Kelleher v The Queen

[1974] HCA 48

High Court of Australia|1974-07-01|Before: Jacobs JJ

View original source

At a glance

Source facts

Court

High Court of Australia

Decision date

1974-07-01

Before

Jacobs JJ

Source

Original judgment source is linked above.

Judgment (121 paragraphs)

[1]

The applicant was represented at the trial by counsel of experience. In his final address to the jury, which was a long and forceful address, counsel dealt at length with the law's emphasis on the necessity for a jury to exercise great caution on any issue involving identification. He referred to mistakes that had been made in past criminal trials when an innocent man had been convicted as a result of a mistake in his identification. He also quoted, but without reference to their origin, classic legal statements of the principles relating to identification. Counsel for the Crown, in his address, also referred at length to the question of identification and warned the jury of the care with which they must approach the issue. In his summing up, the learned trial judge made it quite clear that it was essential to a verdict of guilty that the jury must be satisfied beyond reasonable doubt that the intercourse with the prosecutrix had been had by the applicant without her consent. He said: "The big question is, has it been proved that it was the accused person who had the intercourse with her". Later, his Honour said: "The real issue posed to you by both the Crown and Mr. Roden" (counsel for the applicant at his trial) "is as to whether it has been proved beyond reasonable doubt that the accused man did these things", that is to say, raped Leslie Watson and stole her ring and the television set. His Honour proceeded: "Gentlemen, it is said that jurors should be warned when approaching this question of identification of people, they should be told that there are dangers in it. I echo those words of warning that the law says should be given to a jury but having mentioned that let me say this to you, which is also the law, that you have regard to the circumstances. You should be told that you must approach evidence of identification and this issue of identification with care, but that warning quite obviously depends on the circumstances", and his Honour then proceeded to give illustrations of circumstances where perhaps there was less need for great warnings to be careful. I take his Honour's reference to what the jury should be told, and which he said he echoed, to be a reference to the lengthy and emphatic statement which the counsel for the applicant had made to the jury. I may say at once that it is not a satisfactory method of summing up for a judge simply to endorse what a counsel has said: he should himself express so significant a matter as a warning against a too ready acceptance of evidence by a prosecutrix of the identification of her sexual assailant. But when a judge does take the course taken by his Honour, undesirable and unsatisfactory as I think it is, he does give the weight of his authority to the statement which counsel has made, so that it cannot be said that the judge has not given those warnings or cautions which, having been given by counsel, he has confirmed by such a statement as I have quoted. It seems to me that the trial judge felt that the emphasis made by counsel on the need for caution in this case had been perhaps undue, and that in order to restore balance to the matter he gave the illustrations to which I have made reference.

[27]

The only other point which was raised before the Court of Criminal Appeal was that the learned judge did not warn the jury that it would be dangerous to convict the accused of having raped the complainant on her evidence alone that the accused was a man who raped her. It seems from the judgment of the Court of Criminal Appeal that they thought, as regards the charge of rape, it would have been appropriate to give such a warning. But they considered that on the whole it would be right to apply the proviso in s. 6 (1). For this reason they dismissed the appeal. The complainant's evidence proved that she was raped by the accused and she was then similarly outraged by five others. No doubt they were some of the men to whom the learned judge referred in the passage quoted above. I take the view that the complainant's evidence, that the accused was one of her assailants in the bedroom to which she was dragged, was corroborated. Hers was not the only evidence that proved that she was dragged into the bedroom immediately preceding the rape of herself to which she testified. Evidence was given by a Mr. Harper, who was not one of the gang who intruded upon the two girls in the flat. He gave evidence that the accused, assisted by two of the gang, dragged her off a lounge on which she was sitting in another room of the flat to the bedroom in which the crime occurred. In my opinion the jury could reasonably take the view that Mr. Harper's evidence strengthened the complainant's evidence that the accused was one of those who raped her. In my opinion it is not a flaw invalidating the trial, that the learned judge did not warn the jury that it was dangerous to convict of the crime of rape on the uncorroborated evidence of the complainant, that the accused and the man who raped her are one and the same man. As I have said, in my opinion, there was corroboration and it would follow that the question of applying the proviso does not really arise. I do not think it fit in this case to give special leave to appeal, and I would therefore dismiss the motion.

[58]

It is manifest that in cases where the vital issue is whether the identification of the accused person is certain and reliable the judge must direct the jury with great care. However careful is his general direction as to the onus of proof, the judge will feel it necessary to deal specifically with all the matters relating to identification. In that case, which was one in which the witness who had identified the accused man knew him well, the House of Lords held that a summing up which dealt fairly and amply with the evidence of identification could not be held as a matter of law to be defective if it did not contain a general warning to the jury of the dangers of acting on evidence of visual identification, and their Lordships left it open whether it was necessary to lay down any rule for the guidance of courts in a case in which a witness had identified someone he did not previously know and whom he had had only a limited opportunity to observe. In Reg. v. Long [17] , the Court of Appeal (Criminal Division) held that in such a case a judge is not bound, as a matter of law, to give the jury a specific warning about the danger of convicting on such evidence. However, it seems to me that although it is perfectly true that the adequacy of a summing up can only be decided in the light of the circumstances of the particular case, and that where a warning is necessary no particular form of words is required, it is in practice generally desirable that where the case for the prosecution includes evidence of visual identification by a person previously unfamiliar with the accused, an appropriate warning should be given to the jury, since jurors may not appreciate as fully as a judge may do, or even at all, the serious risk that always exists that evidence of that kind may be mistaken. The failure to give an adequate warning where one is required may have the result that the conviction must be quashed - a course that has been taken in a number of recent cases in Australia: See Reg. v. Gaunt [18] ; Reg. v. Preston [19] ; Reg. v. Boardman [20] ; Reg. v. Maarroui [21] ; Reg. v. Goode [22] ; Reg. v. Harris [23] . If a warning is necessary, the duty to give it will not be satisfactorily discharged by the perfunctory or half-hearted repetition of a formula, and a warning in general terms will not alone be sufficient; the jury should be given careful guidance as to the circumstances of the particular case, and their attention should be drawn to any weaknesses in the identification evidence. 1. (1970) 55 Cr. App. R. 161, at p. 168. 2. (1973) 57 Cr. App. R. 871. 3. [1964] N.S.W.R. 864. 4. [1961] V.R. 761. 5. [1969] V.R. 151. 6. (1970) 92 W.N. (N.S.W.) 757. 7. [1970] S.A.S.R. 69. 8. (1971) 1 S.A.S.R. 447.

[59]

In the present case the learned trial judge not only told the jury again and again that the real question for them to decide was whether the applicant was the person who had committed the offences, but he also gave an express warning of the necessity to approach identification evidence with care. He referred to and reiterated the matters upon which counsel for the applicant had relied in criticism of the Crown case, including the facts that the applicant had previously been unknown to Miss Watson, that she identified him after an interval of time and only after he had been pointed out to her by Lee, and that she appeared to have been mistaken in some details of her evidence, particularly in saying that the applicant was the man who had attempted to drag her away on the first occasion but had been restrained from doing so. Counsel for the applicant submitted criticisms of the manner in which the trial judge warned the jury and of his treatment of the evidence, but on an application for special leave to appeal it is unnecessary to deal in detail with contentions of this kind if the conclusion is reached, as it should be in the present case, that regarded as a whole the summing up did give the jury the requisite guidance on the question of identification. One particular submission that may be mentioned, however, was that the judge should have reminded the jury that no identification parade was held, and informed them of the special possibility of error that exists when a witness whose previous knowledge of the accused man has not made her familiar with his appearance has been shown the accused alone as a suspect and has on that occasion first identified him: see Davies and Cody v. The King [24] . The present, however, was not a case in which an accused man was first identified when he was pointed out as a person whom the police suspected of having committed the crimes in question. When the applicant was seen by Miss Watson outside the court he was standing in a group of people who were not there in connexion with the offences now in question and Lee, in pointing him out, merely referred to him as one of the persons who had been at the flat - as of course he had been. Miss Watson then recognized him as one of the men who had raped her. In these circumstances it would have been a completely meaningless formality to arrange for an identification parade and the additional possibility of error that arises when an identification is prompted by a suggestion that the person identified was the offender did not exist in the present case. The contention that the summing up so defectively dealt with the question of identification that the verdict ought not to be allowed to stand was rightly rejected by the Court of Criminal Appeal.

[61]

It is now established that in cases of rape and other sexual offences in which corroboration is not required as a matter of law the trial judge should, as a matter of practice, warn the jury that it is dangerous to convict on the uncorroborated testimony of the person (whether male or female) on whom the offence was committed, although the members of the jury may act upon that testimony if, after scrutinizing it with great care, and paying heed to the warning, they are satisfied of its truth and accuracy. I have set out the general nature of the warning required, but of course there is no set form of words that must be used. This rule of practice is of comparatively recent origin; its first recognition in a reported decision appears to have been in R. v. Graham [25] . However, it was quickly accepted as peremptory, and by 1919 Isaacs J. was able to say that "the practice is so well established as to have, as was said of the analogous case of accomplices, "almost the reverence of a rule of law" ": Hargan v. The King [26] . The practice appears originally to have been formulated under the influence of the similar rules relating to the corroboration of the evidence of young children and of accomplices, but its real reason and justification would appear to be that which was suggested by Salmon L.J. (as he then was) in Reg. v. Henry; Reg. v. Manning [27] - that it is dangerous to convict on the evidence of the complainant in a sexual case because experience shows that in such cases people do sometimes tell "an entirely false story which is very easy to fabricate, but extremely difficult to refute". If this is the real reason for the practice, it would not, in the absence of authority, be immediately obvious that it is necessary to direct the jury in this way where the only issue at the trial is as to the identity of the offender. Nevertheless, in the case of accomplices, it is very well settled that "evidence in corroboration must be independent testimony which affects the accused by connecting or tending to connect him with the crime", i.e., evidence "not merely that the crime has been committed, but that it was committed by the accused": R. v. Baskerville [28] . The question whether in sexual cases corroboration must be in a material particular implicating the accused was at first regarded as doubtful: R. v. Myro Smith [29] ; Hicks v. The King [30] . Since that time, however, it has been consistently held in the Court of Criminal Appeal in England that a warning of the danger of acting on the uncorroborated evidence of the complainant must be given in cases where identity is the sole issue: Reg. v. Sawyer [31] ; Reg. v. Clynes [32] ; Reg. v. Trigg [33] ; Reg. v. O'Reilly [34] ; Reg. v. Midwinter [35] . These decisions are supported by the authority of the Judicial Committee; in James v. The Queen [36] , Viscount Dilhorne, delivering the judgment of the Judicial Committee, said [37] : "In sexual cases, in view of the possibility of error in identification by the complainant, corroborative evidence confirming in a material particular her evidence that the accused was the guilty man is just as important as such evidence confirming that intercourse took place without her consent." It must now be taken as settled that it is the duty of a trial judge to give an appropriate direction on the subject of corroboration on a charge of rape, although there is no dispute that the woman was raped, and the only issue is whether the accused committed the crime. This direction will still be necessary even if a proper warning has been given to the jury as to the danger of convicting upon evidence of visual identification, because a warning of the latter kind will not be enough to inform a jury of the danger of convicting on the uncorroborated evidence of the complainant in a sexual case. Further, in my opinion, the fact that it was not suggested at the trial that Miss Watson was untruthful, but that she was mistaken, did not render the direction unnecessary. Even if the purpose of the direction as to corroboration is to put the jury on guard against testimony fabricated by a complainant, the mere fact that it is not expressly suggested that a complainant is lying does not render this rule of practice inapplicable; an accused person may say in effect, "The complainant's story is untrue, and since I do not know whether the untruth results from mistake or from malice I will not suggest the latter".

[63]

The question, however, arises whether the warning must be given when there is evidence capable of corroborating the complainant. In Hicks v. The King [38] , this Court held that no warning was necessary in a case in which the evidence of the complainant was corroborated, but the question was not fully discussed; the Court mainly directed its attention to the question whether the complainant's evidence was corroborated, and once that question was answered in the affirmative it was assumed that a warning was unnecessary [39] . Hicks v. The King was decided when the rule of practice now under discussion had only recently become recognized, and although the decision in that case may well have been in accord with opinions held in England at that time, it is opposed to later authorities. InDavies v. Director of Public Prosecutions [40] , it was held by the House of Lords that a warning must be given to the jury that it is dangerous to convict upon the evidence of an accomplice unless it is corroborated, even if in fact there is ample corroboration of the evidence of the accomplice, and that if the judge fails to give such a warning the conviction will be quashed unless the appellate court can apply the proviso to s. 4(1) of the Criminal Appeal Act, 1907 U.K.. The Court of Criminal Appeal in England has applied the same rule to sexual cases, and has held that a failure to give a warning will be fatal to a conviction (except in those rare cases to which the proviso can be applied) even if there is evidence which the jury could accept as corroborative: Reg. v. Clynes [41] ; Reg. v. Trigg [42] ; Reg. v. O'Reilly [43] . In Australia and in New Zealand, it has often been held, in reliance on the English authorities, that it is necessary to give an appropriate warning as to corroboration in sexual cases where there is evidence capable of amounting to corroboration, although different opinions have been expressed as to how far the judge should go in explaining the position to the jury: see, for example, R. v. Baynon and Pitama [44] ; Reg. v. Adams and Ross [45] ; Reg. v. Jansen [46] . In my opinion the mere presence of evidence which the jury might accept and treat as corroborative does not make a proper warning any the less necessary. The jury may regard the independent evidence as untruthful or inaccurate, or of little cogency; if the members of the jury have to decide whether the independent evidence gives rise to an inference that the accused committed the offence - as, for instance, when the accused has made false statements - they may decline to draw the necessary inference; for a judge to decide that the evidence capable of corroborating the complainant does amount to corroboration, and that therefore the jury need no guidance as to the manner in which uncorroborated testimony should be approached, would be to usurp the jury's function. I can see no reason in principle why the fact that there is ample evidence capable of amounting to corroboration should render it unnecessary for a judge trying a sexual offence to warn the jury of the danger of acting upon the complainant's evidence unless it is corroborated, and I can see no reason why the propositions laid down in Davies v. Director of Public Prosecutions [47] should be regarded as inapplicable to the corroboration of the complainant in sexual cases. In my opinion it was necessary for the learned trial judge to give a warning in the present case, notwithstanding that there was evidence capable of amounting to corroboration, and notwithstanding that the question was not raised on behalf of the defence. The Court of Criminal Appeal regarded the judge's direction as defective, and they were right in doing so.

[67]

The circumstances of the present case do not appear to me to make it a proper case in which to grant special leave to appeal. During the course of the evening Miss Watson had many opportunities to observe the applicant, and her identification of him, when she next saw him, was prompt and positive. The applicant admits, not only that he was at the scene of the crimes, but that he played an active part in some of the lawless activities of the gang at about the time the offences occurred. It would have been open to the jury, had their attention been directed to the matter, to regard the applicant's statement to the police, the making of which was not challenged, as strongly confirmatory of the complainant's evidence that he had raped her. At first he prevaricated, suggesting that he had left the flat early and had taken no part in what had later occurred; then, although admitting that Miss Tomkins had been repeatedly raped, he said that he saw Miss Watson having sexual intercourse but, as far as he knew, voluntarily. The jury may well have regarded it as quite incredible that the applicant could have believed that Miss Watson, whose flat had been invaded by a mob of young ruffians, who had tried to escape but had been prevented from doing so, whose male friends had been assaulted and whose female companion had been raped, and whose flat was ransacked, would thereafter have voluntarily engaged in sexual intercourse with one of the mob who had invaded her flat. Further, it was not contested that Miss Watson had been dragged from the lounge-room and raped by a number of men, and the jury may well have thought that it was not possible that the applicant could have been unaware of what had occurred. It was open to the jury to conclude that the applicant denied any knowledge of the rape of Miss Watson because of his complicity in it. Apart from the evidence of Harper, there was therefore evidence of a particularly compelling character to support the evidence of Miss Watson identifying the applicant as one of the men who had raped her. The whole case, including the criticisms made of the evidence of identification, was presented fairly to the jury in the summing up, the only deficiency being the omission to give the requisite warning as to corroboration, and that deficiency was not sought to be rectified at the trial. In all these circumstances I do not consider that this case is of such a special character as to justify this Court in granting special leave to appeal.

[114]

This rule has since been applied in England in the case of accomplices and of sexual offences. Moreover, there has been a growing reluctance to apply the proviso to which I have earlier referred. See Reg. v. Trigg [12] . In a way it is illogical that the proviso can ever be applied. If the rule were actually a rule of law then failure to warn as to corroboration would be an error of law and the conviction would perforce be quashed unless the proviso (in New South Wales to s. 6) could properly be applied. But as the rule is not a rule of law but is one of practice so strong that failure to observe it is a miscarriage of justice even where there is some, perhaps ample, corroboration then it seems strange at first sight that it should be possible to proceed from there and say on an appeal that despite the miscarriage of justice there has been no substantial miscarriage of justice. The miscarriage of justice arising from the failure to warn arises from the fact that if the warning had been given the jury might have concluded that the evidence of the accomplice or the victim of sexual assault stood uncorroborated, that is to say, might have found the whole of the corroborative evidence unacceptable. How then in any case could it be said that, with that possibility, no substantial miscarriage of justice had occurred when without the requisite warning the jury might in fact have convicted without the evidence of the accomplice or the sexual victim being corroborated? The answer must lie in pragmatic rather than logical considerations. The rule requiring the direction had developed after the passing of the Criminal Appeal Act to a degree at which the appeal courts were insistent that the warning should be given in every case, whether there was present in the case corroborative evidence or not. The rule could not historically be stated to be one failure to observe which was an error of law and therefore it had to be subsumed under the provision of the Act conferring the new power to quash in case of miscarriage of justice. If the existence of corroborative evidence which may have been believed could lead to the conclusion that there was no miscarriage of justice and therefore no ground for quashing the conviction the purpose of ensuring that the warning was given could not be achieved. A trial judge in each case would need to make up his own mind whether the warning was or was not appropriate, depending no doubt on his assessment of the strength of the corroboration. Therefore the definitive rule was enunciated applicable in all cases. But that left the case where the corroborative evidence was so strong that a different result on a new trial or indeed on the first trial if the jury had been properly directed could not reasonably be regarded as a probability or a possibility. Clearly the Court of Appeal in England desired to retain the right to apply the proviso in such a case and has done so. See, e.g., Reg. v. Henry [13] where the corroboration was described as "formidable". The English developments would appear to achieve the practical result sought, namely, the insistence that in every case appropriate warnings should be given by the trial judge with power in the appeal court in an appropriate case to allow the conviction to stand despite an inadvertent omission or an inadequate direction at the trial.

[118]

Leaving aside the case of the evidence on oath or otherwise of children of tender years I cannot see that in sexual cases the approach of the Court in Hicks v. The King read in the light of Hargan v. The King [21] is no longer appropriate. I can appreciate that the strengthening of the rules relating to warning on the dangers of convicting on the uncorroborated evidence of an accomplice was a salutary development, but I do not see that the same rules, stricter than those previously thought to apply, should extend to corroboration in sexual cases. Where there is a development of the law in one field it does not necessarily carry over into a similar field. There is not a single branch of the law on the need for corroboration of particular evidence or kinds of evidence. There are different branches which until recent developments could be regarded as having similar characteristics. But when one branch develops, as without so deciding I would think it has in respect of warnings on accomplices, I would need to be satisfied that the same factors which led to that development ought to be carried over into the similar but not identical subject-matter of warnings on corroboration in the case of sexual offences. I will concede that the rule requiring warning in the case of sexual offences remains well based in modern times despite the changes in social conditions and in the status and education of women over the centuries. But I must confess that I am puzzled why there is a greater need for corroboration of identification of an alleged rapist or indecent assaulter than of a murder attempter. Be that as it may, I am not prepared to adopt a stricter rule than was held to exist when Hicks v. The King [20] was decided. I do not understand that case as deciding that a trial judge should not in every case as a matter of practice give an appropriate warning, but if he does not and if there is substantial corroborative evidence the Court of Criminal Appeal should examine the strength of that evidence and must be satisfied that there has been a miscarriage of justice before quashing the conviction. However it need not be satisfied that the jury with the proper direction would "inevitably" have reached the same result, as it would need to be if the proviso were to be applied.