Lee Firkins & Anor v R

On 5 November 2003 Graham Fisher and his wife Carol Fisher were brutally murdered in their home. Each was shot repeatedly with a shotgun and then beaten to death with a blunt instrument. On 19 January 2006 these appellants, who are brothers, were convicted of both murders. On 27 February 2006 they were each sentenced to life imprisonment with a minimum term of 26 years less the time they had spent on remand in custody awaiting their trial. Their applications for leave to appeal against their convictions were refused by the full court on 17 December 2008 (“the first appeal”). Their cases now come again before this court by way of a reference by the Criminal Cases Review Commission (“CCRC”), which takes effect as an appeal against conviction.

The grounds of appeal are focused upon the credibility of a prosecution witness who gave evidence at trial of confessions made to him by Robert Firkins whilst in custody on remand. The identity of the witness concerned has been known to the appellants and their representatives throughout the proceedings in this case, and he identified himself to the court by his true name when he gave evidence. However, in view of the high level of risk to him as a serving prisoner giving evidence for the prosecution, the trial judge was satisfied that it was in the interest of justice to prohibit any publication of the witness’ name. That order was upheld by this court in the first appeal. The witness was accordingly referred to as “Z”. We shall so refer to him in this judgment. We order, pursuant to section 11 of the Contempt of Court Act 1981, that no report of these proceedings may include the true name of Z, or any matter which might lead members of the public to identify Z.

We record at the outset our gratitude for the assistance we have received from the written and oral submissions of counsel including the additional submissions which the court allowed to be submitted after the conclusion of the hearing. We shall not in this judgment mention all the many points which were argued, and we shall summarise the competing arguments briefly, but we have considered them all.

A detailed summary of the facts is contained in the judgment of this court on the appeal heard in 2008 (“the first appeal”). That judgment is publicly available under Neutral Citation Number [2008] EWCA Crim 2981, and it is unnecessary to repeat all that it contains. For present purposes, it is sufficient to refer briefly to the following.

Mr and Mrs Fisher were proprietors of a garage near Wadebridge in Cornwall, and lived in an adjacent bungalow. They were murdered during the evening of 5 November 2003. Their bodies were found the following morning. Mr Fisher’s body was in the kitchen. He had been shot twice with a shotgun and then beaten to death by blows to his head with a blunt instrument, possibly a sledge hammer found at the scene. Mrs Fisher had been shot once in the living room, and then shot twice more as she tried to run away from the bungalow. She too was then battered to death with a heavy blunt instrument, and her blood was found on the handle of the sledge hammer.

Examination of the scene suggested that at least two persons had entered the bungalow. A safe was found open, but with £2,000 in cash still inside it. Drawers and cupboards had also been searched, but it was not known what, if anything, had been taken. There was evidence that someone had tried to dial 999 from the bungalow, but had misdialled.

On the evening of the murder, the appellants, who lived at that time in Weston-super- Mare, had visited relatives who lived in Cornwall about 12 miles from the Fishers’ home. Their case at trial was that they had spent the evening either at the homes of their relatives, or at a local public house the Grenville Arms, before returning to Weston-super-Mare. They denied that they had been anywhere near the scene of the murders.

At around that time, the appellants plainly possessed, or had access to, sawn-off shotguns: it was admitted that they had buried two such weapons, and ammunition, which were found on 29 December 2003 on the beach at Weston-super-Mare. One of those shotguns was of the gauge used in the murders.

The appellants also admitted that in late November and December 2003 they had committed a number of violent crimes, including in the south-west of England. On 29 November, Robert Firkins attacked three men in a public house near to the Cornwall home of the appellants’ relatives: two of the men were punched, and all three were injured three by the use of an ammonia spray. On 2 December Robert Firkins travelled from his home to Essex, where he committed offences of assault occasioning actual bodily harm and possession of a firearm at the time of committing that assault. On 18 December both appellants committed a robbery at a shop in Taunton. On 19 December 2003 Lee Firkins committed a robbery, in which a sawn-off shotgun was produced and fired, at a petrol station at Fraddon, not far from the Fishers’ home. On 20 December, both appellants seriously injured a man in the course of a minor dispute: their victim was slashed with a Stanley knife, sprayed in the eyes with ammonia, tied up and left on remote moorland.

The appellants were arrested on 23 December 2003 for the offences on 18, 19 and 20 December. They were remanded in custody.

Initially, both the appellants were classified as category A prisoners and were held in the same prison as each other. On 30 March 2004, however, Robert Firkins was downgraded to a category B prisoner, and was transferred to HMP Exeter. He there occupied, for a number of days, a cell near to that occupied by Z, who had recently been sentenced to a short term of imprisonment.

On 17 May 2004 Robert Firkins was arrested in connection with the murders. He was interviewed under caution on the following two days, but was not at that stage charged.

On 18 May 2004 the South Wales police arranged for Z to be taken from prison to a police station on 28 May, to be interviewed about offences of theft, deception and driving whilst disqualified.

On 26 May 2004 the television programme Crimewatch broadcast an appeal for witnesses in relation to the murders of Mr and Mrs Fisher. The programme included some details of the murders. A reward of £10,000 had been offered for information leading to arrests.

Two days later, Z was taken to the police station as arranged. In the course of the journey, he said that he had “information about the shooting”. Whilst at the police station he spoke to officers investigating the murders, said that Robert Firkins had confessed to the murders, and expressed his willingness to provide information. He was later returned to the prison.

When seen again on 3 June 2004, Z declined to cooperate with the police because he thought he was going to be refused bail; but having been bailed the next day he again agreed to cooperate. He was interviewed by the police on 5 June, and on 11 June 2004 he made a statement in which he said that he had become friendly with Robert Firkins whilst they were in custody together, and that Robert Firkins had confided in him. He alleged that Robert Firkins had made confessions of serious crime, including a detailed account of how the appellants had gone to rob Mr and Mrs Fisher and had murdered them. Z further alleged that Robert Firkins had later asked that, when released from his sentence, Z should pick up and dispose of something in Cornwall which was connected to the murders. His statement included the following assertion:

“The reason that I am helping the police with this is because I am now going the ‘Christian way’. There is no way I can help Rob with this and live the Christian life.”

The appellants were arrested for the murders on 12 July 2004 and charged the following day.

The trial in the Crown Court at Exeter, before Owen J and a jury, lasted for about 3 months. After a ruling by the judge, the prosecution did not seek to adduce evidence from a witness called Craig Mack who had made allegations implicating Lee Firkins. In very brief summary, the prosecution adduced –

On the evening when the murders were committed, the appellants had with them a mobile phone. It was used at various times before 6.50pm. There was then a period of inactivity until the phone received an incoming call at 8.46pm (“the 2046 call”) from Robert Firkins’ then girlfriend. She gave evidence that Robert Firkins was tearful and emotional during that call. The prosecution case, disputed by the defence, was that cell- siting evidence was consistent with Robert Firkins having received that call whilst travelling away from the scene of the murders.

The principal prosecution witness was Z. He gave evidence of the confessions which he alleged had been made to him by Robert Firkins whilst they were in custody together at HMP Exeter. If accepted by the jury as truthful, accurate and reliable, Z’s evidence provided an ample basis for the jury to be sure of Robert Firkins’ guilt. Robert Firkins denied making any confession to Z, and Z’s evidence at trial was disputed by both appellants. Z was cross-examined in detail, in particular by leading counsel then representing Robert Firkins. He put himself forward as a reformed character who wanted to do the right thing.

Although Z’s evidence related only to Robert Firkins, it also formed an important building-block in the prosecution case against Lee Firkins. As against Lee Firkins, the prosecution relied on the decision of the House of Lords in R v Hayter [2005] UKHL 6, [2005] 1 WLR 605. It was held in that case that in a joint trial, when the evidence against defendant A consisted solely of his own out of court confession, the jury could find A guilty on the basis of that confession and could then go on to find that the fact of A’s guilt, coupled with any other evidence there may be which incriminated the other defendant B, was sufficient to prove B’s guilt. In such circumstances, at the end of the prosecution evidence there would be a case for B to answer.

In the course of the trial the appellants’ counsel made a number of submissions and applications. At the conclusion of the evidence, and before the summing up, Mr Boyce

The judge in his summing up directed the jury that if they found Robert Firkins not guilty they must also find Lee Firkins not guilty, because –

“… the evidence against Lee Firkins, without using the conviction of Robert Firkins against him, is, I direct you, insufficient to find Lee Firkins guilty.”

In what we refer to for convenience as “the first appeal”, applications for leave to appeal against both conviction and sentence were considered by the full court. The appellants (at that stage, applicants) mounted a sustained challenge to the credibility of Z, relying both on the evidence given at trial and on evidence as to things said and done by Z after the trial. Detailed consideration was given to whether the Crimewatch programme to which we have referred could have been the source of some or all of the information about the murders which Z alleged he had been told by Robert Firkins. It was also submitted on behalf of Lee Firkins that evidence of his bad character had wrongly been admitted in order to bolster a prosecution case which was very weak because of the suggested unreliability of Z and the suggested absence of any significant support for Z’s evidence.

This court held that the assessment of Z’s credibility was properly left to the jury, and that other evidence provided some support for Z. In a passage from the judgment of the court which we must cite in full, Hughes LJ said:

“29.

All this demonstrates that detailed analysis of the content of the alleged confession by Robert could be capable of undermining ‘Z’ but might equally have contained some support for him. In the end there was no conclusive material either way. Thus there were proper questions for the jury to determine about it.

30.

Courts are properly cautious of alleged cell confessions. There is obvious scope for invention by a criminal in the hope of advantage. The need for a careful warning by the judge is clear: see Benedetto and Labrador [2003] UKPC 27. The judge gave such warnings in this case, and there is and could be no complaint about their terms. We are prepared to accept that there might arise a case in which the evidence of a cell witness like ‘Z’ was so destroyed that in the absence of any support for it the judge might be driven to the conclusion that no jury could safely convict upon it. More often, however, the evaluation of the witness ought to be left to the jury, with the warning of the need for caution. See for example Stone [2005] EWCA Crim 105. One extremely important consideration in any such case is whether there is any support for the evidence.

31.

This was not a case which depended entirely on ‘Z’. Whether or not the other evidence would have justified a conviction in [Z’s] absence, he certainly did not stand alone. His evidence was capable of being supported by:

(i)

the applicants’ propensity for robbery and for extreme violence, and for criminal behaviour of both kinds in Cornwall in the period of this offence;

(ii)

their possession at the time of sawn off shotguns;

(iii)

their admitted presence in this particular part of Cornwall on the night of the murders;

(iv)

the improbable reason given for their presence: the suggestion that they needed to make a round trip of just under 300 miles to buy a few pounds’ worth of cannabis when both had experience of drug use and lived a few miles from a major city was one which the jury might well conclude was palpably false;

(v)

the evidence that they were not, as they said they were, in the Grenville Arms for the middle part of the evening;

(vi)

the period of telephone silence, likely to be seen as significant because of the potential for a mobile telephone to demonstrate the area where the user is, even if no call is made out; moreover Robert’s girlfriend said that he told her that the telephone had been switched off;

(vii)

the evidence showing that the call at 2046 could well have been taken when travelling away from the scene, and could not have been taken by Robert where he said he took it;

(viii)

Robert’s girlfriend’s evidence that he was distressed and tearful in that call;

(ix)

the improbability of their evidence that, having gone all that way for cannabis, they had been told by their relative Wayne Vicary that he would see if he could get some, had then repaired to the public house to wait to see whether he did, but had then left Cornwall without finding out if he had succeeded and despite being in the area at least until some time after 10pm;

(x)

the evidence of Hawken that Robert had been heard to say, apparently to Lee, ‘what are we going to do about the Wadebridge thing?’;

(xi)

the evidence of Stephanie Best that Robert had spoken of Wadebridge;

(xii)

the evidence of Harper that Robert returned from being interviewed by the police about the murder elated at being told that no further action was then contemplated but distinctly worried about a telephone call and/or a gap in telephone calls;

(xiii)

the evidence of Jay that Robert had spoken of doing the shootings.

32.

There were no doubt arguments, some of substance, to be made about the reliability of several of these pieces of evidence. But those arguments were for the jury. If some or all of these aspects of the evidence were accepted, they meant that ‘Z’s’ evidence by no means stood alone. In those circumstances we are satisfied that the judge was right to leave the evaluation of the evidence to the jury.”

The court went on to consider a number of other grounds of appeal, including a submission that the convictions were unsafe because there was a “lurking doubt” about the appellants’ guilt. The submissions were rejected, and at [67] the court concluded:

“… all the grounds advanced by both applicants fail. Nor do we entertain any lurking doubt about the safety of these convictions. We cannot tell what impact the several witnesses made upon the jury, nor which evidence the jury accepted and which it rejected. But the decision was for the jury. We are quite satisfied that this case did not depend wholly upon ‘Z’. On the contrary, the other evidence in the case was potentially highly significant support. These applications are accordingly refused.”

In March 2010 Lee Firkins made an application to the CCRC, which did not lead to a referral. We need say no more about that. We focus upon the more recent referral, which has brought both appellants before this court.

The CCRC sought expert advice from, amongst others, a consultant forensic clinical psychologist, Professor Craig. Having regard to Z’s previous declining of all offers to

The CCRC considered that the new medical evidence, had it been available, would have significantly assisted the defence at trial. They considered the other evidence, and the thirteen points which this court identified in the first appeal as being capable of supporting Z’s evidence. At paragraphs 177 and 178 of their referral, the CCRC summarised that aspect of the case as follows:

“177.

… Thus, Witness Z’s evidence as it was presented at trial was capable of being supported by other evidence in the case and had to be assessed in this context.

178.

However, if the jury had heard the medical evidence and as a result rejected the evidence of Witness Z, there would be little left for the additional evidence listed by the Court of Appeal to corroborate.”

Having also considered, but not relied on, a number of other issues raised on behalf of the appellants, the CCRC referred their convictions, pursuant to s14(4A) of the Criminal Appeal Act 1995, on the grounds that –

We are grateful to the CCRC for the care with which they have considered this case.

Counsel have referred to a number of matters relating to Z which were not known to the court at the time of the first appeal, either because they had not been disclosed to the appellant’s representatives or because they relate to events which occurred after the first appeal. We do not think it necessary to go into the detail of all those submissions, but two matters must be mentioned.

In August 2003, whilst detained at HMP Cardiff, Z contacted a police officer who was known to him and reported that a fellow prisoner, charged with murder, had approached Z and asked Z to move an item on Z’s release from prison. The other prisoner told Z that the item was something the police were looking for and had not found, and which would be relevant to prove the prosecution case against the other prisoner. This fact was not known to the appellants’ representatives at the time of the trial or the time of the first appeal. It was disclosed to them by the respondent in a note sent in June 2023, pursuant to a direction given by this court at a directions hearing.

In [year] Z, who already had a number of convictions for offences of violence, and another man “D”, were convicted of an offence of murder, and two offences of attempted murder. [circumstances redacted]

At his trial in [year], Z blamed D and an unknown man for the offences, asserted that D had made a cell confession to him, and fabricated an elaborate explanation for the evidence against himself.

The jury convicted both Z and D. [further details redacted]

In addition to those developments, the parties have obtained medical reports which were not before the court in the first appeal, and which add to the expert evidence considered by the CCRC.

We turn to the present appeal.

The CCRC’s reasons have been adopted by the appellants as their first ground of appeal. They seek leave also to argue further grounds. We need not recite the precise terms of those further grounds. It is sufficient to record that, in summary, they contend that Z’s evidence of a cell confession by Robert Firkins, and his stated reasons for giving evidence about that confession, are substantially undermined by the evidence relating to Z’s offence of murder in [year], by the evidence relating to Z’s behaviour in connection with his trial on that matter, and by the evidence relating to Z’s behaviour in connection with his appeal against his convictions (a matter to which we return in paragraph 52 below). They further contend that there was a material failure of disclosure in relation to material relevant to Z’s credibility, and in particular in relation to material showing that in 2003 Z had made to the South Wales police an allegation very similar to that which he made against Robert Firkins: namely, that whilst in

The appellants also apply, pursuant to section 23 of the Criminal Appeal Act 1968, to adduce fresh evidence from a number of expert witnesses; the documents considered by those expert witnesses in preparing their reports; and material, now available to them but not disclosed or not in existence at the time of the trial, relating to Z’s conduct. That material can be summarised as relating to Z’s role as a police informant; his obtaining of a reward in 2009; his commission of murder in [year] and his conduct before and after his trial for murder, all of which it is submitted supports the appellants’ case at trial that Z was a witness who could not be believed.

The appellants submit with regard to all of this fresh evidence that it is capable of belief; that it affords grounds for allowing their appeals; that it would have been admissible at trial, or (in relation to material relating to events after the trial) would be admissible if Z were a witness now; and that there is a reasonable explanation for the failure to adduce it at trial, because it was either not available, had not been disclosed or could not reasonably have been obtained.

The respondent, rightly, does not oppose this application. We have therefore considered, de bene esse, all of the additional material and the following expert evidence:

The expert evidence can most conveniently and fairly be summarised by quoting the joint statements. We quote them in full, but it should be noted that in paragraphs 60 and 77 we refer to submissions of the respondent as to the admissibility of parts of the joint statements.

In their statement dated 17 May 2023, the three psychologists expressed their agreement as follows:

“1.

Z meets the diagnostic criteria for Antisocial Personality Disorder (ASPD), Dissocial Personality Disorder and psychopathy. Additionally, Z also fulfils the criteria for Substance Use Disorder.

2.

The severity of Z’s psychopathy is extreme in the top 1% of psychopaths.

3.

Whilst ASPD is relatively common among the prisoner cohort, relatively few prisoners (less than 10%) are psychopaths.

4.

When considering the very small group of prisoners who might meet the threshold for psychopathy, only a tiny minority would rate as highly on Hare’s Psychopathy Checklist (Revised) as Z.

5.

At the time of the Firkins’ trial in 2005, Z would have met the criteria for severe ASPD and also for clinically and statistically extreme psychopathy.

6.

Further, had Z undergone a forensic psychological examination at the time of the trial, an expert would have come to the view that he met the criteria for ASPD and psychopathy.

7.

Credibility is a subjective matter for a jury to decide upon. We agree that consistency is not necessarily an indication of truth- telling.

8.

The ultimate decisions as to reliability lie with any judges tasked with ruling on the admissibility of evidence and thereafter the jury. What we can provide by way of expert opinion is that, flowing from Z’s diagnoses and the severity of his conditions, Z’s mental disorder (ASPD, psychopathy and substance use disorder) in the context of his particular personality means that he is highly likely to provide unreliable testimony.

9.

Z’s primary objective is self-interest, and for him consistency would occur only if it served his self-interest. Where Z exhibits consistency with regard to his evidence this should not be viewed as an indicator of truthfulness but rather seen as reflecting his own self-interest.

10.

The judge and the jury at the Firkins’ trial were disadvantaged by not having expert evaluation of Z and the effect of his mental disorders and characteristics when considering the question of reliability.

11.

Z is a man for whom conning and manipulation are his modus operandi. Without expert evidence the jury were not advised as to the difficulty in discerning truths from untruths when spoken by Z.

12.

We have carefully considered the content of all the reports in this case and there is nothing that we disagree with save for the clarification that Z is a ‘pathological liar’ and not a ‘compulsive liar’. The two terms which denote different types of lying may have been used interchangeably by Dr Latham in his report for reasons of brevity. Further, as stated above we caveat Dr Latham’s use of the terms ‘credibility’ and ‘reliability’ by stating that during the trial process in our view a consideration of credibility is a subjective assessment on the part of the jury.

13.

In each of the expert reports there are very helpful analyses of how Z’s offending history, his behaviour as documented or observed in the assessments in this appeal is indicative of or informed by his diagnoses. We commend these passages t the court.”

In an addendum dated 13 June 2023 the psychologists further agreed on the following: “1. It is entirely a matter for the court to decide upon the ultimate

In their statement dated 24 May 2023, the two psychiatrists expressed their agreement as follows:

“1.

We agree that the diagnosis of antisocial personality disorder applies to Z.

2.

We agree that Z has a high degree of psychopathy.

3.

We agree that available documentation and the clinical formulation indicates Z is not a reliable person and is prone to lie at times.

4.

We agree that Z would not have been reliable at the time he gave evidence in the trial of R v Firkins but the question, ultimately, as to whether he told the truth is a matter for a jury.

5.

We agree that credibility does contain a subjective element and is ultimately for a judge or jury but his reliability and proneness to lying would inform the assessment of credibility.

6.

We agree that Z’s motivation for giving a false statement is complex and that certainty about this is not possible.

6.1

Dr Latham’s opinion however is that he was not motivated by telling the truth and more likely to have been motivated by some self-serving reason. As above, Dr Cumming cannot be confident about the motivation. This difference in opinion reflects the only area of disagreement.

7.

We agree that the decision-making around whether he was used as a witness was disadvantaged by not knowing about the degree of his likely unreliability at the time of the trial.

8.

We agree that it is reasonable to conclude that the jury would have been at a disadvantage when evaluating the evidence of Z because they had no knowledge of this clinical information and the impact on his credibility.”

In summary, the appellants submit that their convictions are unsafe on the following grounds. First, that if the fresh evidence had been available at trial, Z’s evidence would not have been adduced at all; or would have been excluded by the trial judge; or would have enabled the appellants to make a successful submission of no case to answer; or would have had to be considered by the jury in the light of expert evidence showing cogent reasons why Z should be considered unreliable. Secondly, that evidence relating to Z’s trial in [year] and his subsequent appeal substantially undermines Z’s evidence at the trial of these appellants, and his professed reasons for giving evidence for the prosecution. Thirdly, that the material which has now been considered by the proposed expert witnesses could in itself have been deployed to undermine Z’s credibility.

It was not suggested on behalf of the appellants that the jury would have to have been directed, in the light of the fresh evidence, that they must disregard all of Z’s evidence.

It is convenient to mention first the submissions on behalf of Robert Firkins. Ms Elliott KC and Ms Arshad submit that there is a reasonable explanation for the failure to adduce at trial the expert evidence on which they seek to rely: namely, that the opinions of the expert witnesses are based on prison and medical records which have been disclosed since the trial, and on Z’s conduct after the trial. They submit that the evidence is of a different order from that which was available at the time of the trial, and cannot simply be regarded as “more of the same”. They submit that the evidence shows Z to be a man who is indifferent to the truth and is motivated only by personal gain. They argue that whereas consistency may be relevant in assessing the credibility of someone who is not a psychopath, it cannot be relied on as an indication of truthfulness in the case of a psychopath. They point out that all five experts agree on the severity of Z’s disorder and on the likely effect on his reliability as a witness, those being matters of expert opinion which should be taken into account by a jury deciding whether to accept Z’s evidence.

Counsel point out that the prosecution had intended to adduce evidence from two other men, “T” and “W”, who claimed that Robert Firkins had made confessions to them; but both were abandoned when evidence came to light of their unreliability. It is submitted that if the evidence now available as to Z’s psychopathy had been available at trial, the prosecution would not have been able to put Z forward as a witness of truth. Alternatively, it is submitted that the fresh evidence would have enabled a successful application to exclude Z’s evidence. In the further alternative it is submitted that if Z’s evidence had been adduced, the jury would have had to assess it in the light of the expert evidence and may well have come to a different conclusion.

As to the 13 points identified by this court when giving judgment in the first appeal, counsel submit that these could only ever provide support for Z’s evidence and could not found a case to answer in the absence of Z’s evidence.

The evidence relating to Z’s conviction for the [year] murder is said to be important for a number of reasons. First, the fact that Z was paid to kill someone supports the expert evidence that he is motivated by the fulfilment of his own needs, and undermines his evidence at the appellant’s trial that he was disgusted by the murder of Mr and Mrs

On behalf of Lee Firkins, Mr Wood KC and Mr Lyons advance similar arguments about the effect of the fresh expert evidence, and go on to submit that, if that evidence had been available at trial, it could have enabled both appellants to make a successful submission of no case to answer. It is submitted that the expert evidence showing Z’s lying disposition and lack of credibility is much stronger now than the evidence available to the defence at the time of the trial and the first appeal, and also stronger than the evidence considered by the CCRC when it made the referral. Counsel emphasise that all the expert witnesses give evidence to the same effect, that Z is a pathological liar and that it is dangerous to rely on his evidence. They also emphasise the high level of psychopathy affecting Z, making his disorder a very significant deviation from the norm.

Counsel made further submissions challenging the 13 points identified by this court in its judgment in the first appeal, in particular arguing that no safe inference could be drawn against the appellants from the period of mobile phone inactivity referred to in point (vi).

As to the recent disclosure concerning Z’s allegations about a fellow-prisoner in the summer of 2003, counsel submit that there is an extraordinary level of coincidence between that account and Z’s evidence against Robert Firkins.

Finally, counsel add that the evidence now available would also support the submission, made unsuccessfully at the trial, that the evidence of Lee Firkins’ bad character should not have been admitted because it would unfairly bolster a weak prosecution case.

Mr Boyce KC and Miss Robinson made clear that, even if all the fresh evidence were available, the respondent would still have thought it proper to call Z and to put him forward as a witness of truth. They do not accept that there has been any material failure of disclosure. In particular, Mr Boyce tells us that no psychiatric or psychological evaluation of Z was available to the prosecution either at the time of the trial, or at the time of the first appeal, which suggested that Z was not or may not have been a reliable witness. He notes that when recently seen by Dr Cumming and Dr Latham, Z told each

The respondent relies on the analysis by this court in the first appeal of the evidence against the appellants other than that of Z. It is submitted not only that the evidence supported the evidence of Z, but also that it was sufficient in itself for the jury to convict both appellants, even without the evidence of Z. It is further submitted that the 13 points identified in the first appeal were matters for the jury’s consideration, and collectively meant that the appellants’ bad character could not properly be said to be bolstering a weak prosecution case. Mr Boyce points out that none of the other evidence is any less cogent now than it was at the time of the trial.

Mr Boyce notes that there was, and is, no suggestion that Z’s evidence was not admissible, and no criticism of the judge’s directions of law. Z had been tested in cross- examination by leading counsel then representing Robert Firkins, so that Z’s previous convictions, lying on oath, lying to the police and manipulative character were all before the jury. The jury therefore had to, and did, consider Z’s bad character, possible motivation by self-interest and potential unreliability. The jury’s attention was drawn to inconsistencies between the account given by Z and other evidence, and to the possibility that the Crimewatch programme could have been the source of much of his account. It was put to Z in cross-examination that his evidence was motivated by his desire to obtain the reward mentioned on the Crimewatch programme. It was accordingly clear to the jury, Mr Boyce submits, that they must approach Z’s evidence with caution, and must consider whether Z may be giving false evidence in order to further his own interests.

Mr Boyce submitted that in those circumstances, the issue for this court is whether the admissible fresh evidence is of such a different order from that which was considered by the jury at trial that it renders the convictions unsafe. He submitted that the question should be answered in the negative. He does not seek to dispute the agreed views of the experts, though he points out that some parts of their joint statements go beyond the legitimate boundaries of their role as defined in Pora v R [2015] UKPC 9 (“Pora”). His core argument is that, despite all the expert witnesses have said, Z’s evidence could properly be considered credible and reliable, particularly when it was strongly supported by other circumstantial evidence.

It is submitted that there is a strong public interest in finality of proceedings, and that it will only be in exceptional circumstances that justice requires some flexibility in the application of that rule. Mr Boyce argues that no such exceptional circumstances arise in this case, in particular because there is no direct and compelling nexus between the circumstances of the trial and the circumstances shown by the evidence of post-trial conduct.

We turn to consider some of the authorities cited to us.

In R v Williams and Smith [1995] 1 Cr App R 74, and again in R v Islam [2007] EWCA Crim 1089, it was held that this court may consider evidence of events post-trial which are relevant to the credibility of a witness.

Counsel have invited our attention to a number of decisions concerned with the admissibility of expert evidence to the effect that a defendant suffered from a severe personality disorder which might render a confession unreliable. In R v Fell [2001] EWCA Crim 696 this court said that for such expert evidence to be admissible, it must demonstrate something “well outside the norm” and must be supported by something in the history of the person concerned. In R v Pinfold, R v MacKenney [2003] EWCA Crim 3643, [2004] 2 Cr. App. R. 5 (“Pinfold”), the CCRC referred convictions to this court in circumstances where the prosecution case against the appellant had depended on evidence given by a former co-accused, and expert evidence had later been obtained to the effect that the witness’ personality characteristics were such as necessarily cast some doubt on his reliability. The court reviewed the earlier case law, stated that the approach to admissibility was the same whether the expert evidence related to a defendant or to a witness, and held that the absence of an examination of the person concerned by the expert witness was not in itself decisive in determining admissibility. At [16] Lord Woolf CJ, giving the judgment of the court, continued as follows:

“The court has to determine whether the evidence could be considered credible evidence by the jury as to an abnormality from which the witness suffered at the time of giving evidence and which might mean that the jury would not attach the weight it otherwise would do to the witness’ evidence. The absence of an examination by the expert goes to the weight to be attached to the expert’s opinion and not to the admissibility of that opinion. What a court must be on its guard against is any attempt to detract from the jury’s task of finding for themselves what evidence to believe. The court should therefore not allow evidence to be placed before a jury which does not allege any medical abnormality as the basis for the evidence of a witness being approached with particular caution by the jury. Ultimately, it remains the jury’s task to decide for themselves whether they believe a witness’ testimony.”

The court went on to consider the limits of the expert evidence which might be given when an abnormal disorder was said to render a witness unreliable. At [14], Lord Woolf CJ approved what had been said in R v O’Brien (unreported, January 25, 2000):

“First, the abnormal disorder must not only be of the type which might render a confession or evidence unreliable, there must also be a very significant deviation from the norm shown. … Second, there should be a history predating the making of the admissions or the giving of evidence which is not based solely on a history given by the subject, which points to or explains the abnormality or abnormalities.”

In Benedetto v R, Labrador v R [2003] UKHL 27, [2003] 1 WLR 1545 the Privy Council reflected on the inherent unreliability of evidence by a prison informer of a confession by another prisoner, and the consequent need for a trial judge to be alert to the possibility that the evidence is tainted by an improper motive and to direct the jury to be cautious before accepting the evidence. As we have noted, it is accepted that the judge in this case gave appropriate directions in that regard in relation to the evidence given at trial.

In R (Nunn) v Chief Constable of Suffolk Police [2014] UKSC 37, a case concerned with the extent of the police duty of disclosure after conviction, Lord Hughes JSC (with whom the other Justices agreed) said at [32]:

“The position of a convicted defendant is different in kind from that of a defendant on trial. The latter is presumed innocent until he is proved guilty, as he may never be. The former has been proved guilty. He is presumed guilty, not innocent, unless and until it be demonstrated not necessarily that he is innocent, but that his conviction is unsafe. The defendant on trial must have the right to defend himself in any proper way he wishes, and to make full answer to the charge. The convicted defendant has had this opportunity. The public interest until conviction is in the trial process being as full and fair as it properly can be made to be. After conviction, there is of course an important public interest in exposing any flaw in the conviction which renders it unsafe and in quashing any unsafe conviction, but there is also a powerful public interest in finality of proceedings. All concerned, including witnesses, complainants, the relatives of the deceased and others, have a legitimate interest in knowing that the legal process is at an end, unless there be demonstrated to be good reason for re-opening it.”

In Pora the Privy Council considered the admissibility of fresh evidence from three expert witnesses relating to the reliability or otherwise of confessions made by the appellant. The Board made important observations about the role of expert witnesses which are conveniently summarised as follows in [H5] of the headnote in the report at [2016] 1 Cr App R 3:

“Held … that it was the duty of an expert witness to provide material on which a court could form its own conclusions on relevant issues. On occasions that might involve the witness expressing an opinion about whether, for instance, and individual suffered from a particular condition or vulnerability. The expert witness should be careful to recognise, however, the need to avoid supplanting the court’s role as the ultimate decision-maker on matters that were central to the outcome of the case. The third expert witness had trenchantly asserted that the defendant’s confessions were unreliable and he had advanced a theory as to why he confessed. That went beyond his role. It was for the court to decide if the confessions were reliable and to reach conclusions on any reasons for their possible falsity. It would be open to the expert to give evidence of his opinion as to why, by reason of his psychological assessment of the defendant, the defendant might be disposed to make an unreliable confession but it was not open to him to assert that the confession was in fact unreliable. A report containing such statements could not be admitted as an item of fresh evidence.”

At [27] of the judgment, the Board said this:

“The dangers inherent in an expert expressing an opinion as an unalterable truth are obvious. This is particularly so where the opinion is on a matter which is central to the decision to be taken by a jury. There may be cases where it is essential for the expert to given an opinion on such a matter, but this is not one of them. It appears to the Board that, in general, an expert should only be called upon to express an opinion on ‘the ultimate issue’ where that is necessary in order that his evidence provide substantial help to the trier of fact. As observed above, Professor Gudjonsson could have expressed an opinion as to how the difficulties that Pora faced might have led him to make false confessions. This would have allowed the fact finder to make its own determination as to whether the admissions could be relied on as a basis for a finding of guilt, unencumbered by a forthright assertion from the expert that the confessions were unreliable. In this way it would be possible to keep faith with and preserve the essential independence of the jury’s role, which is to evaluate all the relevant evidence, including both expert evidence and other evidence which the expert may have no special qualification to evaluate.”

We have reflected on the submissions and on the issues raised. Our conclusions are as follows.

We think it appropriate to focus on the proposed fresh expert evidence.

Applying the Pinfold test, we accept the appellants’ submission that the expert evidence would be admissible at a trial. If accepted, it shows that Z suffers from a high level of psychopathy, putting him very far outside the norm, and that features of his personality disorders make it necessary to exercise particular caution before accepting his evidence on any matter.

It is for a jury to evaluate the credibility of a witness, either generally or in relation to a particular matter, and to decide whether his evidence about that matter is truthful, accurate and reliable. It is, of course, open to a jury to be sure that a witness is reliable and accurate about a particular matter even though they are unsure of his credibility and reliability in relation to other matters, or even if they are sure he has given untruthful evidence about some other matters.

A jury will not usually need any expert evidence to assess aspects of human behaviour and motivation with which everyone is familiar: for example, the risk that evidence may be unreliable because it is motivated by a hope of advantage or reward or by antipathy towards another person. Similarly, expert evidence will not usually be necessary, and therefore will not usually be admissible, to identify the general reasons why evidence of cell confessions may be unreliable and must be approached with care. Expert psychiatric or psychological evidence therefore has only a limited role to play in relation to issues of credibility. As to the boundaries of that role, we respectfully

Thus expert psychiatric or psychological evidence is in principle admissible if it is necessary to explain that a witness suffers from a disorder or abnormality which may cause him to give untruthful and unreliable evidence. As Pinfold confirms, the disorder or abnormality must be such as to set the witness well outside the norm, and must be supported by some feature of the witness’ history.

Even when those criteria are met, an expert witness cannot in our view be permitted to give evidence which amounts to telling, or advising, the jury whether or not they should believe a witness. That is so, in our view, even though case law recognises that in some areas of expertise an expert witness may be permitted to give his or her opinion on “the ultimate question”. In the present context – namely, expert evidence relevant to the assessment of the credibility of a witness and of the reliability of evidence or of a confession – the role of the expert is limited to explaining reasons, which are by their nature outside the knowledge and experience of most persons, why the witness may be more likely, or much more likely, than others to give untruthful and unreliable evidence. As was said in Pinfold (in the passage cited at paragraph 64 above), it remains the jury’s task to decide for themselves whether they believe a witness’ testimony. Although we do not need to decide the point, we think it unlikely that there will be any circumstances in which an expert witness could properly opine that a witness was wholly incapable of giving a truthful answer to any question. Within the limits we have indicated, however, the expert witness may properly give his or her professional opinion as to the nature of, the reasons for, and the extent of the risk that, because of the relevant medical factors, the witness may give untruthful or unreliable evidence.

In the present case, we accept Mr Boyce’s submission that some parts of the joint statements which we have quoted at paragraphs 44 to 46 above go beyond the proper boundaries of the experts’ role in relation to issues of credibility. In this regard, we broadly agree with the approach taken in the annotated copies of the joint statements which the respondent had prepared, marking those passages which it was submitted were inadmissible.

Even with those passages removed, however, what remains in the joint statements provides clear evidence that this is an exceptional case: Z suffers from severe psychopathy, rendering him well outside the norm even for the comparatively small number of prisoners who are psychopaths; the particular features of his disorder include a concern solely for his own advantage and an ability to make false statements without compunction or embarrassment; and the nature of his disorder is such that some features of any evidence he might give, such as consistency of account, may not be as safe a guide to whether he is telling the truth as they might be with other witnesses. As was said in one of the reports which we have read, that of Professor Craig dated 27 February 2023 –

“… research indicates that people with high levels of psychopathic traits are better at learning to lie than individuals who show few psychopathic traits, as they are not burdened by social or emotional conventions or reactions associated with honesty.”

We accept that expert evidence is necessary to assist a jury with such matters, which are likely to be outside their knowledge and experience. We accept that the evidence meets the considerations set out in section 23 of the Criminal Appeal Act 1968. Accordingly, we think it right to admit the fresh expert evidence.

The other material which is the subject of the fresh evidence applications is, in effect, material relevant to the opinions of the experts and to the assessment a jury might make of Z in the light of the expert evidence. We bear very much in mind the strong public interest in finality of proceedings. Again, however, the material points to this being an exceptional case. Most strikingly, it shows that Z, having at trial disavowed any interest in the £10,000 reward in relation to the murders of Mr and Mrs Fisher, has not only claimed that reward but has shown himself quite prepared to commit murder [circumstances redacted]. The mere claiming of a reward to which a victim or a witness is entitled will generally not be a compelling reason to doubt his or her testimony; but the fact that a witness was willing to murder [redacted] and doing so, puts the case into a very different category. All of the material now available has to be seen in the context of the fresh expert evidence relevant to a jury’s assessment of Z’s motivation for giving evidence. We therefore think it right to admit the non-expert fresh evidence.

In those circumstances we grant both appellants leave to argue their additional grounds of appeal.

We should add that issues relating to disclosure were continuing at the end of the hearing. The court therefore gave directions and permitted further written submissions. The resultant disclosure revealed, very belatedly, material showing that from February 2006 onwards – very shortly after the convictions of the appellants – Z was pursuing the payment to him of the reward; that Z was very dissatisfied when, after the first appeal, he initially received only £5,000; and that Z thereafter pressed successfully for payment of the full £10,000 referred to on the Crimewatch programme. Such material is obviously relevant to Z’s assertion at trial that he was not interested in the reward, and to a consideration of the prosecution submission to the jury at the trial that Z had nothing to gain by giving evidence, or that any benefit he might gain would be far outweighed by the disadvantages to him. Without making any findings as to how or why this regrettable situation has arisen, the very late disclosure of this further material is another indication that a much fuller picture would now be available to a jury than was considered by the jury at trial.

True it is that those representing the appellants at trial had a great deal of material with which to challenge Z’s credibility and to cast doubt upon his reliability. We recognise that in some cases proposed fresh evidence adds comparatively little to what was available at trial, and can fairly be described, in the convenient shorthand phrase used by counsel, as simply being “more of the same”. In our view, however, this case is exceptional, and such a description is inapt. The expert evidence, if accepted by the jury, identifies Z as one of a very small number of persons suffering from severe psychopathy, with consequences for the reliability of his evidence which a jury could only properly assess with the assistance of expert evidence. It is therefore of a different

We have considered the other evidence which was relied on by the respondent at trial. At paragraph 25 above, we have quoted passages from the judgment in the first appeal, in which the court identified points which were capable of providing support for Z’s testimony and which meant that Z’s evidence did not stand alone. But the court, as is clear from what was said in the judgment at [31], did not find that the other evidence would in itself be sufficient to justify a conviction: rather, the court considered the supportive effect of the other evidence “whether or not [it] would have justified a conviction in [Z’s] absence”. We respectfully agree with, and adopt, that approach. For present purposes, the important point is that the weight to be given to the other evidence was not considered by the jury with the benefit of expert assistance as to the features of Z’s psychopathy and the potential consequences of his disorder for the reliability of his testimony.

In addition, the fresh expert evidence sets the context in which a jury, if all the evidence now available were before it, would have to assess other material relating to Z’s conduct before and after the trial.

For those reasons, we are satisfied that in the very unusual circumstances of this case the public interest in finality of proceedings is outweighed by the public interest in quashing any convictions which are unsafe. In the light of all we have read, and the submissions made to us, we are satisfied that the convictions of both appellants are unsafe.

Accordingly, we grant leave to the appellants to argue those grounds of appeal for which leave is required; we allow their appeals; and we quash their convictions.

Following circulation to counsel of a draft copy of this judgment, we invited written submissions as to any consequential applications. We are grateful to counsel for their responses. We are satisfied that no further or oral submissions are needed. Having considered the arguments on each side, we are satisfied that – notwithstanding the practical difficulties which will arise as a result of the passage of many years since the trial – it is in the interests of justice that both appellants be retried on the two charges of murder. We shall give directions accordingly.

Finally, we are satisfied that there is a substantial risk of prejudice to the administration of justice in the retrial if the submissions made to us and/or the contents of this judgment are reported before or during the retrial. We therefore order, pursuant to section 4(2) of the Contempt of Court Act 1981, that the publication of any of the submissions made during these appeal proceedings, and of this judgment, be postponed until the conclusion of the retrial or further order of this Court. We direct that the respondent must forthwith notify the Criminal Appeal Office when the retrial has been concluded.