Gao Xiong v Public Prosecutor [2026] SGCA 29

Introduction

Gao Xiong (“Appellant”) pleaded guilty to one charge of attempted rape under s 375(1)(a) read with s 511(1) of the Penal Code 1871 (2020 Rev Ed) (“PC”), as well as three charges for criminal trespass under s 447 of the PC. Three charges of unlawful stalking and criminal trespass were taken into consideration (“TIC”) for the purpose of sentencing. A judge sitting in the General Division of the High Court (“Judge”) sentenced the Appellant to six years and six months’ imprisonment, with three strokes of the cane for the attempted rape offence and two weeks’, four weeks’ and six weeks’ imprisonment respectively for the three criminal trespass offences. The sentences for the attempted rape and the most serious criminal trespass charges were ordered to run consecutively, for an aggregate sentence of six years, six months and six weeks’ imprisonment with three strokes of the cane – Public Prosecutor v Gao Xiong [2025] SGHC 260.

CA/CCA 28/2025 (“CCA 28”) is the Appellant’s appeal against his sentence. While the Appellant’s Notice of Appeal refers to his sentence generally, his Petition of Appeal as well as his oral submissions at the hearing focus entirely on the sentence for the attempted rape charge.

This is the first matter before this court involving a criminal attempt offence prosecuted after the Penal Code was amended to abolish the previous statutory one-half sentencing limit for criminal attempt offences – whereas the Penal Code previously provided that the longest imprisonment term that may be imposed for criminal attempt offences was one-half of the longest term provided for the primary offence, it currently provides that attempts with no express punishment provision shall be “punished with such punishment as is prescribed for that offence”. We hence take this opportunity to consider the proper approach in sentencing such offences.

During the appeal, a further issue arose. In pleading guilty to the charges against him, the Appellant, who was legally represented below, accepted the Statement of Facts (“SOF”) without qualification. The Appellant, who is now unrepresented in the appeal, sought to challenge the SOF, claiming that: (a) he was “confused” when he pleaded guilty; (b) the SOF did not accurately reflect the facts; and (c) the victim and others bore some responsibility for the attempted rape. As we explain below, these allegations are baseless and wholly improper. They raise the issue of whether any sanction should be imposed on the Appellant for his conduct in this appeal, and if so, what that sanction should be.

Facts

The matters below were set out in the SOF, which the Appellant had previously agreed to, and in the Judge’s Grounds of Decision (“GD”).

The Appellant and the victim, respectively aged 31 and 22, were People’s Republic of China nationals studying in Singapore. They previously each rented a room in the same condominium unit (“Unit”) together with other tenants. The Appellant left the Unit sometime at the end of August 2023. In early October 2023, the Appellant started to pursue the victim romantically. She repeatedly rejected his advances, but he continued to pester her. As a result of the Appellant’s harassment, which included occasions when he returned to the Unit to look for her, the victim went to stay at a friend’s place for a few days (GD at [9]).

A day before the attempted rape, on 8 October 2023, the Appellant went to the victim’s room in the Unit to speak to her. This was despite the tenancy agent, through whom the Appellant had rented a room in the Unit (“Agent”), having warned him against going to the Unit. The victim spoke to the Appellant briefly but refused to engage further. She also called the police, as did the Agent. The Appellant left the Unit when the police arrived (GD at [10]–[11]). However, he remained at the bus stop outside the condominium and continued to call and send text messages to the victim until the early hours of 9 October 2023. This formed the subject matter of one of his TIC charges, being an unlawful stalking charge (GD at [12]).

Later that day, at around 9pm, the Appellant returned to the Unit. This formed the subject matter of his first criminal trespass charge. The victim was in her room. The Appellant knocked on her door and called out to her. He explained that he had left his bank card in the Unit and wished to apologise for his angry text messages the previous night. The victim rejected the Appellant’s apology and reiterated that the Appellant was trespassing. She also said that she would call the police (GD at [14] and [16]–[18]).

As a result of the victim’s reaction, the Appellant became frustrated. The victim later tried to close the door to her room, but the Appellant prevented her from doing so, before forcibly entering the room and walking towards the victim (GD at [18]–[19]). The victim tried to push the Appellant out of the room, but he pushed against her, causing her to fall backwards to the ground. The Appellant pinned her down and spread her legs, kneeling in between her legs (GD at [20]).

In his anger, the Appellant pulled at the victim’s dress. The victim was not wearing a bra at the time, and the Appellant grabbed at least one of her breasts over her dress. He then removed her panties, and pulled down his pants to reveal his penis, before touching the victim’s vagina (without any penetration). The Appellant admitted that he had done these actions because he wanted to have sex with the victim, to humiliate her, and to “release a bit of anger” (GD at [21]).

The victim struggled against the Appellant and tried to push him back, but to no avail. She screamed as the Appellant continued to touch her vagina with his hand (without penetration) (GD at [22]).

One of the tenants in the Unit, [Q], heard the victim’s screams, and went to her room. Seeing the Appellant pinning the victim to the ground while she struggled and screamed, [Q] dragged the Appellant away from the victim and out of the room. The victim then closed her room door immediately. Afraid of the Appellant, [Q] also went back to his room and locked the door (GD at [23]).

While in her room, the victim began to cry and texted her two other roommates and two friends for help. The Appellant began hitting the wall near the victim’s room door, which was heard by [Q]. The victim called for [Q], who came out of his room. The Appellant stopped hitting the wall. He informed [Q] that he was feeling stressed and asked if he could have some alcohol. [Q] obliged, and the Appellant drank some Yomeishu alcohol (GD at [24]).

After calming down, the Appellant continued to try and speak to the victim from outside her door, but she did not respond. Instead, she got one of her roommates to call the police. [Q] also texted one of their roommates, [Z], about the incident, and [Z] subsequently called the police (GD at [25]).

Following the incident, the victim was examined by a doctor and found to have sustained the following injuries: (a) left inner proximal thigh scratch mark; (b) small bruising over left inner mid-thigh; (c) left anterior knee bruising; (d) left inner mid shin area scratch mark; and (e) right lateral mid-thigh scratch mark.

Arguments below

In the hearing below, the Prosecution sought a sentence of seven to eight years’ imprisonment with three strokes of the cane for the attempted rape charge, based on an application of the framework set out in Public Prosecutor v Andrew Kumaravel (HC/CC 51/2024) (“Andrew Kumaravel”):

(a) Step 1: Apply Stage 1 of the framework for rape offences as established in Ng Kean Meng Terence v Public Prosecutor [2017] 2 SLR 449 (“Terence Ng”) to account for the offence-specific factors.

(b) Step 2: Apply a discount in recognition of the inchoate nature of the offence.

(c) Step 3: Apply Stage 2 of the Terence Ng framework to account for the offender-specific factors.

For context, under Stage 1 of the Terence Ng framework, the court first considers the offence-specific factors including any presence of: (a) group rape; (b) abuse of position and breach of trust; (c) premeditation; (d) excessive force; (e) rape of vulnerable victim; (f) forcible rape of victim below 14; (g) hate crime; (h) severe harm to victim; and (i) deliberate infliction of special trauma, to ascertain the gravity of the offence: Terence Ng at [44]. It then places the offence within one of the following sentencing bands (Terence Ng at [47]):

(a) Band 1: 10–13 years’ imprisonment and six strokes of the cane;

(b) Band 2: 13–17 years’ imprisonment and 12 strokes of the cane; and

(c) Band 3: 17–20 years’ imprisonment and 18 strokes of the cane.

Under Stage 2 of the Terence Ng framework, the court makes the necessary adjustments to account for the offender-specific aggravating and mitigating factors: Terence Ng at [73(c)].

Applying the framework in Andrew Kumaravel, the Prosecution argued that:

(a) At Step 1, the appropriate indicative starting point should be 12 years’ imprisonment and six strokes of the cane, highlighting that the Appellant had caused bruising and scratch marks on the victim’s legs while prying them open.

(b) At Step 2, the appropriate incomplete offence discount is two to three years’ imprisonment and three strokes of the cane, as the Appellant had succeeded in exposing both his and the victim’s genitals, and failed to desist in his actions.

(c) At Step 3, given the stage at which the Appellant pleaded guilty, a discount of 20% should be accorded pursuant to the Sentencing Advisory Panel’s Guidelines on Reduction in Sentences for Guilty Pleas (“PG Guidelines”).

The Appellant, on the other hand, applied the sentencing framework as set out in Public Prosecutor v Khor Khai Gin Davis [2023] SGHC 304 (“Davis Khor”) at [45]. The key difference between the framework in Davis Khor and the framework in Andrew Kumaravel is that the former applies both stages of the Terence Ng framework first before applying the incomplete offence discount, ie, Step 2 is applied after Step 1 and Step 3.

Applying the framework in Davis Khor, the Appellant argued that:

(a) At Stage 1 of the Terence Ng framework, his case should be placed in the lowest end of Band 1, for a starting sentence of ten years’ imprisonment. A six-month uplift should then be applied to account for his TIC charges. Next, at Stage 2, a 20% plead guilty discount should be applied pursuant to the PG Guidelines, to arrive at a sentence of 8.5 years’ imprisonment and six strokes of the cane.

(b) At the final step of the Davis Khor framework, the Appellant sought an incomplete offence discount of 3.5 years and three strokes of the cane, to arrive at a final sentence of five years’ imprisonment and three strokes of the cane. He emphasised that he did not complete the rape offence even though he could have done so.

Decision below

The Judge preferred the sentencing framework in Andrew Kumaravel (GD at [50]), for two reasons. First, the court would, in determining the incomplete offence discount, focus on harm and culpability considerations. Since these are offence-specific considerations, it would be more principled to consider them after the consideration of the offence-specific factors in Stage 1 of the Terence Ng framework. It would, in contrast, be “artificial” to return to consider such offence-specific factors after the offender-specific factors have been considered in Stage 2 of the Terence Ng framework (GD at [45]). Second, the Judge explained that the approach in Andrew Kumaravel was more aligned with the approach contemplated in the PG Guidelines, whereby the plead guilty discount will only be applied after the individual sentence for an offence is decided on a claim trial basis (GD at [47]–[48]).

Applying the Andrew Kumaravel framework, the Judge held that:

(a) At Step 1, the offence fell within Band 1 of the Terence Ng framework, and an appropriate indicative starting sentence would be 11 years’ imprisonment and six strokes of the cane (GD at [51]). She did not give aggravating weight to the injuries suffered by the victim, although she considered the Appellant’s conduct after the act of attempted rape to be aggravating. Specifically, the Judge found that even after [Q] prevented the Appellant from raping the victim, the Appellant remained unrepentant. He continued to harass the victim and [Q] while remaining in the Unit without authorisation, even having the audacity to ask for and drink alcohol in the Unit. Throughout this time, the victim remained fearful. Her ordeal was thus a prolonged one, which lasted beyond the attack (GD at [54]).

(b) At Step 2, considering the factors that: (i) the incomplete offence discount is given to recognise that less harm is caused to the victim, as compared to a case where the offence is completed; and (ii) the incomplete offence discount should reflect the offender’s culpability in attempting to carry out the offence, the Judge found the Appellant’s case to be less serious, and accorded the Appellant a larger discount, than that in Andrew Kumaravel. Specifically, whereas the offender in Andrew Kumaravel received a discount of two years’ imprisonment and three strokes of the cane, the Judge accorded the Appellant a discount of three years’ imprisonment and three strokes of the cane (GD at [56]–[59]). She declined to base the discount to that given in Davis Khor, observing that it was too generous (GD at [60]). The Appellant’s sentence at the end of Step 2 was thus eight years’ imprisonment and three strokes of the cane.

(c) Finally, at Step 3, the Judge declined to accord mitigating weight to the Appellant’s purported clean record and personal hardship. She also did not apply any uplift on account of the TIC charges, which she found to be not specifically aggravating for the purpose of the attempted rape charge. Based on the stage at which the Appellant pleaded guilty, the Judge applied a 20% discount according to the PG Guidelines (“PG Discount”). This resulted in a sentence of six years and six months’ imprisonment with three strokes of the cane for the attempted rape offence (GD at [61]–[64]).

For his three proceeded criminal trespass charges, the Appellant was respectively sentenced to two weeks’, four weeks’, and six weeks’ imprisonment (GD at [70]). The Judge then ordered that the sentence for the attempted rape charge run consecutively with that for the most serious criminal trespass charge, resulting in a global sentence of six years, six months and six weeks’ imprisonment with three strokes of the cane (GD at [72]).

Procedural background

Given the difference in the sentencing approaches for attempted rape offences employed by the Judge and in Davis Khor (see [20] and [22] above), an issue arises as to what the appropriate approach for sentencing attempted rape offences (and more broadly, offences involving criminal attempts) should be. As alluded to above, this appeal is the first matter before this court involving the sentencing of a criminal attempt offence after the amendment to the Penal Code, under which the previous statutory one-half sentencing limit for criminal attempt offences was abolished (see [29] below).

To this end, Mr Chenthil Kumar Kumarasingam and Mr Sui Yi Siong were respectively appointed Independent Counsel and assisting Independent Counsel (collectively, “ICs”) to address the following question: “What is the appropriate sentencing framework that should be adopted for an offence of attempted rape under s 375 read with s 511 of the Penal Code 1871”, and the following sub-questions:

(a) What approach should the court adopt in determining the appropriate sentence for the offence of attempted rape?

(b) What should the maximum and/or minimum adjustment(s) to the offender’s sentence (if any) be on account of the fact that the offence was incomplete?

(c) If adjustments are to be made to the offender’s sentence on account of the fact that the offence was incomplete, how should the court determine the appropriate adjustments?

As the questions of what the appropriate approach to sentencing criminal attempt offences should be and how the appropriate incomplete offence discount should be determined are overarching issues in CCA 28, we deal with them first. We then apply the framework as will be established to determine if the sentence imposed by the Judge for the attempted rape offence was manifestly excessive. Finally, we consider the Appellant’s conduct in this appeal and what sanctions, if any, should be imposed on the Appellant to address that conduct.

For completeness, although the appeal was against the Appellant’s sentence generally, he did not offer any arguments as to why the sentences imposed for the criminal trespass offences were manifestly excessive, or why the Judge was wrong to have run the sentences for the attempted rape offence and the most serious criminal trespass offence consecutively. We see no reason to interfere with the Judge’s decisions on either of these matters, which we consider to be eminently justified. Our decision below therefore deals exclusively with the sentence for the attempted rape offence.

Approach to sentencing criminal attempt offences

Background

Prior to the Criminal Law Reform Act 2019 (Act 15 of 2019) (“CLRA”) coming into force, the maximum imprisonment term for criminal attempt offences was limited to half of that prescribed for the completed offence. This changed on 1 January 2020, when Parliament accepted the recommendation by the Penal Code Review Committee (“PCRC”) to abolish the statutory one-half sentencing limit for criminal attempt offences. By s 167 of the CLRA, inter alia, s 512(2) of the PC was amended to read as follows:

A question thus arises as to how the court should determine the appropriate sentence in cases involving attempts.

The parties’ arguments

The ICs’ opinion

In this regard, the ICs opine as follows:

(a) For the same reasons as those stated by the Judge (see [22] above), the Judge’s approach should be favoured over that taken in Davis Khor. In addition, the Judge’s approach offers a clearer theoretical framework and is likely to reduce the risk of double-counting or missing out relevant factors. However, the incomplete offence discount should, contrary to the Judge’s approach, be expressed as a percentage, as this would allow for a more principled comparison across cases to assess the appropriate size of the discount.

(b) There should be no fixed maximum limit to the incomplete discount accorded, although it would only be in exceptional cases where a very large discount should be given. Parliament had abolished the one-half sentencing limit for criminal attempts to broaden judicial discretion in the sentencing of criminal attempts. Hence, limiting the maximum discount would be an undesirable fetter on judicial discretion. That said, in exceptional cases where a substantial step is taken towards the commission of rape, but no actual harm is caused to the victim, the appropriate discount may be more than 50%.

(c) An attempt would attract a lower sentence than the notional completed offence. This reflects the decreased harm caused in criminal attempts. However, there should be no minimum limit to the incomplete offence discount per se; it would be overly prescriptive to suggest any minimum discount given the varied possible fact patterns in criminal attempts. That said, in the worst type of cases, it is possible to deny an offender any discount.

(d) The precise quantum of the incomplete offence discount should be assessed by reference to three factors:

(i) The reason(s) why the outcome did not materialise, which goes towards culpability: An offender who voluntarily desists in his offending will be significantly less culpable than one who does not. In cases where the offender does not complete the offence due to his physical inability or other external reasons, or where the offender seeks to persist even after his offending becomes thwarted, he should receive a smaller incomplete offence discount or no discount at all.

(ii) How far the offence progressed, which goes towards harm: The earlier the offence ceased, the lower the harm caused, intended to be caused, or might foreseeably have been caused.

(iii) The difference between the actual harm caused and the harm that would or could have been occasioned by the notional completed offence.

The first two factors are interconnected: the earlier an offender voluntarily desists in his offending, the lower his relative culpability.

(e) It would not be unreasonable to accord an incomplete offence discount of 50% or more in appropriate cases. On the other end of the range, in most attempted rape cases where the notional rape offence is almost completed, but the attempt was thwarted, an incomplete offence discount of 15–30% might be appropriate. However, given the paucity of precedents, it would not be appropriate for the court to formulate a framework for determining the precise incomplete offence discount. Instead, it would be more appropriate to let the jurisprudence develop without a formal framework.

(f) Nonetheless, the ICs propose the following framework, in the event we are minded establishing one for determining the incomplete offence discount in attempted rape offences:

The Appellant’s arguments

The Appellant did not file any written submissions on the questions posed to the ICs. During the hearing, he referred to the ICs’ framework set out at [31] and noted that it may have a “huge impact” on his sentence as it proposed a discount of 10%–30%.

The Prosecution’s arguments

Like the ICs, the Prosecution favours the Judge’s approach over that in Davis Khor. This is for the same reasons as those stated by the Judge (see [22] above). However, the Prosecution argues that the said discount should be determined as an absolute value and not as a percentage, because this would facilitate a comparison of the case at hand with the relevant precedents.

The Prosecution also submits that no maximum or minimum incomplete offence discount should be established in cases involving criminal attempts, since that would go against Parliament’s intention of giving the courts the full discretion to determine that discount. Instead, the determination of the precise discount should be left open to judicial discretion and/or the development of case law.

Our decision

We first consider the PCRC’s rationale for recommending the abolition of the statutory one-half sentencing limit for criminal attempt offences. While the PCRC agreed with the principle that a discount ordinarily ought to be applied to sentences for criminal attempt offences (relative to the completed offence), it took the view that the precise discount to be given is more suited for judicial discretion rather than applying an arbitrary one-half reduction: Penal Code Review Committee, Penal Code Review Committee Report (August 2018) (Chairman: Ms Indranee Rajah SC and Mr Amrin Amin) (“PCRC Report”) at p 202 para 32. According to the PCRC, its recommendation was aimed at allowing for the imposition of sentences that more accurately capture the culpability of the offender, notwithstanding that the prohibited outcome did not materialise. There is no reason in principle why someone who attempts an offence is only half as blameworthy as someone who has completed the offence: PCRC Report at p 202 para 34.

With this rationale in mind, we turn to consider the appropriate approach to sentencing criminal attempt offences. While the present case specifically concerns an offence of attempted rape, we see no reason why the same approach cannot be applied across all criminal attempt offences under s 511 of the PC. This is so especially since the CLRA had, on the above-explained rationale, abolished the statutory one-half limit in relation to all criminal attempt offences under s 511 of the PC. Moreover, as will become evident below, the factors which ought to affect the quantum of the incomplete offence discount are consistent across all criminal attempt offences.

The broad approach

On a broad level, we agree with the Judge, the Prosecution and the ICs that, in sentencing criminal attempt offences, the three-staged approach established in Andrew Kumaravel (see [16] above) is preferred over the approach set out in Davis Khor. This is for two main reasons.

First, as the Judge alluded to (GD at [45]; see [22] above), it is more principled to consider the incomplete offence discount, which engages offence-specific considerations, immediately after (and, in a sense, together with) the other offence-specific considerations relating to the completed offence, separately from the offender-specific considerations. As this court explained in Terence Ng (at [37]), it is more principled for a sentencing court to draw a distinction between the offence- and offender-specific factors when sentencing an offender:

Second, as the Judge also alluded to (GD at [46]–[48]; see [22] above), it would better align with the PG Guidelines for the courts to apply the incomplete offence discount before considering the offender-specific considerations in a case. The PG Guidelines recognise that the court should first determine “the sentence that it would have imposed if the accused person had been convicted after trial”, before applying the appropriate reduction to the sentence to account for any plea of guilt. Hence, in relation to a criminal attempt offence, the plead guilty discount should only be accorded after the court derives the sentence it would have imposed had the offender claimed trial. Such a sentence should already encompass the incomplete offence discount.

For these reasons, the court should sentence the offender by determining and applying the appropriate incomplete offence discount after determining the starting point sentence, but before adjusting the sentence to account for any offender-specific factors.

On the issue of whether the incomplete offence discount should be expressed as a percentage or absolute value, we agree with the ICs (see [31(a)] above) that the former would be more appropriate. The concept of a sentencing “discount” connotes that reference should be taken from the original sentence. Indeed, we note that other types of “discounts” accorded under the law are similarly expressed as a percentage of the original sentence. An example is the discount for pleading guilty accorded under the PG Guidelines. This must be so as a matter of logic, for the original sentence provides a reference point against which the quantum of the discount can be ascertained. It would also facilitate a meaningful comparison with other cases and facilitate consistency in sentencing across all criminal attempt offences. Indeed, as will become evident below, the sentencing considerations which are relevant to determining the incomplete offence discount would equally apply across all criminal attempt offences, regardless of the primary offence being attempted. In contrast, determining the discount in terms of an absolute amount will offer no meaningful guidance or comparison for future cases given the difference in starting sentences for different offences and even for the same or similar offences.

Appropriate range of incomplete offence discounts

This leads to the question of whether there should be a maximum and/or minimum incomplete offence discount in cases involving criminal attempt offences. In effect, the issue pertains to what the appropriate range of incomplete offence discounts should be. We take the view that, save in exceptional situations (see [65] below), it would be appropriate for the court to accord some incomplete offence discount.

This is so for the primary reason that the harm caused in a criminal attempt offence would, save in an exceptional situation (see [65] below), be less serious than the harm which would have been caused had the offence been completed. This was explained in Chua Kian Kok v Public Prosecutor [1999] 1 SLR(R) 826 at [31]:

In other words, while it is true that the harm which would have been caused had the offence been completed did not actualise, criminal attempts still present a lower-level harm. That this lower harm should be punished is also consistent with the well-established principle that the court may have regard to the potential harm that could have resulted from an offence, even if it did not materialise: see, eg, Neo Ah Luan v Public Prosecutor [2018] 5 SLR 1153 at [66]–[67].

Consequently, it would be principled to accord some incomplete offence discount in cases involving criminal attempt offences. Indeed, the PCRC agreed with (and the Prosecution and ICs accept – see [31(c)] above) the principle that, all things being equal, an attempt ought not to be punished as severely as the completed offence: PCRC Report at p 202 para 32.

However, the same cannot be said in relation to the offender’s culpability. Where an offender desists in his attempt involuntarily, his culpability could be deemed just as high as an offender who completes the offence. For instance, as noted by the PCRC (PCRC Report at pp 202–203 paras 34–37), in Public Prosecutor v Huang Shiyou [2010] 1 SLR 417 (“Huang Shiyou”) (see [8]), the offender attempted unsuccessfully to insert his penis into the victim’s vagina about ten times before desisting. In such situations, it could be said that the offender’s culpability was not any lower than that of an offender who completes the offence. There is thus no reason to accord some incomplete offence discount on account of an offender’s culpability.

At the other end of the range, we take the view that there should be some upper limit to the incomplete offence discount accorded in cases involving criminal attempts. For one, to hold otherwise would lead to the illogical outcome that offenders who commit criminal attempt offences might theoretically receive a 100% incomplete offence discount and go unpunished. Such an outcome is also wrong in principle, for two reasons:

(a) as explained at [43] above, some extent of harm would inevitably exist in all criminal attempt offences; and

(b) having attempted the offence, and given the actus reus of a criminal attempt under s 511(1) of the PC that the offender has taken “a substantial step towards the commission” of the offence, the offender would inevitably be somewhat culpable.

Determining precise incomplete offence discount

In determining the precise incomplete offence discount to be accorded in criminal attempt offences, we agree with the Judge (see GD at [56]) and (partially) with the ICs (see [31(d)] above) that the court should consider the factors of: (a) the progress of the attempt (“Progress Factor”); and (b) whether the offender voluntarily desisted in his actions and why (“Desistance Factor”). These factors respectively indicate the decreased harm and culpability in a criminal attempt offence (relative to the completed offence), which are the two principal parameters which a sentencing court would generally have regard to in evaluating the seriousness of a crime: Lim Ying Ying Luciana v Public Prosecutor [2016] 4 SLR 1220 at [28]. Indeed, these factors are consistent with those considered by the English and Australian courts in deriving the appropriate incomplete offence discount for criminal attempt offences: see, eg, R v Reed [2021] 1 WLR 5429 at [24]–[25]; R v BI (No 4) [2017] ACTSC 71 at [41]; R v Dawne [2022] ACTSC 64 at [36]–[37].

As regards the Progress Factor, as earlier explained (see [43] above), save in an exceptional situation, criminal attempt offences involve a less serious form of harm (relative to completed offences). As an offender progresses with his criminal attempt, the extent of such harm increases. The incomplete offence discount should therefore correspondingly decrease.

As regards the Desistance Factor, an offender who voluntarily desists out of remorse undoubtedly possesses a lower level of culpability than one who involuntarily desists, ie, was prevented from committing the offence by factors outside of his control. It is thus appropriate that the former receives a larger discount. In determining the actual discount to be accorded, the court should primarily consider the stage at which the offender had voluntarily desisted out of remorse. An offender who voluntarily desists early is substantially less culpable than one who only desists just before he completes the offence. The former should thus enjoy a larger discount.

On the other hand, as earlier explained (see [45] above), the culpability of an offender who desists involuntarily could be as high as a hypothetical offender who completes the offence. In such a situation, no further discount should be given on account of the Desistance Factor in addition to that given on account of the Progress Factor.

It is, however, not a foregone conclusion that an offender who involuntarily desists is inevitably as culpable as an offender who completes the offence. To illustrate, consider a situation in an attempted rape where the offender fails to penetrate the victim’s vagina on his first attempt, and desists immediately thereafter. One could say that the offender’s desistance was involuntary; he could not penetrate the victim’s vagina although he had wanted to. However, his culpability is arguably lower than the offender in Huang Shiyou, who had persisted in attempting to penetrate the victim’s vagina. This factor is also implicitly accepted by the ICs, as evident from how their proposed framework seeks to distinguish between offenders whose attempts were “thwarted” and those whose attempts were “thwarted but [remained] persistent” (see [31(f)] above).

Further, in cases where an offender involuntarily desists early into the attempt, it might be speculative to say that, had he not been stopped, he would not have voluntarily desisted in his actions out of remorse subsequently. It hence stands to reason that, especially when an offender involuntarily desists in his actions early, he ought generally to get some discount (albeit a small one) on account of the Desistance Factor.

For completeness, we do not agree with the ICs that the court should separately consider the difference between the actual harm caused and the harm that would or could have been occasioned by the notional completed offence (see [31(d)(iii)] above). In our view, the difference in harm will already be accommodated within the three-stage approach to sentencing criminal attempt offences as we will set out below. This is so in two aspects: (a) the difference in harm will be accounted for when the court considers the offence-specific factors (which, as will be explained at [55] below, considers the actual steps taken by the offender in his attempt) at the first stage of the analysis; and (b) in some cases, this difference will also be accommodated in the court’s consideration of the Progress Factor and within the discount bands in the framework at the second stage of the analysis.

The approach to sentencing criminal attempt offences

Drawing the threads together, adopting the broad approach taken by the Judge as we endorsed earlier (at [40]), a sentencing court should determine the sentence for a criminal attempt offence in three stages.

(1) Stage 1: Apply the offence-specific factors in relation to primary offence with appropriate adaptations

At Stage 1, the court should determine the starting sentence with reference to the offence-specific factors relating to the primary offence. In doing so however, the court should adapt these factors for the context of a criminal attempt. In other words, the court should only consider the actual steps taken by the offender in his attempt. In this regard, we agree with the Judge that the focus should be on “the act [of the attempt]” [emphasis in original] (GD at [54]). It should not consider any hypothetical scenario where the criminal attempt offence is completed, as this would be speculative. In that regard, we respectfully disagree with the approach taken in Davis Khor, where the court explained (at [49]) that it “first determined what the appropriate sentence would have been if the statutory rape offence had been completed”.

(2) Stage 2: Determine appropriate incomplete offence discount

At Stage 2, the court should determine the appropriate incomplete offence discount. We are of the view that it would be appropriate to promulgate a guideline framework to determine the quantum of incomplete offence discount which should be accorded in criminal attempt offences. In this regard, we respectfully depart from the Prosecution and ICs’ position (see [31(e)] and [34] above). In our view, it is precisely due to the paucity of reasoned decisions concerning s 511 of the PC that it is appropriate for us to provide some guidance. Indeed, we note that even between the few precedents concerning attempted rape offences, there are discrepancies in what was considered an appropriate discount (see [23(b)] above). Further, the absence of a framework would give rise to an approach where the incomplete offence discount is based effectively on a comparison with Andrew Kumaravel. In effect, the court would be employing a “single starting point” approach. Such an approach is, however, inappropriate, for criminal attempts can manifest in vastly different ways: see Terence Ng at [28].

As earlier explained (at [47]), the incomplete offence discount should be determined on consideration of the Progress Factor and the Desistance Factor. Specifically, on an application of the principles enunciated earlier (see [37]–[53]), the court should approach this issue in three steps, with reference to the following framework:

Progress

Desistance

Band 4

Planning phase

Band 3

Limited progress

Band 2

Moderate progress

Band 1

Significant progress

Step 1: Determine appropriate discount based on Progress Factor

30% – 40% discount

20% – 30% discount

15% – 20% discount

10% – 15% discount

Step 2: Determine appropriate additional discount based on Desistance Factor

Involuntary desistance

0% – 10%

Voluntary desistance out of remorse

30% – 40% discount

20% – 30% discount

15% – 20% discount

10% – 15% discount

Step 3: Add up the two percentages (Step 1 + the applicable row in Step 2) to arrive at total discount

By way of illustration, under the framework, the highest discount given (Planning stage; Voluntary desistance out of remorse) will be 80%, while the lowest (Significant progress; Involuntary desistance) will, subject to what we say below, be 10%. This reflects what we discussed above at [43]–[46] that: (a) even where the offender voluntarily desists early and no actual harm as defined by the offence is inflicted, the offender still bears some culpability and some harm will have been caused; and (b) where the offender was on the verge of committing the offence but was prevented from doing so, the harm inflicted would ordinarily not be as serious as that suffered if the offence had been completed. We also note that under this framework, the court would, consistent with the PCRC and Parliament’s intent, retain a wide discretion to determine the appropriate incomplete offence discount in sentencing criminal attempt offences. Indeed, we stress that the bands are not to be rigidly applied and the court has the discretion to depart from them in a particular case, although cogent reasons should be given.

The framework may, however, be inapplicable where the harm and culpability associated with the criminal attempt cannot be said to be lower than those had the offence been completed. In such a situation, no discount should ordinarily be given. This is explained below (at [65]).

We explain our rationale for, and the corresponding features of, the framework.

(A) Step 1: Determine discount for Progress Factor

At Step 1, the court should determine which of the four bands of the Progress Factor the case at hand falls into, before further granulating the discount to arrive at. As earlier explained, this is to account for the decreased (but invariably present) harm in a criminal attempt offence (relative to the completed offence).

Given that criminal attempts manifest themselves in highly varied ways, we have deliberately kept the four bands under the Progress Factor broad and have avoided being overly prescriptive in defining each band: see also Terence Ng at [28], [30], [32], [34] and [36]–[38]. It should be noted that by this point, the offender has been convicted and, by definition, it has been established that he has taken a substantial step towards the commission of an offence, by engaging in an act which is strongly corroborative of an intention to commit the offence (see ss 511(1) and (2) of the PC). The sentencing court should consider the following descriptors of how a criminal attempt falling within each band could manifest:

(a) Band 4 – Planning Phase: The steps which the offender has taken are strictly limited to those which pertain to planning. Examples of situations falling within this band include those where the offender: (i) reconnoitres the place contemplated for the commission of the offence; (ii) procures the materials to be employed in the commission of the offence; or (iii) procures accomplices for the commission of the offence. While it may be rare for cases to fall within Band 4 as offences will more easily be established the closer the offender is to completing the offence, the acts contemplated within this band still sit within the spectrum of an offender’s progress in attempting to commit an offence. We thus consider it appropriate to include this band within the framework. Indeed, we note that under s 511(2)(c) of the PC, it is envisaged that the act of “reconnoitring the place contemplated for the commission of the offence”, which is an act of planning, could suffice to establish the actus reus of a criminal attempt offence.

(b) Band 3 – Limited progress: The substantial step taken by the offender towards the commission of the offence has gone beyond the planning phase, but limited progress has otherwise been made in the attempt.

(c) Band 2 – Moderate progress: The substantial step taken by the offender towards the commission of the offence is one taken well into the attempt, although it cannot be said that the offender’s actions fall just shy of completing the offence.

(d) Band 1 – Significant progress: The steps taken by the offender fall just shy of completing the offence.

How these bands may apply in the context of an offence of attempted rape is discussed at [90] below.

(B) Step 2: Determine discount for Desistance Factor

At Step 2, the court should determine the appropriate further discount to be accorded on account of the Desistance Factor. This is to account for an offender’s potentially lowered culpability, and it should be done by first considering whether the offender had desisted voluntarily out of remorse or involuntarily:

(a) As explained earlier, the former would inevitably demonstrate a much lower culpability than the latter. The offender should thus enjoy a larger discount, especially if he had desisted earlier on in his attempt.

(b) That said, for the reasons stated at [51]–[52] above, even when an offender’s desistance is involuntary, it is not necessarily true that his culpability is as high as an offender who completes the offence. The court should thus consider if he should be given a discount on account of the Desistance Factor. The discount should nonetheless be small, which explains the upper limit of 10%.

(C) Step 3: Determine aggregate incomplete offence discount

Finally, at Step 3, the aggregate incomplete offence discount is determined by adding up the two constituent discounts for the Progress Factor and the Desistance Factor.

(D) Exception

As earlier alluded to, the above framework may not apply to exceptional situations where the harm and culpability associated with the criminal attempt cannot be said to be lower than those had the offence been completed. In such a situation, the rationales respectively underlying the Progress and Desistance Factors are no longer engaged, and no discount should be given. For example, an offender who attempts to steal something from someone (A), but who inadvertently ends up stealing an item of the same or a similar value from another individual (B), cannot be said to have caused a lower form of harm, or to have possessed a lower level of culpability, than an offender who successfully steals from A. Neither the rationale underlying the Progress Factor nor that underlying the Desistance Factor is engaged, and the offender should receive no discount.

Another good illustration of this exception is the case of Public Prosecutor v Mas Swan bin Adnan [2012] 3 SLR 527 (“Mas Swan”). There, the offender had imported diamorphine, thinking it was ecstasy. He was convicted for attempting to import ecstasy. The Court of Appeal declined (at [59]) to give the offender any sentencing discount on the basis that his offence was not completed, because he had in fact imported a more serious Class A controlled drug. This decision can similarly be rationalised on the analysis canvassed in the preceding paragraph:

(a) The Progress Factor is not engaged because it is no longer true that the harm caused by the criminal attempt is lower than that which would have been caused had the offence been completed (cf [43] above).

(b) The Desistance Factor is not engaged because the fact is that the offender had completed the acts constituting a more serious offence. His culpability thus cannot be said to be lower than that of an offender who completes the offence of importing ecstasy.

We digress to observe that we do not agree with the court’s approach in Mas Swan of calibrating the offender’s sentence at Stage 1 of the sentencing process with reference to precedents involving the importation of diamorphine, rather than ecstasy. As the offender was charged with the attempted trafficking of ecstasy, the starting sentence should have been determined with reference to precedents involving the trafficking of ecstasy. It would moreover amount to double counting if, on top of determining the offender’s starting sentence with reference to precedents involving the more serious offence of importation of diamorphine, the court does not grant an incomplete offence discount because the offender had inadvertently engaged in offending conduct associated with a level of harm and culpability which was not lower than those had he completed the offence of importing ecstasy: see Public Prosecutor v Raveen Balakrishnan [2018] 5 SLR 799 (“Raveen”) at [87].

(E) Caning

The above framework does not apply in determining the appropriate number of strokes of the cane to be imposed. The relatively limited range in the percentages stated in the matrix, coupled with the small absolute values of the number of strokes of the cane which could be imposed, would potentially result in an insufficient reduction in the number of strokes of the cane, and a clustering of the number of strokes of the cane imposed for criminal attempts. Instead, the court should consider the appropriate discount, having regard to all the circumstances of the case, especially the Progress and Desistance Factors.

(3) Stage 3: Adjust for offender-specific factors

After determining the incomplete offence discount at Stage 2, the court should, at Stage 3, make the appropriate adjustments to account for any offender-specific aggravating or mitigating factors. In line with the PG Guidelines, the court should then apply the appropriate plead guilty discount and take a last look at the final sentence to ensure that the totality principle is not breached, especially if multiple offences are involved.

(4) Summary

In summary, in sentencing criminal attempt offences, the court should determine the length of imprisonment using a three-stage approach:

(a) Stage 1: Determine the starting sentence, with reference to the offence-specific factors in relation to the primary offence but adapted for the context of a criminal attempt.

(b) Stage 2: Determine the appropriate incomplete offence discount to be accorded based on: (i) the Progress Factor (indicative of harm); and (ii) the Desistance Factor (indicative of culpability), with reference to the matrix at [57] above. In the rare event where the harm and culpability associated with the criminal attempt cannot be said to be lower than those had the offence been completed, no discount should ordinarily be accorded.

(c) Stage 3: Adjust the sentence based on any offender-specific factors and the appropriate plead guilty discount (if applicable).

With this approach in mind, we turn to consider if the sentence imposed by the Judge on the Appellant for the attempted rape offence was manifestly excessive.

Whether the Appellant’s sentence for the attempted rape offence was manifestly excessive

An appellate court will not ordinarily disturb the sentence imposed by a first instance sentencing court unless, inter alia, the sentence is manifestly excessive or inadequate. When a sentence is said to be manifestly excessive or inadequate, it means that the sentence is unjustly severe or lenient and requires substantial alterations rather than minute corrections to remedy the injustice. A high threshold needs to be met before appellate intervention is warranted: Public Prosecutor v UI [2008] 4 SLR(R) 500 (“UI”) at [12]–[13].

Appellant’s arguments

In his Petition of Appeal, the Appellant expresses his dissatisfaction with his sentence for the attempted rape charge, emphasising the prior relationship he shared with the victim, as well as the circumstances surrounding the commission of the offence. In relation to his encounter with the victim at the material time, he states: “At the time, I tried to find words to say, I kept the room door opened, but she hit me continuously causing my behaviour to her. Then I was stopped by another roommates (sic).” He also states that “[t]he descriptions in the charge … [was] incorrect” and that he “didn’t scratch [the victim] and [he did not] think she got hurt as described”.

Subsequently, on 6 April 2026, when written submissions were due, the Appellant sought to tender the Record of Proceedings as his written submissions. When the Registry tried to clarify with him why he had done so and to confirm if he intended to file any written submissions for his appeal, it was informed by the Singapore Prison Service that the Appellant “refused to submit any writing”. However, shortly before the hearing, the Appellant submitted a letter to the court, where he states that he is “dissatisfied with the facts described during the judgment process” and “hope[s] for a retrial and for the witnesses and [the victim] to participate in this”. He also provides the following account of the material facts:

During the hearing, the Appellant repeatedly emphasised his “rejection towards the Statement of Facts”, alleging, among other things, that he was “unable to complete [his] submissions” in the proceedings below. He also explained that he only “accepted the Statement of Facts partly due to personal reasons”, and that he “wasn’t thinking clearly then”. Moreover, he sought tenuously to explain his prior relationship with the victim, as well as the circumstances surrounding the incident. In this regard, he argued, inter alia, that:

(a) his “housemate’s [presumably, [Q]] action enhanced the severity of this issue, and also interrupted [his] apology to the [victim]”;

(b) he felt that the “conflict [presumably, the attempted rape] did not result in severe personal harm”, and relatedly, that “even if this [presumably, the attempted rape] were to happen, it would not have caused [the victim] monumental personal harm”; and

(c) “[the victim’s room door] has caused certain behaviour in [him]”.

When asked if he was “saying that [the victim] was partly responsible for the attempted rape”, the Appellant clarified that his “earlier reply actually had this implied meaning”.

That said, towards the end of the hearing, the Appellant accepted, in his response to the Prosecution’s arguments seeking an enhanced sentence (see [80] below), that he recognised that “this” (presumably referring to the attempted rape) was partly due to him.

For completeness, the Appellant’s father had, on 14 and 30 March 2026 respectively, tendered two sets of letters to the court, pleading for leniency on behalf of the Appellant. We did not take into account the matters raised by the Appellant’s father. Apart from the fact that the Appellant’s father is not a party to the proceedings and is not authorised to represent the Appellant, the arguments which he raised included matters not canvassed in the Petition of Appeal. It is thus impermissible for them to be raised: see s 378(6) of the Criminal Procedure Code 2010 (2020 Rev Ed).

Prosecution’s arguments

In response, the Prosecution argues, citing Public Prosecutor v Dinesh s/o Rajantheran [2019] 1 SLR 1289, that having been represented by counsel and accepting the SOF without qualification before the Judge, it is improper for the Appellant to challenge the facts stated therein in this appeal. The Prosecution also highlights that the Appellant is raising his prior relationship with the victim for the first time without any substantiation, which is improper. In any event, the factors which the Appellant relies on in the appeal do not assist him, respectively, because: (a) the Judge did not place aggravating weight on the victim’s injuries; and (b) any prior relationship which the Appellant may have shared with the victim merely provided the backdrop against which the Appellant had committed his offences and was not relevant to the offences.

The Prosecution further argues that the Judge correctly held that the present case is less serious than Andrew Kumaravel, and justified a larger incomplete offence discount to the Appellant. Overall, the aggregate sentence imposed on the Appellant is consistent with the one-transaction rule and the totality principle.

During the hearing, the Prosecution sought an enhancement of the Appellant’s sentence on account of his conduct in the appeal. To this end, the Prosecution highlighted that: (a) the Appellant is now blaming the victim and is not taking any responsibility for his own actions; and (b) such conduct is a continuation of his behaviour after [Q] had pulled him out of the victim’s room.

Our decision

In our view, applying the facts to the above-established framework, the Appellant’s sentence of six years and six months’ imprisonment with three strokes of the cane for his attempted rape offence was plainly not manifestly excessive and there is no basis to disturb the same. In fact, as we explain, the Appellant’s sentence was somewhat lenient.

Stage 1

At Stage 1, for an offence of attempted rape, it is appropriate to consider the offence-specific factors set out in Terence Ng (at [44]), but in the context of an attempt.

Applying the Terence Ng factors set out at [17] above, the Judge placed the Appellant within the lower end of Band 1 of the Terence Ng framework, with an indicative starting sentence of 11 years’ imprisonment and six strokes of the cane. In doing so, she placed aggravating weight on the fact that the Appellant had continued to harass the victim even after the incident (see [23(a)] above).

The Appellant argues that the victim did not suffer any injuries (see [73] and [75(b)] above). We have no hesitation in rejecting this argument. The victim’s injuries are reflected in the SOF, which the Appellant agreed to and accepted without qualification. We deal with this in detail below – see [99]–[100]. In any event, the victim’s injuries ultimately made no difference to the sentence given that the Judge did not take them into account (GD at [54]).

As to the Appellant’s emphasis of his prior relationship with the victim (see [73] above), that, too, does not assist him. As the Prosecution argues (see [78] above), it is improper for the Appellant to have raised this on appeal given that it was never raised in the proceedings below. In this regard, the present case is analogous to Annis bin Abdullah v Public Prosecutor [2004] 2 SLR(R) 93 (“Annis”) at [32]. In Annis, the appellant appealed against his sentence and filed a motion for leave to adduce fresh evidence on appeal, in the form of an affidavit filed by him: Annis at [5] and [30]. The motion was dismissed: Annis at [39]. The court re-iterated (at [32]) that an application to introduce fresh evidence will only be allowed in extremely limited circumstances and only where the evidence is relevant and reliable, and was not available in the proceedings below. On the facts, the court found, inter alia, that certain documentary evidence appended to the affidavit was available in the proceedings below, while certain assertions in the affidavit contradicted the statement of facts: Annis at [34]–[38].

The Appellant did not file any application to adduce fresh evidence. In any event, it was open to the Appellant to have raised his purported prior relationship with the victim in the proceedings before the Judge. Yet, he did not do so. It is true that in Iskandar bin Rahmat v Public Prosecutor [2017] 1 SLR 505, the Court of Appeal held (at [72], citing Soh Meiyun v Public Prosecutor [2014] 3 SLR 299 at [16]) that an appellate court should generally admit additional evidence which is favourable to accused persons and which fulfils the conditions of relevance and reliability. In other words, when an accused person is seeking to adduce fresh evidence on appeal, the requirement of non-availability is less paramount than the other two requirements of relevance and reliability.

However, the new purported evidence which the Appellant seeks to raise before us is clearly unreliable and irrelevant. The allegation of a prior relationship is contradicted by the matters set out in the SOF, which specifically mentions the victim rejecting the Appellant’s advances and making a police report against him for harassing her. In any event, any purported prior relationship is irrelevant to the Appellant’s sentence for the attempted rape charge. As highlighted in Terence Ng (at [46]), the legal starting point is that the existence of a prior relationship between the offender and the victim can neither be treated automatically as an aggravating factor nor as a mitigating factor. On the Appellant’s best case, the purported prior relationship is at best neutral – the relationship was one-sided, and, prior to the attempted rape, it was the Appellant who had unilaterally sought to make uninvited advances on the victim: see [6]–[7] above.

On the contrary, the Appellant’s continued and prolonged harassment of the victim after the attempted rape was not the only aggravating factor in the present case. The Appellant had also touched the victim’s vagina to humiliate her (see [10] above). In doing so, he had sought to inflict special trauma on the victim. This would also have aggravated the Appellant’s offending (see factor (i) at [17] above). Had the Judge considered this aggravating factor, she could have placed this case within Band 2 of the Terence Ng framework, with an indicative starting sentence of at least 13 years’ imprisonment and 12 strokes of the cane (see [17] above). That said, in the absence of an appeal by the Prosecution, we refrain from disturbing this aspect of the Judge’s decision.

Stage 2

At Stage 2, the Judge awarded the Appellant an incomplete offence discount of three years’ imprisonment and three strokes of the cane, to arrive at a sentence of eight years’ imprisonment and three strokes of the cane (see [23(b)] above). This effectively translates into an incomplete offence discount of about 27%. In our view, on a consideration of the Progress Factor and the Desistance Factor, such a discount cannot be said to be inadequate. It was, in fact, arguably generous.

(1) Progress Factor

In relation to the Progress Factor, the earlier established framework contemplates four bands which track with various degrees of an offender’s progress (see [57] and [62] above). In our view, in the context of attempted rape offences, an offence falling into each of the four bands could manifest in the following manner:

(a) Band 4 – Planning Phase: An example would be one where the offender reconnoitres the place contemplated for the intended commission of rape.

(b) Band 3 – Limited progress: An example would be one where the offender takes some initial steps towards committing the offence, for example enticing the victim to a secluded location.

(c) Band 2 – Moderate progress: An example would be one where the offender has touched the victim in the course of an attempted rape, or where the offender and/or victim’s private part(s) has/have already been exposed before the attempt ceased. For example, in Tan Beng Chye v Public Prosecutor [1965-1967] SLR(R) 131 (at [7]), the offender had attempted to rape the victim in a public area. Before he could complete the offence, a passer-by was alerted, and the offender fled the scene. Throughout the incident, the offender did not manage to remove the victim’s knickers.

(d) Band 1 – Significant progress: An example would be one where the offender has touched the victim’s vagina with his penis or is just about to penetrate her. For example, in Davis Khor (see [52]), the offender had already removed his pants and attempted to push his penis into the victim’s vagina.

The Appellant’s case would fall within Band 2 (discount of 15–20%). He had already exposed both his and the victim’s genitals, and had touched the victim’s vagina with his hand, while restraining her. These actions indicate that the Appellant was well into his attempt, although we do not think it can be said that they fall just shy of completing the offence. He had not, for instance, sought to penetrate the victim’s vagina with his penis. On the facts, considering that the Appellant had touched the victim’s exposed vagina directly “for awhile (sic)” while restraining her with his own penis exposed, he would fall on the lower end of Band 2. A 15% discount would hence be appropriate.

(2) Desistance Factor

As regards the Desistance Factor, the Appellant was stopped by the victim’s flatmate; his desistance was involuntary. While he remained outside the victim’s room and began hitting the wall near the room door (see [13] above), it does not appear that he had sought to persist in his attempt at raping the victim. He did not, for instance, seek to break into the room. That said, he persisted in harassing the victim after he was stopped. A discount of 5% or less would be appropriate on account of the Desistance Factor.

In so deciding, we are cognisant that the Judge had placed aggravating weight on the fact that the Appellant continued to harass the victim even after the incident (see [23(a)] above). It might appear, at first blush, that we have double counted this factor. But we do not think so. At Stage 1, the Appellant’s continued harassment of the victim is aggravating because this caused the victim to suffer a “prolonged” ordeal (see [23(a)] above). In other words, this factor is aggravating on account of the actual harm caused to the victim. In contrast, at Stage 2, the Appellant’s continued harassment of the victim shows that his culpability was not that much lower than the hypothetical offender who completes the offence, given that he was persistent in continuing to traumatise the victim.

(3) Aggregate discount

Overall, under the framework, the Appellant would have received an incomplete offence discount of 15%–20%. The discount of 27% awarded by the Judge cannot be said to be inadequate and was, in fact, generous.

(4) Caning

We are similarly unable to see how the discount of three strokes of the cane awarded by the Judge can give rise to any cause for complaint. We also observe that, had the Judge placed the present case within Band 2 of the Terence Ng framework at Stage 1 (which she would have been entitled to do – see [88] above), the Appellant would in all likelihood have received several more strokes of the cane.

Stage 3

At Stage 3, the Judge applied the PG Discount in accordance with the PG Guidelines, to arrive at a sentence of six years and six months’ imprisonment (with three strokes of the cane). The Judge did not find any offender-specific factor which might be aggravating or additionally mitigating. We note that the Judge did not impose any uplift for the TIC offences, although it would have been well within her discretion to have done so: see UI at [38].

We therefore reject the Appellant’s argument that the sentence imposed on him was manifestly excessive, and we accordingly dismiss his appeal.

Whether the Appellant’s conduct in the appeal warrants enhancing his sentence

That is, however, not the end of the matter. As earlier alluded to (see [4] above), we found the Appellant’s conduct of the appeal troubling in at least two aspects which we elaborate on below. We cautioned him during his submissions that he was not helping himself by conducting the appeal in this way, and when he persisted, we gave him the opportunity to respond to the Prosecution’s submissions on enhancement. We now explain the specific points that concerned us.

First, the Appellant repeatedly sought to qualify his acceptance of the SOF. In doing so, he sought to undermine the integrity and propriety of the proceedings below. A close examination of the record of proceedings below establishes that his allegations of being confused or mistaken in pleading guilty, and not having the opportunity to address the court, are wholly baseless. Indeed, as we took pains to emphasise to the Appellant during the hearing, there are important safeguards that are designed to ensure that all accused persons: (a) understand, and are able to participate in, the proceedings; and (b) fully understand the charges brought against them and the statement of facts prepared by the Prosecution. In the Appellant’s case, there is no doubt that the Judge was meticulous and took all necessary steps to ensure that he fully understood the charges he was pleading guilty to and the SOF, which he accepted without qualification or reservation:

(a) First, a court interpreter was present to ensure that the Appellant understood the charges, the SOF, as well as the exchanges during the proceedings.

(b) Second, the Appellant was legally represented in the proceedings below.

(c) Third, the Judge took particular care to ensure that the Appellant understood the charges and the SOF which he was pleading guilty to, and that he was electing to plead guilty on his own volition. After each charge was read out, she ensured that the Appellant understood the possible punishment which he faced for the charge before taking his plea. When the Appellant appeared unclear, the Judge stood down the proceedings so that the Appellant’s counsel could speak to him to confirm that he wished to proceed to plead guilty:

Following the stand down, the Judge again confirmed with the Appellant that he wished to plead guilty to the charges, and that he understood the nature and consequences of a plea of guilt. Counsel for the Appellant also confirmed the same. After the Appellant admitted to the SOF, the Judge confirmed with his counsel that the SOF had previously been read out to the Appellant, and that he agreed to it. Even for the TIC charges, the Judge directed the interpreter to pass a copy of them to the Appellant, confirmed with the Appellant that the said charges had been read to him and that he agreed to having them being taken into consideration for the purpose of sentencing, as well as confirmed the same with the Appellant’s counsel. To illustrate, we reproduce extracts from the above-described exchange between the court, the Appellant (ie, the “Accused”), and the Appellant’s then-counsel, Mr Tan Jun Hao, Don (“Tan”):

(d) Finally, the Appellant was, upon request, given the opportunity to further address the court through his counsel. Clearly, the Appellant was comfortable with speaking up during the hearing, and took the opportunity to do so:

We have no doubt that these safeguards were more than sufficient to ensure that the Appellant was pleading guilty to the charges against him and admitting to the SOF with a full understanding of the same and on his own volition. In the circumstances, his contentions in the appeal that he was confused and mistaken were plainly dishonest and made in bad faith.

Second, it is evident from the Appellant’s conduct in the appeal that he utterly lacked any insight into or remorse for his offending behaviour. In his Petition of Appeal, letters to the court and submissions, he repeatedly sought to:

(a) shift the responsibility for his offences, including the attempted rape offence, away from himself;

(b) blame the victim for causing him to perpetrate his attempted rape on her;

(c) blame [Q] for intervening, claiming astonishingly that this prevented him from apologising to the victim. This claim was especially dishonest and egregious given the Appellant’s admission that he wanted to rape and humiliate the victim – the Appellant was only prevented from doing so because of [Q]’s intervention; and

(d) downplay the harm which he had caused to the victim by claiming that he did not cause the victim significant harm, or, for that matter, the injuries suffered by the victim.

In our judgment, the Appellant’s conduct plainly constituted an abuse of the court’s process. At the very least, he had conducted his appeal based on arguments which are manifestly groundless and without foundation, and has employed the process of the court in an improper way: Chee Siok Chin v Minister for Home Affairs [2006] 1 SLR(R) 582 at [34]. As this court recently explained in Lian Hoe Heng v Public Prosecutor [2026] SGCA 25 (“Lian Hoe Heng”) at [33]–[34], such conduct can legitimately be considered for the purpose of sentencing, with the result that it can be a basis on which a sentence that might otherwise be untouched by error or manifest inadequacy may be enhanced by the appellate court. This is so as: (a) such conduct demonstrates a clear lack of remorse and engages the need for specific deterrence; (b) others must be deterred from engaging in similar conduct; and (c) there is a need to protect the integrity of the judicial process, which has been impugned by such conduct: Lian Hoe Heng at [36]–[44].

In this regard, there are two bases on which a court may enhance the sentence of an offender who has abused the process of the court (Lian Hoe Heng at [46]; see also Lian Hoe Heng at [47]–[70]):

(a) In a case where the accused person has pleaded guilty, the court can reduce the mitigatory weight accorded to the plea of guilt, in light of the loss of the utilitarian value of the plea and, to a more limited extent, the absence of remorse.

(b) Second, if the court is satisfied that the offender’s conduct reflects a lack of remorse, the court can treat this as an aggravating factor, and enhance either at an individual sentence level or, in the case of a multiple offender, his global sentence, as part of the application of the totality principle.

In our view, it is appropriate and sufficient in the present case to enhance the Appellant’s sentence on account of his abuse of the court’s process, by reducing the PG Discount of 20% which the Judge had accorded him under the PG Guidelines. Under the PG Guidelines, mitigatory weight is given to a plea of guilt on two justifications: (a) the utilitarian justification (in that a plea of guilt saves the resources of the state that would otherwise have to be expended if there were a trial); and, to a lesser extent (b) the remorse justification (in that a plea of guilt can be an expression of genuine remorse and contrition): Lian Hoe Heng at [47]–[51]. By repeatedly and baselessly seeking in this appeal to qualify his admission to the SOF and undermining the proceedings below, and by repeatedly seeking to shift the blame of his offending onto others, including the victim, the remorse justification is diminished (see Lian Hoe Heng at [52]). The PG Guidelines assume that an offender who pleads guilty is remorseful for his conduct and gives credit for the same. On the facts, this premise does not hold. It is therefore appropriate that we reduce the discount accorded to the Appellant for his plea of guilt to account for his abuse.

We consider a reduction of 15% to be appropriate. The Appellant’s PG Discount is accordingly reduced from 20% to 5%. Given that the Appellant’s abuse involved only or largely the attempted rape offence, we consider it appropriate to apply this reduction only in respect of his sentence for that offence. Consequently, his sentence for the attempted rape offence is enhanced from six years and six months’ imprisonment with three strokes of the cane to seven years and seven months’ imprisonment (rounded down) and three strokes of the cane.

Conclusion

It remains for us to thank Mr Chenthil Kumarasingam and Mr Sui Yi Siong for their assistance to the court.