1. For a recent comparative survey of approaches to the topic of transferred malice within Common Law and continental legal systems, see Michael Bohlander, “Transferred Malice and Transferred Defenses: A Critique of the Traditional Doctrine and Arguments for a Change in Paradigm” (2010) 13 New Crim L Rev 555 at 583-607. For a comparative survey of the various approaches in different jurisdictions in the US, see Travis E Robey, “Recent Decisions: The Court of Appeals of Maryland – Criminal Law” (2005) 64 Md L Rev 1098 at 1106-08. There is extensive academic writing in the field, reflecting a variety of approaches, as shown in Section 2 below. Douglas N Husak created a taxonomy of these approaches in his article “Transferred Intent” (1996) 10 Notre Dame JL Ethics & Pub Pol’y 65 [Husak, “Intent”], discussed below. I have also offered a contribution in Shachar Eldar, “The Limits of Transferred Malice” (2012) 32 Oxford J Legal Stud 633 [Eldar, “Malice”], but did not address the combination of transferred malice and offences involving multiple participants, which is the focus of the present article. A review of positions relevant to my present argument appears in Section 2.

2. (1577), 75 ER 706 (QB); 2 Plowd 473 [Saunders and Archer]. Some commentators refer to the verdict in the case of R v Salisbury (1553), 75 ER 152(QB) 1 Plowd 97, 100 [Salisbury] as the original source of the doctrine of transferred malice. See, e.g., William L Prosser, “Transferred Intent” (1967) 45 Tex L Rev 650 at 652; Moore, Michael S, Causation and Responsibility (Oxford: Oxford University Press, 2009) at 204.Google Scholar In this verdict, the defendants were convicted of homicide when, after planning to murder a Doctor Ellis, they ended up killing the servant who attended to him. The instruction given by the court in this case is too short and narrow to be considered the initial formulation of the doctrine of transferred malice: ‘when a man has malice against another, and intends to kill him, and endeavours to put his purpose in execution, and kills one that resists his purpose, it cannot be otherwise construed that by necessity of reason he has malice against all those who would defeat his design, and that he would offer violence to them that would defend the person against whom his malice is directed, rather than desist from his purpose, and therefore if he kills them to whom he had before-hand intended to offer such violence, this cannot be anything else than murder: and so the act declares his intent before, and the malice against the principal begets in himself another malice against those whom he presumes will resist his purpose, which malices are combined one to the other inseparably’. By contrast, the verdict in Saunders and Archer lists cases of accidental miss-aim and of mistaken identity. Moreover, it carves out the rule that malice is transferred from the intended object to the one harmed in practice.

3. Edmund Plowden reported that Lord James Dyer, who headed the panel that handed down the verdict, told him that the publication of the acquittal was delayed to allow Archer to appeal for an amnesty, thereby preventing the ruling from becoming a precedent. See the comments appended to the Saunders and Archer ruling, ibid at 709. The consequences of the acquittal are discussed in Section 4 below.

4. (1555), 73 ER 279; 2 Dy 128b. If we seek an earlier source than Saunders and Archer, supra note 2 to the doctrine of transferred malice, this case seems more appropriate than Salisbury, supra note 2, although a clear rule, such as the one that appears in Saunders and Archer, is not present in this case either.

5. [2011] UKSC 59, [2012] 1 AC 827 [Gnango].

6. It is possible that it was one of the robbers. The verdict omits this detail.

7. The minority opinion acquitted the participants because it was believed that the robbers had no evil intentions toward the woman.

8. The court recognized that the case combines the issues of co-participation and of transferred malice (Gnango, supra note 5 at para 2).

9. The ruling that established the doctrine, Saunders and Archer, supra note 2, explicitly applies it both to cases in which the operation misses its target and to cases of mistaken identity.

10. On the fictitious nature of the doctrine, see Williams, Glanville, Criminal Law: The General Part 2d ed (London: Stevens & Sons, 1961) at 126Google Scholar; LaFave, Wayne R & Scott, Austin W, Substantive Criminal Law (St Paul: West, 1986) at 399Google Scholar; Anthony M Dillof, “Transferred Intent: An Inquiry into the Nature of Criminal Culpability” (1998) 1 Buff Crim L Rev 501 at 506.

11. See Barry Mitchell, “In Defence of a Principle of Correspondence” (1999) Crim L Rev 195.

12. The Law Commission, A Criminal Code for England and Wales, 1988-9, H.C. 299 at 53 (§ 24).

13. The American Law Institute, Model Penal Code, 1962, s 2.03(2)(a) [MPC].

14. Based on the taxonomy devised by Husak, “Intent”, supra note 1 at 69-75.

15. This position is reflected, among other sources, in the following literature: Gross, Hyman, A Theory of Criminal Justice (Oxford: Oxford University Press, 1979) at 102Google Scholar; Prosser, supra note 2 at 653; Kyron Huigens, “Symposium: The Nature, Structure and Function of Heat of Passion Provocation as a Criminal Defense” (2009) 43 U Mich JL Ref 1 at 10–11; Simester, AP, et al. , Simester and Sullivan’s Criminal Law: Theory and Doctrine, 5th ed (Oxford: Hart, 2013) at 165Google Scholar; Card, Richard, Card, Cross and Jones: Criminal Law, 20th ed (Oxford: Oxford University Press, 2012) at 88.Google Scholar The abolitionist position has also been stated within the specific context of events involving multiple perpetrators: CMV Clarkson, “Complicity, Powell and Manslaughter” (1998) Crim L Rev 556 at 559 (‘The accessory is a party to a shared violent venture involving foresight of the death of a human being. All human life is of equal value. Why should the identity of the victim make any difference?’); Smith, KJM, A Modern Treatise on the Law of Complicity (Oxford: Oxford University Press, 1991) at 207Google Scholar: (‘[t]he issue should not be “did A [the accessory] desire or consent to that murder?” but “did A desire or consent to participation in an offence of murder?”’).

16. See Ashworth, Andrew, “The Elasticity of Mens Rea” in Tapper, CFH, ed, Crime, Proof and Punishment: Essays in Memory of Sir Rupert Cross (London: Butterworths, 1981) at 45, 57–58Google Scholar [Ashworth, “Mens Rea”]; Heidi M Hurd & Michael S Moore, “Negligence in the Air” (2002) 3 Theor Inq L 333 at 390.

17. Thus, Husak was reserved in assessing the possibility of deciding between the abolitionist and purist positions: there are no decisive arguments for either position, and ‘all that remains to be said is that reasonable minds may differ in their judgments’. See Husak, “Intent”, supra note 1 at 71, and similarly Peter Westen, “The Significance of Transferred Intent” (2013) 7 Crim L & Phil 321. Moreover, even one who adopts either the abolitionist or the purist position is likely to recognize the need for the doctrine of transferred malice in some circumstances, especially when transferring malice between offences, and when the possibility of attempt liability is not available. For a discussion and illustration of the categories in which the doctrine of transferred malice is likely to survive the assaults of abolitionism and of purism, see Eldar, “Malice”, supra note 1 at 640-49.

18. Occasionally referred to as impersonal or replicated. See Jeremy Horder, “Transferred Malice and the Remoteness of Unexpected Outcomes from Intentions” (2006) Crim L Rev 383; Mitchell Keiter, “With Malice Toward All: The Increased Lethality of Violence Reshapes Transferred Intent and Attempted Murder Law” (2004) 38 USF L Rev 261. Some authors use the type-oriented view of intent to justify the doctrine of transferred malice. See, e.g., Daniel J Curry, “Poe v. State: The Court of Appeals of Maryland Limits the Applicability of the Doctrine of Transferred Intent” (1997) 27 U Balt L Rev 167 at 169-70.

19. For a survey of attitudes toward the level of abstraction of the object of the intent, see Kimberly Kessler Ferzan, “Beyond Intention” (2008) 29 Cardozo L Rev 1147. Note that the purist does not attach an object-specific intent to someone who initially intended to cause general harm without focusing on a given object.

20. The conceptualization of mens rea raises two contradictory intuitions. For a survey of issues that reach a similar dead end, see Christopher Boorse & Roy A Sorensen, “Ducking Harm” (1988) 85 J Phil 115 at 133. It is not surprising that the conceptualization of mens rea is included in this group, as it is based on another question notorious for leading to a similar dead end, that of moral luck. The change in object as a result of missing the target or mistaking the identity of the victim touches upon the question whether it is appropriate to acknowledge the ill luck that caused the intended object not to be harmed and another object to be harmed. For the connection between the doctrine of transferred malice and moral luck, see Kimberly D Kessler, “The Role of Luck in the Criminal Law” (1994) 142 U Pa L Rev 2183 at 2207. Ashworth also linked these topics and based his argument for the purist position on his refusal to acknowledge fortuity in the realization of the act. According to him, the advantage of the purist position is based on the fact that it prevents acknowledging the fortuitous nature of the harm caused to the object in practice. See Andrew J Ashworth, “Transferred Malice and Punishment for Unforeseen Consequences” in Glazebook, PR, ed, Reshaping the Criminal Law: Essays in Honour of Glenville Williams (London: Stevens & Sons, 1978) at 77, 89Google Scholar [Ashworth, “Transferred Malice”]. This argument is not convincing because the purist position acknowledges the fortuity in missing the intended target. We must reluctantly face our inability to reach a clear decision between the levels of conceptualization of intent.

21. For a description of purism as essentially intuitive, see Husak, “Intent”, supra note 1 at 66-67, 69-70.

22. This is also Ashworth’s opinion in “Transferred Malice”: ‘[n]either proposition [framing intentions in general or in respect to the object actually harmed] is deducible from the general principles of mens rea: it depends upon how one chooses to define those principles, and that choice will be influenced by one’s view about the solution of the particular problem presented by the transferred malice situation’. Supra note 20 at 91.

23. See Bohlander, Michael, Principles of German Criminal Law (Oxford: Hart, 2009) at 74.Google Scholar Support for divided purism can also be found in the work of Anglo-American commentators; see for example the position of Ashworth, as it is expressed in the combination of two of his articles: Ashworth, “Mens Rea”, supra note 16 at 57-58, and ‘Transferred Malice’, supra note 20 at 77-78.

24. Eldar, “Malice”, supra note 1 at 636-39.

25. DP = the direct perpetrator of the offence; IP = the indirect or distant participant; A = the intended object of the offence; and B = the actual object that was harmed.

26. I refer particularly to act utilitarianism, as opposed to rule utilitarianism. A utilitarian account of the acts attributed to IP in the above examples leads to equilibrium, as the saving of the life of A is counted against the killing of B. A calculus of rule utility, as any examination of actions that is not based solely on utilitarian considerations, is expected to lead to the condemnation of IP.

27. On the doctrine of innocent or semi-innocent agency see Williams, supra note 10 at 374; The Law Commission, Participating in Crime 100 (Law Com 305, 2007). Any objection to the above analysis on the ground that the offence of murder does not require a direct act on the part of the perpetrator is met in the context of offences that do require such an act, e.g., rape, where the indirect participant directs the rapist to a different victim than the one intended. Yet even in the context of murder, bringing about a homicide by means of another is more accurately dealt with through the laws of complicity and multiple participation than by those concerning the direct causing of harm.

28. We return in Section 4 to the topic of change of object as opposed to a change of modus operandi.

29. Scenarios involving perpetration by means of another similarly emphasize the importance of the specific object. Consider that IP had deflected DP’s hand or misled him (scenarios 1 and 2), not in order to cause him to hurt B instead of A but in order to hurt A indirectly rather than directly (e.g., by causing the explosion of a nearby gas tank, perhaps in order to blur the identity of the perpetrator and make it more difficult to apprehend him). In this case, the object is the same but the modus operandi is different. These scenarios lose their intuitive appeal, and the motivation to impose liability on the indirect perpetrator as a perpetrator by means diminishes significantly.

30. Variables (a)-(c) are mentioned by Smith, supra note 15 at 198. A separate variable is the mental element of the indirect perpetrator: are we going to settle for a factual causal relation between his indirect participation and the harm caused to the different object? Or are we going to demand a mental element of negligence or recklessness in order to convict for harm caused to the different object? Lanham has identified four approaches to the liability of the indirect perpetrator for a change of target by the direct perpetrator, based on: (a) causality, (b) negligence, (c) recklessness, and (d) explicit agreement. See David Lanham, “Accomplices and Transferred Malice” (1980) 96 LQR 110 at 110. Note that the fourth approach is the only one that is unique to joint participation; the others are relevant also to the liability of the individual perpetrator whose act harms someone other than the intended object. This approach, as well as the third one, rejects the possibility of transfer of intent, as both require a subjective mental element regarding the object that was harmed in practice, and therefore they are identical with the purist position or closely resemble it. The first approach allows the unrestricted transfer of intent. The second approach seeks to limit the doctrine in order to strike a balance between it and the requirement that mens rea concur with the physical element of the offence. The second approach is more common in the literature, and since the publication of Lanham’s article, secondary approaches have been formulated based on other touchstones for limiting the scope of the doctrine, as for example, ‘remoteness’, following Horder, supra note 18 or ‘immediate and physical effect’, according to the approach of Glanville Williams, Textbook of Criminal Law, 2d ed (London: Stevens & Sons, 1983) at 181. Below, I refer only to the negligence (or objective foreseeability) approach. I cannot address in full the rationale for choosing the negligence qualifier. For a more detailed discussion, see Eldar, “Malice”, supra note 1 at 649-57. Lanham supports the causality approach. According to him, the liability of the indirect perpetrator should be based on the question whether the harm caused to the object in practice was the direct result of the attempt by the direct perpetrator to consummate the offence against the intended object. Thus, he is critical of the verdict in Saunders and Archer, arguing that the death of the child is a direct result of Saunders’s attempt to poison his wife (ibid at 114). But is that so? Saunders’s failure can be seen as perpetration of murder of the child by means of his (innocent) wife, that is, a new and different offence than the one in which Archer served as accessory.

31. It is possible to argue that the result has changed here, as the intended method was to lead to the awarding of the insurance payment, whereas the method used in practice did not have this result. But concerning the offence at hand (homicide), the result is the same, that is, the death of the wife.

32. Even if the modus operandi of the offence is generally not considered to be substantive, whereas the object of the offence is considered to be such, we should not infer from this fact that the object is the only substantive factor in the offence. In general, the boundaries of the offence are also important, and if the direct perpetrator carries out a different offence from what has been planned with regard to the intended target, the deviation is considered substantive. Thus, if IP gives DP a passkey with the expectation that DP will use it to enter the National Gallery after closing and steal Holbein’s The Ambassadors, whereas DP arrives at the museum during visiting hours and uses the key he received from IP in order to damage the painting, IP is not liable as an accessory to this act despite the fact that the offence was carried out against the intended object.

33. Note that the American Model Penal Code bundles together the change of object with the change of method, and treats them identically. See MPC, supra note 13, s 2.03(2).

34. Saunders and Archer, supra note 2.

35. Ormerod, David, Smith and Hogan’s Criminal Law, 13th ed (Oxford: Oxford University Press, 2011) at 213.Google Scholar This remark may exaggerate the significance of the presence of the direct perpetrator on the scene at the time when the change of object takes place. It may do so by taking for granted whether the change of object was deliberate. If Saunders had not been present during the event, but arranged things so that his wife would pass the apple to his daughter, Archer would be innocent of killing his daughter. And, had Saunders tried to prevent his daughter from consuming the apple, but been unsuccessful in preventing her death, then, despite his presence, the displacement of harm would nevertheless have been accidental, which imposes liability on both Saunders and Archer. It is the combination of Saunders’s presence and his failure to intervene to save his daughter’s life that makes her poisoning a separate intentional act, similar to the case in which Saunders would have used the poison supplied by Archer at another time, in another place, and against a different victim.

36. Michael Bohlander, “Problems of Transferred Malice in Multiple-actor Scenarios” (2010) 74 J Crim L 145 at 150 [Bohlander, “Transferred Malice”]. Contrary to Ormerod, who appears to ascribe exaggerated significance to Saunders’s presence at the scene of the event, Bohlander errs by not ascribing any importance to the question of Saunders’s presence. Granted that from a causal-factual point of view there may be no significance to Saunders’s presence at the scene; nevertheless his presence is likely to transform Saunders’s role from a passive link in the chain of causation into a perpetrator by means of his innocent wife. Thus we may identify three links on the causal chain of the poisoning, by which IP gives the apple to DP, who commits through the agency of W an offence against a new victim. Had the wife told Saunders to pass the apple to the daughter, and he had done so, we would regard this as a new and separate event of direct and intentional killing of the daughter by Saunders. Similarly, given that the apple that Saunders handed to his wife passed to the daughter without his wife’s awareness of the danger, and that Saunders did not prevent this, Saunders can be regarded as an intentional perpetrator of the killing of his daughter by means of her unsuspecting mother. From this vantage point, the case of Saunders and Archer appears similar to the event in which IP hands poison to DP so that he may kill his wife, and DP decides to poison a different person (scenario 6).

37. The distinction between intentional and unintentional deviation is, for example, at the root of Card’s analysis, supra note 15 at 746, and also of Ormerod’s, supra note 35 at 211-13.

38. Hart, HLA & Honore, Tony, Causation in the Law, 2d ed (Oxford: Oxford University Press 1985) at 383–84.Google Scholar They refer to page 203 in their book in a way that can indicate that they nevertheless do agree with the objective foreseeability test for the indirect perpetrator with regard to the deviation made by the direct perpetrator, as on this page they write that ‘a voluntary act which is not reasonably foreseeable … negatives responsibility’. But except for this reference, the discussion focuses on the degree of independence of the decision made by the direct perpetrator. This matter is central in the thinking of Hart and Honore, as expressed in their book. In their opinion, it is not possible to commit an offence by means of a responsible agent because at the point where the liability of the physical perpetrator begins, the causal chain that begins with the actions of the indirect perpetrator ends, and a new causal chain is initiated.

39. This idea is presented also by Williams, supra note 10 at 403.

40. It is important not to confuse the focal points of the test, as did Foster in cases that fall within the category discussed in this section (changing the object of the offence by the direct perpetrator): ‘I believe the following criteria will let the most inquisitive reader into the grounds upon which the several cases falling under this head will be found in turn. Did the principal commit the felony he standeth charged with under the influence of the flagitious advice; and was the event, in the ordinary course of things, a probable consequence of that felony? Or did he, following the suggestions of his own wicked heart willfully and knowingly commit a felony of another kind or upon a different subject?’, Sir Michael Foster, Crown Law, 3d ed (1792) at 372.

41. This is the empirical conclusion reached by Ashworth and Horder after surveying the case law from Saunders and Archer to the present day. See Ashworth, Andrew & Horder, Jeremy, Principles of Criminal Law, 7th ed (Oxford: Oxford University Press, 2013) at 443.Google Scholar

42. Ibid at 189.

43. Admittedly, this alternative is not a common occurrence because normally it is the distant participant who is expected to envision a broader variety of future possibilities than does the direct perpetrator, from whose point of view the realistic alternatives are limited to the circumstances surrounding him at the time of commission. Nevertheless, consider a case similar to scenario 4 (accidental deviation by the direct perpetrator), in which IP sends DP to kill a shopkeeper at a time when the shop is closed and usually occupied by the shopkeeper alone. Upon entering the shop, DP notices a vagrant hiding in the store after business hours; DP proceeds to shoot at the shopkeeper but accidentally hits the vagrant. It may prove easier to illustrate the possibility of higher expectations from the direct perpetrator using cases involving mistaken identification (scenario 5): IP sends DP to burn down A’s house, convinced that DP knows the address. By an unlikely mistake, DP sets fire to B’s house.

44. The liabilities of instigators and aiders are tested derivatively, that is, as derivatives of the direct perpetrator’s liability for the offence. The issue of transferred malice raises a challenge before the doctrine of derivative liability: are the rationales that apply to scenarios of accidental miss-aim or mistaken identification valid in the case of secondary (derivative) intent on the part of the instigator and the aider? Is it possible to derive from the transferring of the intent of the direct perpetrator from one object to another a similar transfer in the intent of the indirect participant as well? In other words, concerning the scenarios involving instigation and aiding discussed in this section, the law must decide whether to derive the mens rea of the indirect participant from that of the direct perpetrator, or to focus the test directly on the intent of the indirect participant. To illustrate, consider that IP provided a gun to DP in order to kill A, and that DP missed A and killed B instead, or alternatively, mistook B for A and killed him. Derivatively, the former deviation represents an accidental miss-aim on the part of DP, and therefore also on the part of IP, whereas the latter is a mistaken identification on the part of DP, and therefore also on the part of IP. By contrast, direct consideration of IP’s intent treats both events as accidents as far as he is concerned, because from his point of view his indirect act against the intended object missed its target, whether DP changed the object by accident or because of a mistaken identification; See Bohlander, “Transferred Malice”, supra note 36 at 148. The situation is different in scenarios where the indirect participant is classified as a perpetrator by means of an innocent agent, e.g., a father who sends his minor son to poison his mother, and the son poisons his sister instead. The doctrine of derivative liability does not apply to the perpetrator by means, and therefore his mental state is tested directly. The father is perceived as having accidentally missed his target whether his son replaced the object of poisoning because of a mistaken identification or because of a missed target.

45. A similar consistency is achieved by abandoning derivative liability on which complicity in the English criminal law is based. It follows that one may use the example in the text to illustrate failures in the method of derivative liability. For a recent analysis of the shortfalls of derivative liability, see Douglas Husak, “Abetting a Crime” (2014) 33 Law & Phil 41.

46. Although we must be open to the possibility that the superior understanding of the indirect participant makes him a perpetrator in the killing of B by means of the semi-innocent agency of the direct perpetrator.

47. Bohlander, “Transferred Malice”, supra note 36 at 151.

48. Gnango, supra note 5.

49. The court accepted that in some cases an affray may base joint perpetration, for example, when there is a preliminary agreement to battle between the parties to the affray (ibid at para 39). Note that the Court demarcated affray and riot (ibid at para 41) and expressed reluctance to extend joint liability for missed operations in the case of large groups, indicating that an increase in the number of participants indeed affects the debate on transferred malice. Ormerod addressed this aspect of the ruling in David Ormerod, “Worth the Wait?” (2012) Crim L Rev 79 at 80.

50. For the details of the verdict, see Section 1 above.

51 An opposing view was taken by the Court of Appeal for Ontario in R v JSR, 2008 ONCA 544, 237 CCC (3d) 305; 239 OAC 42. The evidence at the preliminary inquiry in this case allowed the possibility that the accused was involved in a mutual gun fight, in which one of the bullets directed at the accused strayed and caused the death of a young female bystander. On these facts, the court concluded that the firing of the fatal shot formed part of a joint activity undertaken by the accused and his adversary and may therefore be attributable to the accused.

52. Note that this is not a standard case of encouragement, which falls under instigation or accessoryship to committing an offence, but a case of provocation. See Damian Warburton, “Murder—Whether Secondary Liability by Joint Enterprise Arises in Circumstances of Mutual Conflict Between Defendants” (2011) 75 J Crim L 457 at 461. Richard Buxton pointed out an additional difficulty in attributing accessoryship to murder to Gnango: according to him, even if we accept the ruling of the court that Gnango wanted to assist the Bandana Man, at most it is possible to say that he wanted to assist the Bandana Man in trying to kill him, not to succeed in it. It follows that Gnango’s verdict should not be derived from the consummated offence of killing the passerby. Richard Buxton, “Being an Accessory to One’s Own Murder” (2012) Crim L Rev 275 at 278.

53. Lord Clarke went further and drew a parallel between this case and the verdict in the matter of R v Pagett (1983), 76 Cr App R 279, in which it has been decreed that when a police officer fires at an assailant in reasonable self-defence against deadly force, but misses and kills a passerby, the assailant is convicted of the death of the passerby. See Gnango, supra note 5 at para 83-89. The parallels between these two cases is problematic because in Pagett, the defendant created the event by himself and was responsible for it in its entirety. By contrast, in Gnango the defendant was not fully responsible for the event. Gnango did not coerce the Bandana Man to engage in a shootout with him, whereas the intervention of the police officer in Pagett was the result of a constraint created by the defendant. Indeed, Pagett is an example of perpetration by means of another, whereas Gnango is not, despite the possibility of a causal connection between Gnango’s actions and the killing of the passerby at the hands of the Bandana Man.

54. The Court pointed out this admixture in the case of Gnango, supra note 5 at para 2: ‘The particular areas of criminal law that will have to be considered are (i) joint enterprise; (ii) transferred malice; (iii) exemption from liability where a party to what would normally be a crime is a victim of it’.

55. The minority judge in Gnango, Lord Kerr, was not willing to classify the conduct broadly, as a violent encounter, and therefore did not convict Gnango as a joint perpetrator of the killing; nor did he regard Gnango as the person who caused the Bandana Man to shoot, and therefore he did not convict him either as an accessory to the killing of the passerby. For further opinions concerning this case, see the debate in Graham Virgo, “Joint Enterprise Liability is Dead: Long Live Accessorial Liability” (2012) Crim L Rev 275; Peter Mirfield, “Guilt by Association: A Reply to Professor Virgo” (2013) Crim L Rev 577; Graham Virgo, “Guilt by Association: A Reply to Peter Mirfield” (2013) Crim L Rev 584.

56. Mansell and Herbert’s case, supra note 4.

57. See supra notes 17-22 and accompanying text.

58. Bohlander, “Transferred Malice”, supra note 36 at 157.

59. Buxton, supra note 52 at 279-80. Ormerod suggests an additional consideration in this matter: ‘Is it possible for an intended victim to be liable for assisting his own attempted murder? Consider A and B who are suicide bombers: they intend to blow up a deserted landmark in the depth of night and each other in the process. A’s bomb does not detonate and he survives the blast. Can A, by being an accessory to B’s conduct, be guilty of attempting to murder himself? Such a result would look odd given that if A succeeded in killing B he would only be liable as part of a suicide pact for his manslaughter’. David Ormerod, “Joint Enterprise: Murder – Killing of Bystander by Other Party in Gunfight” (2011) Crim L Rev 151.

60. When the defendant intends harm to a non-particular object, as in the case of the indiscriminate shooter, both levels of abstraction reach the same result, as the defendant’s specific intent includes all objects of the same type. See supra note 19.