D (Fact-finding appeal)

This is an appeal by a mother against the decision of HH Judge Baker on 25 May 2019 in care proceedings concerning her daughter, hereafter referred to as D, that she had suffered an abusive anal injury inflicted by an unidentified male and that her mother knows both who the individual is and that he was responsible for the injury.

At an earlier hearing in July 2018, the judge had found that the threshold criteria in

D’s parents started a relationship in 2013. The father had extensive convictions for violent behaviour and during the relationship was physically abusive to the mother. D was born in November 2014. Following an attack on the mother in 2015, the father was convicted of an offence of wounding and received a 16-month sentence of imprisonment. In January 2017, the father violently attacked the mother again, on this occasion in front of D. As a result of this assault, the mother sustained injuries to her abdomen and a laceration to her liver. Under the father’s influence, however, she initially told medical staff that her injuries had been suffered in a road traffic accident. Subsequently the father admitted the assault and was again convicted and imprisoned.

Following this incident, the mother and D moved away to a different area, and the mother had a brief relationship with another man, S, who also had convictions for offences of violence. That relationship ended when S was sent to prison, and the mother and D returned to their home area. In October 2017, the mother started a relationship with another man, Y. He also had convictions for offences of violence, although there have been no allegations that he has been violent towards the mother. The mother and Y did not live together, but Y was a frequent visitor. It was later alleged that on occasions they had sexual intercourse in D’s presence.

On Monday 22 January 2018, D, then aged three years two months, started attending a new nursery. On the first day, she was taken to the nursery by her mother and Y. Four days later, on Friday 26 January, D attended the nursery for the second time. On that occasion, she was taken to the nursery by her mother alone. It was the evidence of her mother and Y that Y had not been present at the mother’s home on the Thursday night and Friday morning. During the morning on 26 January, a conversation took place between D and L, a member of the nursery staff. L subsequently made the following note of the conversation:

Another record of the conversation, on a form held by the Multi-Agency Safeguarding Hub (“MASH”), reads as follows:

It was later established that this record was incorrect in at least two respects. First, there was no record at the nursery of any staff member asking to look at D’s bottom. Secondly, Y had not accompanied the mother and D to the nursery on 26 January, although he had done so on D’s first day earlier in the week. It is now accepted by all parties that there is no evidence that the mother was accompanied by a man when she brought D to the nursery that morning.

Later that afternoon, D was seen at the nursery by a social worker and two police officers. A typed record of the conversation was prepared by one of the police officers. He recorded that D had been initially shy but quickly became chatty, speaking to the social worker about a toy baby she was holding. In the note, the officer commented that D seemed to be avoiding the social worker’s questioning by talking about the baby. The record continued:

In the first judgment, the judge observed that “it’s fair to say, at the very least, that she does not relate anything that could be safely interpreted as a repeat of what L [the nursery worker] had reported.”

In the light of this conversation, and D’s age, it was decided that she should not undergo a recorded interview under the Achieving Best Evidence procedure.

The following day, D underwent a medical examination conducted by Dr Eleanor Thornton, forensic physician and general practitioner. She found an anal laceration at the 5 o’clock position, with no signs of healing. Genital examination was normal. There was a minor abrasion of the left wrist and bruises on the left thigh and the right shin which the doctor concluded could be attributable to everyday activities. In her report, Dr Thornton made the following observations about the anal injury:

On 1 February 2018, five days after the medical examination, D was spoken to at home by the social worker who had seen her at the nursery, on this occasion with another social work colleague. In a statement signed six days later and filed in the family proceedings, the social worker recorded her conversation with D as follows:

When the mother was asked about the men referred to by D in this conversation as S and M, she initially denied knowing any men by those names. Subsequently, she admitted that she had been in a relationship with a man called S and that there was a man called M in her family, but she had had no contact with him for four years. It is accepted by the mother that she initially did not tell the truth about these two men.

On 6 February 2018, the local authority filed an application for a care order. On the following day, after a contested hearing, D was made the subject of an interim care order and placed in foster care. In the course of the proceedings, the mother filed a statement in which she said that the child had suffered from diarrhoea over the weekend before the incident. She also said that D had fallen in the bath on 25 January 2018 but that she had not noticed any marks on her bottom, nor any blood. D had

An expert report was obtained from Dr Alistair Irvine, forensic physician. He agreed with Dr Thornton’s view that the anal laceration was strongly supportive of recent anal trauma such as penetration, adding “particularly in view of 3-year-old D’s disclosure”. He also stated: “It is also clear, in my opinion, that there is no merit in the mother’s claim that the anal laceration could have occurred when D fell onto her buttocks in the bath”. He concluded:

Following this report, Dr Irvine was asked some supplementary questions. In particular, he was provided with photographs of some of the toys that were in the bath when D was alleged to have fallen onto her bottom and asked whether such a mechanism was a likely explanation for the laceration. He responded:

In its “threshold document” in support of the application, the local authority referred to the history of abusive relationships between the mother and a number of men including D’s father. It asserted that the mother was unable or unwilling to prioritise

The first hearing took place before Judge Baker over eight days in July 2018. The judge heard evidence from 16 witnesses, including L, the nursery worker, the social worker and police officers who spoke to D after the incident, Dr Thornton, Dr Irvine, the mother and Y. In his judgment, having summarised the law and before dealing with the anal injury, the judge considered the evidence about the other allegations and made findings broadly in accordance with the local authority’s threshold document. Those matters do not feature in this appeal and it is unnecessary for us to consider them further.

Turning to the anal injury, the judge started by considering the expert medical evidence which he noted to be broadly consistent and which he accepted. He then considered the statements made by D on 26 January and 1 February. He took into account the fact that

On the other hand, the judge felt able to attach weight to the initial conversation between D and the nursery worker, L. He noted that what D was recorded as saying was a “spontaneous utterance on her part”. There was no suggestion that it was prompted by a specific or leading question. He continued:

What I have found most striking about that initial account is its relationship to what was found on the medical examination. There are elements of it that are difficult to understand, but there are also elements of it that I find it difficult to imagine a child making up uttering by chance. Of course that relates not only to the fact that she refers to bleeding, which we know from the medical experts is a possible consequence of receiving an anal tear, and the comment that relates to the man killing her and her relating that she can’t tell L because ‘the man will kill me’, as she put it. Again, in the context of something having happened that was untoward, in which another adult was involved, that comment makes sense. It is difficult in the context of the whole of D’s comments to L to imagine a circumstance in which a three year old could fabricate attaching [sic] being killed because of something that had happened that had caused her to bleed and complaining that her bottom was sore by unfortunate coincidence.

Then, of course, there is the relationship of that which D said to what was found specifically on the medical examination, which was an injury entirely consistent with penetration by either a body part or an object sufficiently large to cause an anal laceration.

So the internal (by ‘internal’ I mean the internal consistency within the statements themselves) and external consistency between what D is reported to have said and the medical evidence subsequently found is quite striking.”

The judge then considered evidence about the fall in the bath but concluded that it was a relatively inconsequential incident which had only taken on significance subsequently when family members were looking for possible explanations. He stated that he did not believe D’s grandmother’s account of D saying that she had fallen on a Peppa Pig toy, and concluded that the grandmother had fabricated the story to exculpate the mother. He also discounted the possibility of constipation as an explanation, noting that the mother had not reported D suffering from that condition.

The judge then turned to the evidence of the mother and Y. He observed (at paragraph 120):

The judge, however, did not believe the evidence of the mother and Y that Y had not been present at the home on the Thursday night and Friday morning. He noted what he described as “a very clear pattern” in the telephone records which indicated when they communicated with each other. As the judge observed, when they were together they did not need to phone each other. From this evidence, the judge inferred that Y had been present at the home overnight between Thursday and Friday. He accepted that Y left for work on the Friday morning before the mother took D to nursery. The judge said that he was unable to believe the mother about Y’s absence from the home because he was “convinced that the answers she was giving were motivated by her need to protect her relationship with Y”.

In reaching his conclusions, the judge stated first:

Having considered the options for D’s future care, the judge proceeded to make a care order on the basis of the local authority’s care plan for a long-term foster placement with members of her family.

On 16 October 2018, the mother gave birth to her second child, a boy, hereafter referred to as S. His father is Y. The local authority started care proceedings. S has been placed in foster care under an interim care order. He remains in that placement and decisions about his future are awaiting the outcome of these proceedings concerning his older sister.

Meanwhile both parents had filed notices of appeal against the judgment of 24 July. In support of his appeal, Y sought to rely on fresh evidence which he asserted demonstrated that he was not present at the home in the period when the injury was said to have been inflicted. Peter Jackson LJ refused permission to appeal on both applications. In respect of the fresh evidence, he observed that, if any application was appropriate, it would be an application to the trial judge for a reopening of his finding about Y’s whereabouts at the time D sustained the injury so that the trial judge could assess whether the further evidence provided solid grounds for a reopening.

On 21 December 2018, Y filed an application in the family court for the reopening of the finding that he had inflicted the injury on D. The principal evidence upon which he sought to rely was evidence from the police in the form of Automatic Number Plate Recognition (ANPR) data which he claimed demonstrated that he was not present at the mother’s home in the period when the injury was sustained.

On 14 March 2019, Judge Baker considered that application, applying the principles in Re ZZ and Others [2014] EWFC 9 which prescribes a three-stage approach to such applications. At the first stage, the court considers whether it will permit any reconsideration. As Sir James Munby, President, observed in Re ZZ at paragraph 33:

Judge Baker concluded that the first stage was satisfied because the ANPR evidence provided solid grounds for reopening part of the findings made in the July 2018 judgment. As for the second stage, he delineated the extent of the investigation as being confined to “the question of perpetration and to any and all findings dependent on that finding”. Directions were given for the listing of the rehearing over three days in May 2019. During the rehearing, the judge heard oral evidence from Y and the mother. Judgment was reserved and delivered on 23 May.

At an early point in his judgment, the judge made a number of further observations about the ambit of the rehearing. He endorsed the assertion made on behalf of Y that permission had not been given for the reopening of a number of other findings made in the first judgment, in particular (1) that D had sustained an anal laceration caused by an adverse penetrative event, (2) that the injury was inflicted in the previous 24 to 48 hours prior to the examination in the hospital on 27 January 2018, (3) that the assault was perpetrated by a male, and (4) that Y did not bring D to the nursery with the mother on 26 January 2018. The judge continued with these observations:

I mention the above point because of an observation made by Mr Ameen in his position statement prepared for the commencement of this hearing ….

That paragraph gave rise to some lengthy preliminary discussion, not least because the proposed witnesses at this hearing did not include hearing from the experts as to the cause of the injury nor other ancillary witnesses relevant to the issue.

Further, I observed that whilst there is of course an interrelationship between causation and perpetration, in the sense that (a) the court must survey and consider the wider canvas before reaching factual conclusions and (b) there remains the possibility that the totality of the evidence leads to the conclusion (bearing in mind the burden and standard of proof) that there was no possible perpetrator [sic]. However, the logical process for determining findings involves consideration of whether there has been an inflicted injury and thereafter consideration of who perpetrated such an injury. That is the process I set out and evaluated in the context of both the medical evidence and the child’s statements in my original judgment …. Nothing in this application or the fresh evidence provided by Y sheds additional light upon either the medical evidence or the child’s statements and no application before the court establishes any challenge or grounds for reconsidering the findings [in the first judgment].

Such an approach [is] of course the inevitable corollary of the case law relating to the standard of proof. As set out by the Supreme Court in Re S-B (Children) [2010] 1 FLR 1161 by way of an analogous example (see para 12) - it may be unlikely that any person looking after a baby would take him by the wrist and swing him against a wall, causing multiple fractures and other injuries. But once the evidence is clear that that is indeed what happened, it ceases to be improbable. Someone looking after the child at the relevant time must have done it. The inherent improbability of the event has no relevance to deciding who that was. The simple balance of probabilities test should be applied. Likewise, in cases where there is a potential ‘pool’ of perpetrators, consideration of who in the pool was responsible once the court has determined there is an inflicted injury can only be undertaken once it has been determined that there is an inflicted injury. A logical two-stage process is dictated by such an analysis, subject of course to an overview of all the evidence.

In summary therefore, during those preliminary discussions I opined that the rehearing amounted to a determination as to ‘who was responsible?’ and not ‘was there an inflicted injury?’ Having given the matter further thought, I remain of that view that such an approach is correct subject to the caveat that having considered the issue of perpetration (the ‘re-opened’ issue) I must of course once again look at the circumstances as a whole, considering carefully the entire canvas of the case and testing whether there are any grounds on which to reach a different conclusion as to causation.”

The judge set out some relevant points of law, including citations from the judgment of Peter Jackson LJ in Re B (Children: Uncertain Perpetrators) [2019] EWCA Civ 575 and a number of authorities on the issue of credibility. He then considered the fresh evidence about Y’s movements over 25 to 26 January 2018. He reminded himself that, in the original hearing, he had concluded from the telephone records that there was a pattern when the mother and Y were known to be together that text and phone communication was minimal; that there was very little evidence of communication on the evening of 25 January; that he had therefore concluded on the balance of probabilities that they had spent the night together and had therefore lied about his whereabouts during the relevant period. At the rehearing, all parties agreed that the ANPR evidence established that Y did not spend the night at the mother’s home. There was, however, a period of 2 ½ hours in the early evening of 25 January when the APNR evidence demonstrated that Y’s car had been in the mother’s home town and travelling on a route consistent with his visiting her home. It had always been Y’s case that on that afternoon he had taken his car to a car wash in the town where the vehicle had received a full valet service, a process which took about 2 to 3 hours. The car wash is about 10 or 15 minutes’ walk from the mother’s home, but it was Y’s evidence that he did not visit her while the car was being cleaned. He said that he did not want to leave his car with the keys unattended while it was at the car wash. The judge noted that in cross-examination he was unable to account for the fact

The judge noted that the mother’s evidence at the rehearing with respect to the events of that evening was “in similar terms and indeed entirely consistent with the evidence she gave during the first final hearing”. She denied having seen any other man that night or the following morning. She said that she had been on her own with D. She said that she did not believe that D had been abused; that her view was that, when D referred to “the man” when speaking to the nursery staff worker, she had been referring to things that had happened to her in the past when the mother had been in abusive relationships; and that she no longer believed that the injury was caused by the fall in the bath but rather considered it to be the result of constipation. The mother agreed that she had previously lied to protect partners, but denied lying to protect Y or any other unidentified person.

In submissions at the end of the rehearing, the local authority’s primary position was that the court should uphold its previous finding that Y was the perpetrator. Counsel made a number of points in support of the submission, including that there was no evidence of any other man having contact with D, that there was no evidence that Y

In setting out his conclusions, the judge started in these terms:

There are no matters relating to his past life or personal circumstances at the time that provide a background of or any indication towards a sexual interest in children or a propensity to physically abuse or harm children;

There have been no factors identified in his life at the relevant time that indicate the existence of disinhibiting influences (drugs, alcohol, stress or mental health issues);

There is no evidence, either direct or by inference, indicating an individual who has unmet sexual needs;

The child did not name him ….

The child did not subsequently say anything negative about Y (in fact, mildly the opposite) ….

The timescales relevant to when Y could have inflicted the injury to D are at the very cusp of the window of opportunity provided by the medical evidence.”

At this point in his judgment, the judge returned to the question of causation saying:

During evidence in this hearing the mother indicated that she still did not consider the injury to be inflicted and instead proffered a combination of constipation and D’s past experience. The issue of constipation as a differential diagnosis was considered at length during the first hearing. There was no evidence the child was constipated. Likewise, my conclusions about the internal consistency contained within the child’s statement in nursery … and the external consistency between the child’s words and the injury discovered on medical examination remain in my view entirely valid and significantly indicative of the injury being inflicted. Those consistencies go beyond any real likelihood the child has coincidentally transposed her own observations (there being no previous suggestion that in the past D herself has been significantly physically injured by the mother’s previous partners) to herself in a way that notably coincides with the injury she had received.

I remain entirely satisfied, to the appropriate standard, that the findings summarised … above remain extant.”

In reaching his final conclusion about the perpetrator of the injury which he had found had been inflicted, the judge noted that at the first hearing there had been no “pool of perpetrators” beyond the suggestion that the injury might have been inflicted at the nursery, which he had rejected at the first hearing and continued to reject now. This had led to the conclusion at the first hearing that there was no “pool” beyond Y. The judge looked at the basis upon which that conclusion had relied, and observed (at paragraph 66):

It is appropriate for me to set out his conclusions on this issue in full:

A consequence of my finding regarding Y’s presence or lack of it on 25 January also leads to the conclusion that in regard to Y’s presence the mother’s evidence

There is no direct evidence of another male being present in the relevant time period.

The mother denies the presence of any other person and has consistently done so, at the same time she has maintained a consistent account of events leading to the discovery of the injury.

In respect of D’s father’s assault on her, which took place in the presence of the child, she went to great lengths to conceal it and fabricate an alternative explanation ….

The mother lied about continuing contact with D’s father ….

The mother has in the past acted in ways that have not been in the interest of her daughter, prioritising herself over her daughter’s needs and welfare ….

The child has an anal injury which was inflicted.

The child has identified that a man has caused that injury.

The injury was not caused by Y.

When the child subsequently mentioned the name of males about whom she had adverse memories, the mother initially erroneously denied knowledge of such individuals ….

As with Y, likewise with the mother – I have had the advantage of seeing her not only give evidence twice but also in the context of attendance at many hearings. I have to say I was no more convinced of her honesty now than I was in the first hearing, despite giving as much weight as I can to the fact that she was telling the truth about Y’s whereabouts on 25 January [2018]. When she asserted during this hearing that she now accepted my earlier threshold findings I was left in no doubt that her concession was tactical and not an honest proclamation of insight or acceptance. It was my assessment of the first hearing and it is my assessment in this hearing that the mother says what she considers to be in her interests as opposed to any objective or even subjective truth.

I am, on … the balance of probabilities, satisfied that the mother has not told the truth about the period of time she purports to have been on her own with the child on 25 and/or 26 January [2018]. I am satisfied that the presence of injury, the child’s comments and my inability to rely on the mother’s account, especially where that account clashes with her own self interest, enable me to draw the inference that D came into contact with an unidentified male at some point prior to her attendance at nursery and that the mother knows the truth, at the very least, of whom that individual is and that the individual was responsible for the injury to D.

I would emphasise that my conclusion is not solely based upon the logic of ‘someone must have done it’. I have considered carefully whether in fact taken as a whole I must reach a different conclusion as to the cause of the injury (see above). I have also considered carefully the proposition that had I been able to believe the mother when she asserts that D did not come into contact with any other male in the relevant time period but still be of the view that D’s injury was an inflicted injury I would be left with finding that there was an ‘unknown event’ that caused the injury. I am entirely intellectually comfortable with that possible outcome – the court does not always know the answer to every question – and have been careful to consider it in my deliberations.”

The notice of appeal to this court on behalf of the mother contained five grounds of appeal. In oral submissions before the court, Mr Nicholas Goodwin QC (who did not appear at first instance) and Mr Danish Ameen focused their argument on two of those grounds.

First, it was submitted that the judge was wrong not to adjourn the case in order that the question of inflicted injury could be re-opened in full. They acknowledged that this submission faced the difficulty that no formal application was made to the judge on behalf of the mother to reopen the question of causation. They submitted, however, that, when confronted with Y’s new evidence, the judge should have of his own motion adjourned the case to permit further challenge to the paediatric evidence. They contended that, without this, the court deprived itself of the best opportunity to reconsider whether the injury was inflicted at all.

Mr Goodwin and Mr Ameen submitted that this difficulty was compounded by the judge’s decision to compartmentalise the issues of causation (i.e. whether the injury was inflicted) and perpetration. By adopting what Mr Goodwin described as a sequential approach, the judge deprived himself of the opportunity to draw into his evaluation of whether the injury was inflicted an analysis of the likelihood that the only named perpetrator was responsible. It was submitted that this flaw in the judge’s approach was not remedied by his closing observations at the end of the judgment at paragraph 73. Where, as here, perpetration was only pleaded against a single individual and the supportive evidence proved it to be flawed, the resulting absence of evidence that anyone inflicted injury should have led the judge to question whether it was inflicted at all.

Secondly, it was contended that, once the only person thought to have caused injury was exonerated, the judge was wrong not to reconsider the integrity of the child’s comments at nursery. It was pointed out that both medical experts considered the child’s account at nursery to be an important part of the background. It was submitted to this court, however, that the probative value of the child’s statements was, at the very least, debatable. In oral submissions to this court, Mr Goodwin identified a number of features in the notes of D’s initial conversation with the nursery worker which rendered it unreliable as evidence about how she sustained her injury, or who, if anyone, was responsible for inflicting it. There was no specific information about the mechanism or timing, or who was present when the injury was sustained. The child spoke about being “killed”. Mr Goodwin also relied on the fact that D did not mention her mother in this conversation. He submitted that it was unlikely that D would refer to her mother as “the lady”.

Mr Goodwin and Mr Ameen relied on the fact that D did not undergo an ABE interview because of the poor quality of the pre-interview assessment. Mr Goodwin took the court through the subsequent conversations between D and professionals, as cited above, to demonstrate that her accounts were confused. He submitted that no weight could fairly be attached to any statements made by D during her conversation with the social worker in the presence of the police officers. They were simply too incoherent. He cited in particular the various references to a policeman, a green man and a blue man, the visit to the shop, the skateboard, and the account of being kicked. He pointed out that D did not repeat any part of the account recorded as having been given to the nursery worker. It was submitted that this confusion undermined the forensic value of the original account. If the child had made very clear allegations, the judge might have been on firmer ground in declining to reopen the question of inflicted injury. The poor overall quality of the child’s accounts, however, should have led the judge to re-evaluate that question once Y was able to produce evidence supporting his innocence.

The further grounds advanced on behalf of the mother were, to my mind, of secondary importance. It was submitted that, at various points in both judgments, the judge wrongly considered whether the injury had been caused by someone within a pool of perpetrators when the case advanced on behalf the local authority was only against a single individual. It was further submitted that, having found that the mother’s originally discredited account of Y’s absence was, in fact, true, the judge attached insufficient weight to her enhanced credibility when considering whether she had nonetheless lied about the presence of an unidentified third party. Finally, it was contended that the judge attached insufficient weight to the absence of any evidence that a third party had had any access to D during the timescales for the injury.

On behalf of the local authority, Mr Simon Crabtree stressed that at no stage before the judge was it proposed by anyone that the case should be reopened as to whether D had sustained an injury and, if so, the mechanism of injury. The focus was, rather, on the limited application made on behalf of Y. No application was made for a full rehearing and no fresh evidence was put forward as to causation of injury. Nothing occurred at the rehearing in May 2019 to displace the findings as to causation made

It was submitted on behalf of the local authority that the criticism advanced by the appellant that the judge was wrong not to reconsider the integrity of the child’s comments at nursery once Y had been exonerated was unsound. It is contended that the extensive evaluation of the child’s comments in the first judgment did not depend on the identification of Y as the perpetrator of the injuries. It is asserted that all that was gleaned from that exercise was that the perpetrator was a man.

As for what I have described as the secondary grounds advanced by the appellant mother, the local authority contended that none of them had any merit. Contrary to the mother’s assertion, it was submitted that the judge expressly considered whether her credibility was enhanced as a result of Y’s exoneration. As for the arguments advanced on her behalf that the judge attached insufficient weight to the lack of evidence of the presence of any third party in the home, the local authority submitted that the judge was mindful of the absence of direct evidence, but ultimately simply did not believe the mother’s account. Finally, it was submitted that it was incorrect to assert that the local authority only advanced a case against Y. The schedule of findings at the first hearing included an alternative allegation that an unknown person had inflicted the injury and that the mother had concealed his identity. It was submitted that it was only in the light of how the evidence unfolded at the July 2018 hearing that the local authority and guardian advanced the case against Y alone. In any event, the way in which a local authority puts its case in care proceedings does not curtail the findings open to the judge at the conclusion of the hearing.

On behalf of the guardian, Ms Lisa Edmunds, in opposing the appeal, described the case as complex, evolving and evidentially fluid. She pointed out that the judge had dealt with all hearings concerning the relevant issues and submitted that this court should be slow to interfere with his conclusions unless his reasoning was so wrong as to be unsafe. Both judgments were sound and logically coherent. The judge’s navigation of all the evidence, which he assessed and reassessed, allowed him to draw appropriate inferences linked to his assessment of the mother’s credibility which resulted in the findings made.

As to the mother’s contention that the judge was wrong not to adjourn the case to allow the question of inflicted injury to be reopened in full, the guardian contended that this argument fell into two parts: (1) whether the judge can properly be criticised for doing something he was not asked to do, and (2) whether his process for determining findings was the correct approach. As to the first, it was submitted that it was disingenuous to assert that the judge was wrong not to adjourn when no such application was ever made. As to the second, it is submitted that the judge could not have been more alive to the issue whether, if the scope of the rehearing was limited to identifying the perpetrator and Y was successful in setting aside the finding against him, the outcome would create a real risk of further adverse findings against the mother. It was further submitted that, despite the absence of any application, the judge did consider this point as part of the process of reassessing all the evidence. He was

Ms Edmunds submitted that there was no merit in the mother’s argument that the judge was wrong not to reconsider the integrity of the child’s comments once Y had been exonerated. There was no correlation between his exoneration and the comments made by the child. One was not conditional on the other. It was within the judge’s discretion to conclude that Y’s exoneration did not invalidate his conclusions about the child’s comments.

On the other grounds of appeal, Ms Edmunds submitted that it was not open to this court to act as a “remote arbitrator” on the credibility of the mother’s evidence and that the judge was entitled to reach his conclusions on that aspect. Although there was no direct evidence of a third party having access to the child, the judge was entitled to reach his findings having regard to the mother’s history of dishonesty, lack of insight, and failure to protect her child on other occasions, coupled with the clear evidence of the injury corroborative of abuse and the child’s statements.

On behalf of the intervener, Y, Mr Karl Rowley QC (who did not appear at first instance) and Mr Shaun Spencer make two preliminary submissions. First, they stressed that no party was expressly seeking to disturb the judge’s finding exonerating Y. Secondly, Y did not actively seek or support the finding ultimately made by the judge that an unknown man was responsible for inflicting the injury and that the mother was concealing his identity. In those circumstances, Mr Rowley and Mr Spencer confined their submissions to identifying the three options open to the court in the event that the appeal were to succeed, namely:

quash the finding of sexual abuse in toto;

quash the findings against the mother in respect of knowledge of the fact of sexual abuse and the identity of the individual, and substitute a finding that D was sexually abused by an unidentified male in unknown circumstances, or

set the findings concerning sexual abuse aside in their entirety and remit the matter for rehearing by another judge.

It was submitted on behalf of Y that there were problems with all three options. First, although the s.31(2) threshold criteria were satisfied with regard to both children by the other findings made against the mother in respect of her history of abusive relationships, so that it would be open to the judge to make final orders on the basis of those findings alone, it was acknowledged that this option would be unlikely to commend itself to this court. The risk of harm which the mother would present to a child in circumstances where she has condoned, or concealed knowledge of, sexual abuse is of a substantially different nature to that arising out of a history of domestic abuse. Furthermore, since in those circumstances there would be no reason for the welfare decision to be allocated to a different judge, this approach would have the significant complication of fixing the judge with a factual matrix contrary to his own evaluation of the evidence.

The second option would have the merit of retaining Y’s exoneration alongside the finding that abuse had occurred. Again, however, it was acknowledged that this approach was unlikely to commend itself to this court. It was conceded that the extent to which an appellate court may feel able to limit or substitute such findings is likely to be very much more circumscribed. In addition, the potential problem for the judge at the final hearing would be arguably even more pronounced in such circumstances.

It was acknowledged that no such difficulties arose if the matter were to be remitted to a different judge for a rehearing of the sexual abuse allegation. It was submitted, however, that, were such course to be adopted, the logical corollary of the positions now adopted by the other parties – none of whom were actively seeking to disturb the finding exonerating Y – should be that this court, in remitting the matter, should confirm that no finding against Y would be sought by any party at the rehearing. The gravamen of the mother’s appeal has been that the finding of sexual abuse by an unknown person known to the mother is wrong, rather than any suggestion that Y was responsible. It has always been her position that he was nowhere near the child in the relevant period. It was submitted on Y’s behalf that, given the devastation the proceedings have wrought on Y, it would be unconscionable for him to be dragged once more into litigation concerning an act which the evidence demonstrates he cannot have committed and which no party now alleges he did commit.

In oral argument, Mr Rowley rightly submitted that the judge had carefully considered the fresh evidence in reaching his conclusion that Y should be exonerated. He fairly conceded, however, that there was force in Mr Goodwin’s principal arguments, observing in particular that the statements made by D were shot through with difficulties. It was difficult to place significant weight on anything she had said.

This is a difficult and worrying case. I was struck by Ms Edmunds’ description of it - complex, evolving and evidentially fluid. The judge approached his task with great diligence and care and set out his conclusions in a conspicuously thorough and articulate judgment. I am acutely aware of the importance of the principle that the assessment of evidence, and the apportionment of weight to be attached to each piece of evidence, are matters for the judge at first instance and that an appeal court must not interfere with findings of fact by trial judges unless there are compelling reasons for doing so. I have, however, reached a clear conclusion that the judge’s approach to the decision-making process was flawed and that, as a result, his findings cannot stand.

The crucial error in my judgment was his decision to confine the ambit of the rehearing to the question of “perpetrator” when the fresh evidence which was adduced to support the argument that Y did not have an opportunity to injure the child was relevant not only to the identification of a possible perpetrator but also whether the injuries had been inflicted at all. I realise that this may appear a harsh judgment on my part given that no party invited the judge to reopen the issue of causation. But the fact that no party asked him to do so did not absolve the judge from the responsibility of deciding the ambit of the rehearing. These were not adversarial proceedings.

In paragraph 19 of his second judgment, the judge recognised that there was an interrelationship between causation and perpetration, but adopted what he described as the “logical process” of, first, determining whether there had been an inflicted injury and, “thereafter”, identifying the perpetrator. He observed that nothing in the fresh evidence shed additional light on the medical evidence or the child’s statements. In my judgment, however, he should have considered whether the fresh evidence affected the weight to be attached to those other parts of the evidence. By failing to do so, he was compartmentalising his analysis into a consideration of (1) whether there was an inflicted injury and, if so (2) who was the perpetrator.

It is trite law that, when considering cases of suspected child abuse, the court “invariably surveys a wide canvas” as per Dame Elizabeth Butler-Sloss, P, in Re U, Re B (Serious Injury: Standard of Proof) [2004]EWCA Civ 567 and must consider each piece of evidence in the context of all the other evidence. As Dame Elizabeth observed in Re T [2004] EWCA Civ.558:

It is true that at various points in his second judgment the judge reminded himself of the need to consider the wider canvas and the totality of the evidence. Looking at the judgment as a whole, however, it is plain that he restricted the scope of the rehearing to the issue of perpetration and did not conduct any substantive review of the impact of the fresh evidence on the issue of causation.

The importance of not restricting the forensic process in this way was explained in two judgments of Charles J. In A County Council v K D & L [2005] EWHC 144 (Fam) at paragraph 49, Charles J observed:

The exercise of identifying a perpetrator, or the pool of perpetrators, forms part of the exercise of considering whether there was an inflicted injury. In my view, it is important to remember this because it removes or reduces an approach which considers the overall question from the standpoint that someone with the opportunity to injure a child has to show that he or she did not do so …. The correct position is that a medical view as to the most likely cause of injuries is that that cause is established as a real possibility that has to be considered, in all the circumstances of the case, together with the other possibilities, in determining whether a child was the victim of an inflicted injury.

If the assertions of the parents with the opportunity to injure a child that they did not do so are true, a medical conclusion that the most likely cause is inflicted injury would be wrong ….”

In support of his approach in the present case, Judge Baker referred to the decision of the Supreme Court in Re S-B [2009] UKSC 17. The passage he cited (from the judgment of Baroness Hale of Richmond giving the judgment of the court) is in fact a quotation from Lady Hale’s earlier judgment when sitting in the House of Lords in Re

As to the seriousness of the consequences, they are serious either way. A child may find her relationship with her family seriously disrupted; or she may find herself still at risk of suffering serious harm. A parent may find his relationship with his child seriously disrupted; or he may find himself still at liberty to maltreat this or other children in the future.

As to the seriousness of the allegation, there is no logical or necessary connection between seriousness and probability. Some seriously harmful behaviour, such as murder, is sufficiently rare to be inherently improbable in most circumstances. Even then there are circumstances, such as a body with its throat cut and no weapon to hand, where it is not at all improbable. Other seriously harmful behaviour, such as alcohol or drug abuse, is regrettably all too common and not at all improbable. Nor are serious allegations made in a vacuum. Consider the famous example of the animal seen in Regent's Park. If it is seen outside the zoo on a stretch of greensward regularly used for walking dogs, then of course it is more likely to be a dog than a lion. If it is seen in the zoo next to the lions' enclosure when the door is open, then it may well be more likely to be a lion than a dog.

In the context of care proceedings, this point applies with particular force to the identification of the perpetrator. It may be unlikely that any person looking after a baby would take him by the wrist and swing him against the wall, causing multiple fractures and other injuries. But once the evidence is clear that that is indeed what has happened to the child, it ceases to be improbable. Someone looking after the child at the relevant time must have done it. The inherent improbability of the event has no relevance to deciding who that was. The simple balance of probabilities test should be applied.”

Thus the point made by Lady Hale in the passage referred to by the judge in this case was that, in cases where the medical evidence establishes conclusively that a child has sustained non-accidental injuries, someone looking after the child must have been responsible and the inherent improbability of the event has no relevance in deciding who that was. She was not saying that, where the medical evidence taken on its own would lead to a finding, on a balance of probabilities, that a child has sustained nonaccidental injuries, the improbability of someone having inflicted those injuries has no relevance in establishing whether or not the injuries were in fact inflicted nonaccidentally. It is important to note that Lady Hale in Re B did not say that the improbability of an event is irrelevant. On the contrary, as she said at paragraph 70, the inherent probabilities are a factor to be taken into account, where relevant, in deciding where the truth lies.

Where there are various possible causes of an injury, medical evidence considered in isolation may suggest that the injury was inflicted. But if the other evidence demonstrates that it is improbable that any person could have inflicted the injury, the assessment of the totality of the evidence may lead to the conclusion that, on a balance of probabilities, the injury was not inflicted.

The key factors behind the judge’s findings in this case were the medical evidence, the child’s statement at nursery, the accounts given by the mother and Y, and the judge’s assessment of their credibility. When fresh evidence was produced which directly affected his assessment of some of those factors – the accounts given by, and the credibility of, the mother and Y – the judge should, in my judgment, have then carried out a full reassessment of all of the relevant evidence, because the forensic exercise cannot be compartmentalised into causation on the one hand and perpetration on the other. This is demonstrated by the fact that, in identifying factors that led to the ultimate findings against the mother in his second judgment, the judge included the statement by the child which suggested that a man had caused the injury. The child’s statement at the nursery was therefore one of the matters relied on by the judge in reaching his conclusions as to causation in the first judgment and as to perpetrator in the second.

I also accept the submission by Mr Goodwin and Mr Ameen that the judge was wrong not to reconsider the integrity of the child’s comments at nursery. At the point in his second judgment when he briefly returned to the question of causation, the judge referred again to what he described as the “internal consistency contained within the child’s statement in nursery” and the perceived consistencies between the child’s words and the injury. Looking at the totality of the evidence, however, those perceived consistencies are required by the judge’s findings to carry a weight in the balancing exercise which they cannot reasonably bear.

I have set out above in full detail the records of the various conversations involving D on 26 January 2018 and the following days. In assessing the reliability of any individual statement by the child, it was incumbent on the judge to consider the totality of the evidence, including the totality of the child’s recorded statements. In my judgment, in the light of the inconsistent and confusing accounts given by the

It is of particular importance that, when they spoke to D in the nursery shortly after her initial conversation with L, the police and social worker quickly realised that it would not be appropriate to interview the child under the Achieving Best Evidence guidance because of her age and the incoherence of what she was saying. The rationale of the Achieving Best Evidence procedure is well understood. In Re B

The only statement made by D on which the judge attached any weight was the short conversation with the nursery worker who is, of course, untrained in ABE interviews and was not conducting the conversation in accordance with the guidelines. I am not for one moment criticising the nursery worker. On the contrary, she acted entirely properly in telling her manager what had been said and in making a note of the conversation. My observation about the weight to be attached to what the child said is not based on any suggestion that L did anything wrong but rather on the totality of the evidence about statements made by D on 26 January and subsequently. In addition, I accept Mr Goodwin’s submission that there were a number of features about the notes of D’s initial conversation with the nursery worker which undermined the reliability of the child’s statements during that conversation as evidence of how she sustained the injury and, if it was inflicted, the identity of the perpetrator.

It is also significant that the child’s statement to the nursery worker had at least some impact on the medical evidence given to the court. As noted above, Dr Irvine expressed the view in his initial report that the anal laceration was strongly supportive of recent anal trauma such as penetration, adding “particularly in view of 3-year-old D’s disclosure”. It is concerning to read a professional using the word “disclosure” in this fashion. For over 30 years since the Cleveland report in 1987, this has been recognised as an inappropriate use of language that judges have urged professionals to avoid.

If one disregards the child’s statements in her initial conversation with the nursery worker, an important element in the judge’s findings against the mother falls away. This underlines the importance of conducting a full rehearing of the evidence as to causation as well as perpetration. I have not overlooked the fact that the judge was careful to remind himself in both judgments that in reaching his decision he had to consider the totality of the evidence. To that end, at one point in the second judgment, he carried out what might be called a cross-check to see if his conclusion that Y was not the perpetrator affected his original finding that the injury was inflicted. I have looked carefully to see whether this was sufficient but concluded that it was not. In my judgment, once the judge had decided that the fresh evidence produced by Y as to his whereabouts in the relevant period of 25 to 26 January provided solid grounds for a review of his findings, the right course was to conduct a full rehearing of the factfinding exercise. Notwithstanding the judge’s warnings to himself about the importance of having regard to the totality of the evidence, I conclude that the ambit of the rehearing was too narrow for this court to be satisfied that justice was done. To be fair to the judge, it is plain from the lengthy passages of the judgment which I have quoted above that he was aware of the dangers. With the great benefit of hindsight available to this court, however, I consider that, despite the careful warnings the judge gave to himself, the approach which he adopted to the rehearing has resulted in an outcome which is unsafe.

Having reached the clear conclusion that the appeal should be allowed on the principal grounds advanced by Mr Goodwin and Mr Ameen at the hearing, it is unnecessary to consider the arguments about the other grounds of appeal which featured prominently in the written submissions but which are of secondary importance.

What course should this court now take? Mr Rowley’s careful and lucid analysis of the options has been of great assistance, but ultimately and ironically it has led me to conclude that the only fair course is the one his client is most anxious to avoid. There must be a full rehearing of the fact-finding hearing before another judge. All of the issues, including the medical evidence and the evidence about Y’s whereabouts during the relevant period, must be reconsidered.

It was for that reason that at the conclusion of the hearing we indicated that the appeal would be allowed and the matter remitted to the Family Division Liaison Judge, MacDonald J, for a case management hearing to determine allocation. I recognise that this will involve a further delay before decisions can be taken about the future care of D and her brother S. I am confident that MacDonald J will do whatever he can to ensure that the rehearing takes place as soon as possible.

In conclusion, I wish to acknowledge again the great care with which Judge Baker carried out his task in this case. Reluctantly, however, I have reached the conclusion that the appeal must be allowed for the reasons set out above.