This judgment text has undergone conversion so that it is mobile and web-friendly. This may have created formatting or alignment issues. Please refer to the PDF copy for a print-friendly version.

In the GENERAL DIVISION OF

THE high court of the republic of singapore
[2024] SGHC 287
Magistrate’s Appeal No 9043 of 2024
Between
GIL
Appellant
And
Public Prosecutor
Respondent
grounds of decision
[Criminal Law — Offences — Outrage of modesty]
[Evidence — Presumptions — Effect of presumptions in relation to electronic records — Section 116A of the Evidence Act 1893 (2020 Rev Ed)]

This judgment is subject to final editorial corrections approved by the court and/or redaction pursuant to the publisher’s duty in compliance with the law, for publication in LawNet and/or the Singapore Law Reports.
GIL

v

Public Prosecutor
[2024] SGHC 287
General Division of the High Court — Magistrate’s Appeal 9043 of 2024
See Kee Oon JAD
6 September 2024
6 November 2024 
See Kee Oon JAD:
Introduction
1 This was the appellant’s appeal against his conviction in relation to a charge of outrage of modesty of a minor under s 354(2) of the Penal Code (Cap 224, 2008 Rev Ed). The appellant was alleged to have used his hand to touch the thigh and vaginal area of the then-12-year-old victim (the “victim”) when she was at the appellant’s residence for a sleepover with the appellant’s daughter (“AD”), who was then eight years old.
2 The District Judge (the “DJ”) convicted the appellant following a trial and sentenced him to 23 months’ imprisonment and three strokes of the cane. The DJ’s grounds of decision are set out in Public Prosecutor v GIL [2024] SGDC 87 (the “GD”). The appellant only filed an appeal against his conviction. At the hearing of the appeal, he was serving his sentence.
3 I dismissed the appeal on 6 September 2024 after hearing the parties’ submissions. The grounds of my decision are set out below.
The proceedings below
4 The appellant was a teacher at a school in Singapore. The appellant’s wife and the victim’s mother (“M”) were also teachers at the same school, while the victim was a student at the school. The appellant’s family and the victim’s family were close family friends and resided in the same condominium block.
5 On 27 February 2021, the two families went out for dinner together. The victim and AD sought permission to have a sleepover at the appellant’s residence after dinner. The appellant and his wife agreed to this. After dinner, both families headed back to the condominium block. The victim proceeded to the appellant’s home for the sleepover after changing at her own home.
6 The following facts were undisputed. The sleepover took place in AD’s room. AD and the victim sat on the top bunk of the bunk bed in AD’s room, with their heads propped up by pillows against the head of the bed and their legs stretched out in front of them. AD was on the victim’s left, at the side of the top bunk which was closer to the wall, while the victim sat on the side of the top bunk which was closer to the grey fabric railing of the bunk bed. Both of them were under a duvet. Whilst on the bed, AD used the appellant’s phone and the victim used her own phone to design outfits on a mobile phone application called Combyne. The appellant was in AD’s bedroom while the two designed outfits on Combyne. In the course of designing outfits on Combyne, AD would look over the victim’s shoulder and comment on the outfits that the victim was designing. Further, sometime past midnight, the appellant was standing beside the bunk bed at the victim’s right.
7 In summary, the Prosecution’s case at trial was that during the time the victim and AD were designing outfits on the Combyne application in AD’s bedroom, the appellant committed the offence by slipping his hand under the duvet and under the victim’s shorts. He first touched the victim’s thigh skin-on-skin before moving his hand upward into the right leg opening of her shorts and under her panties where he touched the victim’s vaginal area skin-on-skin in a circular motion. The victim was shocked, but she did not inform AD who was beside her of the appellant’s conduct as she did not want to scar AD. Instead, the victim continued to design outfits on the Combyne application and stated a few times that she was tired and that they should go to bed. The appellant eventually stopped touching the victim. According to the Prosecution, the victim did the following after the incident:
(a) Soon after leaving the appellant’s home the next morning, the victim first informed her close friend at the time, who was referred to as “F” in the court below, that something had happened during the sleepover at the appellant’s home. When asked by F if the police could get involved, the appellant responded in the affirmative. The victim also spoke to F on two other occasions about the incident.
(b) On the evening of 28 February 2021, the victim informed her mother, M, that the appellant had touched her, demonstrating this by running her hands up between her legs and around her private area at the front.
(c) The victim made a contemporaneous written record of the incident in her notebook (the “Notebook”). The Notebook was subsequently handed to the vice-principal of the victim’s school.
(d) The victim also wrote a longer account of the incident on a piece of paper which she kept as a record for herself (the “Note”).
8 The appellant’s case at trial was that he had not touched the victim inappropriately. In particular, the appellant made the following arguments at trial:
(a) First, the appellant stated that he could not have committed the offence because he was not tall enough to touch the victim on the upper bunk in the manner the victim described, or that he would not have been able to do so without his elbow sticking out in an obvious and awkward manner.
(b) Second, the appellant pointed to various inconsistencies in the victim’s evidence which he said cast doubt on the reliability of her evidence and credibility as a witness. The inconsistencies related to: (i) where the appellant had been positioned relative to the victim; (ii) where AD had been positioned relative to the victim; (iii) the manner in which the appellant had touched the victim’s thigh; (iv) whether the appellant had touched the victim skin-on-skin or over the surface of her panties; and (v) the duration of the molest. The appellant also offered several reasons as to why the victim might have made a false allegation against him.
(c) Third, the appellant pointed to evidence in the form of data from his smart watch, an Amazfit GTR Smartwatch (the “Watch”), which he said directly contradicted the victim’s account and showed that the appellant was in deep sleep at the time the victim alleged he had committed the offence. Based on the victim’s account, the offence had occurred at about 1.57am. The appellant therefore adduced a report containing the extracted data from the Watch (the “Watch data”) to show that he was asleep at 1.57am, and could not have committed the offence as alleged by the victim. I pause here to explain that the Watch data was admitted into evidence by way of a statement of agreed facts. However, within the statement of agreed facts itself, it was made clear that the expert forensic consultant who extracted the Watch data (one Mr James Tan) was unable to testify about the accuracy or the purport, or significance, of the extracted data.
(d) Fourth, the appellant pointed to the evidence of AD who was in the bedroom at the material time and who testified that she had not noticed anything unusual during the night.
The decision below
9 The DJ convicted the appellant of the charge, finding that the Prosecution had proven its case beyond a reasonable doubt.
10 First, the DJ found that the evidence of the victim was unusually convincing. The DJ noted that the evidence of the victim and the appellant was largely consistent on the events surrounding the incident, apart from whether the appellant had touched the victim. This showed that the victim was a reliable witness whose recollection of the events was accurate (GD at [111]–[112]). The DJ also found that the victim’s evidence was textured and bore a ring of truth when considered against the overall backdrop of the case (GD at [113]). While there were some inconsistencies in the victim’s testimony as well as inconsistencies between her evidence and the evidence of some other witnesses, the DJ found that these were inconsequential and did not affect her credibility (GD at [116]–[136]). The appellant challenged most of the DJ’s findings in relation to these inconsistencies on appeal. The inconsistencies raised by the appellant and my reasons for agreeing with the DJ that these inconsistencies were inconsequential are set out at [34]–[40] below.
11 Second, the DJ found that, while the evidence of the victim was sufficient to prove the charge beyond a reasonable doubt, the victim’s evidence was also corroborated by her contemporaneous reporting of the incident to F and M as well as the written accounts in the Notebook and Note. Further, the victim’s evidence was corroborated in some way by the distress suffered by the victim following the incident as evidenced by her diagnosis of post-traumatic stress disorder (“PTSD”) (GD at [145]–[151]).
12 Third, the DJ found that the appellant had failed to raise any reasonable doubt (GD at [153]–[158]). For example, while the appellant argued that it was inherently improbable that he would have committed the offence in the presence of his daughter, AD, the DJ noted that based on the appellant’s conduct, he had only progressed to touching the victim’s vaginal area after touching a less sensitive area (ie, the thigh) and seeing that the victim did not raise any alarm. Further, it was unsurprising that AD would have been engrossed in playing Combyne as an eight-year-old, and might not have noticed the appellant’s conduct which was happening under the duvet (GD at [158]). Further, while the appellant sought to suggest a motive on the victim’s part to falsely implicate him, the DJ found this to be without basis. The suggestion that the victim had dreamt of the incident was without basis and speculative. The suggestions that she had fabricated the incident because she was seeking attention were again without basis and speculative. There was no evidence to support these claims (GD at [159]–[166]).
13 On sentence, both the Prosecution and the appellant agreed that the sentencing framework in GBR v Public Prosecutor and another appeal [2018] 3 SLR 1048 applied and that the present case fell within Band 2 of the framework (GD at [173]). Applying this framework, and taking into account the offence-specific aggravating factors, the DJ sentenced the appellant to 23 months’ imprisonment and three strokes of the cane (GD at [176]–[182]).
The parties’ cases on appeal
14 On appeal, the appellant contended that the DJ had erred in finding that the victim was unusually convincing, and had also erred in her treatment of the inconsistencies in the victim’s evidence and the Watch data adduced by the appellant. Broadly, the appellant raised the following arguments on appeal:
(a) First, the appellant argued that the victim’s claim that she was molested at 1.57am was contradicted by objective, contemporaneous evidence in the form of the Watch data. The appellant argued that the DJ had erred in requiring evidence to prove the accuracy of the Watch data, and had erred in failing to find that the Watch data had an adverse effect on the victim’s credibility as a witness. In particular, on appeal, the appellant relied on s 116A of the Evidence Act 1893 (2020 Rev Ed) (“EA”) to suggest that the DJ was required to presume that the Watch data was accurate and authentic.
(b) Second, the appellant argued that reasonable doubt arose from various inconsistencies in the victim’s evidence as outlined below:
(i) In relation to the victim’s inconsistent accounts as to how the parties were positioned, the appellant argued that the DJ ought not to have concluded that the discrepancy was inconsequential and therefore not material. Further, the appellant argued that the DJ ought not to have found that the victim’s account in her statement did not make sense.
(ii) In relation to the inconsistency on whether the appellant touched the victim’s thigh in a circular or linear motion, the appellant argued that the DJ ought not to have concluded that the discrepancy was inconsequential and therefore not material.
(iii) In relation to the victim’s inconsistent evidence on the duration of the touching, the appellant stated that the DJ erred in finding that this inconsistency did not affect the victim’s credibility.
(iv) The appellant argued that the DJ erred in finding that no reasonable doubt arose despite the inconsistent evidence by the witnesses on whether the touch was skin-on-skin or over the victim’s clothes.
(c) Third, in seeking to argue that reasonable doubt had arisen in relation to the Prosecution’s case, the appellant sought to compare the facts of the present case with that of Public Prosecutor v BNO [2018] SGHC 243 (“BNO”).
(d) Fourth, the appellant argued that the DJ had erred in taking into account any post-incident distress that the victim suffered since that had no corroborative value.
15 The Prosecution argued that the appellant’s conviction was safe and ought to be upheld. The Prosecution submitted that the DJ had correctly assessed that the victim’s evidence was unusually convincing and corroborated by other evidence led during the trial. The DJ also correctly considered the various inconsistencies in the victim’s evidence and found that these did not affect the victim’s credibility or the reliability of her evidence. Accordingly, the DJ was correct to find that the appellant had failed to raise any reasonable doubt in the Prosecution’s case.
My decision
16 There were broadly two issues for me to determine:
(a) First, whether the DJ had erred in her treatment of the Watch data, and whether s 116A of the EA assisted the appellant’s case.
(b) Second, whether the DJ had erred in her treatment of the inconsistencies in the victim’s evidence or had erred in finding that the Prosecution had proven its case against the appellant beyond a reasonable doubt.
The DJ had not erred in her treatment of the Watch data
17 A key pillar of the appellant’s appeal was the Watch data admitted into evidence. According to the appellant, the Watch data was objective, contemporaneous evidence that he could not have committed the offence at the time stated by the victim, 1.57am on 28 February 2021, because he was in deep sleep at the time.
18 As I had stated at [8(c)] above, the Watch data was admitted into evidence by way of a statement of agreed facts. Among other things, the Watch data contained the raw heart rate data of the appellant, the raw data of the appellant in relation to his state of sleep (including whether he was in deep sleep, shallow sleep or awake) as well as the raw data in relation to the appellant’s step count. On the face of the watch data, the appellant was in light sleep from about 1.12am to 1.54am and deep sleep from 1.55am to 2.03am on 28 February 2021, and switched between light and deep sleep until 7.46am.
19 The DJ, however, did not rely on the Watch data on the basis that no evidence had been tendered to show whether the Watch data was accurate (GD at [130]). It was clear to me that the DJ had not erred in any way in this regard. First, while the parties agreed that the Watch data could be admitted into evidence, the statement of agreed facts made it clear that there was nothing which could be said about the accuracy or significance of the Watch data. Second, the appellant did not adduce any evidence to advance his claim that the Watch data was an accurate reflection of his state of sleep and could be relied on as a record of his activities between 27 February 2021 and 28 February 2021.
The presumptions under s 116A of the EA did not assist the appellant
20 On appeal, the appellant pointed to the presumptions under ss 116A(1) and 116A(2) of the EA. I set out ss 116A(1) and 116A(2) of the EA below:
Presumptions in relation to electronic records
116A.—(1) Unless evidence sufficient to raise doubt about the presumption is adduced, where a device or process is one that, or is of a kind that, if properly used, ordinarily produces or accurately communicates an electronic record, the court is to presume that in producing or communicating that electronic record on the occasion in question, the device or process produced or accurately communicated the electronic record.
Illustration
A seeks to adduce evidence in the form of an electronic record or document produced by an electronic device or process. A proves that the electronic device or process in question is one that, or is of a kind that, if properly used, ordinarily produces that electronic record or document. This is a relevant fact for the court to presume that in producing the electronic record or document on the occasion in question, the electronic device or process produced the electronic record or document which A seeks to adduce.
(2) Unless evidence to the contrary is adduced, the court is to presume that any electronic record generated, recorded or stored is authentic if it is established that the electronic record was generated, recorded or stored in the usual and ordinary course of business by a person who was not a party to the proceedings on the occasion in question and who did not generate, record or store it under the control of the party seeking to introduce the electronic record.
Illustration
A seeks to adduce evidence against B in the form of an electronic record. The fact that the electronic record was generated, recorded or stored in the usual and ordinary course of business by C, a neutral third party, is a relevant fact for the court to presume that the electronic record is authentic.
21 The appellant argued that s 116A(1) required the court to presume that electronic records adduced in evidence were accurate, unless the other party adduced evidence sufficient to raise doubt as to its accuracy. Further, the appellant argued that s 116A(2) required the court to presume that any electronic record was authentic if it was established that the electronic record was generated, recorded or stored by a neutral third party. I did not agree with the appellant’s arguments.
22 As a preliminary point, I observed that the appellant had made no mention of the presumptions under s 116A of the EA in the court below. As such, there was neither any discussion nor submission made below in connection with s 116A. Neither the appellant nor the Prosecution had brought the provision into focus. In my view, this was unsurprising because the purpose of the presumptions under s 116A of the EA was to facilitate the admission of electronic records into evidence. In the present case, parties had agreed to the admission of the Watch data into evidence by way of a statement of agreed facts. The scope or application of s 116A of the EA was therefore a non-issue. The appellant also offered no cogent explanation as to why s 116A of the EA could not have been raised in submissions in the court below if the presumptions were indeed deemed to be relevant.
23 More significantly, however, it was clear to me that the appellant had misunderstood the scope and purpose of s 116A of the EA. First, the plain meaning of the text of s 116A of the EA makes it clear that the presumptions which arise under the provision are meant to facilitate the admission of evidence, and not to relieve parties of their burden to show the reliability of such evidence after it has been admitted. The illustration in s 116A(1) of the EA makes this amply clear: if a party is seeking to adduce evidence in the form of an electronic record produced by a device, the court may presume that the electronic record in question was produced by the device if the party is able to prove that the device is one that, or is of a kind that, if properly used, ordinarily produces that electronic record.
24 Applied to the present case, assuming parties had not admitted the Watch data by consent, and if the appellant intended to adduce evidence in the form of the report containing the raw Watch data, the appellant would have only needed to show that the Watch ordinarily produces such raw data if the Watch had been properly used. Had the appellant done so, s 116A(1) of the EA would have led to the court presuming that the report containing the raw Watch data was an accurate reflection of the data actually captured by the Watch at the material time. This, however, was unnecessary in the present case because there was no dispute that the report containing the raw Watch data was an accurate reflection of the data actually captured by the Watch. What was in dispute was an entirely separate and distinct issue – whether the data actually captured by the Watch was a true and accurate reflection of the appellant’s activities between 27 February 2021 and 28 February 2021, ie, whether the appellant was asleep at the material time. The presumption under s 116A(1) did not provide any basis for the court to further presume that the data captured by the Watch, including data pertaining to the appellant’s state of sleep at the material time, was accurate in any way.
25 Similarly, as made clear by the illustration in s 116A(2) of the EA, s 116A(2) of the EA only gives rise to a presumption that the electronic record is authentic if it is a fact that the electronic record was generated, recorded or stored in the usual and ordinary course of business by a neutral third party. However, it was never in dispute in the present case that the report containing the raw Watch data which the appellant sought to rely on was authentic. The statement of agreed facts made it clear that the Watch data originated from the Watch and was extracted from two sources – the Watch’s Android application, which was accessed using the appellant’s log-in details, as well as the appellant’s account on the watch manufacturer’s website. Again, the presumption under s 116A(2) did not provide any basis for the court to further presume that the data captured by the Watch, including data pertaining to the appellant’s state of sleep at the material time, was accurate in any way.
26 Further, even if it may be said that the purpose of s 116A of the EA is unclear from the wording of the provision, the Parliamentary debates confirm that the presumptions under s 116A of the EA were only meant to facilitate the admission of electronic records into evidence, and not to relieve parties of the burden of proving that the electronic records were reliable once they were admitted into evidence. The purpose of the presumptions under s 116A of the EA was made expressly clear by Mr K Shanmugam, the Minister for Law, during the second reading of the Evidence (Amendment) Bill in 2012, which provided for the introduction of the presumptions under s 116A into the EA:
Let me now explain the key amendments. On computer output evidence, clauses 3, 7, 9, 10, 12 and 13 reform the law on computer output evidence. The current framework for the admission of computer output evidence is found in sections 35 and 36. They were introduced in 1996. Computer technology was then in its infancy. A cautious approach was therefore taken. Currently, short of agreement between parties, computer output can be admitted only if: (i) it is produced in an approved process; or (ii) it is shown to be produced by a properly operating computer which was properly used.
This is a somewhat cumbersome process not consonant with modern realities. With the benefit of experience, we can say now that computer output evidence should not be treated differently from other evidence. Sections 35 and 36 are therefore repealed. In addition, there will be presumptions facilitating the admission of electronic records. For example, where a device is one that, if properly used, accurately communicates an electronic record, it will be presumed that an electronic record communicated by that device was accurately communicated. Sounds a little circular, but it does make sense. Further, documents in the form of electronic records will be treated as primary evidence.
[emphasis added]
27 As was made clear by the Minister for Law, the presumptions were introduced to facilitate the admission of computer output evidence without requiring the parties to show that the computer output evidence was produced in an approved process or was produced by a properly operating computer which was properly used. The presumptions do not, however, do away with the parties’ obligation to lead evidence to show how the computer output evidence is reliable or supports their case.
28 Further, while the appellant relied on Telemedia Pacific Group v Credit Agricole (Suisse) SA (Yeh Mao-Yuan, third party) [2015] 1 SLR 338 (“Telemedia”) and Public Prosecutor v Michael Frank Hartung [2020] SGDC 113 (“Michael Frank Hartung”) in support of his arguments, I found that neither of these decisions actually assisted the appellant:
(a) First, the appellant stated that the High Court’s decision in Telemedia made it clear (at [248]–[250]) that s 116A of the EA “sets out a presumption in favour of the production or accurate communication of electronic records” [emphasis added]. The appellant was correct in so far as there was a presumption as a result of s 116A(1) of the EA that the electronic records accurately communicated what the electronic device (ie, the Watch in the present case) had recorded. I noted that the court in Telemedia expressly cited (at [250]) the purpose of the presumptions in s 116A as stated by the Minister for Law during the second reading of the Evidence (Amendment) Bill in 2012 (as reproduced at [26] above). However, I failed to see how Telemedia supported the appellant’s argument that the DJ was required to presume that the Watch data was a true and accurate representation of the appellant’s activities between 27 February 2021 and 28 February 2021. It was clear to me that Telemedia did not assist the appellant in any way.
(b) Second, the appellant’s reliance on Michael Frank Hartung was misplaced as the issue which arose there related to the admissibility of the chat logs between the offender and an undercover police officer. Again, there was no question in the present case of the admissibility of the Watch data since the parties had agreed to admit the Watch data into evidence. The question in the present case was whether the Watch data accurately represented the appellant’s activities between 27 February 2021 and 28 February 2021. Therefore, Michael Frank Hartung did not assist the appellant in any way.
29 For these reasons, I found that the presumptions under s 116A of the EA did not assist the appellant. The parties had already agreed to the admission of the Watch data into evidence. However, the appellant did not lead any evidence to show that the sleep data of the appellant which was captured by the Watch was reliable. For example, no evidence was led to elaborate on the manner in which the Watch detected whether the appellant was awake or asleep, and whether he was in light sleep or deep sleep. Further, no evidence was led to show that the manner in which the Watch detected whether the appellant was awake or asleep was reliable. On the contrary, the statement of agreed facts explicitly included a caveat that the expert forensic consultant was unable to testify about the accuracy or the significance of the Watch data.
30 While the appellant sought to argue on appeal that the Watch captured information that was involuntary in nature (eg, the appellant’s heart rate and the time when he fell asleep), I failed to see how this fact alone made the Watch data reliable. It is entirely plausible that an electronic device may be able to capture information that is involuntary in nature but using a method which is completely unreliable or flawed. In the absence of any evidence to show the reliability of the manner in which the Watch captures and processes the data to determine whether the Watch user is asleep or awake, the court was not in a position to assess the reliability of the Watch data which had been admitted into evidence.
31 In the absence of such evidence, I found that the DJ had not erred in any way in refusing to rely on the Watch data. The appellant could not expect the DJ to simply accept the Watch data as being a true and accurate representation of the appellant’s activities between 27 February 2021 and 28 February 2021 without any evidence to support this.
The Watch data did not cast a reasonable doubt on the Prosecution’s case even if it was accurate
32 Given that I agreed with the DJ that there was no basis to rely on the Watch data in the absence of evidence on its accuracy, there was no need to consider whether it cast any reasonable doubt on the Prosecution’s case. However, for completeness, I also agreed with the DJ that, even if the Watch data did accurately reflect the appellant’s activities between 27 February 2021 and 28 February 2021, it did not go so far as to cast a reasonable doubt on the Prosecution’s case.
33 As the DJ found, even if the appellant’s case were taken at its highest that he was in deep sleep at 1.57am, the Watch data, if at all, only showed that the victim was wrong about the exact timing of the offence. However, the Watch data, coupled with the appellant’s own evidence, made it clear that there was a 1.5-hour window when the appellant was in AD’s bedroom together with AD and the victim from about 11.27pm on 27 February 2021 to 1.03am on 28 February 2021. As the DJ noted, even if the timing provided by the victim was inaccurate, the Watch data did not go so far as to show that the incident could not have taken place at all. Rather, the Watch data broadly cohered with the undisputed evidence that the appellant was in the bedroom together with AD and the victim as they used the mobile phones to design outfits on Combyne. Therefore, even if the victim may have gotten the exact timing of the offence wrong, this did not inexorably mean that the offence did not happen altogether or that the victim’s credibility was tainted to a point where there was a reasonable doubt cast on the Prosecution’s case.
The DJ had not erred in her assessment of the victim’s evidence and in finding that the Prosecution had proven its case against the appellant beyond a reasonable doubt
34 The next issue I considered was whether the DJ had erred in evaluating the inconsistencies in the victim’s evidence or had erred in any way in finding that the Prosecution had proven its case against the appellant beyond a reasonable doubt.
35 Having considered the parties’ submissions and reviewed the record and the GD, it was clear to me that the DJ had carefully considered each of the inconsistencies in the victim’s evidence and correctly assessed that these did not cast a reasonable doubt on the appellant’s guilt.
36 I accepted that the victim had given contradicting accounts as to how the parties were positioned. In her investigative statement, the victim stated that the appellant was to her left, standing at the side near where her head was and away from the ladder. The victim stated that AD was seated to her right on the top bunk bed and leaning against the wall. In her testimony in court, however, the victim stated that the appellant was to her right, while AD was seated to her left with her legs facing away from the window. While there was an inconsistency, it was clear that this inconsistency was an inconsequential one. The victim had candidly accepted at the trial that she had gotten the positions wrong in her statement and that her account at trial was an accurate reflection of the positions of AD and the appellant. More significantly, the appellant and AD (who were the only other persons in the room) confirmed that the victim’s account at trial about their relative positions was accurate. Therefore, there was no dispute by the time of trial as to their relative positions in AD’s bedroom. Any discrepancy in this regard was, therefore, inconsequential.
37 Second, the victim’s account was inconsistent on whether the appellant touched her thigh in a circular or linear motion. While the victim stated in her investigative statement that the touch on her thigh was circular, she subsequently testified at trial that the touch on her thigh was linear and maintained this account under cross-examination. However, it was clear from both the investigative statement and the testimony in court, as well as the Note and Notebook, that the victim’s evidence was broadly consistent on the key issue – that the appellant first touched the victim’s right inner thigh under her shorts after his slipping his hand under the duvet, before moving his hand up her thigh and under her panties to touch her vaginal area in a circular motion. The appellant’s focus on an inconsistency relating to the type of motion which the victim felt at her thigh was, in my view, a minor and inconsequential inconsistency which did not affect the reliability of her evidence which was otherwise consistent.
38 Further, I did not accept the appellant’s argument that, extrapolating from the DJ’s reasoning, “it would seem that a conviction would stand so long as a complainant is able to consistently maintain an allegation in the broadest sense that she was molested, despite being unable to hold a consistent account regarding the details of how she was indeed molested”. This, in my view, was an overstatement and a mischaracterisation of the present case. Crucially, the appellant’s argument ignored the fact that the victim’s evidence in the present case was largely consistent on a number of important details besides the broad allegation that she was molested – this included who molested her, where the offence occurred, which body parts were intruded and how the appellant committed the offence, as well as what the victim and AD were doing while the offence took place. This was not a case where the DJ simply convicted the appellant on the basis of a broad, unsubstantiated allegation in the broadest sense that the victim was molested. The appellant’s conviction was based on a textured and largely consistent account by the victim on the material issues.
39 Third, the victim had given inconsistent evidence on the duration of the touching. However, the DJ made no error in finding that the victim’s inability to give an accurate estimate of the duration of the touching did not affect her credibility as a witness. As the DJ found, what was clear from the victim’s evidence was that she perceived the touching to have gone on for a long time even though she was unable to tell what the actual duration was (GD at [127]).
40 Finally, there was the inconsistent evidence by the witnesses on whether the molest occurred skin-on-skin or over the clothes. First, I noted that the victim’s evidence on this point was consistent – the molest occurred skin-on-skin. Second, while the two witnesses, M and F, may have provided inconsistent evidence on this issue, I did not think that the DJ erred in her assessment of the evidence:
(a) In the case of M, she clarified in cross-examination that she was personally unaware of whether the molest occurred over or under the victim’s clothes. Rather, her response in her investigative statement was based on an assumption that the molest occurred over the victim’s panties because she did not want it to be a case of skin-on-skin molest – reflecting her own denial of what the victim had experienced. As the Prosecution had highlighted in its submissions, this was completely aligned with M’s testimony that she “wanted it to be a mistake … wanted it to be something innocent … wanted it to be anything other than moving your hands up and down between my child’s legs and touching her on her private area”. I found that the DJ appreciated the full context of M’s evidence in assessing the inconsistency between M’s evidence and the victim’s evidence.
(b) In the case of F, the DJ accorded little weight to this aspect of F’s evidence because F’s first time recounting the events was during the trial, more than two years after the incident (since no investigative statement had been recorded from F). While F stated that she recalled the victim telling her that the molest occurred over clothes at one of two conversations which they had, she was unable to recall any details surrounding this conversation, or elaborate on the details provided by the victim in this regard. In the absence of unambiguous evidence from F on the details of the conversation to support her claim that the victim told her that the molest occurred over clothes, the DJ was entitled to prefer the victim’s clear and consistent account throughout that the molest occurred skin-on-skin.
41 I considered the remaining arguments made by the appellant beyond the inconsistencies in the victim’s evidence and found these to be without merit:
(a) While the appellant argued that the DJ had erred in taking into account any post-incident distress that the victim suffered as corroborative evidence, it was clear from the GD that the DJ had been mindful of the extent to which corroborative weight could be placed on the other sources of evidence. The DJ placed less weight on the fact that the victim had recounted the events to M and F, since these emanated from the victim herself and did not mean that there was more corroborative evidence. However, the DJ placed more weight on the fact that both M and F had independently observed the victim’s distress which aligned with the victim’s diagnosis of PTSD.
(b) While the appellant sought to compare the facts of the present case with BNO, it was trite that each case had to be considered based on its own facts, and that broad comparisons between the facts of the two cases was not particularly meaningful. Further, as the appellant himself recognised in his written submissions, there were significant differences between the facts of the two cases, such as the nature of the inconsistencies, and the findings made by the court in BNO and the present case. For example, the court in BNO found that the evidence of the witnesses seemed to have been rehearsed. No such allegation or finding was made in the present case. In the face of such significant differences, the comparison between the two cases was of no assistance to the appellant.
Conclusion
42 I was of the view that the DJ was correct to find that the victim was unusually convincing and the Prosecution had proven its case against the appellant beyond a reasonable doubt. In particular, I found that the victim’s testimony was largely consistent in relation to the material facts surrounding the incident, including the following:
(a) First, it was the appellant who had molested the victim. This was consistent across the victim’s investigative statement, testimony, contemporaneous disclosure to M and F as well as the Notebook and Note. During cross-examination, the victim maintained that it was the appellant who touched her, and not AD.
(b) Second, the offence took place in AD’s bedroom while the victim and AD were on the top bunk of the bed. This was consistent across the victim’s investigative statement, her testimony, the Note and the Notebook.
(c) Third, the body parts which were intruded upon were the victim’s thigh and vaginal area, and in that order. This, again, was consistent across the victim’s investigative statement, her testimony, the Note and the Notebook.
(d) Fourth, the appellant had touched the victim during the sleepover whilst she and AD were using their mobile phones to design outfits on Combyne.
(e) Fifth, the appellant had touched the victim by slipping his hand under the duvet before touching her right inner thigh and then moving his hand up her thigh and under her panties and touching her vaginal area in a circular motion. This, again, was consistent across the victim’s investigative statement, her testimony, the Note and the Notebook.
43 In my assessment, the DJ had correctly assessed the evidence and found that the Prosecution had proven the charge of outrage of modesty of a minor beyond a reasonable doubt. Her decision was not plainly wrong or against the weight of the evidence, and I saw no reason to disagree with the DJ’s reasoning or findings of fact.
44 For the reasons I have set out above, I dismissed the appellant’s appeal.
See Kee Oon
Judge of the Appellate Division
Tan Jun Yin and Tanaya Shekhar Kinjavdekar (Trident Law Corporation) for the appellant;
Nicholas Wuan and Teo Siu Ming (Attorney-General’s Chambers) for the respondent.
Back to Top

This judgment text has undergone conversion so that it is mobile and web-friendly. This may have created formatting or alignment issues. Please refer to the PDF copy for a print-friendly version.

Version No 1: 06 Nov 2024 (14:54 hrs)