R v Sheldon Sieudath

The provisions of the Sexual Offences (Amendment) Act 1992 apply to these offences. Under those provisions, where a sexual offence has been committed against a person, no matter relating to that person shall, during that person’s lifetime, be included in any publication if it is likely to lead members of the public to identify that person as the victim of that offence.

The applicant is now 42. On 19 July 2023, in the Crown Court at Kingston-upon-Thames before HHJ Lodder (“the judge”) and a jury, the applicant was convicted unanimously of two counts of sexual assault (counts 1 and 2) and two counts of sexual assault of a child under 13 (counts 3 and 4). On 20 October 2023, he was sentenced to a total of 4 years’ imprisonment. He renews his application for leave to appeal against conviction following refusal by the single judge.

Trial counsel, Ms Jemma Levinson, is acting for him today pro bono. We are very grateful to her for her written and oral submissions. We should however say, at the outset that, in our view, her Perfected Grounds of Appeal (“PGA”), which run to 39 pages, are much too long and diffuse. That has tended to obscure some of the stronger arguments which she has put forward with real skill this morning, in support of her renewed application.

There were two complainants “LS” (in respect of counts 1 and 2) and “SB” (in respect of counts 3 and 4). They were at the material time friends of the applicant’s family. LS said she vaguely knew SB because they attended the same school, and she thought she had seen her a few times at the applicant’s address. She said she had last seen SB in about 2012 and that they had had no communication since then. SB said that she did not know LS and could not recall meeting her at the applicant’s address. They both said they did not know about each other’s complaint until it was revealed to them by the police during the investigation.

LS said that, in 2015, when she was 13, she went to the applicant’s house to spend time with his stepdaughter “C”. The applicant was her father’s best friend and the two families were close. One day, when she was about to leave, the applicant asked her to sit next to him. He proceeded to grab her, pull her onto his lap, wrap his arms around her and put his hands down her top and grab her breasts (count 1). According to LS, the applicant continued to touch her breasts and say, “Where did these come from?” In her words he was “playing with her breasts”. She got up and left. He said, “Don’t be a stranger and come back”.

The following day she said the applicant apologised. But later that evening, the applicant came up behind her, hugged her and began to move his hands towards her groin (count 2). She pushed him away and told him that she did not want him to touch her anymore. She left and again he said, “Don’t be a stranger”.

In relation to SB, she said that, in May 2016, when she was 12, she stayed at the applicant’s house for about a week when her mother gave birth to her sister. She shared the bedroom with two of the applicant’s daughters. The applicant would enter her bedroom and touch her thighs and stomach. She asked him to stop but he continued. She was unsure how many times that had happened but said it was more than once.

In addition, she said that the applicant would lift her off the bunk and put her on his lap, so he was behind her. She tried to get away and told him to stop. But she said the applicant repeatedly tried to touch the top of her thighs and described the touching as a “rubbing or stroking sort of gesture up and down”. He would put his hand on the inside of her thigh and she would push it off. He would touch her again and ask her if it was okay. He would then move his hand further up towards her waist and then ask if it was okay, to which she would reply “No”. If the touching took place when she was wearing her school uniform, then the touching would be under her skirt and on top of her tights. If she was wearing some other clothing, it would be on top of that clothing. There was never anyone else in the house when that happened.

SB disclosed the abuse to a number of people in 2018, which led to two meetings with Safeguarding Leads at her school and disclosure to the police. However, SB did not further co-operate and was not ABE interviewed at that time. LS’s mother said her daughter disclosed to her what had happened in August 2019. LS was ABE interviewed. It was only after SB was informed by the police about LS’s allegations in 2019, that SB decided that she would support the prosecution.

The applicant denied touching either complainant. The Defence Case Statement did not expressly contend that there had been collusion between LS and SB, (although the word “collusion” was used in a later heading in respect of a disclosure request). But at paragraph 10 it pointed out that:

“The false allegation of [SB] was then followed by a false allegation by [LS] in circumstances where [SB] and [LS] are, despite an initial denial, known to each other.”

The applicant was to say that he saw SB standing outside LS’s house on 30 March 2021. This allegation was denied by both complainants.

There are three grounds of appeal. First, it is said that the judge intervened excessively and descended into the trial arena. Secondly, it is said that comments made by the judge during his summing-up effectively withdrew the issue of collusion from the jury’s consideration, and he made his negative view of the defence case clear. Thirdly, it is said that the judge’s cross-admissibility direction was deficient.

We set out our view on these grounds below. We consider that each of them is a separate and stand-alone ground. For reasons which will become apparent, we take them in a different order and we start with ground 2.

Active collusion occurs where two or more complainants have or may have got together and made up false allegations against the defendant. Even if there was no concoction, but a complainant had or may have learned about the allegation of others and been influenced by that knowledge, consciously or unconsciously, when making their own allegations, that may also allow a jury to rule out cross-admissibility. That is contamination. In either circumstance, an appropriate direction to the jury is necessary. However, there must be some factual basis for the assertion of collusion or contamination, and these allegations, which are serious, must be properly and fairly put to the relevant complainants.

The complaint under Ground 2 is that the judge began his summing-up of the facts in a way that suggested that the defence of collusion was unarguable. It is said that his summing-up on this point was in the style of a prosecution speech and that he effectively withdrew this topic from the jury.

The single judge said that he considered that the judge’s direction to the start of his summing-up of the facts was “perfectly adequate to remedy the impression given in his opening remarks”. The single judge went on to say:

“Whether or not it was a live issue that SB and LS knew each other, collusion goes significantly further and is altogether different, and the defence had no positive evidence of collusion. It is therefore debateable whether collusion was a live issue in reality.”

In our view, there are a number of reasons which make ground 2 unarguable.

First, there was no cogent evidence that the complainants had known each other at the time of the relevant events. LS thought she might have known SB, but SB was sure she did not know LS. True it is that there was some conflict in LS’s evidence as to when and how many times she may have met SB, but given the passage of time between the relevant events, the interviews and the oral evidence, that was unsurprising. In any event, the defence case was positively to the effect that LS was wrong about when she met SB. That all created an extremely slender thread on which to hang allegations of collusion or contamination.

Secondly, both complainants denied absolutely any suggestion that they had been in contact more recently. On this point, we should address the submission that was made in writing that in some way the applicant’s evidence (that he saw SB on LS’s driveway on 30 March 2021, although he was uncertain about the year), was supported by the agreed evidence from his employer that his route took him past LS’s home every Tuesday. That is incorrect. The applicant’s employer was giving evidence about the applicant’s route home: he was not giving evidence that the applicant had seen SB that day. Both girls denied any such meeting. That was therefore a matter for the jury.

Given this paucity of evidence, it was perhaps unsurprising that when the judge started summing-up the facts, identifying matters that appeared not to be disputed, he said:

“It’s not disputed that separately two girls who are not friends – no-one asserts that they were; certainly not at that time – say that this man sexually assaulted them at different times. Indeed, the Defence positive assert that [LS] and [SB] cannot have met when [LS] thinks that perhaps she had met [SB]... Everyone is agreed that neither of [LS] or [SB] spoke to the other before they made their statements to the police… The applicant is unable to offer any explanation as to why it would appear completely separately, it’s a matter for you, these girls should make these allegations against him.”

He went on to say: “… there is no suggestion of collusion here. It’s not suggested that these two girls have got their heads together.”

As to the vast majority of the specific matters identified by the judge in this passage, it is difficult to say that any of them were, on their own, incorrect. LS and SB were not friends; there was no evidence to the contrary. It was part of the defence case that LS and SB could not have met when LS thought they had. Neither complainant spoke to the other before they were interviewed. Both denied meeting more recently although that was, of course, subject to the applicant’s point about the sighting on the driveway which was a matter for the jury. It was not suggested, at any stage, that the complainants had got their heads together in some way to frame the applicant.

That said, we accept that the impression given by that compressed start of the summing-up of the facts suggested that all of those matters were formally agreed, and they were not. So that was properly raised by Ms Levinson with the judge, in the absence of the jury, during the summing-up, and the judge addressed that position. He said:

“... I’m reminded that, of course, the Defence don’t know whether the girls did speak to each other or knew each other because they have no information one way or the other, so you should bear that in mind...”

In our view, the judge therefore made it clear that those matters were not formally agreed, and that was sufficient to address the potentially misleading impression given at the start of the summing-up. It might, as Ms Levinson has said, have been better presented to the jury but that is not, in our view, an arguable ground of appeal. That is therefore the third reason why we reject ground 2.

There is a final reason why we consider that this ground of appeal is unarguable. Collusion or contamination involves much more than simply two people knowing one another. For that to be an issue even for the jury to consider, there has to be at least some credible evidence that the relevant witness either got together and agreed to give a false statement to the police or, at the very least, may have been unconsciously affected by the statement of the other. There was no evidence of that in this case, as the single judge rightly pointed out. It was not put to the two complainants that they got together to make false allegations against the applicant. All that Ms Levinson had to go on was the fact they were at the same school, albeit 2 years apart and there had been some inconsistency in LS’s recollection as to when and where she had met SB before. That was a long way from amounting to even an arguable case of collusion or contamination.

To the extent, therefore, that the judge gave the impression that there was little, if anything, in that part of the case, that was, in our view, understandable. So, in those circumstances, we do not agree with Ms Levinson’s submission that collusion was an important factual issue in the case. The absence of any cogent evidence to support it meant that, in reality, it was not an issue at all, as the single judge emphasised.

For the reasons that we have set out, we consider there is nothing in ground 2. It is not arguable.

The leading authority to R v Freeman and Crawford [2008] EWCA Crim 1863; [2009] 1 WLR 2723, which confirms that evidence may be cross-admissible in one or both of the following ways:

As to propensity, the example of a cross-admissibility direction given in the Crown Court Compendium in these terms:

“D is charged in count 1 with a sexual assault on W1 and in count 2 with a sexual assault on W2. The prosecution evidence on count 1 is (a) the account given by W1 and (b) a video recording which the prosecution say was made by D as he/she committed the offence. The prosecution evidence on count 2 is only the account given by W2. D claims that W1 and W2 have concocted false accounts and denies that he/she is the person shown in the recording. I have already told you that you must consider each count separately. However, if but only if, you are sure the person shown in the recording of events in count 1 is D and that count 2, committed that offence, you should next consider whether that shows that D has a tendency to commit offences of the kind charged in count 2.”

The judge’s direction on cross-admissibility in this case was as follows:

“[The applicant] is charged with four counts of sexual assault against LS and SB. Each has described the manner in which he lifted them onto his lap, touched under clothing but not onto skin and asked, ‘Is this okay?’ Both were wearing school uniform and as friends of his stepdaughter were staying or visiting [the applicant’s] home. I’ve already told you that you must consider each count separately; however, if, and only if, you are sure that either [SB] or [LS’s] description is true...and [the applicant] did behave in this way, you should then consider whether that conclusion shows he has a tendency to commit sexual assaults on young girls. If you do conclude that [the applicant] has such a tendency, then you may take this into account when you are deciding whether he is guilty of other counts where his behaviour is in dispute. Bear in mind that even if you find that [the applicant] has a tendency to commit this particular kind of offence, it does not follow that he is bound to do so on other occasions. So if you are sure that [the applicant] does have a tendency to commit offences of the type charged, this is only part of the evidence against [the applicant] and you must not convict him wholly or mainly on the strength of it.”

It is said that this direction was inadequate because it did not summarise the evidence relied upon by the defence. It is also said that the judge wrongly introduced the concept of “striking similarities” in his directions.

The single judge thought it was not reasonably arguable that the direction on cross-admissibility was deficient.

In our view, the direction on cross-admissibility was properly formulated to meet the facts of this case. It was very close to the example of a propensity direction in the Crown Court Compendium. We have already said that there was no cogent evidence of collusion or contamination, so that did not require to be identified in this direction. It was not necessary for the judge to summarise the evidence relied upon by the defence on this aspect of the case, because he was not giving a coincidence direction. In any event, all the defence evidence was properly summarised elsewhere in the summing-up. It did not need to be repeated here. Instead, the judge focused on there being two different allegations by two different women, and focused on the issue of propensity not coincidence. That was of benefit to the applicant. It is always better for a defendant not to be the subject of a coincidence direction because that does not require the jury to be sure of the evidence of one complainant before considering the evidence of the other. The propensity direction does.

As to the reference to “striking similarities”, that was not in the judge’s legal directions, which were given before counsel’s speeches at the end of the trial. The reference to “striking similarities” was a comment made by the judge when he was summing-up the evidence. Since he had made it plain that matters of fact were for the jury, it seems to us that that comment, although inadvisable, had no effect, and cannot have influenced the jury’s approach.

In our view, there was nothing wrong with the cross-admissibility direction. It appears to us to be an attempt to put the collusion point in another way. For the reasons that we have given, Ground 3 of the proposed appeal is therefore unarguable.

In R v Hamilton (113) Sol Jl. 546, Lord Parker CJ stated that whether judicial interventions would give ground for complaint was a matter of degree. He said that:

“…the interventions which give rise to a quashing of a conviction are really three-fold: those which invite the jury to disbelieve the evidence for the defence, which is put to the jury in such strong terms that it cannot be cured by the common formula that the facts are for the jury…the second ground giving rise to a quashing of a conviction is where the interventions have made it really impossible for counsel for the defence to do his or her duty…and thirdly, where the interventions had the effect of preventing the prisoner himself from doing himself justice and telling the story in his own way.”

More recent civil cases have emphasised that the old convention, whereby the judge sat in silence throughout the trial is, at least to some extent, a thing of the past. As the Court of Appeal made plain in Southwark LBC v Kofi-Adu [2006] EWCA Civ 281; [2006] H.L.R. 33, “nowadays, of course, first instance judges rightly tend to be very much more proactive and interventionist than their predecessors”. But the subsequent criminal cases demonstrate that there remain clear and proper limits on such proactivity and interventions, particularly in jury trials. Interventions which are hostile, or invite the jury to disbelieve the defence evidence, may lead to the quashing of a conviction: see R v Hulusi and Purvis [1974] 58 Cr App R 378, CA, recent followed in R v Jauvel (2018) EWCA Crim 787.

In Michel v The Queen [2009] UKPC 41; [2010] 1 WLR 879, there were repeated interventions by the judge. The court moved away from the three elements of the Hamilton test to focus more generally on the fairness of the trial. Lord Brown of Eaton-Under-Heywood said:

“27 There is, however, a wider principle in play in these cases merely than the safety, in terms of the correctness, of the conviction. Put shortly, there comes a point when, however obviously guilty an accused person may appear to be, the appeal court reviewing his conviction cannot escape the conclusion that he has simply not been fairly tried: so far from the judge having umpired the contest, rather he has acted effectively as a second prosecutor. This wider principle is not in doubt.”

The facts in Michel were extreme. Lord Brown made clear the ways in which the judge had allowed the trial to become unfair:

“34 …Indeed, it does not entitle him [the judge] to conduct the hearing in any way different from that ordinarily required of a judge at trial. Of course he can clear up ambiguities. Of course he can clarify the answers being given. But he should be seeking to promote the orderly elicitation of the evidence, not needlessly interrupting its flow. He must not cross-examine witnesses, especially not during evidence in-chief. He must not appear hostile to witnesses, least of all the defendant. He must not belittle or denigrate the defence case. He must not be sarcastic or snide. He must not comment on the evidence while it is being given. And above all he must not make obvious to all his own profound disbelief in the defence being advanced.

35 Regretfully the commissioner’s interventions during this trial breached each one of those canons…”

There are a number of authorities which stress the particular importance of the judge not intervening during the evidence in chief of a witness, particularly that of the defendant: see R v Rancoli [1998] Crim. L.R. 584. In R v Perren [2009] EWCA Crim 348, this court said:

“We do not suggest that any intervention in the course of evidence in chief, other than by way of clarification, must render a conviction unsafe. However, there are good reasons why a judge should be particularly careful about refraining from intervening during a witness’ evidence in chief, except in so far as it is necessary to clarify, to keep the evidence moving on and, if necessary, to avoid prolixity or irrelevance. The first is that it is for the prosecution to cross examine, not for the judge, the second is that the right time for the prosecution to examine is after a witness has given his evidence-in-chief. It would be unthinkable for prosecuting counsel to jump up in the middle of a witness’ evidence in chief and seek to conduct some hostile examination. This is not merely in order to preserve an orderly trial. There is a more important, fundamental reason. A jury will inevitably form a view of each witness as the case goes along. As the witness is giving his or her evidence in chief, so the jury will be absorbing that account and forming their own impression of the witness.”

It will, of course, always depend on the facts. We note the recent decisions of this court in R v Inns [2018] EWCA Crim 1081 and R v Binoku [2021] EWCA Crim 48, in which the appeals against conviction were based on the judge’s interventions. In both cases, the Court of Appeal accepted that, at times, the judge had appeared to cross-examine the witnesses, and that that was not the judge’s function. However, in both cases, they concluded that the judge’s departure from good practice was not so gross, or so persistent or so irremediable that the trial was unfair.

There are numerous complaints in the present case about the judge’s interventions. Some of these are general in nature, such as the allegation that the judge showed hostility to the defence case and undermined the applicant, and some are very specific, going to particular questions that the judge asked a particular witness. The complaints are not separately numbered, which has made identification more difficult than it might otherwise have been. The single judge considered this complaint generally.

The single judge took the view that there was insufficient substance in ground 1, that is to say the intervention for them to be reasonably arguable. He said that he considered the interventions to be largely for the purposes of clarifying the evidence. He said that although the judge may have approached the line of what is permissible, he did not transgress it. He also pointed to one passage which he said had the appearance of cross-examination but considered that the questions were legitimate because the defendant had contradicted himself at an earlier stage. The single judge concluded:

“It was unfortunate that the judge went as far as he did, but again I do not consider that he went so far as to have rendered the trial unfair.”

We recognise that ground 1 is by far the most significant ground of this renewed application. We therefore consider the individual points, for want of a better means of categorisation, by reference to the paragraphs in the PGA which deal with the individual witnesses. However, we can make some general points first. First, we are satisfied that the judge’s questions did not generally demonstrate hostility or a desire on the part of the judge to undermine the defence. On the basis of the transcript, the questions were phrased in courteous terms. So, we reject the general criticisms at paragraph 6 and 7 of the PGA, and at paragraph 8, the judge embarked on a course that was designed to assist the prosecution.

Secondly, we consider that the vast bulk of the interventions fell into three generally unexceptionable categories. The first emanated from his desire to ensure that the trial was properly managed: see, for example, the passage of paragraph 68 of the PGA, where the judge was intervening to ask Ms Levinson not to lead her witnesses, seemed to us to be a proper exercise of the judge’s function.

The second category was a genuine desire on the part of the judge for clarification. The third was where he was asking questions that originated from the jury. Those categories of questions too were plainly justified. That was what the judge was there to do.

We then turn to the relevant paragraphs of the PGA dealing with the separate witnesses. One element of ground 1 relates to the evidence to SB’s mother (who was a prosecution witness) and the two safeguarding leads of SB’s school, JG and KS, who were perhaps surprisingly called on behalf of the applicant. As to SB’s mother, we cannot see how the complaints made about the judge’s treatment of her evidence at paragraphs 45 and 46 of the PGA can be of any relevance to the appeal. Although it is now said that she was giving evidence about the potential window of opportunity that the applicant had or did not have to assault SB, her mother was in hospital at the relevant time, so was unable to give any cogent evidence as to what actually happened. We are satisfied that none of the extracts from the transcript at paragraph 45 and 46 of the PGA were of assistance to the prosecution or anticipated points that were to arise in the cross-examination of the defence witnesses. They were in any event of peripheral relevance.

The evidence of the two safeguarding witnesses, JG and KS, (to whom disclosure was made by SB in 2018) did not generally help the applicant’s case, so despite the fact that there were differences between what SB may have said to them and what she said in court, we are still not entirely clear why they were called on the applicant’s behalf. Of course, because they were defence witnesses, they could not be led or cross-examined, which created another difficulty for Ms Levinson. Since KS had made the relevant notes of the meetings and JG had not, the judge said it was sensible for KS to give evidence first. The criticism of his direction to that effect at paragraphs 56 to 59 of the PGA is therefore misplaced. That was a case management matter for the judge. The defence does not have the right to call witnesses in the order it wants if that sequence will needlessly add to the length of the trial.

Thereafter, we are satisfied that the judge’s questions of KS were largely clarificatory. There were issues as to the scope and accuracy of her notes, and the judge was simply seeking clarity on a number of matters. We therefore reject the criticisms at 60 to 63 of the PGA. We have reached the same conclusion about the judge’s questions of JG, whose difficulties in recollection were exacerbated by the absence of any contemporaneous notes. We therefore reject the criticisms at paragraph 64 to 67 of the PGA.

All that said, we accept, as we turn to consider the crucial issue of the judge’s interventions in the evidence of the applicant and his wife, that this was a case where the judge intervened more than is usual or advisable during the evidence of those three witnesses. That therefore is the background to the remaining paragraphs of the PGA which address ground 1.

The criticisms of the judge’s interventions in respect of the applicant and his wife can be found at paragraphs 47 to 54 of the PGA (the applicant) and paragraphs 69 to 75 of the PGA (the applicant’s wife). In our view, for the reasons outlined briefly below, we consider that it is arguable that the judge’s interventions in respect of these two witnesses crossed the relevant line and give rise to an arguable ground of appeal. Although the Full Court may conclude, as they did in R v Inns [2018] EWCA Crim 1081 and R v Binoko [2021] EWCA Crim 48, that the interventions did not ultimately affect the safety of the applicant’s convictions, that issue will need to be fully argued out.

In respect of the applicant, some of the judge’s lengthy interventions occurred during his evidence-in-chief. That is to be avoided for the reasons noted in R v Perren [2009] EWCA Crim 348. In relation to the interventions during his cross-examination, we consider that, although they were couched in courteous language, the interventions were, at times, akin to cross-examination. For example, in relation to the applicant’s answers in the police interview, the judge’s questioning ran to over five pages of the transcript, from 55D to 60D. On at least one occasion during this exchange, it is also arguable that the judge strayed into matters that were privileged. Whilst we understand that this line of questioning arose from a question from the jury (the point made by the single judge) these interventions cause us some concern. We consider that the points made at paragraphs 47 to 54 of the PGA therefore give rise to an arguable ground of appeal. Ms Levinson said that they were her strongest ground of appeal and we agree.

Although the interventions in respect of the applicant’s wife were less extensive, we consider that they too give rise to an arguable ground of appeal. Again, some of the judge’s questioning occurred during her examination-in-chief. We also accept that it is arguable that the judge suggested that the witness may not have had such a good recollection as she claimed in a way that might have been perceived as hostile. Again, therefore, we consider that the points made at paragraphs 69 to 75 of the PGA also give rise to an arguable ground of appeal.

For these reasons therefore, whilst we reject many of the submissions made under ground 1 as unarguable, we accept that, against a background where the interventions during the evidence of the other witnesses were greater than is usual or advisable, the interventions identified in paragraphs 47 to 55 (applicant) and 69 to 75 (applicant’s wife) of the PGA give rise to an arguable ground of appeal.

Accordingly, therefore, this renewed application for permission to appeal is successful but only in the terms referred to in the previous paragraph. The audio file of the oral evidence of the applicant and his wife should be made available to the Full Court. Moreover, we would respectfully urge Ms Levinson to refine the submissions that she makes in respect of those interventions; in our view, some of the points arising out of them are manifestly better than others.