Judgments - Regina v. Smith (On Appeal From The Court of Appeal (Criminal Division))

(back to preceding text)

    The theoretical basis for the distinction is that provocation is a defence for people who are, as Professor Ashworth put it, "in a broad sense mentally normal:" see [1976] C.L.J. at p. 312. If they claim that they had abnormal characteristics which reduced their powers of self-control, they should plead diminished responsibility. There is a clear philosophical distinction between a claim that an act was at least partially excused as normal behaviour in response to external circumstances and a claim that the actor had mental characteristics which prevented him from behaving normally: see Sir Peter Strawson, Freedom and Resentment, in Free Will (Watson, ed. 1982) at pp. 64-67.

    The difficulty about the practical application of this distinction in the law of provocation is that in many cases the two forms of claim are inextricably muddled up with each other. A good example is the recent New Zealand case of Reg. v. Rongonui (Court of Appeal, 13 April 2000, unreported). The accused was a woman with a history of violence against her, suffering from post-traumatic stress disorder. The alleged provocation was that a neighbour she was visiting to ask for help in babysitting her children had produced a knife - not in a threatening way, but sufficient to make her lose control of herself, seize the knife and stab the neighbour to death. The Court of Appeal agreed that it was very difficult in such a case to distinguish between the gravity of the provocation (the accused's previous experience of violence making the mere production of a knife a graver provocation than it would be to someone who had led a more sheltered life) and the accused's capacity for self-control which had been affected by the psychological stress of the violence she had suffered. Tipping J., giving one of the majority judgments which held that the New Zealand statute on provocation (section 169 of the Crimes Act 1961) mandated the application of the distinction, said that it required "mental gymnastics." Thomas J., who thought that the statute did not have to be construed so rigidly, said that most trial judges had seen:

    "the glazed look in the jurors' eyes as, immediately after instructing them that it is open to them to have regard to the accused's alleged characteristic in assessing the gravity of the provocation, they are then advised that they must revert to the test of the ordinary person and disregard that characteristic when determining the sufficiency of the accused's loss of self-control."

    Professor Stanley Yeo, in his recent book Unrestrained Killings and the Law (1998) at p. 61 points out that the reason why jurors find the distinction so difficult is that it:

    "bears no conceivable relationship with the underlying rationales of the defence of provocation. . . . The defence has been variously regarded as premised upon the contributory fault of the victim and, alternatively, upon the fact that the accused was not fully in control of his or her behaviour when the homicide was committed. Neither of these premises requires the distinction to be made between characteristics of the accused affecting the gravity of the provocation from those concerned with the power of self-control."

    Besides these practical difficulties in explaining the distinction to the jury, I think it is wrong to assume that there is a neat dichotomy between the "ordinary person" contemplated by the law of provocation and the "abnormal person" contemplated by the law of diminished responsibility. The Act of 1957 made a miscellany of changes of the law of homicide which can hardly be described as amounting to a coherent and interlocking scheme. Diminished responsibility as defined in section 2 ("such abnormality of mind . . . as substantially impaired his mental responsibility for his acts and omissions . . .") is a general defence which can apply whatever the circumstances of the killing and was introduced because of what was regarded as the undue strictness of the defence of insanity. Provocation is a defence which depends upon the circumstances of the killing and section 3 was introduced, as I have suggested, to legitimate the consideration by juries of "factors personal to the prisoner." If one asks whether Parliament contemplated that there might be an overlap between these two defences, I think that the realistic answer is that no one gave the matter a thought. But the possibility of overlap seems to me to follow inevitably from consigning the whole of the objective element in provocation to the jury. If the jury cannot be told that the law requires characteristics which could found a defence of diminished responsibility to be ignored in relation to the defence of provocation, there is no point in claiming that the defences are mutually exclusive.

    There are in practice bound to be cases in which the accused will not be suffering from "abnormality of mind" within the meaning of section 2 ("a state of mind so different from that of ordinary human beings that the reasonable man would term it abnormal": Reg. v. Byrne [1960] 2 Q.B. 396, 403) but will nevertheless have mental characteristics (temporary or permanent) which the jury might think should be taken into account for the purposes of the provocation defence. The boundary between the normal and abnormal is very often a matter of opinion. Some people are entirely normal in most respects and behave unusually in others. There are people (such as battered wives) who would reject any suggestion that they were "different from ordinary human beings" but have undergone experiences which, without any fault or defect of character on their part, have affected their powers of self-control. In such cases the law now recognises that the emotions which may cause loss of self-control are not confined to anger but may include fear and despair. Professor Ashworth, who argued in 1976 that diminished responsibility and provocation were logically mutually exclusive, was cautious enough to say ("The Doctrine of Provocation"[1976] C.L.J. 292, 314) that it was "difficult to shed all one's misgivings about whether the law actually operates in this way." I think not only that this scepticism was justified but also that section 3 prevents the judges from trying to force cases into logical dichotomies.

    There is however one really serious argument in favour of the distinction between characteristics affecting the gravity of the provocation and characteristics affecting the power of self-control. This is the claim that, despite all its difficulties of application, it is the only way to hold the line against complete erosion of the objective element in provocation. The purpose of the objective element in provocation is to mark the distinction between (partially) excusable and inexcusable loss of self-control. As Lord Diplock said in D.P.P. v. Camplin [1978] A.C. 705, 717, the conduct of the accused should be measured against "such powers of self-control as everyone is entitled to expect that his fellow citizens will exercise in society as it is today." If there is no limit to the characteristics which can be taken into account, the fact that the accused lost self-control will show that he is a person liable in such circumstances to lose his self-control. The objective element will have disappeared completely.

    My Lords, I share the concern that this should not happen. For the protection of the public, the law should continue to insist that people must exercise self-control. A person who flies into a murderous rage when he is crossed, thwarted or disappointed in the vicissitudes of life should not be able to rely upon his anti-social propensity as even a partial excuse for killing. In Stingel v. The Queen (1990) 171 C.L.R. 312, for example, the accused was obsessively infatuated with a woman who had terminated their relationship. He became a stalker, following her about. She obtained a court order restraining him from approaching her. One evening after a party he found the woman in a car with another man. According to his own account, they were having sex. He went back to his own car, fetched a butcher's knife and came back and killed the man. His evidence conformed to the standard narrative which the legal requirement of "loss of control" imposes on such defences:

    "I was all worked up and feeling funny. It was like I was in a rage, almost to the stage where I felt dazed. It was like I really didn't know what happened until the knife went into him."

The High Court of Australia held that the judge was right to withdraw the issue of provocation from the jury on the ground that such conduct could not raise even a reasonable doubt as to whether the objective element in the defence had been satisfied. I respectfully agree. Male possessiveness and jealousy should not today be an acceptable reason for loss of self-control leading to homicide, whether inflicted upon the woman herself or her new lover. In Australia the judge was able to give effect to this policy by withdrawing issue from the jury. But section 3 prevents an English judge from doing so. So, it is suggested, a direction that characteristics such as jealousy and obsession should be ignored in relation to the objective element is the best way to ensure that people like Stingel cannot rely upon the defence.

9. The English cases.

    The first important English case after D.P.P. v. Camplin [1978] A.C. 705 was the judgment of Lord Lane C.J. in Reg. v. Newell (1980) 71 Cr.App.R.331. He interpreted section 3 as meaning that the jury can be directed to take into account personal characteristics of the accused in relation to both the gravity of the provocation and the degree of self-control which could reasonably have been expected. It is true, as Lord Goff of Chieveley pointed out in Luc Thiet Thuan v. The Queen [1997] A.C. 131, 141-144, the Lord Chief Justice adopted the construction which had been given to a somewhat different statute in New Zealand. He approved a passage in Reg. v. McGregor [1962] N.Z.L.R. 1069 in which North J. had said:

    "The offender must be presumed to possess in general the power of self-control of the ordinary man, save in so far as his power of self-control is weakened because of some particular characteristic possessed by him."

    But the course of the law in New Zealand has been a rather tangled story, as the judgments in Reg. v. Rongonui (Court of Appeal, 13 April 2000) reveal. I have already said enough to explain why I think that the construction of section 3 adopted by the Court of Appeal was in this respect correct, independently of any support which might be obtained from New Zealand. It is therefore inappropriate for me to undertake any analysis of the New Zealand cases or comment upon the construction which the courts have given to their statute. Nor can any direct assistance be obtained from Australia and Canada, where the objective standard remains a matter of law for the judge.

    The construction adopted in Newell 71 Cr.App.R. 331 was followed by Lord Taylor of Gosforth C.J. in Reg. v. Ahluwalia [1992] 4 All E.R. 889 , a case of a battered wife. He said that characteristics relating to the "mental state or personality of an individual" such as the fact that a battered wife was suffering from post-traumatic stress disorder, could be taken into account. It is true that he recorded counsel for the appellant as having described this as a characteristic which the jury "might think might affect the gravity of the provocation". The same comment may be made about Lord Taylor's later judgment in Reg. v. Dryden [1995] 4 All E.R. 987. In that case the accused was convicted of murder after he had shot and killed a planning officer who was engaged in demolishing his bungalow pursuant to an enforcement notice. There was psychiatric evidence that the accused had developed an obsession about his planning problems. The Court of Appeal said that the obsessiveness and eccentricity of the defendant should have been left to the jury as "mental characteristics" which they should take into account. In neither case, however, did Lord Taylor suggest that the jury should have been directed to have regard to these characteristics only insofar as they might have affected the gravity of the provocation and not insofar as they may have affected the accused's power of self-control. No doubt this omission was for the very good reason that, on the facts of both cases, no jury would have understood what such a distinction meant.

    Finally, in Reg. v. Campbell [1997] 1 C.App.R. 199 Lord Bingham of Cornhill C.J. affirmed the principle of the earlier decisions, which he said represented "a judicial response, born of everyday experience in criminal trials up and down the country, as to what fairness seems to require."

    My Lords, in the face of these views of three successive Lord Chief Justices, I would be most reluctant to advise your Lordships to turn back such a strong current of authority unless it was clearly inconsistent with the statute. But I do not think it is. On the contrary, it seems to me to reflect a realistic appreciation of what the statute has done.

10. Guiding the jury.

    My Lords, I think that some of the concern about the recent trend of authority in the English Court of Appeal has been due to the assumption that unless the judge can direct the jury that certain characteristics of the accused are legally irrelevant to the objective element in the defence, the jury may receive the impression that the law actually requires them to take such matters into account. The effect would be to encourage juries to find provocation on inappropriate grounds. Obviously, my Lords, there is always the risk that a jury may do so. That is the risk which Parliament took when it gave the jury an unfettered right to give effect to its own opinion on the objective element. But it considered that risk less likely to cause injustice than to confine the jury within the rules of law which had been developed about the notional characteristics of the reasonable man. In any case, I think that much can be done to reduce that risk if judges guide juries on this issue in a way which fully takes into account the difference which section 3 has made to their respective roles.

    Before 1957 the judge had to direct the jury as to whether, if they found that some act had caused the accused to lose his self-control, that act was "capable" of amounting to provocation. It would be so capable if the judge considered that a rational jury could find that it satisfied the objective element. If he did not, he would withdraw the issue by telling the jury that there was no evidence upon which they could properly find that the accused had acted under provocation. If, therefore, the judge left the issue to the jury, he would do so in terms which conveyed to them that they could rationally find that the objective element was satisfied.

    The effect of section 3 is that once the judge has ruled that there is evidence upon which the jury can find that something caused the accused to lose self-control (compare Reg. v. Acott [1997] 1 W.L.R. 306), he cannot tell the jury that the act in question was incapable of amounting to provocation. But that no longer involves any decision by the judge that it would be rational so to decide. For example, in Reg. v. Doughty [1986] Cr.App.R. 319 the Court of Appeal held that the judge had been wrong to direct the jury that the crying of 17 day old baby, which had caused its father to kill it by covering its head with cushions and kneeling on them, could not constitute a provocative act. Section 3 said that the jury were entitled to take into account "everything both done and said." I respectfully think that this construction of the Act was correct. But that does not mean that the judge should tell the jury that the crying of the baby was, in the traditional language, capable of amounting to provocation. This would give the jury the impression that the judge thought it would be rational and in accordance with principle to hold that the crying of the baby constituted an acceptable partial excuse for killing it. The point about section 3 is that it no longer matters whether the judge thinks so or not. He should therefore be able simply to tell the jury that the question of whether such behaviour fell below the standard which should reasonably have been expected of the accused was entirely a matter for them. He should not be obliged to let the jury imagine that the law now regards anything whatever which caused loss of self-control (whether an external event or a personal characteristic of the accused) as necessarily being an acceptable reason for loss of self-control.

    11. The reasonable man.

    The main obstacle to directing the jury in a way which does not give such a false impression is the highly artificial way in which courts and writers have attempted to marry two discordant ideas: first, the old formula that the provocation must have been such as to cause a "reasonable man" to act in the same way as the accused and, secondly, the rule in section 3 that no circumstances or characteristics should be excluded from the consideration of the jury. They have done so by telling the jury that certain characteristics are to be "attributed" to the reasonable man. By such a combination, they have produced monsters like the reasonable obsessive, the reasonable depressive alcoholic and even (with all respect to the explanations of Lord Goff of Chieveley in Reg. v. Morhall [1996] 1 A.C. 90, 98) the reasonable glue sniffer. Nor does it elucidate matters to substitute "ordinary" for "reasonable." Quite apart from the question of whether the jury can understand what such concepts mean, it is bound to suggest to them that obsession, alcoholism and so forth are not merely matters which they are entitled in law to take into account but that, being "attributed" to the reasonable man, they are qualities for which allowances must be made.

    So, for example, in Reg. v. Humphreys [1995] 4 All E.R. 1008 there was a good deal of discussion as to whether "attention seeking" and "immaturity" were "eligible characteristics" in the sense that they were to be attributed to (in that case) the reasonable woman. The Court of Appeal decided that they were. Similarly in Reg. v. Dryden [1995] 4 All E.R. 987, which I have already mentioned, the question was framed as being whether the obsessiveness and eccentricity of the defendant were "mental characteristics" which the jury should attribute to the reasonable man. Professor M. J. Allen, in the article to which I have referred in [2000] Journal of Criminal Law 216, 239, says with some force that this decision, "endorsing obsession as a characteristic to attribute to the reasonable man should sound an alarm bell for all sexual partners." If Dryden's obsession could be attributed to "the reasonable man," why not Stingel's?

    My Lords, the concept of the "reasonable man" has never been more than a way of explaining the law to a jury; an anthropomorphic image to convey to them, with a suitable degree of vividness, the legal principle that even under provocation, people must conform to an objective standard of behaviour which society is entitled to expect: see Lord Diplock in Camplin [1978] A.C. 705, 714. In referring to "the reasonable man" section 3 invokes that standard. But I do not think that it was intended to require judges always to use that particular image, even in cases in which its use is more likely to confuse than illuminate. When Keating J. in Reg. v. Welsh (1869) 11 Cox C.C. 336, 339 borrowed the mot juste which Baron Alderson had used in Blyth v. Birmingham Waterworks (1856) 11 Exch. 781, 784 to define negligence, he did not imagine that he was changing the law. He merely thought he had hit upon a felicitous way of explaining it. Whether he was right is perhaps questionable. Even before the Act of 1957, there had been expressions of doubt about the extent to which it really was a helpful way to explain the notion of objectivity in the particular context of provocation. The jury may have some difficulty with the notion that the "reasonable man" will, even under severe provocation, kill someone else. But, my Lords, whatever the force of the earlier criticisms, the value of the image has been hopelessly compromised by the Act of 1957. This may not have been foreseen, just as many did not foresee the effect which the Act would have upon the concept of the reasonable man and the abandonment in Camplin [1978] A.C. 705 of the law laid down in Mancini [1941] A.C.1 and Bedder [1954] 1 W.L.R. 1119. But it seems to me now, since Camplin, impossible to avoid giving the jury a misleading, not to say unintelligible, account of the law when particular characteristics, sometimes highly unusual and even repulsive, are welded onto the concept of the reasonable man. I do not find it surprising that nine judges who gave written evidence to the House of Lords Select Committee on Murder and Life Imprisonment (HL Paper 78-III Session 1988-89) said that the reasonable man test was "logically unworkable, or [rendered] the defence almost ineffective if it were strictly applied by juries."

    My Lords, I do emphasise that what has been rendered unworkable is not the principle of objectivity which (subject to the changes noted in Camplin) section 3 was plainly intended to preserve, but a particular way of explaining it. I am not suggesting that your Lordships should in any way depart from the legal principle embodied in section 3 but only that the principle should be expounded in clear language rather than by the use of an opaque formula.

    In my opinion, therefore, judges should not be required to describe the objective element in the provocation defence by reference to a reasonable man, with or without attribution of personal characteristics. They may instead find it more helpful to explain in simple language the principles of the doctrine of provocation. First, it requires that the accused should have killed while he had lost self-control and that something should have caused him to lose self-control. For better or for worse, section 3 left this part of the law untouched. Secondly, the fact that something caused him to lose self-control is not enough. The law expects people to exercise control over their emotions. A tendency to violent rages or childish tantrums is a defect in character rather than an excuse. The jury must think that the circumstances were such as to make the loss of self-control sufficiently excusable to reduce the gravity of the offence from murder to manslaughter. This is entirely a question for the jury. In deciding what should count as a sufficient excuse, they have to apply what they consider to be appropriate standards of behaviour; on the one hand making allowance for human nature and the power of the emotions but, on the other hand, not allowing someone to rely upon his own violent disposition. In applying these standards of behaviour, the jury represent the community and decide, as Lord Diplock said in Camplin ([1978] A.C. 717), what degree of self-control "everyone is entitled to expect that his fellow citizens will exercise in society as it is today." The maintenance of such standards is important. As Viscount Simon L.C. said more than 50 years ago in Holmes v. D.P.P. [1946] A.C. 588, 601, "as society advances, it ought to call for a higher measure of self-control."

    The general principle is that the same standards of behaviour are expected of everyone, regardless of their individual psychological make-up. In most cases, nothing more will need to be said. But the jury should in an appropriate case be told, in whatever language will best convey the distinction, that this is a principle and not a rigid rule. It may sometimes have to yield to a more important principle, which is to do justice in the particular case. So the jury may think that there was some characteristic of the accused, whether temporary or permanent, which affected the degree of control which society could reasonably have expected of him and which it would be unjust not to take into account. If the jury take this view, they are at liberty to give effect to it.

    My Lords, I do not wish to lay down any prescriptive formula for the way in which the matter is explained to the jury. I am sure that if judges are freed from the necessity of invoking the formula of the reasonable man equipped with an array of unreasonable "eligible characteristics," they will be able to explain the principles in simple terms. Provided that the judge makes it clear that the question is in the end one for the jury and that he is not seeking to "impose a fetter on the right and duty of the jury which the Act accords to them," the guidance which he gives must be a matter for his judgment on the facts of the case.

12. The burden of proof.

    The burden is upon the prosecution to disprove provocation. This means that the prosecution must satisfy the jury that a version of the facts in which the accused was provoked could not reasonably be true. But the decision as to whether, having regard to the objective principle, those facts should count as sufficient provocation to reduce the offence to manslaughter has nothing to do with the burden of proof. The jury either think it does or they do not. It is irrelevant that they may think that a different jury could have taken a different view.

    13. Conclusion.

    In my opinion the judge should not have directed the jury as a matter of law that the effect of Smith's depression on his powers of self-control was "neither here nor there." They should have been told that whether they took it into account in relation to the question of whether the behaviour of the accused had measured up to the standard of self-control which ought reasonably to have been expected of him was a matter for them to decide. For the above reasons and those given by my noble and learned friends Lord Slynn of Hadley and Lord Clyde, I would dismiss the appeal.


My Lords,

On 16 November 1996 the respondent killed James McCullagh in the course of a heated argument. There was evidence that the respondent was suffering from a depressive illness and that that illness rendered him more disinhibited than would otherwise have been the case. His defence included submissions on provocation and diminished responsibility but it is only the matter of provocation which is raised in this appeal. More particularly the question in this appeal concerns the application of section 3 of the Homicide Act 1957 to the factual circumstances of the case.

    Section 3 provides:

    "Where on a charge of murder there is evidence on which the jury can find that the person charged was provoked (whether by things done or by things said or by both together) to lose his self-control, the question whether the provocation was enough to make a reasonable man do as he did shall be left to be determined by the jury; and in determining that question the jury shall take into account everything both done and said according to the effect which, in their opinion, it would have on a reasonable man."

    For present purposes it is to be noted that the section did not re-state the law of provocation, but subject to what it contained left the matter to rest upon the common law. One important change was the express recognition that provocation could be constituted by words alone, or by words along with acts. In addition to that it regulated the respective functions of the judge and the jury. The earlier part of the section describes the responsibility of the judge. It is for the judge to determine if there is any evidence upon which the jury can find that the accused was provoked to lose his self-control. The evidence may be slender but if it is sufficient for the jury to make the finding then the latter part of the section comes into play. The latter part of the section requires that it is for the jury, not the judge, to decide whether the provocation was enough to make a reasonable man do as the accused did. The judge is not entitled to withdraw the matter of provocation from the jury on account of any view which the judge may take on the reasonableness of the accused's behaviour. If there is evidence of the fact that the accused was provoked to lose his self control the reasonableness of the accused's action is matter for the jury.

continue previous