LADY JUSTICE SMITH: There are before the court two appeals against conviction for offences of sexual abuse which have been referred to this court by the Criminal Cases Review Commission ("CCRC"). On 14th November 2000 at Leeds Crown Court, after a trial before Harrison J and a jury, the appellant H was convicted by majorities of ten to one of two counts of indecent assault (counts 1 and 2) and four counts of rape of his daughter, JH (counts 3, 4, 5 and 7). He was sentenced to twelve years' imprisonment. He was acquitted of count 6, also an allegation of rape.

On 12th January 2001, at Bradford Crown Court, after a trial before Harrison J and a jury, the appellant G was convicted of five counts on indecent assault and three counts of rape of JH. He was sentenced to eight years' imprisonment.

In July 2002 the Court of Appeal (Criminal Division) dismissed both men's appeals against conviction following a joint hearing.

G died in prison in August 2002.

Both cases were referred to the CCRC, which, in September 2004, referred them to this court under section 9(1) of the Criminal Appeal Act 1995, on the basis of fresh evidence said to be admissible pursuant to section 23 of the Criminal Appeal Act 1968, as amended. The referred appeals were heard together on 30th June 2005. At the commencement of the hearing the court gave leave under section 44A of the Criminal Appeal Act 1968 for Mrs M.G., the mother of the deceased appellant G, to pursue the appeal. At the conclusion of the hearing, H's appeal was allowed in respect of all six counts of which he had been convicted and a retrial was ordered on all counts. The appeal of G was adjourned pending the outcome of the retrial of H. We now give reasons for those decisions.

Because there will be a further trial, we shall confine ourselves to an account of the facts which is just sufficient to explain our reasons.

JH was born in September 1979. She claimed to have memories going back to a time before she was 3 years old. In her police statement, she recounted in some detail an account of an event when her father had been angry with her and had taken her upstairs to her bedroom. She would have been just 3 at the time. She suggested that the way her father had carried her, with his hand between her legs, was indecent. However, this incident did not figure as an allegation on the indictment.

Counts 1 and 2 alleged that JH's father had indecently assaulted her when she was aged 4 or 5. She described two incidents, both of which took place in her parents' bedroom. In her witness statement made at the age of 19, she gave a very detailed narrative account of an incident in which her father made her masturbate him. She included much detail about the circumstances leading up to the incident and about her own emotional reaction to what had happened. In evidence at trial, she said only that her father had made her masturbate him on several occasions.

In respect of the second allegation, her witness statement contained a detailed narrative account of how her father had touched her vagina, then inserted his finger into it and then progressed to kissing it. Her account again included considerable detail about the surrounding circumstances, including a recollection of what she and her father were wearing. In evidence, she said only that she could remember her father touching and kissing her vagina on a number of occasions and sometimes putting his finger inside it.

Counts 3, 4 and 5 all alleged rape when JH was between the ages of 8 and 10. In her statement, she gave very detailed narrative accounts of three specific incidents of sexual intercourse, one of which took place in her parents' bedroom, one at her grandparents' house and one at her father's place of work. In evidence, her account of the first such incident was general and described what usually happened when her father raped her.

In respect of counts 1 to 5, the detailed accounts in the police statement were not adduced in evidence. There was no real inconsistency between what JH said in evidence and what she had said in her witness statement. No doubt counsel for H would not have wished to explore the detail given in the statement as there would have been a danger that the detail would have been accurately repeated and would have impressed and convinced the jury.

In respect of count 6, JH described in her statement how, when she was about 10, her father took her to a room in a building to which she had never been before. There she saw four men whom she did not know. She gave a detailed description of the layout of the room and the furniture. She also described each man and his clothing, in two cases descending into some detail. She then described, again in detail, how her father had had intercourse with her in the presence of the four men and how, after that, her father had forced her to have oral sex with him, had ejaculated into her mouth and had told her to swallow. The men watching were laughing. She then described how she was touched by some of the other men and had to masturbate one of them.

In evidence her account differed in two important respects from her previous statement. She said that she had not had to have oral sex with her father and had no recollection of ejaculate in her mouth. Also she added an allegation that the other men had inserted objects into her vagina, which had not previously been mentioned.

In respect of that count, JH was cross-examined closely about the differences between her police statement and her evidence, and, as we have said, H was acquitted on that count.

In respect of the seventh count, JH said that on numerous occasions up to the age of about 11 her father would rape her in his bedroom at home. She described what would usually happen on such occasions. She said that intercourse then ceased, but that inappropriate, sometimes indecent, touching continued until she was about 17.

There was a good deal of evidence about the events of JH's teenage years. She began to harm herself from the age of about 13. Her brother, her only sibling, to whom she was very close, died when she was 15. She was deeply upset. In the ensuing years, she became closely involved with a Pentecostal church. In due course she began to receive counselling.

Despite all these difficulties, she did well academically at GCSE and A' level, and obtained entrance to medical school. At the age of 18, she left home and went to live with another family in a different part of the country. Soon afterwards, in November 1998, she made a formal allegation of sexual abuse by G, who had been her music teacher at junior school. A month later she made a statement against her father, H. She was 21 when she gave evidence at trial.

On arrest, H, who was of good character, answered all questions and denied any improper conduct. He gave evidence at trial in accordance with his answers in interview. The defence called an expert, who expressed the opinion that JH's allegations were the result of false memory syndrome.

G was a peripatetic music teacher who visited JH's school. From September 1989, when she was just 10 and in her third year at the school, JH had weekly music lessons from G, initially as part of a group and then eventually on a one-to-one basis. She alleged that, during such lessons, G began to touch her inappropriately. One day, he made her sit in a chair and masturbate him. Then he forced her to have oral sex, ejaculated in her mouth and told her to swallow it. This, she said, happened on a number of occasions and was covered by counts 1 and 2 on the indictment. She also alleged that on several occasions G inserted coloured objects into her vagina. This, she said, was very painful.

Counts 4 and 5 alleged indecent assaults when she was in the fourth year at school, one of which entailed licking her vagina.

Counts 6 to 8 alleged rape, also during her fourth year at the school. In her statement, JH gave detailed descriptions of three specific incidents of rape, the first of which took place during the winter of that year. She said that penetration was extremely painful. Such incidents, she said, had taken place on more occasions than she could remember. She said that G had stopped visiting the school at the Easter of her fourth year. The abuse then stopped.

When arrested, G, who was of good character, answered all questions in interview, denying any impropriety and gave evidence at trial to the same effect. The defence did not call the expert on false memory syndrome, but did call a number of character witnesses. As we have said, G was convicted on all counts.

No complaint was made by either appellant about the conduct of his trial or the summing-up. H's appeal relied to a large extent on a submission that the jury appeared to have accepted the expert evidence of false memory syndrome in respect of count 6 and it was therefore worrying that they had convicted on other counts. Various inconsistencies in JH's evidence were pointed out, and it was submitted that the evidence could not safely be relied upon without support from an independent source. The court rejected that argument.

G's appeal relied mainly on criticism of the tactics adopted by G's trial counsel, who had been replaced. The court rejected that argument. However, at the end of the judgment, Buxton LJ said that the court had anxiously considered whether there was a lurking doubt about the safety of both convictions because it was a striking coincidence that J had been sexually abused by two men over the same period of time in circumstances where it was accepted that there was absolutely no connection between the two men. There was also a number of similarities between JH's evidence in respect of both men. However, the court concluded that the allegations in the two cases were not sufficiently similar to give rise to a fear that one or other or both had been fabricated or were otherwise unreliable. The court was not prepared to interfere with the conclusions of two separate juries, reached on the basis of impeccable directions and after the most careful of trials.

In the present appeals counsel sought to introduce the expert evidence of Professor Martin Conway, an academic psychologist, now Professor of Cognitive Psychiatry at the University of Leeds. For about 25 years, he has worked and published extensively in the field of memory formation and development. He has a specialist interest in autobiographical memory.

In a report, which comprised answers to questions asked of him by the CCRC, he had expressed the view that some parts of JH's evidence should be regarded as unreliable. Mr Richardson QC, who appeared for the Crown, quite rightly objected to those parts of his evidence as being inadmissible on the ground that such evidence usurped the function of the jury. However, as Mr Goss QC for H submitted, there was within the report an explanation of the formation of childhood memory and the features usually to be found in memories of childhood experience, which amounted to true expert evidence. They were capable of providing a jury with information on matters that would be outside their usual knowledge and experience. Mr Goss submitted that this evidence would potentially assist a jury in its approach to the reliability of JH's evidence, particularly the evidence of her memories of very early childhood.

We agreed to hear Professor Conway de bene esse in order to see whether, after oral explanation of his views and cross-examination, his evidence amounted to expert evidence which would, if available, have been admissible at the trial of H and that its relevance to the issues in the case was such that it might afford grounds for allowing the appeal.

It had been conceded by the Crown that, in other respects, the requirements of section 23 of the Criminal Appeal Act 1968 were satisfied: the evidence was credible and there was a reasonable explanation for the defence's failure to call Professor Conway at the trial. Although much of his work was in the public domain in 2000, Mr Richardson accepted that it was not well known.

We do not propose to set out the whole of Professor Conway's evidence. The gist of it was that memories of early childhood are qualitatively different from memories of later events. Adults cannot usually remember events of early childhood so as to be able to give a coherent narrative account. They may remember an event, and sometimes a visual image, but the recall will be fragmentary, disjointed and idiosyncratic. This period in early childhood of which the adult will have an impoverished memory is called the 'period of childhood amnesia'. Usually childhood amnesia extends to the age of about 7. Adult memory of events relating to later childhood becomes gradually richer, more detailed and more organised.

In the course of his research, the professor had never come across a person who had been able to provide a detailed narrative account of an event that had taken place at the age of 4 or 5. A child of 4 might well remember something that had happened when he was 3, but by the time he was 7 or 8 he would have forgotten it and it would not be recaptured. It was possible that, if a child was regularly reminded about an event which occurred when he was very young, he might retain the memory of it. If the child had a traumatic experience, one would expect that, as an adult, he or she might recall a few - usually three or four - intrusive and disjointed features of the event. However, where the childhood event was merely unpleasant, such as a painful medical procedure or an accident such as falling off a bicycle, the adult might well remember the salient or central feature of the event but would not remember the surrounding or extraneous details. He would not be able to give an accurate and reliable narrative account.

Professor Conway's explanation for this state of affairs was that, during the first five years of life, the frontal lobes of the brain were in a very rapid state of change and development and material was not retained in the memory. He said that, so far as he knew, all psychologists working in the field of memory formation agreed with what he had said about the effect of childhood amnesia, although not all agreed with his view about why it occurred.

It was Professor Conway's opinion that, if evidence of an event said to have occurred at an early age was very detailed and contained a number of extraneous facts, it might well be unreliable. It may be that the central feature of the event had indeed occurred, although research showed that, when seeking to remember life events from childhood, it was quite common for people genuinely to believe that they could remember events which were known not to have taken place.

In an adult account of a childhood event some surrounding detail might well be accurate if it was derived from conceptual knowledge, which are things that the adult knows about his childhood, such as which school he went to, where he lived, or the layout of his home. However, details which do not come from conceptual knowledge may well be false or unreliable. In effect they may well have been added on to the true memory at some later time.

Professor Conway considered that a narrative account of an event which was said to have taken place during the years of childhood amnesia should be treated with caution, especially if it contained a number of details that were extraneous to the central feature of the event.

Professor Conway also said that the effect of giving an account which contained details was to enhance its credibility to the ears of the listener. Research showed that listeners responded differently to two accounts of essentially the same event. If the account included a detail, the listener was more likely to believe and accept it than if that detail were omitted. The detail might not make any difference to the information being conveyed; it might be quite without value, but its effect would be to enhance the listener's perception. What this came down to, although Professor Conway did not articulate it in quite this way, is there is a danger that if a witness gives an account of a childhood event which contains detail about which he feels confident but may well be unreliable because of childhood amnesia, the listeners (namely a jury) might find the account more convincing than they safely should.

We came to the conclusion that Professor Conway's evidence was true expert evidence, suitable for admission at a criminal trial, in that it provided information likely to be outside the knowledge and experience of the jury. We also considered that, in the exceptional circumstances of this case, where JH had provided quite remarkably detailed accounts of events which she claimed had taken place at the ages of 3, 4 and 5, his evidence was relevant and was capable of affording a ground for allowing the appeal of H, in that it might affect a jury's view of JH's reliability as a witness. We decided, therefore, to admit the evidence.

There was also fresh evidence put before the court in the form of documents found within JH's post trial medical records. JH has undergone psychiatric treatment since 2001. A letter of referral written in 2003 by a Dr B seeking to arrange psychotherapy for JH with Miss G contained an account of JH's childhood abuse. It appears from this letter that JH had said things to Dr B that were inconsistent with her earlier evidence and that one statement made to Dr B was demonstrably untrue. A report of the first two consultations between JH and Miss G likewise appear to show that JH has given her at least an exaggerated or overdramatised account of past events.

Mr Richardson agreed this evidence "for what it was worth", as he put it. He accepted that it was credible, relevant to the issues and was capable of affording a ground of appeal. Self-evidently it had not been available at the trial. We agreed to receive the evidence under section 23.

Mr Goss for H submitted that if, hypothetically, the evidence of Professor Conway and the medical records had been available at the trial, it was likely that the jury would have taken a different view of JH's reliability as a witness. Her evidence in relation to the very early allegations in counts 1 and 2 might have been seriously undermined. The effect might have extended to the later counts. When these new factors were taken into account, in addition to the inconsistencies in her evidence that were established at the trial and the inherent implausibility of her being abused by two completely unconnected men at the same time, the result might well have been an acquittal on all counts.

Mr Swift QC, who appeared on behalf of G, accepted that JH's evidence against G could not have been affected by Professor Conway's evidence about childhood amnesia. The allegations against G related to a time when JH was 10 and 11. However, he submitted that, if it could be shown that JH's evidence against H had been unreliable, there might well be an effect on the jury's view of the case against G. Moreover, the medical records cast doubt on JH's reliability as a witness generally, even though she had not said anything to her doctors about abuse by G. Indeed the fact that she had omitted to mention him at all might be of some significance. Mr Swift submitted that the new evidence rendered G's convictions unsafe.

In respect of H, Mr Richardson submitted that, even with this additional material, the case against H was overwhelming. JH had obviously been an impressive witness. The jury had also heard H at some length and had convicted him. There was no reason to fear that the result would have been different even if the new material had been available.

However, he agreed with the proposition put to him by Wakerley J that the effect of Professor Conway's evidence was such that it would have been incumbent on the judge to warn the jury that, if they accepted the evidence, they should approach JH's evidence with special caution. This would be because the detailed narrative accounts that she had given of the early incidents would, if Professor Conway's evidence were accepted, be inherently unreliable and yet might well sound particularly convincing. Mr Richardson accepted that an appropriate warning would be similar to the warning that is usually given in identification cases, where the judge warns the jury of the danger that a witness might be very convincing, because he or she was quite sure that the identification was correct, and yet could still be mistaken. He also accepted that if, such a warning had been necessary but had not been given, that might well render a conviction unsafe.

Very fairly and realistically, Mr Richardson accepted that when, in addition, the fresh evidence of the medical records was brought into consideration, this court might feel real doubt about the safety of H's convictions. In respect of G, Mr Richardson submitted that the new evidence was of lesser effect than in the case of H and that the verdicts in his case were safe. However, he accepted that, unless and until it was known how a jury would react to the new evidence if adduced by H, it would not be possible to say what effect the new evidence might have in respect of G.

In the case of H, it is our view that, if the new evidence had been put before the jury, it is possible that they would have reached different conclusions and would have acquitted H of more or even all counts. We cannot say that they would have done: the new evidence falls far short of showing that JH was not abused. But we are persuaded that it might have made a crucial difference, particularly when considered together with the inconsistencies in evidence that were noted at the trial and the striking coincidence that JH should have been abused at the same time by two men completely unconnected with each other. We have concluded that the appeal should be allowed and that the convictions on counts 1 to 5 and count 7 of the indictment should be quashed.

We granted Mr Richardson's application for a retrial. Provided that the prosecution is satisfied that it is not against JH's interests that she should have to give evidence again, as to which independent advice will be taken, we consider that it is in the public interest that H should be retried for what are very grave allegations of abuse.

So far as G is concerned, the basis for allowing his appeal is less clear. Professor Conway's evidence is not irrelevant to G's case but is of less direct relevance than to H's case. The evidence of the medical records is also of less direct relevance to G, although it is by no means irrelevant. Bearing in mind that G is dead, that there is no possibility of a retrial and therefore no urgency to resolve his posthumous appeal, we decided to adjourn our decision pending the outcome of the retrial of H. We consider that the jury's reaction to Professor Conway's evidence, the view that they take of JH in the light of that evidence and the warning which we anticipate the judge will have to give are matters that will affect the strength of the appeal in G's case. Mr Swift did not seek to dissuade us from this course, unusual though it is. Accordingly, G's appeal stands adjourned.

We would not wish to leave this case without sounding a note of caution about the introduction of evidence of the kind given by Professor Conway in this case. It will only be in the most unusual of circumstances that such evidence will be relevant and admissible at the trial of allegations of child abuse. The evidence would be relevant only in those rare cases in which the complainant provides a description of very early events which appears to contain an unrealistic amount of detail. That, in the experience of this court, does not happen often.

The principles set out in R v Turner [1975] QB 834; 60 Cr App R 80 CA should be kept firmly in mind. Expert evidence is only admissible when it is likely to assist the jury on a topic which falls outside its ordinary experience. A witness's ability to remember events will, absent the special considerations arising from the period of early childhood amnesia, ordinarily be well within the experience of jurors. We would not wish it to be thought that the introduction of evidence such as that heard from Professor Conway will be helpful in any but the most exceptional case.

MR RICHARDSON: My Lady, this morning I saw the complainant in consultation at 8 o'clock, as I indicated last evening. Her present intention is that she would wish the case to proceed. However, when matters were explained to her, she thought it wise, as did I, that at least she should think about the matter.

LADY JUSTICE SMITH: I think that is very wise indeed.

MR RICHARDSON: She is going to see me, consequently, next Thursday morning. She has agreed additionally to see an independent psychiatrist, but such will not be instructed until we know her personal view next Thursday. I have undertaken to my learned friend that we will notify the defence solicitors soon as maybe thereafter.

May I say this? If a decision is made before the indictment is drafted that the case is not to proceed, I will ask that the matter be restored before your Ladyship so that the appeal could simply be quashed and you would rescind the retrial order. If, however, the matter has already gone to the Crown Court, it will be a matter for the Crown Court to dispose of in the appropriate way.

LADY JUSTICE SMITH: Of course if the matter were not to proceed, this court would have to reconvene in any event to consider the appeal of G.

MR RICHARDSON: Indeed. My Lady, then on that occasion the H appeal can be dealt with, if needs be.

LADY JUSTICE SMITH: I am very grateful, Mr Richardson.

Wakerley J said that in the event that there is not to be a retrial it would be very helpful to the court if we were to have a skeleton argument from you --

MR RICHARDSON: Certainly.

LADY JUSTICE SMITH: -- in respect of G and of course I imagine Mr Swift will wish to address us as well.

MR RICHARDSON: Certainly.

LADY JUSTICE SMITH: But we would be grateful for your assistance.

MR RICHARDSON: Yes, certainly.