1. [1975]3All E.R.446(C.A.).

2. The choice of the masculine here is deliberate: in this and all the other cases I will consider, the accused was a man. As with Blaue, in many of the cases I will consider, the victim was a woman. I chose the masculine to refer to the wrongdoer and the feminine the victim to reflect this pattern.

3. Supra note 1 at 450.

4. Ibid, at 450.

5. H.L.A. Hart & Tony Honoré, Causation in the Law, 2d ed. (Oxford: Clarendon Press, 1985) at 361.

6. [1963]Qd. R. 157 (C.C.A.).

7. [1969] S.A.S.R. 141 (S.C.A.).

8. (1971), 56 Crim.App.R. 95 (C.A.).

9. I will argue below that the intuitions of fairness that underlie the rejection of these two answers rest upon the fact that the adoption of either of these answers would violate what I have called the principle of equality.

10. See Eric Colvin, Principles of Criminal Law, 2d ed. (Toronto: Thompson, 1991) at 78–90, and Glanville Williams, Textbook of Criminal Law (London: Stevens and Sons, 1983) at 378–401. I will stick with “proximate cause.”

11. For a sceptical look at this claim in the context of tort law, see Wex S. Malone, “Ruminations on Cause-In-Fact” (1956) 9 Stanford L.R. 60.

12. See supra note 5 at 109–29, Richard W. Wright, “Causation in Tort Law” (1985) 73 California L.R. 1735, and Richard Wright, “Causation, Responsibility, Risk, Probability, Naked Statistics, and Proof: Pruning the Bramble Bush by Clarifying the Concepts” (1988) 73 Iowa L.R. 1001.

13. Jill Presser argues that the substantial cause test represents what she calls a threshold test approach to causation, namely one where the distinction I have drawn between cause-in-fact and proximate cause is not relied upon. Instead, the courts set a threshold above which causation is taken to be satisfied. See Jill Presser, “All for a Good Cause: The Need for Overhaul of the Smithers Test of Causation” (1994) 28 C.R. (4th) 178. This is certainly a plausible way to look at things, insofar as, as I have suggested, the substantial or operating cause test distinguishes the sort of causal contribution represented by proximate causation from that represented by cause-in-fact merely in terms of degree. Nonetheless, I think the line ought to be drawn here as well as with the approach represented by the reasonable foreseeability test, insofar as the satisfaction of the minimal threshold represented by the but-for test is taken by both approaches to be a necessary condition for further causal inquiry (save, again, in cases of causal overdetermination).

14. R. v. Smith, [1959] 2 Q.B. 35 at 42–3.

15. Hallet, supra note 7 at 149.

16. Ibid. I will argue below that reliance on such provisions shows that elements of foresight creep into the substantial or operating cause test.

17. Supra note 1 at 450.

18. Supra note 10 at 84.

19. Ibid.

20. Supra note 6 at 168.

21. The old case to which Stephenson L.J. refers here is R. v. Beech (1912), 7 Crim.App.R. 197. The facts of Beech are very similar to the facts of Roberts. In Beech a man broke into his victim’s house, and while he kicked and beat at her door, she jumped from a window, sustaining serious injuries. At 200, Darling J., writing for the English Court of Criminal Appeal, reasoned that “[n]o one can say that if she jumped through the window it was not the natural consequence of the prisoner’s conduct.” Roberts represents the interpretation of the standard in Beech in terms of reasonable foreseeability: rather than ask “was the victim’s action the natural consequence of the of the accused’s actions?” we ask “was it reasonably foreseeable that the victim would do what she did?”

22. Swpra note 8 at 102.

23. However, in its report on codifying Canadian criminal law, the Law Reform Commission of Canada appealed in part to the standard of reasonable foreseeability.

2(6) Causation. Everyone causes a result when his conduct substantially contributes to its occurrence and no other unforeseen and unforeseeable cause supersedes it.

Law Reform Commission of Canada, Report 31: Recodifying Criminal Law (Ottawa: Law

Reform Commission of Canada, 1987) at 27.

24. (1977) 34 C.C.C. (2d) 427 (S.C.C.).

25. Ibid, at 435.

26. On this issue see Allan Manson, “Rethinking Causation: The Implications oí Harbottle” (1993) 24 C.R. (4th) 153 at 158–9. Some have thought this test to be over-inclusive. Writing for the Alberta Court of Appeal, McClung J.A. characterized the de minimis test as a one of “sweeping accountability,” inconsistent with “principles of fundamental justice,” in the language of s. 7 of the Canadian Charter of Rights and Freedoms. R. v. F.(D.L) (1989), 73 C.R. (3d) 391 (Alta C.A.), per McClung J.A. Section 7 of the Charter provides that “[ejveryone has the right to life, liberty and security of the person and the right not to be deprived thereof except in accordance with the principles of fundamental justice.”

27. (1993), 24 C.R. (4th) 137 (S.C.C.).

28. R. v. Cribben (1994), 17 O.R. (3d) 548 (Ont. C.A.). Writing for a unanimous court, Arbor J.A. argued that, insofar as the actus reus of murder and manslaughter is the same, the same causal test—whatever it turns out to be—must be applied to both.

29. Supra note 13 at 186.

30. The criminal law is loathe to partition blame. The idea is not that the application of the doctrine of novus actus interveniens may result in attributing say 25% responsibility to A and 75% responsibility to B. It is rather that A may have to answer for assault, whereas B may have to answer for homicide, in the case that after A assaults C and B takes advantage of C’s state to kill C.

31. [1959] 2 All E.R. 193.

32. Supra note 6 at 168.

33. Supra note 5 at 136.

34. These approaches need not be mutually exclusive. In fact, courts sometimes argue in the alternative, as did the English Court of Appeal in R v. Pagett (1983), 76 Crim.App.R. 279. Pagett used a woman as a shield as he shot at two police officers who were in pursuit of him. The police officers returned fire and the woman was killed. The Court dismissed Pagett’s appeal from his conviction for manslaughter in the death of the woman. After cautiously endorsing Hart and Honoré’s view (supra note 5), Goff L.J. held, at 289, that it was an “almost self-evident proposition” that “a reasonable act performed for the purpose of self-preservation, being of course itself an act caused by the accused’s own act, does not operate as novus actus interveniens.” This proposition follows, on the Court’s reasoning, from the fact that—as we saw in Roberts—A may be held responsible for the injuries his victim sustains while attempting to escape from A. Goff J.L. argued at 289 that “one form of self-preservation is self-defence; for present purposes, we can see no distinction in principle between an attempt to escape the consequences of the accused’s act, and a response which takes the form of self-defence.”

Further, the police officers were acting in accordance with a legal duty, another factor which on Goff L.J.’s view supported the conclusion that their actions did not constitute novus actus interveniens. On either account, it is immaterial whether the police officer’s actions were a reasonably foreseeable consequence of Pagett’s actions. Instead, the salient feature of these actions is that they were, in Goff L.J.’s words, “caused by the accused’s own act.” Alternatively, me Court held at 291 that

if, as the jury must have found occurred in the present case, the appellant used Gail Kitchen by force and against her will as a shield to protect him from any shots fired by the police, the effect is that he committed not one but two unlawful acts, both of which were dangerous—the act of firing at the police, and the act of holding Gail Kitchen as a shield in front of him when the police might well fire shots in his direction in self-defence.

This suggests that the fact that the police officers’ actions did not constitute novus actus follows from the fact that it was reasonably foreseeable that they would return Pagett’s fire.

35. See Don Stuart, Canadian Criminal Law, 3d ed. (Toronto: Thompson Publishing, 1995) at 132–33.

36. Supra note 24 at 437–38.

37. R. v. Nicholson (1926), 47 C.C.C. 113.

38. Supra note 24 at 437–38.

39. Ibid, at 437.

40. Supra note 1 at 450.

41. Ibid, 450

42. Supra note 10 at 88–89.

43. Supra note 5 at 361.

44. Ibid.

45. We can certainly do so here, insofar as the omission in this case would be an omission to perform a legally recognized duty.

46. Supra note 5 at 361.

47. See supra note 36 and accompanying text.

48. Supra note 10 at 88–89.

49. Supra note 7 at 149.

50. Ibid. at 150.

51. The point I am making is supported, by analogy, by the argument in Beech, supra note 21, that “natural consequences” should be interpreted to mean “reasonably foreseeable consequences.” See supra note 23.1 note that Hart and Honoré, supra note 5 at 254–84, explicitly reject this view, and regard the standard of reasonable foreseeability as an alternative to ordinary rules of causation.

52. I set aside, as well beyond the scope of this paper, the question of what due regard for the interests of others consists in. In the context of tort law, the question may be rephrased in terms of asking what the standard of care imposed by the duty of care consists in. Judicial opinion and academic commentary on this issue is considerable. My point here (thankfully!) does not require settling the issue.

53. It may seem morally out of place to consider whether it would be fair to ask Hallet to consider, insofar as we are asking whether it would be fair to ask too much of him when deciding where to leave the body of the man he violently assaulted. Imagine different facts then, where what we are asking is whether we could fairly ask a good Samaritan to guard against the chance of tidal waves when deciding whether to move Hallet’s victim before going for help.

54. As Rawls suggests, this distinction—whose roots are likely in Kant’s distinction between hypothetical and categorical imperatives—is reflected in ordinary language. “We say ‘[tjheir proposal was perfectly rational given their strong bargaining position, but it was nevertheless highly unreasonable, even outrageous’.” John Rawls, Political Liberalism (New York: Columbia University Press, 1993) at 48. Rather than define it directly, Rawls develops the concept of reasonableness in terms of the qualities a reasonable person would have; these are in the end all motivated by the desire to interact with others on terms of equality. I borrow my formulation of Rawls’ contrast from Arthur Ripstein, to whom I am also indebted for the idea that Rawls’ contrast can help illuminate the content of the idea of reasonableness as it is used in objective standards in tort and criminal law. See Arthur Ripstein, “Self-defence and Equal Protection” (1996) 57 U. Pitt. L.R. 685 at 689. For an application of Rawls’ distinction in the context of tort law, see Gregory Keating, “Reasonableness and Rationality in Negligence Theory” (1996) 48 Stanford L.R. 311.

55. I say “aspects of wrongful conduct” to highlight the fact that one and the same action can give rise to both criminal and civil proceedings, for example, assault in criminal law and battery in tort.

56. Supra note 54. Ripstein defends the objective standard in the doctrine of self-defence on these terms of analysis: it is, on his view, the only acceptable standard because it alone distributes the liberty and security interests of the parties to the violent confrontation on terms of equality.

57. See, for example, Lange v. Hoyt, 114 Conn. 590, 159 A. 575 (1932).

58. Donald Galloway, “Causation in Criminal Law: Interventions, Thin Skulls and Lost Chances” (1989) 14 Queen’s L.J. 71 at 78.

59. Galloway continues “…and therefore unforeseeable,” and hence constituted novus actus inter-veniens. As we will see, on my view we have no business evaluating the reasonableness of Woodhead’s belief, so I here set aside the question of whether the inference from unreasonableness to unforeseeability on which Galloway relies is, as he seems to suppose, analytic. For what its worth, I disagree. On this issue see Arthur Ripstein, Equality, Responsibility and the Law (Cambridge: Cambridge University Press, forthcoming) at ch. 5.

60. Supra note 58 at 78.

61. Ibid.

62. So the common law held until the notorious case of/?, v. Daviault, [1994] 3 S.C.R. 63, where a narrow majority of the Supreme Court of Canada held that self-induced extreme intoxication could be a defence to sexual assault, inasmuch as the defendant was in a state of automatism when the assault occurred. The legislature quickly intervened, and codified what was (probably) the common law all along.

63. R. v. Rabey (1980), 54 C.C.C. (2d) 1 (S.C.C.), aff’g (1977), 37 C.C.C. (2d) 461 (Ont. C.A.).

64. Thanks to Joseph Ellin for suggesting this defence.

65. See Dubé v. Lobar, [1986] 1 S.C.R. 649 (S.C.C.).

66. One may respond that the better analogy in the case of physical thin skulls is with the defence of contributory negligence. But the moral considerations cited here forbid equating medical conditions which predispose one to greater injury with contributorily negligent actions no less than equating them with voluntarily assumed risks.

67. This is true even on the most strict deterrence theory. To say that we punish murder more severely than petty theft because we are more concerned to deter the former than the latter is to pass judgment on their relative severity.

68. The point here is not tied to thin skull beliefs. One may substitute for Blaue, a case where some physical condition caused the defendant to die of internal bleeding despite receiving treatment which would have saved an average person.

69. Supra note 7 at 149.

70. Thanks to Nathan Brett and Ruth Opochinsky Dick for raising this objection.

71. So Kant argues with respect to the duty of truth-telling in his notorious “On the Supposed Right to Lie from Benevolent Motives” in Joram Graf Haber, ed., Absolutism and its Consequentialist Critics (Lanham, MD: Rowman and Littlefield, 1994) at 15–19.