murder.

The third interview was later the same evening, between 9.45 and 10.20 pm, with only DI Morgan in attendance. The questions again focussed on whether Hannon had come home with him and where the fish key ring had come from. He again confirmed that the questionnaire was correct in saying that Hannon had come home with him, and when told that Hannon was saying that he had not, said that Hannon was lying. He was again told that the key ring belonged to the victim, but said he could not remember where he had got it from. He said that he had been very upset by this interview and could not sleep at all that night.

The custody record appears to show that the appellant’s first wash facility after his arrest was 37 hours later, at 8 am on the morning of 25 April.

The fourth interview began at 2 pm on that day, 25 April, and lasted until 3.30. It was conducted by DS Falconer and DC Dixon. The appellant said: “I know it looks as if I killed her and I want to make you believe that I didn’t but I can’t, can I?” And then, about the fish key ring, “You’ve no idea how I’ve sat and thought. Thought about that fish thing really. That proves it against me, doesn’t it?” DS Falconer said: “That fish was Carol Wilkinson’s…What were you doing with it?”, and the appellant said: “That’s it. I don’t know do I? I don’t know where I got it from, whether I found it, had it given, bought it or what…If I had told them where I had got it from it would not have done me any good, they would just have called me a liar…There’s no point in me trying to make something up about where I got it from because if I did I’d have to lie wouldn’t I, then I’d be in trouble.” “You’re in trouble now.” “Yes, but I’d be in even more trouble if I started to lie about things.” “Do you know what that sounds like to me?” “What?” “It sounds like you are trying to tell me that you got it off Carol Wilkinson.” “I did nothing of the kind.” There was no dispute about the interview thus far, but the appellant’s evidence was that the police officers were shouting at him. There was then a discussion about the appellant’s alibi and the fact that Hannon (and Pamela Ward) were not supporting his account of their being together.

There then came what on the Crown’s evidence was a critical moment, a turning point at which the appellant started for the first time to make admissions. “I’ve lied about something.” “What’s that?” “The fence, I jumped over it.” “Go on.” “When I saw her.” “Saw who?” “Her, Carol.” The appellant had previously denied seeing the victim on her route at any time that morning. “Did you kill her?” “I must have done…well, I did see her. I jumped over the fence at the back of our house the second time I saw her…I sort of got up in front of her. We’d a ‘tiff’ and that’s all I can remember.” “A tiff?” “A struggle.” “How did you kill her?” “I can’t remember.” “Did you get astride her and butt her? She’d bruises on the top of the arms as if somebody might have done that to her.” “I don’t know. It’s sort of blank.” “Let’s get this straight, are you saying that you killed Carol?” “I must have done.” But the appellant said that he added: “You won’t take no for an answer.” “What you are admitting is murder, do you realise that?” “Yes, when I bought the Telegraph that night, and read what I’d done, I couldn’t believe it. We’d just had this struggle, you see, and I thought there might be a fuss but nothing like that. You see, if I’d known just what I’d done, I might have got a better alibi, but there didn’t seem much point until it was too late.” The appellant denied saying these things, as distinct from accepting them under pressure when they were put to him.

There was then a discussion as to what should happen next. DS Falconer said that he would like the appellant to tell DI Mould exactly what happened, but that it was up to the appellant whether he spoke about it, wrote a statement, dictated a statement, or said not another word. DI Mould was sent for. DS Falconer said that there was silence in the room for the next ten minutes, while DI Mould was awaited. The appellant, however, said that DS Falconer said that “it wasn’t as bad as [you] might think, there could be one of three charges, murder, manslaughter or sexual assault.”

The appellant’s evidence about this interview, taken from the judge’s summing up, were:

“I was shouted at. I first denied that I had anything to do with the killing and later I said I must have done it and I said this because of the pressure by Mr Falconer. I asked him for a solicitor, I was told that I would get one when they had finished. I was shaking, shifting on the chair, moving my hands around, I had not done it…I told him that I had not done it but he just would not take no for an answer…”

and the judge continued: “and that really is the essence of his contention here, that he was being pressurised and being thus pressurised he simply accepted all that was being put to him. In other words, the whole note of that interview, do you understand, was distorted…”

The fifth interview was timed between 4.00 and 6.20 pm, in the presence of DI Mould and DS Falconer. It led to the making of the written statement. The appellant said that the interview began with a promise of bail and a visit from his wife as an inducement to speak, and a threat of violence if he did not. The interview then went through the morning’s events starting at the beginning and ending with the attack on the victim. For the first time, the appellant accepted that Hannon had not been with him. The Crown’s evidence was that he said: “I saw her again in Ravenscliffe Avenue…I followed her, I was just going to get a date, chat her up…I walked up in front of her and then I waited for her by the wall…Well, I just put my arm round and tried to kiss her, just kiss her and there was a ‘tiff’…a struggle. Then she ran over the wall to the field. She was screaming. I just wanted to talk, get a date. Then there was a struggle, it’s blank, it’s blank. Just running, just running…The fish, I remember the fish. It was hanging from her bag, I took it…” “Anything about the struggle that you remember?” “The stone, I remember the stone.” “What about the stone?” “I just remember a stone.” That was the only reference to any weapon.

The appellant’s evidence was again that he was only accepting under pressure what was suggested to him.

The appellant then agreed to dictate a statement, which DS Falconer wrote. This is a document of five pages. It begins: “I Anthony Steel wish to make a statement. I want someone to write down what I say. I have been told that I need not say anything unless I wish to do so, and that whatever I say may be given in evidence.” At the foot of the last page he wrote: “I have read the above stament and I have ben told that I can correct alter or add enything I wish this statement is true I have made it off my own free will”. The statement refers to drawing a route on a map . The critical part of the statement reads:

“I stopped there near to a stone wall. There were a few stones out of the stone wall just there where I stopped…As she just got by me she looked at me and I just put my arm out…and I tried to kiss her, but she pulled away…We had a bit of a tiff (Q A tiff?) a struggle you know. She sort of got away from me and went over the wall…screaming…and I just wanted to explain that I just wanted a date. I jumped over the wall after her and chased her...I caught hold of her like this (Indicated with his left arm making a gesture as if grabbing someone from behind)…I was trying to stop her struggling…I was panicking because she was screaming and she sort of fell onto the floor as we were struggling…We were just struggling on the floor, struggling on the floor, just struggling…I kept on saying that I just wanted to talk to her. Then suddenly she just wasn’t moving. I remember looking at her eyes were sort of open…I don’t know but that key ring was hanging off her bag I just took that…”

There was no mention of use of any stone or other weapon.

The appellant’s evidence about this statement was that it was not dictated by him, rather it was put to him and he just agreed. As for the map, he said he drew a route on the police officers’ instructions, and DS Falconer put two crosses on it himself. He kept on being told that he would get bail and a visit from his wife. He said that he felt very upset signing something which he knew was not true. The passage added at the end was dictated to him. As for reading the statement, that was not true: he read the first three lines, did not take it in, and threw the statement back on the table. The police said that he read out the whole of the first page.

The sixth interview was between 7.15 and 7.40 pm the same evening, and was requested by DCS Hobson. DS Falconer was also present. The judge said this:

“There has been, as I understand it, some criticism of that…So far as that Chief Superintendent was concerned he was, you may think, entitled to take the view that he ought to ask some questions. Don’t forget, this was a situation where 18 months and more had passed since that girl died, and as you and I know, with the possible exception, and it is for you to judge that, of that key ring with the fish, apart from that there was nothing, and there is nothing to connect this young man with that murder save only what he has said about it. It is right, therefore, you may think that great care should be taken in that regard…”

The interview began with the superintendent saying there were one or two things he wanted to clarify. The appellant was asked to go over the matter again. There came a point when the superintendent said: “When you chased her over that wall you wanted to rape her, didn’t you and when she refused you hit her with the boulder?” The appellant replied: “No, I can’t remember, no.” He was also asked whether he had taken off the victim’s bra and pants. He said, he could not remember. “Do you realise that you have killed Carol?” “Yes, I must have, but I can’t remember.” He was asked about the fish, and he said it looked nice, so he took it. He described the bag as dark, sort of black or brown (but that was not a description of the victim’s bag which was an exhibit in court). Then he was asked: “Carol was killed by someone hitting her with a large boulder and there would be blood about. Did you have any blood on your hands?” The appellant said: “I don’t remember hitting her with a boulder, and I didn’t have any blood on me.” “Do you realise that you killed Carol?” “Yes.” And finally: “Is everything in your statement true?” “Yes.”

The appellant’s account of this interview was that he was sick of being interviewed by the police and just wanted to be left alone. He was giving answers which were untrue about things that in fact he knew nothing about in order to get the whole thing out of the way.

We have already dealt above with the visit of DS Falconer to the appellant’s cell later that night and the conflict of testimony relating to that.

The fish key ring which the police in interviewing the appellant represented as having belonged to the victim and which therefore played a major role in his questioning rather faded in significance by the end of the trial. It was made of a plain gold coloured metal and was made an exhibit, and a number of witnesses were asked about it. In the end, their evidence was equivocal. Carol Wilkinson’s younger brother, Paul, said that he had seen her with an identical key ring. Three brothers of her fiancé, Kevin Best, gave varying evidence: Brian Best said that he had seen her with a similar ring, but that it had coloured stones for its eyes; Neil Best spoke of a similar key ring to that exhibited, and so did Terry Best. However, the fiancé, Kevin Best himself, said that he had never seen anything like the key ring in her possession. Mr Stowell, a work colleague of the victim likewise had never seen such a key ring in her possession. Miss Broadbent, another work colleague who saw her every day at work and in particular at meal times, said the same and also pointed out that there was nowhere to attach the key ring to the victim’s bag. Mrs Alldred, who said that she saw Carol and Kevin frequently, remembered a key ring of similar shape, but she described it as silver with blue, yellow and green colours on its back and upper part. When shown the exhibit she said it was not the victim’s; but later also said that it could be the same.

In his summing up, the judge on occasions described the key ring as, subject to the jury’s view, capable of giving some support to the appellant’s confession. On another occasion he referred to it, among other pieces of evidence, as “consistent with this defendant being her assailant, but no more. They of themselves, you may think, would be insufficient to prove his guilt” (at 5F).

The judge therefore went on to say (at 5F/6A):

“So it is that the matter becomes even simpler, and it is this. In the end do you accept what the Police say he said to them as truth? If it is you may think it amounts to a complete confession and he is guilty. If you do not, if you think that what he is alleged to have said is really, or even may be, if you are in doubt, no more than a produce of a sinister conspiracy, for that is what it amounts to, by those policemen who put words into his mouth that would convict him of murder, if you think that may be true then he must be found not guilty…At the end of the day you may think that is really the sole question, “Am I sure that what he said he said of his own free will and truthfully?””

The judge made two points in that passage. The main one was that the jury’s view about the confession was the critical, and indeed the sole, question on which guilt or innocence should be tested. The subsidiary point was that the defence’s case that the confession was involuntary and unreliable amounted to an allegation of a sinister conspiracy on the part of the police.

These two themes were repeated or reflected elsewhere in the summing up. Thus the judge went on, just after he summarised the appellant’s defence, to say (at 20G):

“The question you will have to resolve is simply this. Were the answers which he gave and the contents of the statement that he made true? If the answer to that is a sure yes, as I have said, you may think that is conclusive. If the answer is no or we are not sure then equally it should be conclusive in his favour, not guilty.”

Again, towards the close of his summing up, the judge said this (at 47B):

“If you are satisfied so that you are sure, and this applies to all the interviews, that what the Police have said is correct, or substantially correct, why then, you may think, that is conclusive of the case, against the defendant. If, on the other hand, you are not so satisfied and you think that what the defendant says, even may be true, though you are not sure, you are left in a state of doubt in other words, why then there really is no evidence against this young man.”

As for the subsidiary theme, there is this passage (at 22C):

“It is right to say that if what he says is true or you think may be true, well then there has been a most shocking and cynical conspiracy against him, a dreadful thing. A conspiracy that goes so far so that the Defence case is to saddle him with the crime of the enormity of murder, That is what it comes to.”

That presumably reflects the way in which the defence case was put.

The appellant was born on 26 May 1956. He was therefore 22 years old at the time of his arrest and 23 at the time of trial. On 7 August 1979 Dr Terence Kay, a consultant forensic psychiatrist, reported on the appellant’s mental condition at a time when he was detained in Leeds prison. There was no evidence of any mental disorder and he was fit to plead. The report’s conclusion, however, contains this passage:

“Although Steel’s psychological testing indicates his Intelligence Quotient is only 67, I believe this is an underestimate and clinically he is of dull normal intelligence.”

An IQ score of 67 in fact falls within the range (below 70) which is now called “extremely low” (Mr Burdett also describes it as the “learning disabled range”) and was formerly called mentally retarded or described as the mental handicap range. Dr Kay does not state what led him to his conclusion that the score was an underestimate, but it may have been the general findings mentioned elsewhere in his report that the appellant “spoke easily and gave a good account of himself”.

That the appellant was not intelligent, however, appears to have been discussed at trial. The judge mentioned it in his summing up in these terms (at 21E):

“It is said of him, and I say this without, I hope, giving offence, that he is not very bright intellectually. Well, you have seen him and you have heard what has been said about him. I repeat, it is for you to judge it and nobody else.”

There was no psychiatric evidence about him at trial, however.

We have already referred to the judge’s rulings (a) that the further interviews after the written statement were “in no way oppressive to this defendant” even upon the assumption that there had been a breach of the Judges’ Rules in that respect; (b) that the appellant had not been refused the services of a solicitor during the course of the investigations; and (c) that Mr Taylor could not give evidence as to what the appellant had said to him when he met him for the first time at 8.01 pm on 25 April 1979. It is necessary to say something more about those rulings and the grounds of appeal for which application was made to this court in 1981.

It is not clear from the judge’s rulings (a) and (b) whether there was an application generally to exclude all the police evidence of the interrogation and statement or only that relating to the period after the confession. It seems on the whole as though ruling (b) related to the whole interrogation, and ruling (a) related in particular to what followed the written statement. To judge from what was subsequently said in this court, however, it may be that bound up in these rulings there was also consideration of a general submission that there had been such oppression as should lead to the exclusion of the police evidence as a whole. As for ruling (b), since the appellant did not give evidence on a voire dire, the judge had in any event no alternative but to find, as he did, that “I have no evidence before me that there was such a refusal” of the services of a solicitor.

Thus in the course of the application to this court in 1981, Lord Lane had to consider a ground of appeal that the judge had erred “in rejecting the defence submission that the evidence of the interviews by police officers after he had made his voluntary statement ought to be excluded”. That appears to be related to the judge’s ruling (a). In the course of dealing with that ground, however, this court also said this (at 29):

“Quite apart from that, the situation was this. There was a submission made in the absence of the jury – a trial within a trial – that these statements, and particularly the written statement, had been extracted from the Applicant by means of oppression or possibly breach of the Rules or possibly unfairness. The Applicant did not see fit to give evidence on the voire dire, thus so far as the oppression allegation was concerned the judge had only one set of facts (those presumably given by the police) on which to come to his conclusion. It is not surprising therefore that he came to the conclusion that he did, namely that oppression had been disproved…”

On this basis, there had been a more general submission.

Reverting to ruling (a), we note that this court dealt with the ground of appeal relating to that ruling, after setting out the relevant passage from the Judges’ Rules, in these terms (at 26/27):

“The Detective Chief Inspector [sc Superintendent] in charge of this case had before him the alleged written confession, parts of which I have already read. He had also the knowledge of the sort of person it was who was alleged to have made the statement. If, without further ado, he had charged the man then no doubt that would have been the subject of criticism. If he took steps to ensure that he was not charging someone who had made a false confession, then again he was going to be subject to the very criticism to which he has been subject in this court. This court is very far from saying that the Detective Chief Inspector was wrong in not immediately charging this Applicant after the written confession, but assuming that he was wrong and assuming that he should have charged the defendant immediately, and even assuming that having charged him then it was improper to go and see him to ask him any further questions, it seems to us nevertheless that the evidence contained in the statement itself, coupled with everything else…was ample evidence for the jury to come to the conclusion that they did.” [emphasis added]

It would seem from the sentence emphasised above that there was some evidence that DCS Hodson had entered upon the sixth interview partly out of concern to satisfy himself that the appellant, whom the police themselves may have regarded as not all that bright (cf the judge’s comment to the jury at para 48 above), had given a voluntary statement.

The application also put forward a ground of appeal related to the judge’s summing up to the effect that the choice for the jury was between the police account of the confession and the defence case of “a sinister conspiracy”. That ground was rejected. This court pointed out that it was part of the defence case that at least some parts of the confession had been made up and did not emanate from him at all, and continued (at 31):

“In those circumstances it seems to us that the judge was at the very least entitled to point out to the jury, in clear terms without fudging the issue, the suggestion which was being made against the police; namely that it was a cold cynical conspiracy to convict an innocent man of murder.”

Ms Tunstall’s first report was written following an assessment of the appellant in prison on 18 March 1996 over a period of 4½ hours. His IQ score was 74 (borderline to the lowest range, which begins below 70), his reading age 10 years 7 months, his basic number age 7 years 11 months, his suggestibility score (on the Gudjonsson suggestibility scale form 2, GSS2) was 17 equivalent to the 98^th percentile, which is within the abnormal range which begins at the 96^th percentile, and his compliance score (on the Gudjonsson compliance questionnaire, GCQ) was 14 equivalent to the 94^th percentile, which is just outside the abnormal range. Ms Tunstall said it was difficult to evaluate the significance of the 7 point discrepancy between the IQ score of 74 and that of 67 found by Dr Kay. The GSS2 and GCQ tests were described by Ms Tunstall as follows:

“The distinction between suggestibility and compliance as these traits are defined and measured by the Gudjonsson scales is that suggestibility requires personal acceptance by the subject of the information provided or requests made, whereas compliance refers to the tendency of an individual to accede to propositions, requests or instructions put forward by others even though he may privately disagree with what is being put to him.”

Ms Tunstall described in her report the appellant’s account of the pressures brought on him in interview. She commented:

“Mr Steel’s allegations concerning the pressures which were placed upon him during the interviews are not supported by the records of the interviews contained in the witness statements of the police officers. The interviews however were not tape recorded but were contemporaneously hand recorded and in my opinion there is some reason to question whether the records contained in the witness statements constitute a full and comprehensive record of the interviews. Some interviews appear to me to contain very much less material than one would expect given the length of time they occupied. In addition there are very substantial disparities in the quantities of material yielded by different interviews and by different parts of the same interview. I have set out on a separate sheet the calculations upon which I have based this assertion…”

Ms Tunstall concluded as follows:

“Mr Steel has a Full Scale IQ of 74, is educationally retarded and abnormally suggestible and is thought to have been potentially vulnerable in the context of the police interviews.

Further sources of vulnerability at the time of the interviews were his youth, his lack of previous experience of being questioned under caution and the fact that he had no access to legal advice…”

Ms Tunstall also pointed out that measures of interrogative suggestibility were not developed until the mid 1980s and that this information would therefore not have been available at the time of trial. The same appears to be true of the test for compliance. This is a point also made by Lord Bingham of Cornhill CJ giving the judgment of this court in King at 405C, where, speaking of 1985/6 he described tests for suggestibility and compliance as a “new and embryonic science”.

Ms Tunstall interviewed and assessed the appellant for a second time on 15 May 2001, for a period of some 4 hours. The results were written up in her further report dated 30 May 2001. On this occasion the appellant’s full scale IQ score was only 65, his reading age was 10 years 9 months, his suggestibility score was 14, equivalent to the 90^th percentile, and his compliance score was 17, equivalent to > 95^th percentile. The IQ score fell into the “extremely low” range (below 70), previously described as the mentally retarded or mental handicap range. The suggestibility score on this occasion was just below the abnormal range but was on its borderline. The compliance score on this occasion was in the abnormal range (above the 95^th percentile).

In her second report Ms Tunstall discussed the differences in her findings in 1996 and 2001. As for IQ, she concluded that the appellant was unlikely to have deteriorated intellectually, as was moreover shown by the IQ sub test scores, and that the deterioration in score was rather due to the different versions of the WAIS test adopted on the two occasions: in 1996 the test was WAIS-R and in 2001 it was WAIS III. The previous test had become outdated and was regarded as having overestimated IQ. As for suggestibility and compliance, whereas the 1996 tests had shown the appellant to be abnormally suggestible and borderline to abnormal on compliance, the 2001 tests now reversed those findings which thus became borderline to abnormal on suggestibility and abnormally compliant. Ms Tunstall discussed these findings in these terms:

“The present scores therefore provide no evidence of abnormal suggestibility but do provide evidence of abnormal compliance. Very small changes in the actual scores have brought this about and in my opinion it would be consistent with the test evidence to perceive Mr Steel as functioning near the borderline in the case of both suggestibility and compliance.

“In my report dated 11.4.1996 I pointed out that measures of interrogative suggestibility had not been developed at the time of Mr Steel’s trial and the same is true of measures of compliance.

“The results obtained in the present assessment have not caused me to change the opinions expressed in my report dated 11.4.1996. It remains my view that Mr Steel was potentially vulnerable in the context of the police interviews by reason of his level of intelligence, now estimated to fall within the mental handicap range, and his levels of suggestibility to which I would now add compliance. Other potential sources of vulnerability listed in my earlier report which I repeat here were his youth at the time, his lack of any previous experience of being interviewed under caution and his lack of any access to legal advice until after the police interviews had been concluded.”

Ms Tunstall went on to comment on DCS Hobson’s evidence about the sixth interview concerning the appellant displaying no signs of stress or distress and as seeming relieved. She pointed out that the relief of bringing a highly stressful period of interrogation to a close could result in a demeanour which was consistent for a false as well as a true confession.

On 14 July 2001 the appellant was interviewed and assessed by Mr Burdett in turn. He had before him Ms Tunstall’s two reports. His full scale IQ score for the appellant (using WASI) was 64, very close to Ms Tunstall’s second report finding of 65. The suggestibility score placed the appellant in the 99^th percentile (abnormal), and the compliance score placed the appellant in the 88^th percentile (at the upper, ie more compliant, end of the average range). Mr Burdett expressed the following opinions. As to IQ, he agreed with Ms Tunstall that the 1996 WAIS-R test would throw up slightly higher scores. He appears to have considered it significant that the 1979 and 2001 scores were almost identical and stated that Dr Kay’s opinion on the 1979 score should be discounted as being outside the area of his expertise. As to suggestibility, he commented on the statistically significant difference between the scores found by Ms Tunstall in 2001 and by himself:

“I feel that this difference in score may well be associated with examiner variables, and it seems feasible that Mr Steel might be more likely to give way to a male examiner. This may have implications for Mr Steel’s behaviour when he was interviewed by the Police, if his interrogators were mainly men.

“Overall, Mr Steel’s scores show a strong tendency towards suggestibility and to agreeing with the interrogator when put under interrogative pressure. Any confession must be considered to be of dubious validity.”

As for compliance, Mr Burdett referred to the wide and significant variation in the three scores recorded in the three reports, but commented that “all the results suggest a strong tendency towards obedience to authority”. He went on to point out that Gudjonnson had himself warned that the test might be unsuitable for subjects whose IQ was below 70.

Mr Burdett’s final conclusion was as follows:

“Mr Steel is functioning within the learning disabled range. He is also highly suggestible to interrogative questioning and any confession obtained from Mr Steel would be unsafe. If Mr Steel were interviewed by the Police at the present time, he would need to have a responsible adult present at all times during the interview.”

We have to consider whether to receive this new evidence. Similar evidence was received by this court in King (see below). Its reception has not been opposed by the Crown. We have no hesitation in finding that it should be received as meeting the requirements of subsections (a), (b), (c) and (d) of section 23 of the Criminal Appeal Act 1968.

King was convicted of murder in 1986. He was 22 at the time. The murder was of a widow in her home; money was stolen. King did not apply for leave to appeal. His application for leave to appeal out of time was heard by this court in 1999. Time was extended and he was granted leave to appeal. He was also granted leave to adduce the evidence of a report from Ms Tunstall. His defence at trial was that the confession he had given to the police had been obtained under pressure and as a result of his accepting the suggestions of the interviewing officers. There was no forensic evidence to link him with the crime. It appears that he had been cautioned in respect of theft, but not in respect of murder. The Police and Criminal Evidence Act 1984 (PACE) had been enacted but not yet brought into force, but the Act and the new Codes were being implemented on a pilot basis by several police forces, including the force in Northumbria which was investigating the widow’s murder. King was not arrested, but interviewed 10 times during a period of 12 hours. His co-accused’s solicitor had referred to him as a “bit daft” and as a result a social worker attended at one interview; but otherwise he was on his own. He had the benefit of no legal advice and it does not appear that the opportunity to consult a solicitor was offered until just before the tenth interview, by which time full admissions had been made. At trial, the picture presented of him was of a man of low intelligence, but not unusually so. There were at that time three relevant medical reports in existence. One described him of dull average intelligence, a second assessed him as in the lower end of an average range of intellectual abilities, and the third recorded a full scale IQ score of 89 (low average range).

The case against King at trial rested solely on his confessions, which were made during interrogations “when the rules in force at the time were clearly breached” (at 401D), for instance he was not advised of his right to receive legal advice until after he had made extensive admissions. It is not entirely clear whether this was a reference to the PACE rules then under trial, or to the prior rules. Mr Mansfield addressed the court on the basis that reference was made to the PACE rules. I am not sure, for Lord Bingham went on to say:

“We were invited by counsel at the outset to consider as a general question what the approach of the Court should be to a situation such as this where a crime is investigated and a suspect interrogated and detained at a time when the statutory framework governing investigation, interrogation and detention was different from that now in force. We remind ourselves that our task is to consider whether this conviction is unsafe. If we do so consider it, section 2(1)(a) of the Criminal Appeal Act 1968 obliges us to allow the appeal. We should not (other things being equal) consider a conviction unsafe simply because of a failure to comply with a statute governing police detention, interrogation and investigation, which was not in force at the time. In looking at the safety of the conviction it is relevant to consider whether and to what extent a suspect may have been denied rights which he should have enjoyed under the rules in force at the time and whether and to what extent he may have lacked protections which it was later thought right that he should enjoy. But this Court is concerned, and concerned only, with the safety of the conviction. That is a question to be determined in the light of all the material before it, which will include the record of all the evidence in the case and not just an isolated part. If, in a case where the only evidence against a defendant was his oral confession which he had later retracted, it appeared that such confession was obtained in breach of the rules prevailing at the time and in circumstances which denied the defendant important safeguards later thought necessary to avoid the risk of a miscarriage of justice, there would be at least prima facie grounds for doubting the safety of the conviction – a very different thing from concluding that a defendant was necessarily innocent.”

That passage was cited to this court on this appeal. At any rate, by the time of King’s trial PACE was in force, and Lord Bingham pointed out that no application was made under its sections 76/78: instead the reliability of the confessions was challenged.

The critical feature in King’s appeal was the new evidence, received by this court, of Ms Tunstall’s psychological report. That gave King a full score IQ of 78, ie lower than had been thought at the time of trial (but still above the lowest range), and scores for both suggestibility and compliance in the abnormal range. By itself, the new information about King’s IQ “would not be of major significance”, for there was nothing to prevent his IQ being accurately assessed at the time of trial and it “is not suggested that he was mentally handicapped or severely retarded or subnormal, even though he was on the borderline of subnormality” (at 405A). However, that factor was put together with the more significant evidence, not available at trial, concerning suggestibility and compliance. As Lord Bingham said (at 405E):

“There is, however, now evidence that the appellant was significantly less intelligent and more vulnerable than was understood at the time, and abnormally ready to accept what was put to him. That is exactly what the appellant had claimed had happened…”

On that occasion King’s appeal was opposed by the Crown. Lord Bingham concluded by giving full weight to the Crown’s arguments, but ultimately rejected them (at 405G/406D):

“Mr Thorn Q.C., who has represented the Crown and made submissions with notable skill and moderation, submits that nothing in truth has altered. Some of the answers given by the appellant to the police officers, he submits, betrayed a knowledge of detail which is inexplicable on any basis other than guilt. He points out that the jury saw and heard the appellant and had every opportunity to form an impression on him. He tells us, and we have no reasons to doubt, that the appellant performed very badly as a witness. He contends that the delay in presenting this challenge undermines his credibility. He furthermore points out, quite rightly, that no criticism is or could be made of the trial judge or those representing the appellant at the trial.

“These are fair points properly made and this is not on any showing a case where it could be said that the innocence of the appellant is established. But we are concerned not with innocence but with the safety of the conviction. The Crown’s points, although properly made, do not confront the real problem: whether, in a case which depended solely on the reliability of the appellant’s confessions, improperly obtained, the appellant is now scientifically shown, as he could not have been shown at the time, to be highly abnormal in respects directly related to the reliability of the confessions and in a way which throws doubt on their reliability.

Had the defence had the benefit of this new expert evidence, there would have been strong grounds for seeking the exclusion of the confession evidence under sections 76 and 78 of the Police and Criminal Evidence Act 1984, and possibly even section 77. Had the jury heard the confession evidence, but also heard the expert evidence of Mrs Tunstall, it would in our judgment have been very hesitant indeed to convict the appellant on the strength of his uncorroborated and retracted confessions, and rightly so. In the light of this new evidence we feel bound to regard the appellant’s conviction as unsafe and we accordingly quash it.”

As for the psychological evidence, that case and this, despite differences of degree, are remarkably parallel. Whereas King was not intellectually abnormal, he was suggestible and compliant to an abnormal degree. In this case, the appellant is on the borderline of abnormality in both suggestibility and compliance (with various figures putting him on both sides of that borderline for both tests) but he is also intellectually abnormal. It may be thought that overall these differences are not critical. What may be different, however, is that, subject to further consideration, it has not been established, as this court considered that it had been in King, that the confessions had been “improperly obtained”. That factor has led to Mr Mansfield framing his submissions in two ways.

Mr Mansfield’s primary submission was that the new evidence made the appellant’s conviction unsafe, and that was irrespective of any breach of the Judges’ Rules or of the absence of any modern safeguards thought necessary for the protection of a suspect and the avoidance of a miscarriage of justice. His alternative submission was that the conviction was in any event unsafe once such breaches or the absence of such safeguards were taken into consideration.

In support of his primary submission, Mr Mansfield emphasised that the new evidence showed that the appellant, unknown to the jury or any other participant at his trial, was a far more vulnerable person in the context of his interviews than had been or could have been appreciated at the time and that, in a case which depended essentially on his confession, his conviction must therefore be regarded as unsafe. For the Crown, Mr Kelsey-Fry felt unable to proffer an argument to the contrary.

We agree. Although, as in King, the evidence of the appellant’s abnormally low IQ could have been established at trial, it is the combination of his borderline abnormality in terms of suggestibility and compliance and of his unforeseen abnormally low IQ which rendered him particularly vulnerable to interrogation; and this would be so irrespective of the appellant’s allegations concerning violence and the threats of violence, inducements, and the refusal of access to a solicitor. It is not possible to regard the conviction as safe when the essential issue, indeed the sole issue on which the jury were asked to judge the appellant’s case was whether his confession was voluntary and true or not. In so judging him, the jury would have to rely on the appellant’s own evidence from the witness box without the advantage of the new expert evidence as to his vulnerability when alone in the police station during lengthy interviews. If in those circumstances the jury might have found it difficult to believe how the appellant could confess to a crime he did not commit, they would not have found it any easier to accept, or to be rendered doubtful about, his more specific allegations as to how he was treated during his interrogation. If, however, they had had the advantage of hearing the new evidence, it cannot be postulated that they would not have looked at the whole of the appellant’s evidence in a new light.

The judge, whose summing up was a model of fairness, was at a similar disadvantage. So were counsel at trial. Without appreciating how potentially vulnerable the appellant was, any of them was entitled to present the conflict of evidence between the police witnesses and the appellant in stark terms, reflected in the judge’s summing up, between on the one side an entirely straightforward interview process and on the other side a sinister conspiracy. Presented, however, with the new evidence, it becomes apparent that a third alternative becomes possible, which is that, without any abuse of the process, a highly vulnerable suspect was not recognised for what he was.

It is precisely because of the dangers of such vulnerability that certain safeguards have been built into the process. Of course, it has never been permissible to use violence, threats or inducements to extract a confession. In other respects, however, the rules have changed to reflect an increasing appreciation of what is necessary for safeguards against the risk of a miscarriage of justice. At the time of the interviews and trial in this case, the test for the admissibility and relevance of a confession was that it was voluntary, defined in note (e) to the Judges’ Rules as not being obtained “by fear of prejudice or hope of advantage…or by oppression”, see R v. Prager (1972) 56 Cr App R 151. Nowadays a new test found in section 76(2)(b) of PACE is that of reliability. As the note in Archbold, 2003, at para 15-268 points out, a confession might be inadmissible under section 76(2)(b) without any impropriety. At the time of the interviews in this case, under paragraph (c) to the preamble to Judges’ Rules “every person at any stage of an investigation should be able to communicate and to consult privately with a solicitor”: at that time, however, it appears to have been the view that there would be no breach unless the suspect had requested the services of a solicitor, and even in case of breach it would seem that there was uncertainty as to its importance: see Archbold, 40^th ed, 1979, at para 1390. Nowadays, however, the right of access to legal advice and to being apprised of that right is dealt with in detail in PACE and its Codes and its importance underlined (R v. Samuel [1988] QB 615 at 630). Whether the appellant had in fact asked for a solicitor was of course a point in issue in this case: but it is not something on which the new evidence bears; and it is not possible to say that there had been a breach of paragraph (c), although Mr Mansfield urged us to infer that there had, for instance from the speed with which Mr Taylor appeared once the interrogation had been concluded with the sixth interview.

In this connection, however, the fact that the appellant was what used to be called mentally retarded or mentally handicapped is of particular importance. Its importance was recognised under the Judges’ Rules, for Rule 4A read (see para 1391a of the 1979 Archbold):

“If it appears to a police officer that a person (whether a witness or a suspect) whom he intends to interview has a mental handicap which raises a doubt as to whether the person can understand the questions put to him, or which makes the person likely to be especially open to suggestion, the officer should take particular care in putting questions and accepting the reliability of answers. As far as practicable, and where recognised as such by the police, a mentally handicapped adult (whether suspected of crime or not) should be interviewed only in the presence of a parent or other person in whose care, custody or control he is, or of some person who is not a police officer (for example a social worker).”

Nowadays, there is a straightforward prohibition in the Codes on the interview of a mentally handicapped person in the absence of an appropriate adult, which goes beyond the qualified language of Rule 4A (“If it appears to a police officer…where recognised as such by the police”), and in addition section 77 of PACE requires the judge to warn the jury of a special need for caution before convicting an accused in reliance on his confession where the case against him depends wholly or substantially on that confession and the court is satisfied that he is mentally handicapped and the confession was not made in the presence of an independent person.

On behalf of the Crown, Mr Kelsey-Fry was himself inclined to concede, indeed to volunteer, that an inference could be drawn that in this case the police had realised that the appellant was not a normal suspect, but was vulnerable, and that this inference was to be drawn from the circumstances of DCS Hobson’s desire for the sixth interview following the statement. In the judge’s summing up, this interview was spoken of as prompted by the superintendent’s caution about a confession being made 18 months after the murder when (subject to the key ring) there was nothing to connect the suspect with the crime. In the application to this court in 1981, however, Lord Lane spoke, as has been noted above (at para 52), in terms which suggest that there was also concern on the superintendent’s part about the vulnerability of the appellant. On this basis, a significant breach of the Judges’ Rules could be shown, if that were necessary, and one that related to the new evidence.

In our judgment, however, the need to show a breach of the rules then in force, although a factor in King and no doubt not one to be disregarded, was not a condition precedent to the success of that appeal, nor to the success of this. Rather, the background of the rules then in force and, as Lord Bingham put it, of “circumstances which denied the defendant important safeguards later thought necessary to avoid the risk of a miscarriage of justice” is what gives to the new evidence its power to affect the safety of the conviction in question. It is because of concern for the vulnerability of the mentally handicapped, underlined in more modern provisions but already well demonstrated in Rule 4A, that the new evidence links directly to the safety of the appellant’s conviction.

This was a separate ground of appeal. In the light of our decision that the conviction must in any event be regarded as unsafe, it is unnecessary to make this ground determinative. However, we are minded to agree with Mr Mansfield’s submission and with Mr Kelsey-Smith’s concession that, in the face of DS Falconer’s evidence concerning what we have called the seventh interview, Mr Taylor’s evidence of what the appellant said to him at their first meeting was relevant and admissible in rebuttal. That evidence did not go to the truth of the appellant’s case that his confession was untrue, as the judge could have made clear to the jury, but to the dispute on the evidence of DS Falconer and the appellant as to what each had said to the other at the seventh interview. The Crown relied on DS Falconer’s evidence of what the appellant had said to him to support the inference that it was Mr Taylor who had caused the appellant to retract his confession. It seems to us strongly arguable that in such circumstances the defence was entitled to adduce the evidence of Mr Taylor that it was the appellant himself who told Mr Taylor that he wished to retract his confession. That is an issue independent of the truth of the confession.

We mention here one other matter into which it is unnecessary to go in detail. Following the reference of this appeal by the CCRC, the Crown instructed Mr Mark Bates, a specialist advisor to the Forensic Science Service, to conduct a review into the death of Carol Wilkinson to assess whether or not forensic science could make any contribution to the issue. His report is dated 9 September 2002. The answer is that forensic science could throw no new light on the subject. However, Mr Mansfield relied on two observations made by Mr Bates. The first related to areas of bloodstaining on the stones of the roadway and the wall. The appellant said in his confession that his contact with the victim in the roadway was an attempt to kiss her, which she rebuffed (the tiff) before climbing over the wall into the field on the other side. The second related to evidence before the court to the effect that the victim had been struck on the back of her head and that scratches on her face suggested that it had been in contact with a firm surface. Mr Bates commented:

“This would suggest that Carol was face down and struck with the boulder (rather than face up and struck against it) and does not concur with Anthony Steel’s confession in which he states he noticed she wasn’t moving and that her eyes were open suggesting she was face up.”

In conclusion, we regard the appellant’s conviction as unsafe and therefore allowed his appeal. Although the facts and legal circumstances of this case are not identical with those of King, we think that the essence of this court’s reasoning and approach in that case guides our conclusion in this.