Royall v The Queen

[1991] HCA 27

High Court of Australia|1986-11-16|Before: McHugh JJ, Dawson JJ

View original source

At a glance

Source facts

Court

High Court of Australia

Decision date

1986-11-16

Before

McHugh JJ, Dawson JJ

Source

Original judgment source is linked above.

Judgment (309 paragraphs)

[1]

The applicant's case at the trial, as expressed in his unsworn statement, was that, after the deceased had entered the bathroom to take a shower, he had become concerned for her. He had heard a thump on the wall. He said that he knew that the deceased had used amphetamines over a long time and that she suffered from epilepsy which caused her to faint or become unconscious. This knowledge, it was suggested, was the foundation of his concern for the deceased and of the claim that she took her own life by voluntarily jumping out the window. The applicant forced the lock with a knife and banged against the door until it opened. He said: "When I finally got it open she was going out the window." There was evidence that, in his record of interview, he said that, when the bathroom door opened, "I just saw the back of Kelly going out the window. She just jumped out." The applicant denied that any violence was done to the deceased or that he touched her in the bathroom. Moreover, his case was that he did not cause her death or intend to injure her.

[2]

The case was left to the jury on the footing that it was for them to determine whether the applicant had caused the deceased's death in any of the three ways suggested by the Crown. The trial judge instructed them that, if they were satisfied that the applicant caused the deceased's death, they should consider whether he had the requisite intent. His Honour stated that the requisite intent would be satisfied by an intent to kill, an intent to inflict grievous bodily harm or reckless indifference to human life. In making this statement his Honour did not differentiate between the three ways in which the Crown suggested that the applicant had caused the deceased's death; in other words, the jury was left with the impression that, if they were satisfied on the issue of causation in any one of the three suggested ways, the requisite intention would then be satisfied by an intent to kill or an intent to inflict grievous bodily harm or reckless indifference to human life.

[9]

It is convenient to deal initially with the question of causation. The applicant submits that neither the trial judge nor the Court of Criminal Appeal defined adequately the act done by the applicant which was alleged to have caused the death of the deceased, thereby creating difficulties in determining the issues of causation and intent, in particular the coincidence of act and intent. Ordinarily there is no occasion for a trial judge to spend much time on the identification of the act causing death, but there is a "logical and practical necessity to isolate that act, for it is of it, and it alone, that one or more of the several specified conditions or concomitants must be predicated if the terms of s. 18 are to be satisfied", to repeat the words of Barwick C.J. in Ryan v. The Queen [1] . In Ryan the trial judge's directions were deficient in that they failed to isolate the particular act or acts which the jury might identify as the cause of death. In that case there was room for argument about what was the act which caused death. Different considerations arose for determination in ascertaining whether Ryan's state of mind satisfied the requirements of s. 18, depending upon which act was identified as the cause of death. The Crown case was that Ryan went to rob a service station. While he had his finger on the trigger of a loaded, cocked gun pointed at the deceased's back, with his other hand he tried to find a cord in his pocket. The deceased made a sudden movement, Ryan stepped back and the gun discharged, killing the deceased. If pressing the trigger was identified as the act causing death, the question was whether Ryan willed that act and intended to kill or inflict grievous bodily harm or whether it was an unwilled reflex movement. If, however, presentation of the gun was identified as the act causing death, the question was whether Ryan knew that in the circumstances the involuntary discharge of the gun was probable: see per Barwick C.J. [2] ; and note Reg. v. Crabbe [3] (where the test applied by Barwick C.J. in Ryan was discarded in favour of that just stated).

[19]

Generally speaking, an act done by a person in the interests of self-preservation, in the face of violence or threats of violence on the part of another, which results in the death of the first person, does not negative causal connexion between the violence or threats of violence and the death. The intervening act of the deceased does not break the chain of causation. But the governing principle has been expressed in a variety of different ways. The trial judge's direction was based on that approved by the Full Court of the Supreme Court of New South Wales in Reg. v. Grimes and Lee [12] (where the deceased jumped to his death from the window of a railway carriage in order to escape from the accused who had assaulted and robbed him). The key element in the chain of causation, as contemplated by the direction, is that an accused's conduct creates in the mind of the victim a well-founded and reasonable apprehension of danger as a result of which the victim takes steps to escape leading to his or her death. According to this view, it is enough that the victim's apprehension of danger is well-founded and reasonable; there is no requirement that the steps taken to escape should be reasonable. However, in England, formulations have been applied from time to time which pay more attention to the mode of escape chosen by the victim which results in death, no doubt with a view to emphasizing the causative link between the accused's conduct and the death of the victim. Thus, it has been held that a reasonable act performed for the purpose of self-preservation in attempting to escape the violence of the accused does not destroy the causative connexion with the accused's violence:Reg. v. Pitts [13] ; Reg. v. Pagett [14] , per Robert Goff L.J. In this context, "reasonable" means reasonable in the light of the accused's conduct and the apprehension of danger which it induced in the mind of the victim. A variant on this formulation is to say that the victim's act which results in death is caused by the accused's violence if the act is "the natural consequence" of that violence, something that is very likely to happen: Reg. v. Beech [15] , per Darling J. (where the victim was injured after jumping out of a window to escape from a man whom she had reason to fear). The natural consequence test has been explained as a test which poses the question whether the victim's act "was something that could reasonably have been foreseen as the consequence of what [the accused] was saying or doing": Reg. v. Roberts [16] , per Stephenson L.J. So if an act is so unexpected that no reasonable person could be expected to foresee it, then it is a voluntary act on the part of the victim which breaks the chain of causation between the accused's conduct and the ultimate injury: Roberts [16] .

[60]

The basic proposition relating to causation in homicide is that an accused's conduct, whether by act or omission, must contribute significantly to the death of the victim: Reg. v. Pagett [40] . It need not be the sole, direct or immediate cause of the death. However, when the death is not caused directly by the conduct of the accused but by something done by the victim or by a third person in response to the conduct of the accused, there is a question whether the chain of causation has been broken. If the response is a reflex or automatic reaction to the conduct of the accused, the chain of causation is not broken. It is also well established that the death may be held to have been caused by an accused's conduct though the direct or immediate cause of death is a final fatal step taken by the victim where that step is taken in an attempt to preserve himself or herself from physical harm which unlawful conduct on the part of the accused has induced the victim to fear, provided the victim's attempt at self-preservation is reasonable having regard to the nature of the accused's conduct and the fear it was likely to have induced (or, as I would prefer, provided the attempt is proportionate to that conduct and the fear it was likely to have induced): Reg. v. Pitts [41] ; Reg. v. Grimes and Lee [42] ; R. v. Curley [43] ; Director of Public Prosecutions v. Daley [44] . In such a case, the taking of the final fatal step is not a novus actus interveniens breaking the chain of causation: Pagett [45] , per Robert Goff L.J. On the other hand, where the victim's attempt at self-preservation is not reasonable (or proportionate), the chain of causation is broken and the victim's death is not treated as having been caused by the accused's conduct. Subject to a qualification presently to be mentioned, the question whether the chain of causation is broken by the victim's taking of the final fatal step is a question of fact to be answered by reference to the objective circumstances. Nevertheless, an accused cannot be held criminally responsible for a death that has been caused in fact by his conduct if the final fatal step taken by the victim was neither foreseen nor reasonably foreseeable. Foresight or reasonable foreseeability marks the limit of the consequences of conduct for which an accused may be held criminally responsible. In Roberts [46] , Stephenson L.J. said (in a case of an assault occasioning bodily harm):

[75]

The second and third hypotheses clearly directed the jury's attention to the reasonableness (or proportionality) of the avoiding action which the victim may have taken. The third hypothesis does not direct the jury's attention to the test of reasonable foreseeability of the possibility that the victim might jump out of the window in an attempt to save her life, but the hypothesis postulates conduct on the part of the accused which induces "a well-founded and reasonable apprehension that she would be subjected to such further violence as would endanger her life". In circumstances where the window offered the only possible avenue for escape from the accused's further violence, and in a case where the accused is found to have acted either with reckless indifference to the victim's life or with an actual intent to kill or to inflict grievous bodily harm, it would be perverse not to find that the accused could reasonably have foreseen that she might attempt to escape by jumping out the window. It may be that, if the third hypothesis be true, the accused neither intended the deceased to jump out the window nor foresaw that she might do so but that state of mind is not inconsistent with his guilt of murder. Provided it was reasonably foreseeable by the accused in the circumstances postulated in the third hypothesis that the victim might jump out the window, the jury was entitled to be satisfied that her death was caused by the accused's conduct in one or other of the ways mentioned in the first three hypotheses. If the external elements of the offence were thereby proved, the jury could properly convict of an offence of culpable homicide and, if any of the mental elements specified in s. 18(1)(a) were also proved, of murder. It was immaterial that there might have been a disconformity between the mental element (intent to inflict grievous bodily harm or reckless indifference to human life) and the actual steps in the chain of causation between the accused's conduct and the victim's death.

[223]

The difficulty with the Court of Criminal Appeal's analysis is that it does not reveal why the verdict was safe and satisfactory if the jury found that the applicant attacked the deceased in the bathroom with intent to kill or with reckless indifference to her life or if the jury found that the third of the three "possibilities" was the cause of the deceased's fall. The Court of Criminal Appeal was required to make its own independent assessment of the evidence to ascertain whether the verdict was safe and satisfactory: see Chidiac v. The Queen [5] . The Court of Criminal Appeal, however, has found some only of the possible ways in which the jury may have reached their verdict to be safe and satisfactory, and has expressed no opinion as to whether a verdict, found in one of the other ways left to the jury, would be safe and satisfactory. Nevertheless, the Crown contended that a verdict of guilty may be safe and satisfactory even though the jury had an issue left to them which they could not safely consider. To support this proposition, the Crown relied on Chamberlain v. The Queen [No. 2] [6] . In Chamberlain [7] , Gibbs C.J. and Mason J. held that, in considering a charge of murder, the jury could not safely act upon the basis that the evidence established that certain blood was foetal blood. Nevertheless, their Honours held that the verdict of guilty was not unsafe or unsatisfactory because there was other evidence to support that verdict. Their Honours reached this conclusion despite the trial judge's direction to the jury that they could find that the blood was foetal blood. But, with great respect, if there is a real possibility that the jury may have acted upon an unsafe basis in reaching their verdict, I think that the verdict must be regarded as unsafe and unsatisfactory. Indeed, the case where a verdict may have been based on evidence or decided on an issue upon which the jury could not safely act seems to me to be the paradigm of an unsafe or unsatisfactory verdict. The remarks of Gibbs C.J. and Mason J. in Chamberlain were authority for the course which the Court of Criminal Appeal took in the present case. But in my view no verdict in a criminal trial can be regarded as safe and satisfactory if the jury may have reached their verdict on evidence or by reasoning which is unsafe or unsatisfactory. In so far as any remarks in Chamberlain may suggest or imply the contrary, they are in my opinion wrong and should not be followed.

[263]

Judicial and academic efforts to achieve a coherent theory of common law causation have not met with significant success. Perhaps the nature of the subject matter when combined with the lawyer's need to couple issues of factual causation with culpability make achievement of a coherent theory virtually impossible. But there is little hope of obtaining a coherent theory of causation if the principles of causation in criminal cases are significantly different from those in the civil law. No doubt the object of the civil law is not the same as the object of the criminal law. But both areas of law use principles of causal responsibility to limit liability for the consequences of wrongful acts. Tort and contract law do so, inter alia, by rules of remoteness of damage which are based on notions of justice and morality: cf. Overseas Tankship (U.K.) Ltd. v. Morts Dock & Engineering Co. Ltd. (The Wagon Mound [No. 1]) [42] . Criminal law does so by reference to rules which are based on notions of moral culpability. Speaking generally, the broad principles of causation applicable in civil cases should be equally applicable in criminal cases. The law of negligence, where issues of causation have arisen frequently, has used the doctrines of reasonable foreseeability and novus actus interveniens to limit responsibility for negligent acts and omissions which are causally connected with injury suffered: Mahony v. J. Kruschich (Demolitions) Pty. Ltd. [43] . Reasonable foreseeability has been used to limit the liability of a tortfeasor because "it does not seem consonant with current ideas of justice or morality that for an act of negligence, however slight or venial, which results in some trivial foreseeable damage the actor should be liable for all consequences however unforeseeable and however grave, so long as they can be said to be "direct" ": The Wagon Mound [No. 1] [44] . The novus actus interveniens doctrine has been used to limit the liability of a tortfeasor because he or she is perceived as having no moral responsibility for damage when a fully-informed actor, not acting under constraint or pressure flowing from the tortfeasor's actions, has intervened and produced that damage even though it would not have occurred but for the tortfeasor's act or omission. For the same reasons, in a criminal case, a person should not be held liable for a wrongful act or omission which has caused harm in a "but for" sense if that harm was the product of a novus actus interveniens or was not a reasonably foreseeable consequence of the act or omission. It goes almost without saying, however, that a person should be held liable for harm which is causally linked with his or her conduct and which he or she intended should be brought about by that conduct: cf. Hart and Honoré, Causation in the Law, 2nd ed. (1985), p. 79.