Pairing Section 300(c) Murder with Section 34 Common Intention:
The Case for a “Co-terminous” Approach
Alan Tan Khee Jin*
(forthcoming article, Singapore Law Review Vol. 38)
Since the Court of Appeal’s pronouncement on accomplice liability in Daniel Vijay, doubts have arisen over the exact mens rea required of the secondary offender for a section 300(c) murder conviction. Several courts have suggested that he must have intended or known that the injuries inflicted by the primary offender were sufficient in the ordinary course of nature to cause death. This article disagrees with that approach, for it unjustifiably confers a more favourable test on the secondary offender. Instead, a “co-terminous” approach should be adopted that subjects all assailants to a consistent section 300(c) test.
I. Murder and Common Intention under section 34: An Overview
Prior to the enactment of the new section 308A on group crimes, prosecutors in Singapore regularly invoked section 34 of the Penal Code to impose constructive liability on secondary offenders or accomplices for violent crimes committed “in furtherance of the common intention of all.” Thus, in cases where death occurs and the primary offender is charged with section 300 murder, the secondary offender can also face a murder charge under one of the limbs of section 300, read with section 34.[1]
In the seminal case of Daniel Vijay s/o Katherasan & others v. PP[2], the Court of Appeal clarified the meaning of section 34’s key phrase that imposes constructive liability—“in furtherance of the common intention of all”. To convict the secondary offender, the Court held that the prosecution must prove that he had shared in the primary offender’s intention to commit the requisite crime. Thus, in a straightforward “single crime” situation where A and B plan to kill the victim with B standing guard and A committing the actual killing, section 34 imposes constructive liability on B even if he had had no actus reus for the killing. In other words, section 34 serves to construct the actus reus and impute it to B, so long as there was a common intention (the mens rea) on both A’s and B’s part to kill.
In the so-called “twin crime” situation, A and B set out to rob the victim but A, the primary offender, formed the intention (either pre-meditated or on the spot) to kill the victim in the course of the robbery. Here, B, the secondary offender, can be convicted of section 300(a) read with section 34 if he can be shown to have had the intention to cause the victim’s death, even if he had had no actual physical role (the actus reus) in the killing. As long as he had participated in the initial crime of robbery, he can be convicted for the collateral crime of murder if he and the primary offender shared in a common intention to kill. Conversely, if his only intention was to rob, B cannot be convicted of murder.
Under the earlier test expounded by the Court of Appeal in Lee Chez Kee v. PP[3], it would suffice to convict B in the above scenario if the prosecution can prove that he had subjective knowledge of the likelihood of A killing the victim. Hence, this is a more favourable test for the prosecution since it need not meet the high bar of proving B’s intention to cause death. Instead, all that needs proving is the lower mens rea of knowledge or foresight of the likelihood of death being caused. For instance, if B had known of an earlier hostility between A and the victim and that A had brought along a lethal weapon, B can be convicted of section 300(a) murder read with section 34 despite his not having had the intention to cause death. Here, B’s actus reus would have been participation in the initial act (the robbery). He would not have had the actus reus of causing the victim’s death since the act—the lethal blow—was committed by the primary offender. Again, the successful invocation of section 34 serves to construct this actus reus for B.
It must be emphasised that the Court of Appeal, in formulating the Daniel Vijay test, did not expressly overrule Lee Chez Kee. In fact, the Court did not view the two tests to be vastly different. Chan Sek Keong CJ, in delivering the judgment in Daniel Vijay, saw common intention as, in principle, a more exacting requirement than subjective knowledge.[4] That said, Chan CJ went on to observe that in certain circumstances, subjective knowledge may be evidence of the existence of a particular intention.[5] Hence, the secondary offender’s subjective foresight of the likelihood of the primary offender committing the collateral offence may be viewed by the court as evidence of the shared intention to commit that offence. In this way, intention can be inferred from that subjective knowledge.
At the outset, it is critical to note that section 34 is also relevant in cases like PP v. Miya Manik[6] where death ensued in a melee and there is insufficient evidence to show who had delivered the fatal blow (or the co-assailants could have escaped arrest). Thus, all the offenders could each have delivered blows, with the sum total of blows causing the victim’s death but without conclusive evidence of the contribution of each party’s blows. In yet another difficult scenario exemplified by PP v. Azlin binte Arujunah,[7] death could have been caused by cumulative injuries arising from the acts of different assailants in a series of separate incidents. The major challenge in each of these cases is to divine the common intention of the parties.
II. The Trouble with Section 300(c)
Reading section 34 with section 300(a) is straightforward. The true difficulty with section 34 only arises when it is paired with section 300(c), where the primary offender acts with the “intention of causing bodily injury to any person, and the bodily injury intended to be inflicted is sufficient in the ordinary course of nature to cause death”. The established case law following the seminal Indian Supreme Court case of Virsa Singh v. State of Punjab[8] has consistently interpreted the latter part of section 300(c) in an objective fashion. This means that as long as the primary offender, A, intended to cause a bodily injury, it would not matter if he did not know that the injury he inflicted was sufficient in the ordinary course of nature to cause death.
Hence, as long as the injury is one that would ordinarily be sufficient to cause death, A would be liable. For instance, if A slashes the victim in the leg with a parang or chopper and the victim bleeds to death, section 300(c) is attracted as long as A can be proved to have intended to inflict that bodily injury (it not being accidental or unintentional or some other kind of injury being intended).[9] It would not matter if A never intended death, or did not know that the injury would be fatal. Indeed, if he did so intend death or know of the likelihood of death, his act would have been covered by section 300(a) (intention to cause death) or section 300(b) (such bodily injury as he knows would be likely to cause death).
While this interpretation of section 300(c) serves commendably to avoid overlaps with the other limbs of section 300 (particularly section 300(b) that requires knowledge of the likelihood of death), it is difficult to reconcile with the retributivist principle that requires the accused to be punished strictly for his moral culpability, and no more. In essence, section 300(c) can potentially convict for unintended or unforeseen consequences going beyond the accused’s subjective moral culpability. As such, it can only be justified (if at all) by the principle of utilitarianism—thus, no one should have the license to go around slashing his victims’ legs,[10] and one takes whatever consequences (including death) that arise from that act. This would provide both specific and general deterrence established by the principle of utilitarianism.
From the retributivist perspective, section 300(c) is thus wholly out of kilter with the other limbs of section 300 that prescribe a higher degree of mens rea for moral culpability. Again, the lowered mens rea in section 300(c) (effectively a form of constructive liability for unintended consequences) can only be justified by strict utilitarianism. Indeed, this construction of section 300(c) may be reminiscent of the now-discredited common law adage that a man is to be taken to intend the natural consequences of his act. As it turns out, the courts in Singapore (and elsewhere in the Penal Code jurisdictions) have long struggled with these uncomfortable implications of section 300(c), particularly when the accused intended a lesser injury but death nevertheless ensued.
One such case is the leading Singapore authority of Lim Poh Lye v. PP.[11] There, Chao Hick Tin JA in the Court of Appeal held that if the offender intended to inflict what, in his view, was an inconsequential injury, where, in fact, that injury is proved to be fatal, he would be caught by section 300(c) for murder.[12] Thus, if the injury caused was clearly intended but the offender did not realise the true extent and consequences of that injury, he would be liable.[13] However, if the accused only intended to cause a particular minor injury, which injury would not, in the normal course of nature, cause death, but, in fact caused a different injury sufficient in the ordinary course of nature to cause death, section 300(c) would not be attracted.[14] Here, Chao JA was clearly stepping away from the dictum of Yong Pung How CJ in the earlier case of Tan Joo Cheng v. PP[15]—that even if the accused intended a relatively minor injury, he would be caught by section 300(c) if death eventuated.
Another important fine-tuning arose in the High Court case of PP v. AFR.[16] Thus, if the cause of death is remote and not within reasonable contemplation, section 300(c) would not be made out. Indeed, this position has since been underlined and confirmed by the prosecution itself after the Attorney-General’s Chambers (“AGC”) released a rare public statement in December 2017 in the wake of the famous case of a young woman, Annie Ee, whose death had been caused by abuse at the hand of her housemates. The victim had been hit with a large roll of shrink wrap, and later died from acute fat embolism ensuing from her wounds. The prosecution explained why murder charges were not brought:
The evidence given by the forensic pathologist was that Ms Ee's death was caused by acute fat embolism. This was an unusual occurrence that would not have ordinarily resulted from the injuries inflicted by Pua and Tan ... As Pua and Tan did not intend to cause Annie's death, and the injuries they inflicted would not ordinarily cause death, the offences of murder and culpable homicide cannot be proved against them.[17]
As seen from the above, section 300(c) is controversial enough for the sole offender simply due to the possible imposition of liability for unintended consequences going beyond her moral culpability. The courts, and the prosecution, have thus bent over backwards to restrict the operation of section 300(c) to cases where the accused intended to inflict (and so inflicted) injuries that would ordinarily have caused death. In other words, if he intended to inflict injuries that were relatively minor and that would not ordinarily cause death, section 300(c) would not be satisfied. That death still ensued in such cases could then be viewed as a remote or unusual occurrence. This is a valiant but fundamentally still unsatisfactory attempt to blunt the anti-retributivist edges of section 300(c). Inevitably, the approach necessitates an enquiry into the seriousness of the injuries, ie whether they were objectively serious enough to ordinarily lead to death.[18]
III. Invoking Section 34 with Section 300(c)
A. An “Elevated Mens Rea” Approach?
In the event, the pairing of section 34 with section 300(c) for the secondary offender(s) poses even thornier problems. Daniel Vijay was just such a case. The overriding policy concern here is that if the primary offender can be convicted on a lowered mens rea under section 300(c) for what is in essence an unintended consequence, extending such liability to the secondary offender would be even more troubling given that he would typically have no actus reus in causing fatal injuries. Indeed, apart from the melee cases, the secondary offender would likely be one or several steps removed from the primary offender’s act of causing fatal injuries.[19] Indeed, his only actus reus would have been participation in the initial crime (of robbery, rape, extortion, etc.) before things got out of hand with the primary offender.
For this very reason, the Singapore courts have also displayed caution (though only relatively recently) in convicting the secondary offender under section 300(c) read with section 34. This would have been especially true in the period before 2013 when the death penalty was still mandatory for all section 300(c) convictions, both for the primary and secondary offenders. In Daniel Vijay itself, the Court of Appeal found that the secondary offenders did not share in the primary offender’s intention to inflict fatal injuries on the victim, despite their knowing that a weapon had been brought to the scene and that violence would be used to overpower the victim in order to facilitate the robbery.[20] As explained above, Chan CJ shifted the test from one of subjective foresight (Lee Chez Kee) to one of intention to commit the collateral act.
In PP v. Ellarry bin Puling,[21] the first case in the High Court to apply the Daniel Vijay test, Chan Seng Onn J interpreted the application of section 34 to section 300(c) as demanding a “high degree of specificity” in the secondary offender’s intent before there could be conviction.[22] In other words, the prosecution had to show that the secondary offender intended to inflict the specific injury that was actually inflicted that caused death. This has led to the observation that the mens rea test for the secondary offender has effectively been elevated to that of section 300(a), ie he had to be shown to have the intention to cause death. In fact, Chan J. himself alluded to such elevation.[23]
In the last year or so, there have been two High Court decisions—PP v. Azlin binte Arujunah and PP v. Miya Manik—professing to employ such “elevated mens rea” to the secondary offenders. In fact, these are the first judgments to have explicitly required the secondary offenders to intend (or know) that the injuries inflicted were sufficient in the ordinary course of nature to cause death. In PP v. Azlin binte Arujunah, Valerie Thean J, in acquitting the two co-accused for section 300(c) murder of their child, found that:
… in order for constructive liability to be imposed under s 300(c) read with s 34 of the Penal Code, the offenders must share a common intention to cause s 300(c) injury, and not any other type of injury, meaning that the fact that the injury is sufficient in the ordinary course of nature to cause death must be intended. [24]
Incidentally, Thean J was also the judge in PP v. Miya Manik, a judgment released just two days before Azlin bte Arujunah. In Miya Manik, Thean J acquitted the accused on similar grounds that common intention to inflict a section 300(c) injury had not been proven beyond reasonable doubt. In addition, she stated that “the fatal nature of the injury had also to be commonly intended.”[25] Interestingly, in neither case was it made explicit that the prosecution had failed to show that the secondary offenders intended or knew that the injuries would be sufficient in the ordinary course of nature to cause death. Instead, both cases were decided on the more rudimentary basis that there was no proof to begin with of a common intention to inflict the kind of injuries that would attract section 300(c).[26] The acquittals for section 300(c) then led to the accused in both cases being charged and convicted for causing grievous hurt under section 326 read with section 34.
In other words, there did not appear to be a need to elevate the secondary offenders’ mens rea to that of section 300(a) to dispose of the cases. That said, it may well have been a relevant consideration in Thean J’s mind—that the accused in both cases (viewed as secondary offenders) would need to be shown to have the elevated mens rea for conviction, and that the prosecution had therefore failed to prove common intention on the whole.[27] In any event, to intend (or know) that the injury was sufficient in the ordinary course of nature to cause death goes against the established grain of section 300(c) jurisprudence. In other words, if Thean J is correct, the test for the secondary offender has become different from and more exacting than the corresponding section 300(c) test for the primary offender. Where was the authority for this view?
In reaching her conclusion in both Azlin bte Arujunah and Miya Manik, Thean J drew support from several passages in Chan CJ’s judgment in Daniel Vijay. These passages purportedly sought to distinguish between a section 300(c) test applying to the primary offender simpliciter and that applying to secondary offenders.[28] In particular, Thean J sought to rely on the following passage from Chan CJ’s judgment:
[W]e are of the view that he [the secondary offender] should not be made constructively liable for the offence of s 300(c) murder arising from the actual doer’s criminal act unless there is a common intention to cause, specifically, a s 300(c) injury, and not any other type of injury.[29]
Chan CJ’s short-form reference to “specifically, a s 300(c) injury” is enigmatic. To understand what he meant by this phrase, it is apposite to reproduce his views in Daniel Vijay:
… Different policy considerations apply when imputing direct liability for murder and when imputing constructive liability for that offence. … It does not necessarily follow that the Virsa Singh interpretation of s 300(c), which is applicable to the actual doer, is or should be equally applicable to a secondary offender, especially where the secondary offender did not inflict any injury on the victim at all. In other words, as a principle of criminal liability, it may not be unjust or unreasonable to hold the actual doer liable for s 300(c) murder by applying the Virsa Singh test since (as just mentioned) he was the one who inflicted the s 300(c) injury sustained by the victim. However, it may not be just or reasonable to apply the Virsa Singh test to hold a secondary offender constructively liable for s 300(c) murder where he had no intention to do the specific criminal act done by the actual doer which gave rise to the offence of s 300(c) murder, and also did not subjectively know either that that criminal act might likely be committed or that that criminal act would result in s 300(c) injury to the victim.[30]
Earlier, Chan CJ had observed:
In this respect, it is not sufficient, in our view, for s 34 to apply if the secondary offender merely has subjective knowledge that the victim might likely suffer an injury (or, for that matter, if the secondary offender shares a common intention with the actual doer to inflict an injury on the victim), and that injury is subsequently shown to be of a type which is sufficiently serious to amount to s 300(c) injury.[31]
Chan CJ further observed:
In the context of s 300(c) injury, a common intention to cause such injury is substantially the same as a common intention to cause death by the infliction of the specific injury which was in fact caused to the victim since s 300(c) injury is, by definition, injury that is sufficient in the ordinary course of nature to cause death.[32]
Then, Chan CJ concluded:
In contrast, in the former case (ie, where a secondary offender is charged with murder under s 300(c) read with s 34), because of the express words “in furtherance of the common intention of all” in s 34, it is necessary to consider whether there was a common intention among all the offenders to inflict s 300(c) injury on the victim (the inflicting of such injury being the criminal act which gives rise to the offence of s 300(c) murder). This is a critical distinction to bear in mind.[33]
Now, these passages do reveal that Chan CJ did have in mind a more exacting mens rea requirement for the secondary offender when reading section 300(c) with section 34.[34] After all, he was at pains to explain why different policy considerations should apply to a secondary offender, given that he did not inflict the fatal injury. There is also a hint that the Virsa Singh approach, while drastic, could be fairly applied to the primary offender since it would not be unjust or unreasonable to do so. But it would be wholly different, in Chan CJ’s reckoning, where the secondary offender is concerned. This reluctance must be seen in an important context—at the time, the secondary offender would also suffer the mandatory death penalty for a section 300(c) conviction.[35] Here, it is submitted that Chan CJ’s and Chan J’s concerns in Daniel Vijay and Ellarry bin Puling respectively can be traced to the overriding discomfort with the secondary offenders receiving the mandatory death penalty.
Again, we need to go back to what Chan CJ could have meant by his truncated term “s 300(c) injury” that appears numerous times in Daniel Vijay.[36] In his judgment,[37] he had first defined this as “bodily injury which was sufficient in the ordinary course of nature to cause death (hereafter referred to as “s 300(c) injury)”. That, in itself, reveals little. While Chan CJ did go on to express caution about applying the Virsa Singh approach to the secondary offender,[38] nowhere in the Daniel Vijay judgment is it stated that the secondary offender had to intend or know that the injury inflicted by the primary offender was sufficient in the ordinary course of nature to cause death.
In other words, it is not at all clear that Daniel Vijay is authority for applying the “elevated mens rea” approach to the secondary offender. As stated above, Chan CJ dispensed with the facts in Daniel Vijay by holding that the requisite test was intention—and that the two secondary offenders never shared in the primary offender’s intention to inflict bodily injury. Much like in Azlin bte Arujunah and Miya Manik, whether the secondary offenders intended or knew that the injury was sufficient in the ordinary course of nature to cause death was not directly material to the holding. In any event, it is submitted that Chan CJ’s cautiousness in applying section 34 with section 300(c) lies more with the problems and possible injustices inherent in section 300(c) itself (for which see below) as well as the then-mandatory death penalty.
Even if Daniel Vijay were authority for elevating the secondary offender’s mens rea, it is doubtful that this is the correct approach. To require a higher or more specific mens rea on the part of the secondary offender is to make it exceedingly difficult for the prosecution to secure his conviction. The overwhelming policy for such an approach is, of course, the recognition that the secondary offender is typically—by way of actus reus—removed from the act of causing death. The supposition then, is that his mens rea must exceed that of the primary offender before he, too, can be found guilty of section 300(c) murder.
Such an approach would take care of and make sense in the very rare category of cases like Chia Kee Chen v. PP[39] where the secondary offender was actually the mastermind who had recruited the primary offender and planned the operation to beat up the victim severely. Even if it was unclear whether he delivered the fatal blows, the secondary offender in such cases can still be comfortably convicted for section 300(c) read with section 34 under the “elevated mens rea” approach. Indeed, based on the evidence, his mens rea would exceed the primary offender’s in that he either intended or knew it likely that death would ensue (hence, having a high degree of specificity in his mens rea).
But what about the majority of group crimes that do not involve a mastermind secondary offender? In these cases, the elevation of the secondary offender’s mens rea to approximate section 300(a) will make it almost impossible for the prosecution to secure conviction. This is simply because it is exceedingly difficult to prove an intention to cause death beyond reasonable doubt. In fact, it would be difficult even in a mastermind case.
In cases involving a melee or where the causative contribution of each party towards death cannot be determined, there is no reliable way to tell the primary offender apart from the secondary offender, much less to separate and apply different mens rea tests to each. Even in the “plain vanilla” “twin crimes” where A and B set out to rob but A ends up causing death (B being the mere follower), it does not make sense to demand separate legal tests, to the point of requiring an elevated section 300(a) or section 300(b) mens rea for the secondary offender. If nothing else, requiring that the secondary offender intended or knew that the injuries inflicted by the primary offender would be sufficient in the ordinary course of nature to cause death renders section 300(c) otiose in the light of the existing section 300(a) and section 300(b). What then could be a more satisfactory solution?
B. The “Co-terminous” Approach
The policy reason for favouring the secondary offender who is more removed from the victim’s death is an understandable one. However, resorting to an effective section 300(a) test tips the scales too much in the secondary offender’s favour, making his conviction almost impossible to secure for a section 300(c) charge. Instead, the test to be adopted should be what I term a “co-terminous” one. The secondary offender’s mens rea should be co-terminous with the primary offender’s section 300(c) mens rea, no more and no less.[40] In other words, all that is required to convict the secondary offender is for the prosecution to show that he, too, shared in the primary offender’s intention to cause some bodily injury to the victim. That that bodily injury is one sufficient in the ordinary course of nature to cause death need not be intended by or known to the secondary offender. This would be the familiar Virsa Singh objective approach that we have seen employed in section 300(c) jurisprudence, making the same test applicable to both primary and secondary offenders.
This “co-terminous” test will more justly settle those cases involving melee killings or where it is difficult to assign causation to each offender’s blow(s). In such cases, all the accused persons—be they primary or secondary offenders—should be found guilty of section 300(c) read with section 34 if it can be shown that they all shared in the intention (either pre-meditated or formed on the spot) to inflict bodily injury on the victim. That the injury subsequently proves fatal is a separate objective enquiry that has no bearing whatsoever on either offender’s mens rea. Indeed, consistent with Lim Poh Lye, if the injury caused was clearly intended by both the primary and secondary offender but neither realised the true extent and consequences of that injury, they should both be found liable.[41]
The “co-terminous” approach would thus find both the primary and secondary offenders guilty (or innocent) simultaneously, ie with consistency of result. Here, we must remember that sentencing can now be tailored to the role that each played—the automatic death sentence for section 300(c) is a relic of history. Each offender can be sentenced to the extent of the individual role played and his overall moral culpability, taking into account any mitigating circumstances.
In fact, the “co-terminous” approach is also more consistent with the sentencing principles for secondary offenders laid out in Michael Anak Garing v. PP.[42] In that case, the Court of Appeal held that in deciding whether a secondary offender acted in blatant disregard for human life (and thus meriting the discretionary death penalty under section 300(c)), the relevant factors would be his mental state at the time of the attack and his actual role or participation in it. To elevate the mens rea required of the secondary offender to that of section 300(a) would be to wholly negate the 2012 amendments to the Penal Code (that made the death penalty discretionary for section 300(c)) and the sentencing principles of Michael Anak Garing. Were the mens rea requirement elevated to such a threshold, every secondary offender found guilty under section 300(c) read with section 34 would necessarily exceed the threshold set out in Michael Anak Garing. In every such case, the result can only be the death penalty—why then would Parliament have made the death sentence discretionary?
The “co-terminous approach” does have some judicial support, even if it is not labelled as such.[43] It was seemingly adopted in Chia Kee Chen, where the relevant enquiries were what fatal injury was dealt to the victim and whether the assailants shared a common intention to inflict the injuries in question.[44] At no point did the Court of Appeal consider whether the accused intended or knew that the injury was sufficient in the ordinary course of nature to cause death.[45] Nor did the Court directly address whether a higher mens rea was required. There was no such need as the accused was the mastermind. He would likely have satisfied both the “elevated mens rea” and “co-terminous” tests.
Elevating the secondary offender’s mens rea to section 300(a) would not make sense generally. Why should one offender benefit from a more favourable test compared to the rest? Who should this test be applied to, when the primary offender(s) are unknown or are at large? Even if all the assailants have been apprehended, there could still be evidential difficulties in determining who delivered the fatal blow or what contributions each party’s blows made to the victim’s death. Logically, all of them—whether primary or secondary offenders—should benefit from the same favourable test that approximates section 300(a), and all will likely be acquitted. So where does that leave section 300(c) in group crimes?
C. Back to the Real Problem
This leads us to the nub of the problem. The original sin for the current legal quandary on secondary offenders does not reside in section 34 nor Daniel Vijay. The test in Daniel Vijay requiring the prosecution to prove that the secondary offender intended whatever the primary offender intended (or “shared in the common intention”) is arguably sound, and entirely consistent with the “co-terminous” approach. The real problem is, and has always been, with section 300(c). It is precisely because section 300(c) imposes a form of constructive liability that the secondary offender risks section 34 constructive liability for section 300(c) constructive liability. In other words, constructive liability twice over?
It is submitted that this quandary lies behind the tortured interpretation of the short-form label “s. 300(c) injury” used by the Court of Appeal in Daniel Vijay. The High Court in Azlin bte Arujunah and Miya Manik has now chosen to read “s. 300(c) injury” to require the secondary offender to intend or know that the injuries inflicted by the primary offender were sufficient in the ordinary course of nature to cause death, possibly to distinguish the section 300(c) offence from that of grievous hurt. As argued above, this artificially creates separate tests for the primary and secondary offenders and goes against the long-standing section 300(c) jurisprudence. In fact, nowhere in Daniel Vijay is it clearly explained that “s. 300(c) injury” means the secondary offender has to intend or know that the injuries inflicted by the primary offender would ordinarily cause death. Again, if nothing else, this interpretation renders section 300(c) otiose in the light of section 300(a) and section 300(b).
Another way to imagine the problem with section 300(c) is to compare its ambit with the offence of voluntarily causing grievous hurt resulting in death. By definition, the moment a primary offender intends to inflict and does inflict serious injuries (of the kinds that would ordinarily lead to death), liability for section 300(c) would already have coalesced or been consummated.[46] This is because the fact that the injuries are sufficient in the ordinary course of nature to cause death is a separate enquiry. For secondary offenders, the issue should be no different—if the “co-terminous” approach applies, liability crystallises the moment the intended injuries are inflicted. It is therefore artificial to acquit the secondary offenders of section 300(c) read with section 34, only to amend the charge and convict them of grievous hurt read with section 34. Indeed, one way to get around this artificiality is to insist on “elevated mens rea” to distinguish section 300(c) murder from grievous hurt simpliciter. But this only belies the true problem that is section 300(c).
The true solution is a legislative, not judicial, one. Section 300(c) should simply be abolished for being out of step with the other limbs of section 300. What is left would be section 300(a), for the most part, since section 300(b) and section 300(d) cover rare factual instances.[47] If the prosecution were made to use section 300(a) to prove murder in all instances, then the courts will simply have to inquire if the secondary offender(s) shared in the intention to cause death. No one would need to benefit from an elevated or more favourable mens rea test, which is exactly the kind of legal gymnastics that the current situation compels. The different kinds of accomplice liability cases can all be accommodated by combining section 34 (as interpreted by Daniel Vijay) with section 300(a)—the mastermind, melee, and plain vanilla accomplice cases.[48]
D. Reverting to the Lee Chez Kee test?
Be that as it may, section 300(c) lives. There is possibly a third way to settle the relationship between section 34 and section 300(c). This is to abandon the Daniel Vijay approach altogether and restore the Lee Chez Kee test for section 34—viz. it suffices for the prosecution to prove the secondary offender’s subjective knowledge of the likelihood of the primary offender committing the collateral act. But this does not really solve the problem—in fact, it simply begs the enquiry again: subjective knowledge of the likelihood of what act? That the primary offender was likely to inflict serious/fatal section 300(c) injuries? But to maintain absolute fidelity with the section 300(c) jurisprudence, there must strictly be no enquiry as to the secondary offender’s knowledge of the likelihood of death. Injuries that are “sufficient in the ordinary course of nature to cause death” already connote an objective likelihood; it does not make sense to have a subjective knowledge of the likelihood of an objective likelihood.
In other words, knowledge of the likelihood of death under section 300(c) must remain irrelevant for both the primary and secondary offenders. Under the Lee Chez Kee approach, all the prosecution has to prove is that the secondary offender had some subjective knowledge of the likelihood of the primary offender inflicting an injury (although not a minor one), without more. The problem is that this swings the pendulum too much in the other direction, this time to the prosecution’s favour. All the prosecution needs to prove is that the secondary offender had a suspicion of an injury being inflicted (to reiterate, it cannot be a minor injury, but neither need there be a suspicion of serious injury or death). The mere fact that a weapon is brought would likely trigger this suspicion, and conviction would be too easily obtained.
IV. Conclusion
In sum, there are not two, but at least three possible approaches in interpreting section 300(c) read with section 34. They are:
(a) The “elevated mens rea” approach—the secondary offender benefits in that the prosecution must effectively prove section 300(a) or section 300(b) mens rea beyond reasonable doubt. This is the most favourable position for the accused, but least favourable for the prosecution. Put another way, it does not make sense to charge both assailants for section 300(c) read with section 34 when the secondary offender (if he can be distinguished from the primary offender) is effectively being tried for section 300(a) or section 300(b). For the reasons argued above, this should not be the preferred test.
(b) Resurrecting the Lee Chez Kee test—the prosecution has to prove that the secondary offender subjectively had knowledge or foresight of the likelihood that the primary offender(s) would inflict bodily injury (and the bodily injury is objectively sufficient in the ordinary course of nature to cause death). This is the least favourable approach for the accused, and the most favourable for the prosecution since it does not have to prove intention in any form, simply subjective knowledge. Thus, any knowledge or even suspicion of a weapon being brought along by the primary offender would most likely suffice to convict.[49]
(c) The “co-terminous mens rea” approach—the prosecution has to prove that the secondary offender had (or shared in) the intention of causing some bodily injury, but need not show further that he intended death or knew that that injury was one that would ordinarily cause death. Nevertheless, the prosecution still has the burden to prove intention to cause bodily injury (just not intention to cause death).
My contention here is that the “co-terminous” approach effects the most just and appropriate balance in the adversarial criminal process. The prosecution still has to prove the high mens rea of intention, though only the intention to inflict a bodily injury. Any alternative approach risks tilting the balance too much in favour of either the accused or the prosecution. While these are all imperfect fixes to a problem that is necessarily section 300(c)’s making, the “co-terminous” approach makes virtue out of necessity in best approximating the fairest balance between the prosecution and defence.
V. Epilogue
Although I have some sympathy for the reconsideration of Daniel Vijay and the resurrection of the Lee Chez Kee subjective foresight approach for general application to all crimes (this is fodder for another article), my chief concern here is with the Lee Chez Kee test’s application to section 300(c) murder—again, underlining the fact that the problem is truly section 300(c)’s.
* Professor, Faculty of Law, National University of Singapore. I thank my clever students Ryan Leong Lup Mun and Kai Kiran Gosian (NUS Law Class of 2024) for sharing their incisive thoughts in class and with me. All errors remain mine.
[1] Penal Code (Cap 224, 2008 Rev Ed Sing), s 34: “When a criminal act is done by several persons, in furtherance of the common intention of all, each of such persons is liable for that act in the same manner as if the act were done by him alone” [emphasis added].
[2] [2010] 4 SLR 1119 (CA).
[3] [2008] 3 SLR(R) 447 (CA).
[4] Daniel Vijay, supra note 2, at para 87.
[5] Ibid, at para 89. Nowhere is this clearer than on the facts of Illustration (f) set out by Chan CJ at para 168, providing for the court to infer intention from knowledge.
[6] [2020] SGHC 164.
[7] [2020] SGHC 168.
[8] 4 AIR 1958 SC 465.
[9] Ibid, at para 12.
[10] Cf Vivian Bose J’s own words in Virsa Singh, ibid, at para 13, explaining the rationale for section 300(c) in the Indian Penal Code: “No one has a license to run around inflicting injuries that are sufficient to cause death in the ordinary course of nature and claim that they are not guilty of murder. If they inflict injuries of that kind, they must face the consequences.”
[11] [2005] 4 SLR(R) 582 (CA).
[12] Ibid, at para 23.
[13] Ibid.
[14] Ibid, at para 22.
[15] [1992] 1 SLR 620 (CA). See also the wealth of scholarly ink spilled on the section 300(c) problem, including M. Sornarajah, “The Definition of Murder Under the Penal Code” [1993] SJLS 1, Stanley Yeo, “Academic Contributions and Judicial Interpretations of Section 300(c) Murder”, Singapore Law Gazette (April 2004); Jordan Tan Zhengxian, “Murder Misunderstood: Fundamental Errors in Singapore, Malaysia and India’s Locus Classicus on Section 300(c) Murder” [2012] SJLS 112; and this author, Alan Tan Khee Jin, “Revisiting Section 300(c) Murder in Singapore” [2005] 17 SAcLJ 693.
[16] [2011] 3 SLR 653 (HC).
[17] Ng Huiwen, “Death of Annie Ee: AGC explains why couple who abused her were not charged with murder” (18 December 2017), online: The Straits Times <www.straitstimes.com/singapore/death-of-annie-ee-agc-explains-why-couple-who-abused-her-were-not-charged-with-murder> [emphasis added]. The original statement is no longer available on the AGC website.
[18] See Chao Hick Tin JA in Lim Poh Lye, supra note 11 at para 46, doubting the relevance of seriousness from the accused’s perspective. Chao JA is correct, since the correct enquiry is into objective seriousness—it is immaterial whether the accused subjectively appreciated if the injuries were serious.
[19] But it is not at all clear that this policy consideration is supportable. After all, the very point of section 34 is to ascribe criminal liability to a secondary offender who lacks the actus reus for being removed from the primary offender’s collateral acts. In other words, that is the precise nature of constructive liability.
[20] Daniel Vijay, supra note 2, at para 148.
[21] [2011] SGHC 214 (HC).
[22] Ibid, at para 74.
[23] Ibid, at para 100.
[24] Azlin bte Arujunah, supra note 7, at para 97 [emphasis added].
[25] Miya Manik, supra note 6, at para 78.
[26] In Azlin bte Arujunah, the problem was that there were four incidents that led to the fatal cumulative scald injury, and it could not be established that both accused shared in a common intention to inflict that cumulative injury, or to inflict the injuries in each of the four individual incidents. In Miya Manik, it was found that there was no evidence of a common intention to inflict serious injury. In Thean J’s words, “the assailants simply wished to demonstrate their force without going so far as to inflict fatal wounds”, ibid, at para 105.
[27] It is more likely than not that such elevated mens rea was—in Thean J’s view—dispositive of both cases, thus opening the way for grievous hurt convictions instead. If the same injuries that constituted the offence of grievous hurt were co-terminous with that constituting section 300(c), the only way to distinguish the two offences would be to insist on an elevated mens rea for section 300(c) murder. But a fundamental problem with this interpretation is that it renders section 300(c) otiose in the light of section 300(a) and section 300(b), even if for the secondary offender only.
[28] Azlin bte Arujunah, supra note 7, at para 95.
[29] Ibid, at para 91 [emphasis in original], referencing Daniel Vijay at para 76. This passage actually appears at para 145 of Daniel Vijay, not para 76 as stated by Thean J.
[30] Daniel Vijay, supra note 2, at para 76 [emphasis added].
[31] Ibid, at para 74 [emphasis added].
[32] Ibid, at para 146 [emphasis added].
[33] Ibid, at para 167. This passage was also relied upon by Thean J in both Azlin bte Arujunah and Miya Manik, at paras 93 and 77 respectively.
[34] There are other paragraphs in Daniel Vijay illustrating Chan CJ’s concern—apart from paras 76 and 145, see paras 74, 146 and 167, also referenced by Thean J in Azlin bte Arujunah.
[35] Parliament amended the Penal Code in 2012 such that with effect from 1st January 2013, only section 300(a) murder attracted the mandatory death penalty. For section 300(b), section 300(c) and section 300(d) murder, the court has discretion to impose either the death penalty or imprisonment for life and caning.
[36] Supra note 34.
[37] Daniel Vijay, supra note 2, at para 49.
[38] In Daniel Vijay, at para 146, Chan CJ provided two examples where section 34 read with section 300(c) would not be attracted for secondary offenders—where the assailants had a common intention to give the victim a good beating or to disfigure him. In either case, if death occurs, the secondary offenders would not be liable (see also the scenarios at para 168 referenced by Thean J in Azlin bte Arujunah, supra note 7, at para 96). However, Chan CJ did not state clearly if the primary offenders would be liable in those scenarios. If the “co-terminous” approach were used, as this article argues below, neither the primary nor secondary offenders should be found liable since the intended injury would not be one that would ordinarily lead to death to begin with.
[39] [2018] 2 SLR 249 (CA).
[40] It appears this is the position taken by the prosecution in Azlin bte Arujunah, supra note 7, at para 94.
[41] Lim Poh Lye, supra note 11, at para 23.
[42] [2017] 1 SLR 748 (CA). I credit my students, Ryan Leong Lup Mun and Kai Kiran Gosian, for this astute observation.
[43] See Chan CJ in Daniel Vijay, supra note 2, at para 168, giving various scenarios where section 34 and section 300(c) might or might not apply. Illustrations (b) and (c) would be consistent with the “co-terminous” approach, even though it remains unclear what Chan CJ meant by causing “s. 300(c) injury”. It is submitted that the participants need not intend or know that the injury would be sufficient in the ordinary course of nature to cause death. See also Imran bin Mohd Arip v PP [2021] 1 SLR 744, where the Court of Appeal, at para 135, felt that “in other words, it would not suffice if A’s mental state falls short of the mental state of B” (B being the primary offender). The Court did not anticipate A’s mental state to exceed that of B, suggesting that the expectation was simply to match it in a “co-terminous” fashion. The co-terminous approach also seems to have been adopted by the Court of Appeal in Michael Anak Garing, supra note 42, at para 55.
[44] Supra note 39, at para 87.
[45] It should be noted that Thean J in Azlin bte Arujunah, supra note 7, at para 102, read Sundaresh Menon CJ in Chia Kee Chen, ibid, at para 89 as requiring “a common intention to inflict the particular s. 300(c) injuries”. Again, this is neither here nor there. Like Chan CJ, Menon CJ never went on to insist that the secondary offender had to know or intend that the injuries were sufficient in the ordinary course of nature to cause death.
[46] This is reminiscent of the approach of the Court of Appeal in Wang Wenfeng v. PP [2012] 4 SLR 590 characterising section 300(c) as an offence that inherently satisfies coincidence of actus reus and mens rea.
[47] Resort could also be had to section 299 (culpable homicide not amounting to murder) and sections 322, 325 and 326 for voluntarily causing grievous hurt (which includes death).
[48] Admittedly, the multiple-incident cases like Azlin bte Arujunah will continue to vex. But that problem persists regardless of which test is employed.
[49] See also the formulation in the new section 308A—“… a deadly weapon, or anything which, used as a weapon of offence, is likely to cause death”—which suggests that the fact that a weapon is a deadly one or one likely to cause death is a strictly objective enquiry that is independent of the accused’s mens rea.