RECYCLED MALICE

Abstract

The criminal law doctrine of “transferred malice” has been much discussed. What has gone comparatively unnoticed is the phenomenon of “recycled malice”. For example, those who endorse transferred malice would hold that, if D tries to shoot V, and the shot misses and hits T, D’s intention to hit V is “transferred” to T, and a completed offence against T is constructed. But many legal systems that endorse transferred malice also allow D to be convicted of an attempted offence against V. In other words, D’s intention to hit V can apparently be used multiple times. Once this phenomenon is noticed, a question arises over its justification and limits. This article argues that no convincing justification for recycling mens rea exists.

I. Introduction

Consider the following hypothetical scenario:

One Shot, One Hit: D shoots at V, intending to cause him injury. The shot misses V and hits T, a bystander, injuring her. ¹

In systems that recognise the concept of transferred malice, ² D is straightforwardly liable for a non-fatal offence against the person in relation to T. If one takes the idea of transferred malice seriously, ³ D’s mens rea relative to V is, by dint of legal fiction, “transferred” to T and combined with the occurrence of the actus reus relevant to that form of mens rea to form a completed offence against T. Although, as noted below, opinions appear to differ on this front, this “transfer” might occur even if it was unforeseeable that the actus reus would befall T. ⁴

The language of a “transfer” here might suggest that there is one “unit” of mens rea that can be reassigned, by dint of legal fiction, from V to T. ⁵ Admittedly, some early descriptions of the doctrine suggest exactly this. ⁶ For instance, Blackstone talks about how the law takes D’s “intent … [and] transfers from one to the other”. ⁷ Notably, and in common with other early sources, Blackstone makes this comment in the context of examples where D misses the target, V. ⁸ Perhaps these writers meant to imply that this was the only situation in which transferred malice operated. On such views, D’s mens rea might be thought to be “used up” in constructing liability for the offence against T. ⁹ That would seem to exclude the possibility of any offence against V.

Contemporary views differ on this point, however. In many jurisdictions, as well as a completed offence against T, D can additionally be found liable for attempting to cause the injury that was meant to be caused to V. ¹⁰ Very many theoretical discussions of transferred malice assume that liability for both offences can be made out on the facts, too. This suggests that D’s intention to harm V can be “used” at least twice to find liability, even in basic cases like One Shot, One Hit. ¹¹ I will call this idea “recycling” malice. Although this phenomenon has been noted at times, ¹² it has never been subjected to proper analysis. This paper addresses this gap in the literature. ¹³

Once the potential for recycling malice is recognised, a question arises over its limits and justification. Consider a second hypothetical scenario:

One Shot, Two Hits: D shoots at V, intending to cause him injury. The shot hits V, injuring him, and passes through V and into T, a bystander, injuring her. ¹⁴

What has changed from One Shot, One Hit is an aspect of the actus reus: the bullet has injured V, rather than missing her. If the mens rea picture remains unaltered, presumably D’s mens rea can again be “recycled” to reach the conclusion that D is liable for a completed non-fatal offence against V and a completed non-fatal offence against T.

Indeed, following the same logic, it would appear that a defendant could be liable for as many instances of a relevant actus reus that the conduct accompanied by mens rea causes. Consider a third hypothetical scenario:

Neighbours: Following a neighbourly dispute, D sets fire to V’s home in the middle of the night, intending to kill her. The fire takes hold and V dies. V’s overnight guests, T1, T2 and T3, are also killed in the blaze. D had no idea that T1, T2 and T3 were present. ¹⁵

If malice can be recycled three times, then D is liable for four murders. And so on, until causation is exhausted.

Transferred malice is meant to reflect widely held intuitions about “practical” ¹⁶ and “rough” ¹⁷ justice in cases like One Shot, One Hit. This coherence with intuition is one of transferred malice’s alleged “strengths”, ¹⁸ and is often ¹⁹ about the only express argument that is marshalled in favour of recognising the doctrine legally. ²⁰ But intuitions are more likely to vary when considering One Shot, Two Hits, and Neighbours (or so I have found), and so have the conclusions reached by courts when they notice the potential problem in this context. ²¹ If one is unhappy with two convictions for completed offences in One Shot, Two Hits, or four completed offences in Neighbours, but remains wedded to transferred malice as a doctrine, one seems to have two options:

1. (1) establish a non-arbitrary way of denying liability for both completed offences in One Shot, Two Hits, and all four completed offences in Neighbours, whilst endorsing liability for an attempt and a completed offence in One Shot, One Hit; or

2. (2) give up on the idea of recycling mens rea and convict D only of the completed offence against T in One Shot, One Hit.

Below, I examine arguments in favour of the doctrine of transferred malice and conclude that no such non-arbitrary line can be drawn to support (1). I will also argue that (2) would lead to counter-intuitive results, when – as already noted – its chiming with common intuitions is one of transferred malice’s alleged strengths. I will then briefly explain that defenders of the view that “transferred malice” cases are simply cases of an immaterial variation in the facts surrounding the commission of an offence (so called “abolitionists”) face similar challenges if they propose recycling malice as many times as the actus reus allows. These challenges mean that only “single use abolitionism”, where a “unit” of mens rea can only be married with one instantiation of an actus reus, is defensible. Such “single use abolitionism” nevertheless faces the same quandary: why would it be fair labelling to convict D of an offence against T, whilst ignoring the intended wrong to V? It will be suggested that no compelling answer can be provided to this question. In sum, then, abolitionist views do not offer an unproblematic home for those who have, hitherto, endorsed the transferred malice doctrine.

II. Assessing the Potential Justifications for Recycling Malice

In this section, I contend that no account of transferred malice supports the recycling of malice, and so those who remain wedded to the idea of transferred malice as a doctrine (and so are unwilling to give it up) are forced into denying attempts liability in a case like One Shot, One Hit, liability for two completed offences in One Shot, Two Hits, and liability for four completed offences in Neighbours. I will further argue that this approach will lead to counter-intuitive results unless one backs the solutions that those who reject a doctrine of transfers of malice ²² champion, or one endorses some sort of “abolitionist” account of such cases (looked at in Section III). Either way, transferred malice requires to be abandoned. ²³

A. Liability Gaps

One historical justification for transferred malice concerns the absence of a generalised law of attempts. ²⁴ Consider Coke’s articulation of the justification of transferred malice:

[H]e who has the ill and felonious intent shall be punished for it, for he is as great an offender, as if his intent against the other person had taken effect. And if the law should not be such, this horrible and heinous offence would be unpunished; which would be mischievous, and a great defect in the law. ²⁵

If the answer in One Shot, One Hit is that, if there is no offence against T, there is no offence committed by D at all, Coke is correct that this over-privileges moral luck and seems disastrous in terms of the criminal law’s ability to respond meaningfully to D’s conduct. ²⁶

This justification for transferring malice is, of course, inapplicable in modern times: generalised laws of attempt are nowadays commonplace. ²⁷ This fact also tells against recycling malice: the criminal law does not over-privilege moral luck if it imposes liability for an attempt to harm V, and D can still be liable for an offence against T if D possesses relevant (separate) mens rea (for instance, recklessness regarding harming T). ²⁸ D might, of course, lack such distinct mens rea, and thus be acquitted in relation to the harm suffered by T. But that is only problematic if one holds that the harm against T must be responded to by the criminal law, even in the absence of mens rea relative to that specific instance of harm. Such an argument remains to be made convincingly, as will be seen repeatedly below.

It might be thought that a liability gap argument would be stronger in a context where the mens rea state at issue is not intention, but some other form of fault such as recklessness. Here, many Anglo-American jurisdictions will deny liability for an inchoate offence based on recklessness regarding a particular consequence. ²⁹ Consider another hypothetical scenario:

Target Practice: D is aiming his bow and arrow at a target, aware of the substantial risk that he might hit V, who is standing nearby. D shoots the arrow, misses the target, and hits T, whom D had failed previously to notice.

In the absence of a generalised offence of reckless endangerment (let alone negligent endangerment or injury), D will not be liable for any offence in this scenario unless malice – recklessness regarding causing harm to V in this example – transfers. This potential liability gap might be thought to give us good reason to recognise an offence against T, so that D is convicted of something. But why should this something be a completed offence against T, rather than an offence of reckless (or even perhaps negligent) endangerment? Why, in other words, is transferring recklessness, rather than recognising a distinct offence premised on reckless endangerment, the answer? It is true that there are sound reasons to be cautious about the creation of such generalised reckless endangerment offences. ³⁰ But there are significant reasons (as will be seen further below) to be suspicious about transfers of fault. ³¹ Furthermore, even if an answer could be given to the question of why transferring recklessness is superior to a generalised offence of reckless endangerment being created, it will not provide any ammunition for the recycling of malice. If the idea is to convict D of “something”, that need not self-evidently be offences against V and T.

A variation on the theme of liability gaps concentrates not on the absence of liability for attempts or endangerment, but on the inability, without transferred malice, to convict D of an offence of mens rea against T. As William Prosser puts the point: “The early criminal cases were understandably preoccupied with mens rea, moral guilt, and the obvious fact that if the defendant was not convicted there would be no one to punish for the crime [against T].” ³² The view Prosser is sketching here is doubly unconvincing. First, and most obviously, it presupposes that the conclusion of the absence of transferred malice is the absence of a conviction for an offence against T. It may, however, be entirely possible to convict D of a crime relative to T, if D also had a relevant form of mens rea in relation to T. For instance, if D was subjectively reckless regarding killing T, or grossly negligent regarding T, most systems of criminal law would find a way to convicting D of T’s manslaughter, or some equivalent offence, if he killed T whilst trying to kill V.

The second difficulty with the view outlined by Prosser is that it is question-begging, for it presupposes that T has been the victim of “a crime”, when that is precisely what is under dispute. Consider two versions of One Shot, One Hit. In the first, D is reckless regarding causing harm to T. Here, transferred malice is not necessary for us to find that a “crime” has been perpetrated against T. In the second, however, D is not even negligent with regard to causing harm to T. Here, there is a compelling case for saying that no “crime” has in fact been perpetrated relative to T (subject to the alternative view outlined in Section III): actus non facit reum nisi mens sit rea. ³³

The situation relative to T can be contrasted usefully with the situation regarding V. Nobody can dispute that there has been a more than merely preparatory act towards harming V, and that D intended for that harm to be visited on V. If T is removed entirely from the scenario, that attempted crime against V remains visible.

In the liability gap arguments surveyed thus far, the impetus is for D to be convicted of “something”. But losing the ability to do “something” is vague. There is, however, a more specific cost, at least in a subset of crimes. In murder, for example, transferred malice avoids D evading a mandatory “life sentence” in the UK’s jurisdictions. ³⁴ Were the law to reject transferred malice in cases like One Shot, One Hit (where D attempts to kill V, misses, and kills T) D’s liability would be limited to attempted murder and manslaughter, and only a discretionary life sentence would be available to the court. If the mandatory penalty for murder is, nevertheless, unjust (as mandatory penalties often plausibly are), this is not a strong argument for forcing it upon a defendant via the doctrine of transferred malice. ³⁵ Note also that, if D is liable for T’s manslaughter, the trial judge can still (at least in England and Wales) pass a life sentence if that is warranted. Indeed, the same is true (at least in England and Wales) of attempted murder, ³⁶ and so refusing to transfer malice would not prevent the punishment being life imprisonment, if appropriate circumstances surround T’s killing. The type of argument under discussion is also inapplicable in contexts where there is no mandatory penalty. Indeed, in England and Wales, transferred malice’s relevance would, if this argument about mandatory penalties is followed, be limited largely to murder. ³⁷

Of course, in reality, attempters do not tend to receive the same penalties as completers, even if they are in theory available. Manslaughterers tend to receive determinate (i.e. non-“life”) sentences. ³⁸ But if one dislikes those outcomes, it would appear that reform of the sentencing of those offences (or more generally, as discussed in the next subsection) should be argued for. ³⁹ It is not clear why transferred malice is an appropriate tool for achieving such reform, particularly given the other problems noted in this article. If transferred malice is unnecessary for this purpose, recycled malice is, a fortiori, unnecessary.

B. Equal Culpability

Arguments about “avoiding” mandatory sentences lead naturally to a second major tranche of arguments for recognising transferred malice: that D is equally culpable whether it is V or T who is harmed. ⁴⁰ Therefore, D’s punishment should be identical. After all, D has shown the relevant type of fault for the completed offence, and has brought about the relevant type of harm. D’s culpability should be reflected, the argument goes, in liability for a completed offence against T.

A less popular, but similar, account defends transferred malice on the basis that D is, through the application of that doctrine, denied the opportunity to “profit from [their] own wrong, error or mistake”. ⁴¹ In One Shot, One Hit, it might be thought that, if D is not liable for injuring T, he has “profited” from his own error in missing V; he is liable “only” for an attempt and perhaps, if separate mens rea (for instance, recklessness regarding harm to T) can be proved relative to T, a less serious non-fatal offence against the person. Accordingly, the theory goes, it is necessary to convict D of wounding T, to minimise the impact of luck on his liability, at least in punishment terms.

The difficulty with both arguments is that, once again, it is difficult to see why transferred malice is the answer, rather than some more straightforward alternative. One straightforward reform would reach the same result by convicting D only of the attempted offence against V. Remove T from the equation: D shoots at V. At that point, as moral luck sceptics have made clear repeatedly, we arguably know all that we need to know about D’s culpability, if we understand that to refer only to D’s faulty reasoning and lack of sufficient concern for others’ interests. ⁴² If punishment should just be a matter of culpability, and this is what culpability consists in, then D should be punished equally whether or not V is injured. ⁴³ The fact that T is injured is, on this account, neither here nor there, unless D had distinct culpability (warranting distinct punishment) relative to the harm risked regarding T. On such views, transferred malice reveals only an irrelevant detail.

Perhaps, however, defenders of transferred malice recognise that a luck-neutral criminal law is very unlikely to be implemented. To them, transferred malice might be a “second best” option, that gets us closer to the luck-neutral ideal, albeit only in a narrow range of cases where D’s action misfires. The defender of such a position must agree with the argument outlined above: the issue is really one about the role of luck in sentencing, and not about a foundational doctrine of culpability. Transferred malice is a consolation prize, rather than something worth defending intrinsically. That is why transferred malice, in this context, should not be viewed as the superior answer to resolving the more general question about the role of luck in sentencing.

Any positive argument for transferred malice is thus going to have to find a home in one of the accounts of punishment that allows differential sentencing for inchoate and complete offences. For instance, some theorists argue that we have good reason to reduce punishment for attempted rather than completed offences, because we are relieved at D’s failure (and think this point worth marking). ⁴⁴ Or perhaps we should not reflect only culpability in cases of unsuccessful attempts because to do so would at least risk the State being “cruel”. ⁴⁵ Such relief and leniency might be argued to be less apparent (or entirely absent) and less appropriate (or entirely inappropriate) in a case where, although D failed to harm V’s interests, she managed to harm T’s interests in an identical manner. We might be relieved that D missed V, but we will not be relieved to hear that T befell exactly the same harm. We might agree that it risks cruelty to D to punish her as though she hit V, until we learn that she hit T in an identical way. Transferred malice allows these points to be noted in a way that requiring independent proof of fault relative to T would likely prevent, because that fault is likely to be of a “lower” order than the one D possessed relative to V. For example, the resulting conviction is more likely to be one involving recklessness or some form of negligence, than one involving intention. ⁴⁶

Once again, however, it is not clear that this conclusion follows. Firstly, proof of independent fault relative to T would present an independent ground for punishing D, and would thus alleviate the “discount” for D’s having failed in the attempt to harm V’s interests. This would, it appears, allow the law to reach the same sort of result that transferred malice would push towards. It would do so in a way that does no violence to regular principles of mens rea, and would accord with (I assume) common intuition. Secondly, it is not clear why a sentence could not take account of the fact that D caused harm to others without formally requiring proof of a completed crime against T.

In short, whatever the “correct” answer is regarding the punishment for inchoate and completed offences (if there is a “correct” answer), that answer does not seem to justify the transferred malice doctrine.

The prospects of an equal culpability argument assisting in a defence of recycling malice are even dimmer. D will not “luckily” escape full punishment for the offence against V, assuming that is made out. ⁴⁷ Although it is not always clear what “culpability” means, it would be odd for a theory of culpability not to recognise, in cases like Neighbours, that it is more culpable to intend that one’s action kills four people, rather than one person. ⁴⁸ For instance, if one’s view of culpability is that it is about the demonstration of a lack of concern for a specific interest, a person shows more contempt for the value of life if she is willing to kill four people without justification or excuse rather than just a single person. Although it is hard to be sure about such matters, presumably it is more culpable, in cases like One Shot, Two Hits, to intend bodily injury to two people, rather than one person. The same conclusion presumably follows in One Shot, One Hit.

Accordingly, if recognising culpability was what transferred malice was about, then it would seem that the recycling of malice is outright prohibited: “the purpose of transferred intent is not to multiply criminal liability, but to prevent a defendant who has committed all the elements of a crime (albeit not upon the same victim) from escaping responsibility for that crime”. ⁴⁹

For the same reason that we do not usually charge “successful” murderers with murder and attempted murder, we should not – if we are serious about punishment reflecting culpability – seek to convict D of the attempt to murder V and the completed murder of T. ⁵⁰ To do so would, illegitimately, suggest that D has two “units” of culpability, rather than the one that in actuality exists. ⁵¹ Again, D has shown insufficient regard for a human life (V’s) in attempting to kill V, but this fact tells us nothing about D’s regard for an additional life (T’s).

An alternative argument would be that the discussion above is misconceived in its focus. It would hold that it is the aggregate culpability of an attempted murder and a manslaughter, for instance, that is equivalent to the culpability required for murder. ⁵² Accordingly, D is “culpable enough” to be convicted of murdering T. But such accounts do not convince. Firstly, the main point of transferred malice is that no independent fault needs to be proved in relation to T, which means that the prosecution is not required to prove that the separate ingredients exist before they can be aggregated together in the manner proposed. Even assuming that transferred malice is limited to foreseeable mishaps (a point returned to below), manslaughter often requires “more” than civil negligence (a concept that turns on considerations of foreseeability). ⁵³ Secondly, any such aggregate account appears to suggest that there is “enough” culpability to support liability being imposed for a completed offence regarding T. But if this is a genuine exercise in addition, then the fault involved in the attempted murder is, indeed, “used up” in constructing the culpability required for a murder conviction. There is, in other words, no possibility of recycling mens rea.

C. Marking Harms and Wrongs

Culpability is, then, something of a dead end for defenders of transfers and recycling of malice. Perhaps the theory underlying such doctrines is instead one concerned with marking harms and wrongs, insofar as these can be separated from concerns of culpability. ⁵⁴ The doctrine of transferred malice allows us to mark the harm that befell T, without having to seek independent mens rea relative to that harm. ⁵⁵

This account founders, but not for the reasons one might initially offer. It might be objected, for instance, that – outside of regulatory and quasi-regulatory contexts – the criminal law presumptively prohibits the criminalisation of bare harm. Actus non facit reum nisi mens sit rea does not mean that any mens rea will do: the mens rea must typically “correspond” to the elements of the actus reus. ⁵⁶ A defender of transferred malice would respond that such correspondence is ensured by the limits placed on that doctrine: D’s mens rea can only “transfer” to an instantiation of the actus reus of the same offence. ⁵⁷ Critics of transferred malice cannot help themselves to the assumption that correspondence does not work like this, because D’s mens rea must relate to a specific individual.

The problem is, instead, that the criminal law is, at least ostensibly, concerned with wrongful harms. ⁵⁸ We might quibble over whether D has necessarily wronged T in transferred malice cases, ⁵⁹ because that depends on one’s conception of wrongdoing. If wrongdoing means acting contrary to the balance of reasons (ignoring excluded reasons, for the moment), ⁶⁰ then it is possible to say that D had reason not to shoot in One Shot, One Hit because of the chance of hitting T (even if, on some views, D was unaware of T’s presence). ⁶¹ Other theorists conceive of wrongs as breaches of duties, and would demur if a duty of D to T ⁶² was imposed in situations where the resulting harm to T was unforeseeable. ⁶³ The basic thought is that D could not hope to comply with such a duty, and thus the duty demands the impossible (and should not be recognised).

Theorists critical of transferred malice often allege that the doctrine covers unforeseeable harms. ⁶⁴ Despite this, real cases of such phenomena are rarely, if ever, cited. Andrew Ashworth wonders ⁶⁵ about the case of Agnes Gore, ⁶⁶ where D poisoned V’s medicine, intending that V would consume it and die. V took some of the medicine and became sick, as did another person, T, who tried the medicine. T returned the medicine to the apothecary and complained about its quality. To protect his reputation, the apothecary consumed the medicine and died. Ashworth asks whether this precise course of events could have been foreseen by Gore, but it is not clear why this is the correct question. Is the better question not whether it is foreseeable that someone other than V would consume the electuary and be killed? ⁶⁷ Of course, there is no clear answer to this question, but it does give rise to doubts about the limits of transferred malice.

Some theorists assume that unforeseeable harms would not be covered by the doctrine. ⁶⁸ This caution is wise. Even if one might quibble about Gore, the harm that befell T in real cases of transferred malice was typically readily foreseeable to a person in D’s position. In such cases, there might be the foundation of a duty on D towards T, and a conception of negligence regarding the harm to T, and that would be enough to suggest, on this type of account of wrongdoing, that D has wronged T. ⁶⁹ The fact that D was at fault regarding an identical harm relative to V would, I assume, support the argument that D’s unleashing the relevant risk on T was unjustified. ⁷⁰

What good can come, however, of marking this negligence through conviction for a crime that requires intention or (subjective) recklessness? To do so inflates D’s culpability, relative to T, in a way that misrepresents the situation, all (the argument goes) in the name of marking (accurately) what happened to T. ⁷¹ If negligently harming others is thought sufficiently culpable to warrant such conviction, such an offence should be created. Although the fact that D had mens rea regarding V might be a useful factor in assessing whether such negligence is made out on the facts, and might indeed be a useful factor to use to limit the reach of such negligence in the criminal law, the doctrine of transferred malice does not seem to follow smoothly from these facts.

There are further senses in which it might be alleged that transferred malice misrepresents D’s wrongdoing. If D has risked foreseen or foreseeable harm to T without adequate justification, and the harm has arisen, D has, in Antony Duff’s terms, endangered T. ⁷² But transferred malice often allows us to take the attack against V and act as though it was directed against T. The significance of the attacks versus endangerments distinction is, as with many of the core concepts in this area, contested. ⁷³ There is a doubt, then, over whether the criminal law should have any truck with it. But, if it should, the failure to note the distinction is yet another argument against transferred malice.

The above discussion provides reasons to doubt that transferred malice is supportable on the basis that it marks wrongs appropriately. The doctrine seems to allow for the marking of bare harms that would otherwise not concern the criminal law, or the miscommunication of wrongs that would otherwise concern the criminal law, and there is, perhaps unsurprisingly, yet to be a convincing case put forwards for doing these things.

Further difficulties for transfers of malice arise when consideration turns to the recycling of malice. Initially, the recycling of malice might be thought to be a good thing, because it could mark a wrong that a bare transfer, where mens rea is “used” up, would fail to. Although it is controversial whether putting others at risk is a harm, ⁷⁴ it is uncontroversial that putting others at risk, without adequate justification, is a wrong. Allowing recycling of malice would mark this point. It would, however, mark that point repeatedly (and, accordingly, in excess of D’s culpability) where multiple instances of the actus reus materialise, as in One Shot, Two Hits, and Neighbours. We cannot, in other words, properly mark wrongs and culpability if we allow the recycling of mens rea.

A defender of transferred malice might demur at this point. As Patrick Tomlin points out, a theory of criminal responsibility that marked the harm (and wrong) done to T, but not to V, would be deficient because it would “erase one of [the] victims from the moral description” of what D did. ⁷⁵ This, then, could be an argument in favour of recycling D’s malice: the wrongs (and potentially the harms) done to V and T can be marked. But this argument for recycling depends on the soundness of the reasons for transferring malice, and those have been doubted above, and are – tellingly – doubted by Tomlin.

Tomlin contends that V has an interest in having the attack against her noted authoritatively by the criminal law. ⁷⁶ If the criminal law concentrates only on the completed crime against T, it unfairly labels (or rather fails to label) what D did to V; it deletes the victim-centred wrong against V. ⁷⁷ Although fair labelling tends, in criminal theory, to be thought of as primarily of concern to defendants, it is plain that V also has an interest in having D’s attempt against her labelled (and labelled accurately). ⁷⁸ Additionally, the public has an interest in the label applied to D’s conduct being accurate, rather than the product of fiction. ⁷⁹ And those claims are plausibly stronger than T’s in having a crime constructed against her.

These arguments provide good reasons to recognise the crime that has been perpetrated against V in preference to utilising D’s mens rea in relation to V to construct a crime against T. Again, then, the view that D should be found liable for attempting to cause harm to V, and that any offence against T should be looked at separately, is superior.

One can imagine accounts of transferred malice that seek to overcome these difficulties by disputing the claim that it is always inappropriate to prefer T over V in terms of the criminal law’s authoritative judgments. For instance, in relation to One Shot, One Hit, it could be argued that it is appropriate to mark the full crime against T, rather than the attempt against V, because T has been more seriously affected by D’s conduct than V. In One Shot, Two Hits and Neighbours, however, each individual T is not affected more severely than V, and so the crime against V should be marked. This view seems, however, to merely repeat the error noted above of concentrating on the harm to T, rather than the wrong to V.

Another response to Tomlin, which again would distinguish One Shot, One Hit from the other cases discussed above, would point out that what occurred to T (serious injury) is closer to what D intended to happen than what happened to V (no injury). It is, accordingly, more appropriate for the criminal law to focus on what happened to T, via the doctrine of transferred malice. ⁸⁰ This seems, however, to concede that what we are concerned with is not the reality of what D did, and the wrong against V, but instead with a fictionalised account of what D did to T. This violence to fair labelling should not be done without strong justification, and it has been doubted above that transferred malice has any such foundation.

D. Recycling Malice: Conclusion

The main arguments in favour of transferred malice, insofar as these go beyond bare intuition, have been discussed. The conclusion is that no account of transferred malice convincingly supports recycling malice. In consequence, defenders of transferred malice, and convictions for completed offences regarding third parties, seem forced to conclude that D is not liable for an attempted offence against V in One Shot, One Hit, and is not liable for a completed offence against V in One Shot, Two Hits, or in Neighbours.

These conclusions are, however, deeply counter-intuitive, and – as noted above – transferred malice is often defended on the basis that it accords with intuition. One thing everyone seems to agree on is that D attempted to cause harm to V. Removing T from the scenario entirely demonstrates the correctness of that position: the offence against V would remain plainly visible. Accordingly, ignoring that attempt and convicting only of the harm to T in One Shot, One Hit is unrepresentative of reality in a way that should concern us. V is removed from the formal legal narrative of what D did. ⁸¹ There is no convincing explanation why, in One Shot, Two Hits, for example, it would be the offence against V that would be passed over in favour of constructing an offence against T. Differences in sentencing practice are not sufficient reason, as argued previously.

If recycled malice cannot be defended, transferred malice becomes even more suspect, and there is more cause to abandon that doctrine.

This does not, however, mean that one is bound to accept the argument that malice can never be recycled. Writers have, for a long time, doubted whether there is in fact any “doctrine” of transferred malice. Instead, there is just a legally immaterial mishap along the way to the commission of an actus reus accompanied by the appropriate form of mens rea. In the next Section, it will be seen that such views could conceivably support the recycling of malice, but fall prey to many of the objections above. They thus fail to provide an unproblematic alternative theoretical home for former endorsers of transferred malice.

III. Recycling for “Abolitionists”

A common argument, which Douglas Husak has termed “abolitionist”, ⁸² is to the effect that there is no “doctrine” of transferred malice, even in cases like One Shot, One Hit. On such views, there is actus reus and there is mens rea: a human being has been injured in the relevant way, and the defendant intended to injure a human being in that manner. The fact that the human being was a different one to the person D intended to injure makes, on this view, no legal difference. ⁸³

On such views, there is, it is maintained, no need to transfer malice. In theory, there is no need to recycle malice. One can imagine a brand of “abolitionism” that sets the ambit of the defendant’s criminal liability as the actus reus: every time that the actus reus occurs against a viable target (for instance, a human being, or property belonging to another person), concurrent mens rea relating to the relevant subject matter of the actus reus results in liability. In result crimes (which transferred fault is most often relevant to), causation seems to be the limit on liability. “The intention follows the bullet” ⁸⁴ as far as the relevant causal path goes. ⁸⁵

On such views, there is no problem in convicting D of both attempted wounding and wounding in One Shot, One Hit, and no problem with two convictions for wounding in One Shot, Two Hits, or four convictions for murder in Neighbours. There is also no problem in convicting D of three counts of arson when he recklessly sets fire to property X, and the fire spreads to properties Y and Z.

Of course, core difficulties that told against the recycling of malice under justifications for transfers of malice apply with equal force here. D is culpable for X, and held liable for, and punished for, X+Y. Y may outstrip X in severity massively. To continue the example of a fire spreading, consider where D sets fire to some newspapers to keep warm one night, expecting them to eventually extinguish themselves on some concrete. The fire spreads to a nearby plastic bin, and then to a shop, causing £1 million worth of damage. ⁸⁶ Assuming that the newspapers were “property belonging to another”, ⁸⁷ a fairly trivial act of arson seems to become multiple serious instances of arson. ⁸⁸ The lack of concern for others’ property rights that D showed in his choice to start the fire is very far removed from the lack of concern for others’ property rights that a decision to set fire to the buildings would have demonstrated. Once they get going, this type of “abolitionist” argument seems difficult to stop. ⁸⁹ The fact that the recycling of fault is not noted as expressly as it is by defenders of transferred malice is hardly a merit of such views.

It is not clear, on such “abolitionist” views, what justifies punishing D to a greater extent than in a situation where he succeeds only in bringing about what his mens rea stretched to. Granted, the trial judge will often (mandatory sentencing regimes aside) be able to make the sentences for each offence run concurrently, in a manner that, ultimately, results in D’s punishment being the same as it would have been for one crime. But D’s record will misrepresent his culpability, which brings up another objection: this variety of “abolitionism” also runs into problems of potential unfair labelling, the further the actual outcome is from what D intended or anticipated. ⁹⁰ Until causation is exhausted, culpability can seek out relevant instances of an actus reus. If, ultimately, the justification for this kind of “abolitionist” view is that we should mark the consequences of the defendant’s acts, the objections to transferred malice (discussed above) once again apply with equal force. ⁹¹ If one wishes to increase D’s punishment, and the severity of one’s labels, one should have to prove independent culpability relative to the harms caused by D’s behaviour.

Ultimately, then, the main advantage this variety of “abolitionist” view has over transferred malice is that it can, coherently, explain why multiple convictions can follow in the hypothetical cases discussed above. Bare coherence is not enough to overcome the doubts about this brand of “abolitionist” view in such scenarios, and it should be noted that such “abolitionism” works only by largely defining away the relevant problem.

This unsatisfactory form of “abolitionist” thought can be contrasted with another, which might – given the theme of recycling – be branded “single use abolitionism”. On this type of account, D’s intention to harm a person (for example) can only be utilised once, after which it is irrelevant to D’s criminal liability. Such an account faces the same dilemma as the defender of transferred malice: is D’s mens rea regarding V employed to find liability relative to the actus reus against V, or is the actus reus against T to be married with the relevant mens rea? If the choice is to “use” the mens rea against T, some account needs to be given of why this is appropriate.

One suggestion could be that, where various actus reus elements are provable, prosecutors should elect to prove the one that most closely fits D’s mens rea. ⁹² Accordingly, in One Shot, One Hit, the prosecutor has an attempted section 18 offence’s actus reus (against V) and a completed section 18 offence’s actus reus (against T). D’s intention was to cause grievous bodily harm, and the actus reus regarding T more closely reflects this intention. Accordingly, the prosecutor should elect to prove a section 18 offence relative to T. In the other cases discussed above, where V suffered the harm intended by D, there is then no imperative to prefer an offence against T.

Once more, however, this account seems to be overly concerned with the harm done to T, rather than the wrong done to V, in cases like One Shot, One Hit. Although the “single use abolitionist” is not engaging in the same fiction as the defender of transferred malice, this is again largely because the problem has been defined away, rather than solved. To convict D of a completed offence against T, whilst ignoring the attempt against V, is to once more engage in unfair labelling.

In sum, “abolitionism” does not provide an unproblematic home for the erstwhile defender of transferred malice.

IV. Conclusion

It has been contended that the arguments for transferred malice do not support the recycling of malice. Accordingly, supporters of transferred malice, at least if they aspire to any sort of intellectual coherence, must either accept that a “transfer” “uses up” D’s mens rea, and counter-intuitively, and problematically, ignore D’s offence against V, or accept the arguments of the critics of transferred malice and find D liable for an offence against V and seek independent fault relative to the actus reus that befell T. “Abolitionists” will need to either bite the bullet that their account inflates D’s culpability and punishment through “recycling”, or face the same dilemma regarding the preference of T over V.

Hypothetical scenarios are often neat, and the problem identified in this article risks being seen as theoretical. Furthermore, it might be thought that transferred malice is a niche doctrine, with little real-world application. ⁹³ In closing, it is useful to consider two relatively-recent, real-world cases that show what is at stake, and to allow readers to test more fully their intuitions. ⁹⁴

Arkan Ali, Hawkar Hassan and Aram Kurd conspired to blow up their convenience store to claim insurance money. ⁹⁵ They did not trust Viktorija Ijevleva, who knew of the plan, to keep quiet about it. The conspirators arranged for Ijevleva to be in the premises when they exploded. Ijevleva was killed in the blast, as intended. The explosion was, however, so violent that it destroyed the residential property above the shop. Inside were Mary Ragoobar, her sons Shane and Sean, and Shane’s girlfriend Leah Reek, who was visiting for the evening. All four were killed.

Ali, Hassan and Kurd were convicted of five counts of murder.

If the intention to kill Ijevleva can be recycled, then the conspirators are straightforwardly liable for murder in relation to every death their actions caused. If, however, no such recycling can take place, and a defendant’s mens rea is “spent” if his intended purpose is achieved, the conspirators’ multiple convictions for murder are more problematic. Under English law, the defendants would need to be proved to have intended to kill or cause grievous bodily harm to the occupants of the residential property, ⁹⁶ and that would seem unlikely. They presumably did not view the occupants’ deaths as a success condition of their insurance scam (compare Ijevleva’s death). ⁹⁷ That would appear to leave the prosecution having to sustain a difficult argument to the effect that it was virtually certain that the blast would at least seriously injure the occupants, and that the conspirators appreciated this fact. ⁹⁸ Otherwise, the prosecution’s main arguments would be that the defendants should be liable for the manslaughter of the occupants of the flat, perhaps on the basis of an unlawful and dangerous act, ⁹⁹ or on the basis of gross negligence. ¹⁰⁰ Such manslaughter convictions appear likely to be returned, and – on the argument presented above – should have been returned.

In a second, real-world example of the problem explored in this article, Nathaniel Grant chased Roshaun Bryan into a convenience store, and fired two shots, intending to kill him. ¹⁰¹ Bryan was unhurt, but one bullet hit and paralysed five-year-old Thusha Kamaleswaran, and the other hit and wounded Roshan Selvakumar. Grant was convicted of attempting to murder Bryan and, using the doctrine of transferred malice, wounding Kamaleswaran and Selvakumar with intent to cause grievous bodily harm. ¹⁰² Assuming that the Court of Appeal was correct to reject Grant’s argument that an intention to kill does not include, logically, an intention to cause grievous bodily harm, and that mens rea can be recycled, the resulting convictions are capable of explanation. (Indeed, Grant perhaps got off lightly, as recycling his intention to kill Bryan could presumably have resulted in three convictions for attempted murder. ¹⁰³ ) But what is still required is an explanation as to why this does not involve illegitimate “double counting” of Grant’s culpability.

A rejection of the idea of recycling malice, and full invocation of transferred malice, results in two convictions for attempted murder (maximum sentence: life imprisonment), and one conviction for malicious wounding (maximum sentence: five years’ imprisonment) – assuming that the prosecution could prove recklessness regarding the causing of some harm to other people within the shop. ¹⁰⁴ This deeply counter-intuitive outcome would differentiate between Grant’s liability for the injuries sustained by Kamaleswaran and Selvakumar, and for no good reason. A rejection of transferred malice, in addition to a rejection of recycled malice, would result in a conviction for attempted murder, and two convictions for malicious wounding (making the same assumption about recklessness regarding the causing of some harm). Again, it is submitted that these were the appropriate conclusions, despite any intuitive pull towards increasing the severity of the offences against the two shot persons, particularly when one was a young child.

This is an area replete with intuition-based judgments, and ultimately “knock down” arguments should not be expected. ¹⁰⁵ It is nevertheless suggested that the puzzle of whether malice can be recycled gives yet further reason to reject both transferred malice and “abolitionism”.