THE QUEEN v PHILIPPE SOSSONGO

The appellant’s convictions for murder and two linked offences of violence have been referred to the Court by the Criminal Cases Review Commission (“CCRC”). He is now 19 years old. The grounds for referral are that since his conviction in June 2017 fresh evidence has come to light to show that at the time of the offence, and at the time of his trial, he suffered from undiagnosed Autism Spectrum Disorder (ASD) and Attention Deficit Hyperactivity Disorder (ADHD) which, had they been known about at the time, might well have affected the jury’s assessment of the case and therefore render his convictions unsafe.

The murder and other offences took place on 18 November 2016. The appellant was then 14 years old (born 22 July 2002). He was convicted on 19 June 2017 after a trial before HHJ Leonard QC and a jury. The appellant was still then 14 years old.

Arising from the same violent incident as the murder (count 1), the appellant was also convicted of wounding with intent (count 2) in respect of a second victim and attempted wounding with intent (count 3) in respect of a third victim. His co-accused Vanquer Muanza was convicted on the same three counts. The other co-accused, M, was acquitted on the same three counts. The appellant was convicted as a secondary party. It was not suggested that he had taken any physical part in the three stabbings.

On 24 July 2017 the appellant was sentenced for the offence of murder to be detained at Her Majesty’s pleasure with a minimum term of 14 years less the days he had spent remanded in custody awaiting trial. No separate penalty was imposed on the other two counts, their criminality being reflected in the minimum term for murder.

The appellant was refused leave to appeal against conviction by the single judge and on 23 May 2018 his renewed application for leave was refused by the Full Court: [2018] EWCA Crim 2168.

It was only when the appellant was subsequently assessed in July 2018 by a clinical psychologist at the institution where he was serving his sentence that the diagnosis of ASD was made. A fuller report was obtained in September 2019 from Dr Arthur Anderson, a consultant clinical and neuropsychologist. He concluded that the appellant had suffered for many years from mild ASD and from ADHD.

In the light of the important discovery of these hitherto undiagnosed conditions, leave to appeal out of time against sentence was sought and granted. The appellant’s appeal against sentence was one of three otherwise unconnected cases heard together by the Court in December 2019 which raised issues about the proper approach to sentencing offenders suffering from autism or other mental health conditions or disorders: R v PS; Abdid Dhir; CF [2019] EWCA Crim 2286; [2020] 2 Cr App R (S) 9. The Court’s guidance foreshadowed the Sentencing Council Guideline “Overarching principles: Sentencing offenders with mental disorders, developmental disorders, or neurological impairments” which came into force on 1^st October 2020.

In allowing the appeal against sentence, the Court was satisfied that Dr Anderson’s report met the criteria in section 23 of the Criminal Appeal Act 1968 and should be admitted as fresh evidence. The Court accepted the submission that the diagnosis of ASD and ADHD was relevant to the assessment of the appellant’s culpability. The conditions significantly reduced his culpability in the joint enterprise. For this and other reasons not presently relevant, the Court reduced his minimum term from 14 years to 10 years.

Because the appellant was still then only 17 years old, the reporting restrictions imposed by the trial judge pursuant to section 45 of the Youth Justice and Criminal Evidence Act 1999 remained in force until the he attained the age of 18. As he is now 19 years of age those statutory restrictions no longer apply.

At the hearing of the sentence appeal in December 2019 the Court was informed that the matters raised in Dr Anderson’s report were also relied on in support of a pending application to the CCRC. Following a very thorough investigation, including the obtaining of a further report from Dr Anderson, the CCRC referred the conviction for murder back to the Court on 26^th February 2021.

In short, the grounds of appeal advanced by Mr Emanuel QC and Mr Solley on behalf of the appellant are that, in the light of the fresh evidence in Dr Anderson’s reports the convictions are unsafe because the jury were unaware that he suffered from any mental disorder and were unaware of the effects they would have had on his behaviour at the time of the stabbings and on his presentation at trial. It is contended:

On behalf of the Crown, Mr Orchard QC accepts that the Court should receive the fresh evidence of Dr Anderson, which is not challenged. Indeed, Dr Anderson’s conclusions are confirmed and supported in a report from Dr Ian Cumming, a consultant forensic psychiatrist instructed by the Crown. In short Mr Orchard submits, however, that evidence of the appellant’s ASD and ADHD, had it been known to the jury, could have made no difference to their verdicts. He submits that the convictions are safe.

The prosecution case was that the appellant was affiliated to a gang in Hendon. He denied this. His co-accused, Muanza, was affiliated to the gang, as was another young man Ahmed Cabdulagdir, who fled after the stabbings and was never arrested. The other co-accused, M, was not affiliated to the gang. The appellant’s older brother (aged 18) was also affiliated to the gang. He and the appellant shared a bedroom. Four knives were found in the bedroom. The prosecution relied on this as bad character evidence against the appellant.

On 18 November 2016 the appellant travelled with Muanza, Cabdulagdir and M by taxi to Harrow, a distance of about seven miles. He was the youngest of the four. The prosecution case was that the purpose of the journey was to find and attack members of a rival gang.

To demonstrate that this was the purpose the prosecution relied on evidence of three previous incidents of violence between the gangs at which the appellant had been present. The first was on 13 June 2015 when another member of the gang was stabbed. The appellant, then aged 12, had been present on that occasion as one of a group of nine schoolboys. The co-accused Muanza had also been present. The appellant admitted swinging a bag at another boy on that occasion but only in self-defence.

The second incident took place on 30 August 2016 outside a McDonalds in Harrow. Muanza, Cabdulagdir and the appellant were present. There was a confrontation in which Cabdulagdir was stabbed. Muanza was involved in the violence. The appellant was not.

The third incident took place on 16 November 2016, two days before the murder. Another young man who was a friend of the appellant and Muanza and also affiliated to their gang, was attacked and found unconscious by police officers having sustained serious head injuries.

The prosecution alleged that the murder and the other stabbings were in retaliation for the attack two days earlier.

At about 4.30pm on Friday 18 November 2016 a taxi was booked from the co-accused M’s phone. It was booked as a return trip to Harrow. The taxi picked up Muanza, Cabdulagdir, M and the appellant. The taxi stopped first in Harrow town centre. The prosecution said the appellant remained in the taxi whilst the other three went looking for a victim. The appellant eventually got out of the taxi and joined the other three when they returned to the taxi a few minutes later. The prosecution alleged that the appellant was keeping a lookout and “minding” the taxi, ensuring an escape route.

The taxi drove on to Eastcote Road in South Harrow. The prosecution case was that within seconds of the taxi arriving the other three got out and were scouting the area for potential victims. Again, the appellant remained in the taxi. The other three separated and walked in a pincer movement. They crossed into Parkfield Road and went out of sight of CCTV cameras for 25 seconds, during which time the victims in counts 2 and 3 were stabbed. The third victim, Hussein Ahmed, was fatally stabbed in the back (count 1). CCTV showed Muanza with a knife in his hand shortly before the fatal stabbing. It also showed Muanza and M each wearing a latex glove.

The appellant had got out of the taxi shortly after the others, telling the taxi driver he wanted to urinate. He then remained some distance away outside a shop called “Best Foods”. CCTV confirmed that the appellant was standing outside Best Foods some 70 metres away from the violence that was taking place out of his view. Cabdulagdir ran towards him. The appellant ran back with him to the taxi. The other two joined them and the taxi took them all back to Hendon.

The prosecution case was that the appellant had again been keeping a lookout and minding the taxi to ensure an escape route. He knew violence was to happen. There was also some suggestion that the appellant was minding a bag belonging to Muanza which was left in the taxi, the implication being that the knife (or knives) and latex gloves may well have been carried to the scene in this bag.

The appellant’s case was that he was not present when any of the victims were attacked. He denied acting as a lookout or minding the taxi to effect any escape. He was simply standing outside Best Foods. At no stage was he in possession of a knife or any weapon. He did not know that any of those travelling in the taxi with him were in possession of weapons or surgical gloves. He denied being in the gang.

He admitted being present during the incident at McDonalds on 30 August 2016 when Cabdulagdir was stabbed. He was aware there had been an argument which turned into a fight but denied being involved. He denied that he went with the others in the taxi seeking revenge. He had no idea that the taxi journey had anything to do with that incident. There was no discussion or plan to get into a fight or commit acts of violence. At some point during the journey Cabdulagdir said he wanted to “squash some beef” but did not elaborate. The appellant thought this meant resolve a problem by discussion.

When they arrived at South Harrow he told the others he did not want to be involved in squashing any beef as it was not his problem. He remained in the taxi for a short time when the others left. He got out initially intending to urinate. He walked to the corner to see where the others had gone and stopped outside Best Foods. He was standing there when he saw Cabdulagdir running towards him. He denied seeing anyone with weapons, seeing any gloves at any stage in the cab, or knowing that anyone had been stabbed. He discovered this only when he was interviewed by the police three days later.

The appellant was represented at trial by Queen’s Counsel and junior counsel. After the trial they confirmed in correspondence that they had no inkling that the appellant suffered from ASD or ADHD. By contrast the co-accused M, who was a year older than the appellant, had been diagnosed with ADHD and traits of ASD. For that reason, an intermediary had been appointed to assist him. The intermediary sat alongside M in the dock, between M and the other two defendants (Muanza and the appellant).

Much of the background evidence was the subject of narrative formal admissions, including the previous incidents of violence on 13 June 2015, 30 August 2016 and 16 November 2016. Very little if any of the live evidence implicated the appellant. In the course of oral submissions counsel confirmed our assumption that there was very little cross-examination on behalf of the appellant of any of the prosecution witnesses.

In the case of the appellant, therefore, the key evidence was his own testimony. The credibility of his denials was the central issue. The jury had to assess his credibility as a witness and the plausibility of the explanations he gave for his presence at the scene and his understanding (or lack of it) of what was afoot.

Muanza, who was convicted of murder as the stabber of the deceased, did not give evidence, although his case was that he was acting in lawful self-defence.

M gave evidence and called character evidence. He also called a clinical and forensic neuropsychologist, Dr Halsey, to explain the impact of his psychological problems. Dr Halsey had interviewed him a few weeks before the trial. M had been involved with the Child and Adolescent Mental Health Service (CAMHS) since 2013 and had been receiving medication for ADHD since 2014. Dr Halsey explained to the jury in some detail how his ADHD would have affected him from day to day and might affect his presentation in the witness box. Dr Halsey said, for example, that when faced with situations which were dynamic, fast moving and complex, when decisions had to made very quickly and often with little information, an individual with ADHD is more likely to find that difficult compared with someone without ADHD. Such an individual would not stop to think about the consequences of their actions and might often find themselves simply getting into trouble and doing things and acting in ways that, on reflection, appear ill-founded. As well as ADHD, Dr Halsey had also noted “autistic type traits” in M’s behaviour.

Dr Halsey explained to the jury that overall M’s ADHD and ASD traits would lead to a very restricted capacity to consider the consequences of his actions, and the consequences of other people’s actions. Dr Halsey explained that people with autism are likely to attribute a very simplistic explanation for other people’s behaviour in quite an unquestioning manner, without understanding that people might be motivated by a range of different thoughts and emotions. People with autism will always tend to struggle in situations that require an ability to grasp the more nuanced aspects of inter-personal behaviour.

We have studied the transcript of the appellant’s evidence and were taken by counsel in their written and oral submissions to certain passages. In his evidence-in-chief the appellant gave a reasonably coherent account of his actions and intentions, generally in monosyllabic terms under skilful examination by his leading counsel. He was cross-examined at length by counsel for M, who sought to distance M from the appellant and Muanza. He was also cross-examined at length by prosecuting counsel. It is apparent even from the transcript that on occasions the appellant did not understand the question, at least at first. The pervading tone and purpose of the cross-examination, quite legitimately, was to convey to the jury the impression that the appellant’s denials of knowing involvement in a revenge attack were simply not credible.

Early on in his summing up of the facts the judge gave the jury the following direction, at page 33C-G:

“I remind you from the outset that you are dealing with young witnesses and young defendants. You must bear in mind when considering the evidence that they have given… any difficulties that they may have had in understanding or their ability to express themselves. Keep that in mind. You also know the difficulties which M has with ADHD and perhaps autism. You know that the court has taken steps to ensure that he has been able to follow the trial and give evidence to the best of his ability. And, of course, he has had someone with him throughout the trial to assist with the explanations and who has monitored the questions that were asked of him and intervened to ensure that they were in a form which he would be able to answer. None of that you will hold against him; quite the reverse. You should bear in mind, when assessing his evidence, the difficulties he will have had. You also heard him give evidence over a prolonged period and are in a very good position to judge the effect his ADHD, in fact, had on his ability to express himself and that there may have been pauses in his evidence, for instance, which only signify that he was assimilating the information before answering another question.”

The judge reminded the jury of Dr Halsey’s evidence, explaining the ways in which it might assist the jury.

Applying the judge’s directions on joint enterprise, the jury had to be sure:

The judge directed the jury that an agreement to act together need not have been expressed in words. It may be the result of planning, or it may be a tacit understanding reached between them on the spur of the moment. An agreement may be inferred from the circumstances. He directed the jury that those who commit crime together may play different parts to achieve their purpose. It does not require the defendant to be in sight of the stabbing. One may wield the knife, another may guard his back, or ensure that he has an escape route, or provide support by adding to the weight of numbers in the group, or encourage him to commit the offence, so long as the elements in (i) and in (ii) or (iii) are established.

In relation to “past behaviour” (i.e. bad character), the judge directed the jury that they had to be sure that the appellant was a member of the gang. In deciding this they could take into account the views of the specialist police unit (set out in the formal admissions) that he was associated with or affiliated to the gang but must give “what weight you think right” to his denial of gang membership. They should look for any other evidence which supported or undermined the views of the specialist unit. The prosecution relied on the appellant’s involvement in or presence at the incidents of violence on 13 June 2015 and 30 August 2016.

In relation to the four knives found by the police in the bedroom the appellant shared with his older brother, the judge directed the jury that if they were sure the appellant had possession of those knives, or at least had access to them, then that may amount to evidence that he had a propensity to use knives. There was no suggestion that any of the knives could be connected with the stabbings on 18 November 2016. The appellant had said in evidence that he did not know who left the knives in the room and he never took them out on the street. The judge gave appropriate directions about the limited support that any such propensity could provide for the prosecution case.

In deciding whether the appellant was party to a planned retaliation attack the jury had to examine carefully his explanations for a number of matters:

On occasion in the course of summing up the facts the judge identified key issues the jury might wish to consider. For example, in relation to the incident on 30 August 2016 the judge said, at page 42 C:

“The prosecution suggest that it is just not believable that a group of young friends who were present when one of them was stabbed would not talk about it afterwards. They suggest that he has tried to hide from you what he knew about it and the effect that incident had on their subsequent decision to go armed to the same area of Harrow as this incident happened and then go searching in South Harrow. So that is the sort of thing where you have to look at both sides and see where it leads you.”

And at page 49H the judge said:

“Again an issue that you may want to consider is which, if any, of the defendants you are sure got into the cab knowing what the purpose of the trip to Harrow was about and which of them may have been taken on that trip without knowing what was going to happen and who would, therefore, have been potential spectators to violence to which they were not involved.”

Dr Anderson’s first report of 22 September 2019 was based on an examination of the appellant in custody on 11 September 2019. He confirmed that the appellant suffers from mild ASD and ADHD. His symptoms would have been present for many years and impaired his social functioning. The appellant (although 17 years old at the time of the examination) had the emotional maturity of a 13-year-old. He had severe difficulties with social interaction skills. These disorders interfered with his logical decision-making and consequential thinking skills. They made him vulnerable to manipulation by others. He is suggestible and very likely to comply with requests and instructions and statements that people would ordinarily reject.

His combined deficits negatively affect consequential thinking skills and cause him to react impulsively to situations and not display appropriate logical decision-making skills. He will rapidly scan the facts of a “decision tree” and miss salient problem factors. The appellant’s combined deficits interfere with his interpretation of events as they proceed and can result in misperception. He tends to rely on others to lead activities. He masks his deficits by following others and is naive in his conception of the world around him.

The combination of ASD and ADHD allows for a general level of confusion and poor thinking skills. The poor processing skills caused by ADHD would have interfered with his logical decision-making ability, memory formation and retention and judgement. There is a clear indication of inflexible thinking. His ASD makes him quite literal in his thinking. His ADHD leads to poor consequential thinking:

“…This substantially contributes to his impulsive acting out when he feels he has been threatened and his inability to resist poor leadership from others. His logical reasoning skills are too weak to process consequential pathways when he believes he is under threat. He then acts out emotionally and impulsively at such times.”

In his further report of 6 December 2020, in answer to specific questions posed by the CCRC, Dr Anderson expressed the opinion that the appellant’s ASD and ADHD would have had a profound effect on his presentation in court. His poor level of verbal communication and monosyllabic tendency could appear to have indicated evasiveness. The jury would have seen him as evasive and uncooperative when in fact he would have been simply struggling to concentrate and attend to the evidence in answering complex questions about his behaviour and intentions in a high stress environment. He was not medicated during the trial, in contrast to his co-accused M. A registered intermediary would have assisted in the appellant’s case. He is highly suggestible and would easily be influenced and led in his evidence, especially in cross-examination.

ASD and ADHD would cause him to formulate decisions without testing the potential consequences of his actions and without questioning what happens underneath. This causes a misunderstanding of the emotions and intentions that result in attributing simplistic explanations to other people’s behaviour without questioning it. He is always a follower of others. He relies and relied on others to make decisions for him. There is an indication that he followed others whom he respected in a very literal sense, without processing their intentions and testing the consequences of his own actions.

Dr Anderson’s conclusions are broadly supported by Dr Ian Cumming, consultant forensic psychiatrist in his report of 5July 2021 following an examination shortly before. He agreed with the diagnosis of ASD. He also found evidence of ADHD, although it was not as well made out. ADHD can tail off with time and Dr Cumming was assessing him aged 18. He agrees that ADHD would have been present at the age of 14.

As for the appellant’s likely presentation at the trial, Dr Cumming explained that someone suffering from ASD may process information differently from a normal individual, which can lead to misinterpretation by the jury. Someone with ASD may react to questioning without emotion, giving the impression that they lack feeling or compassion. Dr Cumming explains that the failure to detect ASD and ADHD is not uncommon; it may remain undetected even late into adulthood.

Having read the transcript of the appellant’s evidence, Dr Cumming observed that in general transcripts can give a misleading impression by suggesting that an individual functions better than they do:

“…Overall I found that he gave evidence well both in chief and in cross-examination. I would tend to consider that… in terms of his age, his degree of learning and having ASD he would have coped poorly with jumping from topic to topic as seemed to occur at a few points in cross-examination. Chronology can also be an issue with those who have ASD…”.

In response to the prosecution case that it is an overwhelming inference that the appellant was aware of what the others were planning and the purpose of the journeys, Mr Emanuel submitted that key to the jury’s consideration of these issues was an assessment of what they would have expected the appellant to have understood was happening, his relationships with the others, his explanations for why he was there, and what he thought was going on. He submitted that without knowing that this 14-year old boy suffered from ASD and ADHD, with all the consequences described by Dr Anderson, the jury were deprived of vital information necessary to make a fair and accurate assessment of the appellant and his case.

He submitted that this same gap in the jury’s knowledge would have affected their assessment of the bad character evidence. His alleged possession of knives and his alleged gang membership should have been assessed in the light of the appellant’s limited ability to see the bigger picture and make connections between events. This would have required an additional warning as part of the bad character direction.

Mr Emanuel further submitted that the prosecution case against the appellant as a secondary party was hardly compelling. There was no need for the appellant to “mind” the taxi; a return trip had been booked anyway. Nor did he remain with the taxi. He was standing 70 metres away from the violence, out of sight and hearing. He was in no position to act as a lookout, still less to lend any physical assistance by weight of numbers. The jury must have rejected the appellant’s explanations for his conduct, in ignorance of his ASD and ADHD.

The second ground of appeal raised the separate issue of the likely impact of the Appellant’s ASD and ADHD on his presentation during the trial, and particularly his evidence from the witness box. Mr Emanuel submitted that had these conditions been diagnosed before trial the appellant, like his co-defendant M, would have had the benefit of an intermediary. Like M, he would have had the benefit of expert medical evidence and a full direction from the judge. Mr Emanuel submitted that although the transcript suggests that the appellant performed reasonably well in his evidence, the transcript cannot convey the full picture which the jury will have formed, including body language, eye contact, facial expressions, visual emotional responses, and the way he answered questions.

Mr Emanuel submitted that the jury would have been left with the impression, now known to be quite wrong, that in respect of the appellant they did not need to exercise the same caution and make the same allowances as they did for M. They would have been left with the false impression that the appellant, in contrast to M, was “normal”. In fact, the appellant’s conditions were more severe; in M only “traits of ASD” had been diagnosed in addition to his ADHD.

Mr Emanuel submitted that had the jury known of the appellant’s ASD and ADHD it is impossible to say that their verdicts would have been the same. In ignorance of the true position, and through nobody’s fault, the appellant did not have a fair trial. His convictions are unsafe.

Mr Orchard submitted that the convictions are safe. There was a compelling case against the appellant which depended on many strands of evidence including that he was affiliated to the gang. One of the knives in the bedroom was found under his mattress. He had been present at the previous incidents of violence on 15 June 2015 and 30 August 2016. He could not have forgotten them. He must have known about the attack on his fellow gang member and friend on 16 November. He must therefore have known the purpose of the trip to Harrow two days later and was a willing party to the plan for retaliation. After school that afternoon he had been in phone contact with Muanza and Cabdulagdir. He even changed his trousers at one stage to avoid identification.

Mr Orchard submitted that the position of M was different and his acquittal explicable. He was not a member of the gang. He had not been present at or involved in the earlier incidents. He was of good character and called character evidence. Although the jury knew nothing of the appellant’s ASD and ADHD the judge gave the jury a general warning, at the start of the passage we have already quoted at [33], about the evidence of young witnesses and difficulties they might have in expressing themselves.

Mr Orchard submitted that the appellant’s account and his denials were demonstrated to be incapable of belief. He had no difficulty in comprehending and dealing with questions in the witness box. In the transcript of his evidence Mr Orchard took us to examples where the appellant had even corrected counsel on certain details (for example a street name in South Harrow). Mr Orchard submitted that the jury was perfectly able to assess his ability to answer questions and whether he was likely to be lying to them. He had the benefit of representation by experienced Queen’s Counsel, well able to intervene (as he did on one occasion) if the appellant appeared to be confused by a question.

On several occasions in recent years this Court has considered cases where fresh evidence had emerged only after conviction that the defendant was labouring under some undiagnosed mental impairment at the time of the offence and at trial: for example, R v Thompson [2014] EWCA Crim 836; R v Grant-Murray and others [2017] EWCA Crim 1228; R v Lammar Gordon [2018] EWCA Crim 1555; R v Roddis [2020] EWCA Crim 396. In some the Court has received the fresh evidence and quashed the conviction. In others the Court has declined to do so.

It must not be thought that the mere fact of a subsequent diagnosis, after trial, of a mental disorder such as ASD or ADHD will necessarily result in a successful appeal. Everything depends on the facts of the case. The task of the Court is to identify the issues at trial to which the fresh evidence would have been relevant, and to assess the likely impact of the fresh evidence on those issues: see R v Grant-Murray, at [53].

In the present case there is undisputed fresh evidence of the appellant’s ASD and ADHD. We are satisfied, pursuant to section 23 of the Criminal Appeal Act 1968, that it is in the interests of justice to receive the unchallenged fresh evidence of both Dr Anderson and Dr Cumming in their reports.

In the light of this evidence, the immediate issues for the Court are whether, without this evidence, the jury were able to make a fair and reliable assessment of (i) the appellant’s credibility as a witness, and (ii) his involvement in the relevant events on the night of the murder and previously.

For the reasons which follow, we do not consider that the jury was able to make that fair and reliable assessment in ignorance of the fresh evidence.

The jury had to assess the actions and intentions of a 14-year-old boy whose role was said to be as a secondary party to the murder and the other stabbings. In the light of the fresh evidence, his ASD and ADHD may well have had an impact on his ability to recollect and process relevant information and interpret the actions and intentions of his co-accused. This could have been critical evidence because the prosecution case was that it was beyond belief that he did not associate the previous incidents with the purpose of the trip to Harrow, and beyond belief that he was not so involved in the gang that he was not part of the plan to exact violent revenge for the attack on a gang member two days earlier.

As the Court accepted in allowing the appellant’s sentence appeal, at para [42], the acquittal of the co-accused M demonstrated that there cannot have been a joint intention to kill shared by all who were in the taxi before the stabbings. It follows that the jury’s focus in the appellant’s case must have been on his actions and intentions once the others had got out of the taxi in South Harrow. To make good the case that the appellant’s role at that stage was to act as lookout and to “mind” the taxi, the prosecution relied heavily on the improbability of his denials in evidence. That was the successful thrust of the prosecution’s cross-examination of the appellant. Had the jury known of his ASD and ADHD we cannot be confident that they would have reached the same conclusion.

Furthermore, we think the fresh evidence in relation to the appellant’s consequential thinking could also have had an impact on the jury’s sure conclusion (necessary to convict of murder) that he intended that the victim of the stabbing should suffer really serious bodily injury.

Quite separately from the potential impact of the ASD and ADHD on his actions and intentions around the time of the stabbings, we have considered the likely impact of the appellant’s undiagnosed ASD and ADHD on his presentation at trial. The impression he created on the jury was crucial. His evidence was all the more important because it was almost the only “live” evidence which bore on his case. We have studied the transcript of his evidence. There were occasions when he plainly did not understand or follow the question. Usually counsel or the judge intervened, or the questioning moved on to another topic. But the transcript alone cannot convey the overall impression the appellant will have made on the jury through demeanour, body language and the like. When credibility is so much in issue, the jury’s assessment of the evidence of a witness is based on more than the words spoken.

In cross-examination prosecuting counsel on several occasions followed up his question with “Really?”, and on occasion, when the appellant did not immediately respond, said “You have heard the question?”. On other occasions, prosecuting counsel responded to the appellant’s answers by asking, “Are you telling the jury the truth?” All of this will have added to the impression that the appellant was being evasive or shifty when it may simply have been a feature of his ASD or ADHD. We also bear in mind Dr Cumming’s evidence that the appellant would have coped poorly with jumping from topic to topic in cross-examination. All this may have created an unfairly negative impression, in the light of the fresh evidence.

The contrast (or apparent contrast) between the appellant and M would not have been lost on the jury. The first part of the direction we have quoted at [33] above applied to the appellant and to M in equal measure as Mr Orchard emphasised. But the direction continued: “…. You also know the difficulties [M] has had with ADHD and perhaps autism…”. This may have given the jury false reassurance that in the case of the appellant, by contrast, the jury did not have to be concerned that he was at some additional disadvantage, whereas in fact, as we now know, he was. The distinction between M and the appellant, and the affinity between the appellant and Muanza, would have been accentuated further by the physical separation of M from the other two defendants in the dock through the interposition of the intermediary sitting alongside him. The judge obviously thought it necessary to give such a direction about M’s disorders; that such a direction had value; and that it was necessary in order that the jury could assess M’s evidence fairly. Had the appellant’s conditions been known, he too would have benefitted from a similar direction. He did not and he was thereby unfairly disadvantaged.

Moreover, had the appellant’s ASD and ADHD been diagnosed before trial he would have had the benefit of an intermediary, putting him on an equal footing with M. Like M, the Appellant would also have called expert evidence explaining the impact of his condition.

It is also likely that the judge’s direction on bad character would have been expanded to make it clear that the appellant’s ASD and ADHD might affect the factual issues the jury had to resolve to support the prosecution case. His alleged possession of knives and his alleged gang membership would have had to be assessed in the light of his limited ability to see the bigger picture and make connections between events. In the direction on bad character, quoted at [37] above, the judge told the jury to give “what weight you think right” to the appellant’s denial of gang membership. The jury may have thought it appropriate to give it far less weight had they been aware of the impact of his ASD and ADHD.

It is significant that M, whose ASD and ADHD were known to the jury, was acquitted. His position was undoubtedly different in that there was no suggestion that he was a gang member, or that he had been present at the previous related incidents, and he was of positive good character. It is therefore impossible to say that the only explanation for his acquittal was his ADHD and ASD. Equally, however, it is likely that his ADHD and ASD weighed with the jury in acquitting him. After all, like Muanza (the stabber) M had put on a latex glove; and he had accompanied Muanza and Cabdulagdir to the scene of the stabbings. To that extent, on the face of it, M was more heavily incriminated than the appellant in the unlawful activity once the taxi stopped in South Harrow.

For all these reasons we are unable to accept the submission advanced by Mr Orchard that, despite the fresh evidence of the appellant’s ASD and ADHD, these convictions are safe. It is important evidence that the jury should have been able to consider, as they did in M’s case, and it might have been decisive.

The appellant’s conviction for murder together with his convictions for wounding and attempted wounding both with intent are unsafe and must be quashed.

We have received written submissions on the question whether there should be a retrial. Having considered those submissions we have concluded that the interests of justice require a retrial and that the public interests outweigh those of the appellant.

We direct that publication of this judgment be postponed until the conclusion of the proceedings in the Crown Court.

We make the following further consequential directions: