BRS v The Queen

It was announced that the appeal was allowed and that reasons for judgment would be published later.

BRS was charged on six counts of homosexual intercourse with H, BRS being a teacher and H being his pupil under the age of eighteen years. After a trial before the District Court of New South Wales, BRS was convicted on five of those counts and acquitted on one count. The Court of Criminal Appeal dismissed his appeal against conviction. This appeal, brought by special leave, raises for consideration the admissibility of the evidence given by one of H's fellow pupils, W, and, in the light of that evidence, the sufficiency of the directions given by the trial judge to the jury.

BRS was one of the masters at a boys' boarding school. His headmaster regarded him as "one of the most outstanding primary school teachers that I have known". He had a fine reputation as a teacher and a carer for the boys in his charge. He had duties as a tutor in one of the school houses which required him to supervise a dormitory and look after the young boys who were accommodated there. H was one of the boys in his house.

H had come to the school in 1985 at the age of eleven after his parents had divorced. H was living with his mother on her country property. He was a big boy and an aggressive boy, looking for attention. He got into fights at school. He says he "wasn't really mates with anyone". He encountered some academic difficulties. In 1986, BRS came to the school and was helpful to H: "He would talk to you and give you support and give you a hand." H used to visit him in his room and get help with his school work. In the May holiday of that year, BRS was invited to spend some time at H's mother's country property. On an occasion after this holiday, according to allegations subsequently made by H, the first significant act of physical intimacy occurred. BRS cuddled him on the bed in BRS's room. At that time, H must have been approaching his thirteenth birthday.

As time passed, H alleged that a sexual relationship developed between them. First, BRS masturbated in H's presence, later he and H were on the bed and he masturbated H, then the relationship progressed to mutual fellatio and, after BRS had taken H to a concert and a nightclub and had given him a quantity of a liqueur to drink, he anally penetrated H and H anally penetrated him. Thereafter sexual contact between them was frequent until H left the school in 1988. The headmaster and H's parents decided that he would do better at a country school. H alleges that sexual activity in some form occurred twice with BRS after H left the school: once on H's mother's property during a visit by BRS and once in BRS's room at the school. A feature of H's allegations against BRS was that, when masturbating or having anal intercourse, BRS used a lubricant, KY jelly, which he kept in a drawer next to his bed, and he kept an old yellow towel under his bed to wipe up after masturbating.

H has a younger brother who was sent to the same school. He says that, in 1991, when his mother told him that BRS was organising an overseas tour for some students, he told her that BRS had touched him and that the younger brother should be kept away from BRS. The mother evidently spoke to the chaplain at the school who spoke to the headmaster who spoke to BRS. BRS admitted cuddling H but firmly denied any sexual impropriety. No specific allegations were made by H until 1993 when his father, who had come to know something of his allegations against BRS, went to the police.

W had been a friend of H at school. He and two other friends from school had come to H's mother's property for a week's holiday with BRS in January 1987. After H left school, he telephoned these boys or saw them on only one or two occasions. W was called to give evidence that he had visited BRS in 1988 to tell BRS about his, W's, first heterosexual experience. BRS had asked W if he "felt like masturbating with him in his room". BRS told him that he kept a yellow towel under his bed to clean himself up after masturbation. He offered W some KY jelly, which was beside the bed, if he wanted to use it. W masturbated on the bed in the room while BRS had gone behind a wall in the room saying that he felt like masturbating as well. This evidence was led without objection as counsel for the defence accepted that the evidence of W was admissible to rebut the evidence of good character on which BRS intended to rely.

Throughout the trial, the Crown contended that W's evidence was admissible as corroborative of H's evidence. In the course of his final address to the jury, the Crown prosecutor said this:

[I]f you found [W] to be truthful, reliable and accurate and that these events occurred between [W] and the Accused, that doesn't prove on its own that [BRS] did these acts with [H]. You don't sort of do it in reverse, you don't say, Well I'm satisfied that [W is] telling the truth therefore he's guilty of the offences with which he's charged concerning [H].

I submit [W's evidence] does tend to confirm and link [BRS] to the towel, to the "KY" jelly, but perhaps most importantly it meets some argument if you like and provides you with a bit of a picture on the other side of the coin as to the type of person [BRS] was.

[I]f you accept that then obviously [BRS] had a predilection and a liking for the company of young boys and for masturbation in their company.

Counsel for the defence objected to the use of W's evidence for any purpose other than rebutting the good character which BRS put in issue. He told the jury that H's evidence was "entirely uncorroborated". He said that W's allegation was -

not a homosexual allegation at all as he himself admits, but the accusation is evidence before you in reputation [semble repudiation] of the Accused's claim that he is a person of good character. That's all it does; it doesn't corroborate the charge of course, if anything it shows that [BRS] was not homosexual because there was nothing homosexual whatever about what [W] alleges. Counsel for the defence attacked the prosecution case, submitting that there was an absence of corroboration and delay in the making of the complaint, that the circumstances in which H made the complaint to his father raised doubts about the truth of the complaint and that an accused person faced great difficulty in dealing with matters that had allegedly occurred years before the trial.

Unfortunately, the trial judge did not formally rule on the question whether W's evidence was or could be regarded as corroborative of H's evidence until, after the summing up was complete, counsel for the defence sought a direction that there was no corroboration of H's evidence, "whether by W or anybody else" and that the jury could take the absence of corroboration into account in considering their verdict. His Honour commented that he could not see "anything in the evidence which is even capable of being corroboration". He added:

When it comes to the question of things like "KY" and the towels, I can't see they are any more corroborative or capable of being corroboration than the fact that it was a [bed] room. After that observation, the Crown Prosecutor sought to persuade his Honour that W's evidence was corroborative of H's evidence and asked his Honour to "explain what corroboration in the eyes of the law is and then simply identify those matters that I submit are capable of being corroborative of [H's] evidence". His Honour was unconvinced: "No", he said, "I don't propose to do that." And there the matter rested. Counsel for the defence did not press his request for a direction about the absence of corroboration for which he had contended. He had succeeded in defeating the prosecution's foreseeable request for a redirection that W's evidence, if accepted, was capable of corroborating H's evidence.

The evidence of W did not show BRS to have been homosexual but, if accepted, that evidence did show that BRS encouraged one of the boys in his care to masturbate in his presence, on his bed, with KY jelly being offered as a lubricant and with a yellow towel under the bed available for use. There being nothing to suggest concoction between W and H, was W's evidence capable of strengthening H's evidence that BRS encouraged H to masturbate in his presence, on his bed, using KY jelly as a lubricant and with a yellow towel under the bed available for use? Only one answer is possible. Lord Reid said in R v Kilbourne [1] :

There is nothing technical in the idea of corroboration. When in the ordinary affairs of life one is doubtful whether or not to believe a particular statement one naturally looks to see whether it fits in with other statements or circumstances relating to the particular matter; the better it fits in, the more one is inclined to believe it. The doubted statement is corroborated to a greater or lesser extent by the other statements or circumstances with which it fits in. And this Court said in Doney v The Queen [2] : The essence of corroborative evidence is that it "confirms", "supports" or "strengthens" other evidence in the sense that it "renders [that] other evidence more probable" [3] . However, where a sexual offence against a minor is charged, the minor's evidence is not corroborated by evidence that shows merely that, on one other occasion, an offence of the same kind was committed against another minor. But if the manner in which the offences were committed is strikingly similar, the evidence of the other offence may well be admissible as corroboration. Thus in R v Kilbourne [4] , Lord Hailsham of St Marylebone LC approved a passage in the judgment of the Court of Appeal from which the appeal to the House of Lords was brought [5] : In the present case each accusation bears a resemblance to the other and shows not merely that the appellant was a homosexual, which would not have been enough to make the evidence admissible, but that he was one whose proclivities in that regard took a particular form. Further, the evidence of each boy went to rebut the defence of innocent association which the appellant put forward: this by itself made the similar fact evidence admissible [6] . We have had no doubt that the evidence of one group of boys could properly be taken into account by the jury when considering the counts relating to the other group. But for what purpose since only relevant evidence is admissible? What, for example, did Gary's evidence prove in relation to John's on count 1? The answer must be that his evidence, having the striking features of the resemblance between the acts committed on him and those alleged to have been committed on John, makes it more likely that John was telling the truth when he said that the appellant had behaved in the same way to him. Here, the similarity of the circumstances attendant on his encouragement of W to masturbate with the circumstances of his encouragement to H to do so is so striking that W's evidence was strongly corroborative of H's evidence in that particular. 1. [1973] AC 729 at 750. 2. (1990) 171 CLR 207 at 211. 3. R v Kilbourne [1973] AC 729 at 758, per Lord Simon of Glaisdale. 4. [1973] AC 729 at 737; see also R v Boardman [1975] AC 421 at 440. 5. [1972] 1 WLR 1365 at 1369-1370. 6. See R v Chandor [1959] 1 QB 545 at 550, per Lord Parker CJ.

Of course, evidence that is corroborative must confirm the evidence to be corroborated in a particular material to the commission of the offence charged. W's evidence was evidence of facts similar to some of the facts deposed to by H, albeit not facts constituting an offence similar to the offences charged. However, the corroborative evidence need not be direct proof of the actus reus [7] although, when corroboration of an accomplice's evidence is sought, it can be found only in evidence that confirms or tends to confirm the accused's involvement in the events as related by the accomplice [8] .

1. R v Baskerville [1916] 2 KB 658 at 665, 667. 2. Doney v The Queen (1990) 171 CLR 207 at 211.

it may be asked, what is corroboration? It does not mean that all the material facts have to be proved by independent evidence. What it does mean is well expounded in the following passage from the last edition of Russell on Crimes [10] : - "The confirmation need not extend to every part of the accomplice's evidence, for there would be no occasion to use him at all as a witness, if his narrative could be completely proved by other evidence, free from suspicion. But the question is, whether he is to be believed upon points which the confirmation does not reach. And if the jury find some part of his evidence satisfactorily corroborated, this is a good ground for them to believe him in other parts as to which there is no confirmation. So far all the authorities agree; the only point on which any difference of opinion has been supposed to exist relates to the particular part or parts of the accomplice's testimony which ought to be confirmed." Where corroboration is sought for the evidence of a complainant in a case involving sexual intercourse, it is not necessary that the corroborative evidence should tend directly to prove the act of intercourse. In R v Parish [11] , Ritchie J speaking for the Supreme Court of Canada approved what had been said in the British Columbia Court of Appeal: I think evidence which may be corroboration of the evidence of a female person in such a case is evidence which may, in law, be considered by the jury as evidence of a material particular implicating the accused in the commission of the crime alleged. A particular is material in this sense if it may, in the opinion of the jury, show or tend to show that the testimony of the female person that the offence was committed and committed by the accused is true, thus being relevant to the issue which the jury is called upon to decide. That issue in this case was simply whether or not there was an act of sexual intercourse. To be capable of being considered corroborative, evidence need not in itself prove the guilty act. 1. (1911) 13 CLR 619 at 671. 2. 7th ed, p 2287. 3. (1968) 68 DLR (2d) 528 at 533.

In the present case, BRS's assertion to the headmaster that his cuddling of H was an innocent caring of a young boy was rebutted by H's description of the conduct to which the cuddling led. H's evidence of the way in which sexual relations developed between himself and BRS was an important element in the body of proof of the actual acts charged. Evidence which corroborated or strengthened H's evidence in respect of his first sexual experience with BRS was thus probative of a fact material to the offences charged. Evidence need not tend to prove a "similar fact" offence in order to be corroborative; it is sufficient to constitute corroboration that the evidence should strengthen the evidence to be corroborated as to a fact on which proof of guilt depends. There is no set formula nor are there rigid categories of corroborative evidence. As Lord Hailsham said in R v Boardman [12] :

The rules of logic and common sense are not susceptible of exact codification when applied to the actual facts of life in its infinite variety. 1. [1975] AC 421 at 452.

The facts alleged by H as to the early sexual dealings between himself and BRS were not the facts in issue in the case but they were facts that were critical to proof of the facts in issue and the corroboration provided by W's evidence was corroboration in a most material particular.

If the trial judge had not only refused the Crown Prosecutor's application for a redirection on corroboration but had redirected the jury that W's evidence did not amount to corroboration, the jury would have had its attention focused on the significance of W's evidence and on the contest between the evidence of BRS on the one hand and H and W on the other. Whatever doubt might have been engendered in the jury's mind about the truth of the allegations made after a lapse of years by H, who had been a troublesome pupil, against a school master of high general reputation might have been diminished or dispelled by a redirection drawing their attention to the evidence of W which attributed serious impropriety on the part of BRS in dealing with W. When his Honour declined the Crown Prosecutor's request for a redirection, the appellant escaped the risk of a redirection that might have severely damaged his prospects of acquittal and which could not have improved those prospects.

Contrary to this view, McHugh J is of the opinion that there was "a real chance that the jury not only remembered and accepted the evidence of W but concluded that his evidence combined with that of the complainant made it highly probable that the appellant was a homosexual who preyed on his pupils and had committed the acts with which he was charged". For that reason, although his Honour thinks that the absence of a redirection was the most advantageous course from the appellant's point of view, he would allow the appeal. I respectfully accept the theoretical possibility which leads to this conclusion. But, for my part, I can perceive no greater advantage which the appellant could have obtained in the trial than an absence of any reference by the trial judge to W's evidence. Once it is seen that W's evidence was corroborative of H's evidence and tended to show that BRS had acted in a grossly improper manner with W, any reference to W's evidence would have tended to destroy the defence attack on H's credibility and would have reduced substantially the force of the defence argument that it was dangerous to convict on H's evidence when there was such a long delay between the alleged offences and the making of any specific complaint about them. Any reference to W's evidence, even though the trial judge thought that it was not corroborative, would have tended to give the lie to BRS's contention that he was an innocent carer of the boys in his charge.

In these circumstances, the absence of a direction about the use to be made of W's evidence was incapable of affecting adversely the prospects of acquittal of BRS. Accordingly, I am unable to conclude that there was any miscarriage of justice.

I would have dismissed the appeal. I dissented from the opinion of the majority pursuant to which the order was made allowing the appeal and ordering a retrial.

This appeal concerns the admissibility of evidence at a criminal trial, the use to which admitted evidence may be put and the responsibilities of the trial judge in directing the jury as to the use they make of the evidence.

The appellant was charged on six counts of sexual offences under the Crimes Act 1900 NSW (the Act). All counts related to the same person, a schoolboy who was thirteen years of age at the time of some of the offences alleged and fourteen years of age at the time of the others.

Counts 1-4 were under s 78N of the Act which reads:

A male person who, being a schoolmaster or other teacher, or a father, or stepfather, has homosexual intercourse with any male person of or above the age of 10 years, and under the age of 18 years, being his pupil, son or step-son, shall be liable to penal servitude for 14 years.

Counts 5-6 were under s 78K of the Act which reads:

A male person who has homosexual intercourse with a male person of or above the age of 10 years, and under the age of 18 years, shall be liable to penal servitude for 10 years.

At the time of the appellant's trial, s 405C of the Act distinguished between a prescribed sexual offence and a sexual offence other than a prescribed sexual offence [13] . An offence under s 78K was a prescribed sexual offence; an offence under s 78N was not (s 4). In the former case the judge was not required by any rule of law or practice to give a warning to the jury that it is unsafe to convict on the uncorroborated evidence of the victim. In the latter case nothing affected the operation (if any) of any rule of law or practice requiring the giving of such a warning.

1. Section 405C was repealed by Sch 1 of the Evidence (Consequential and Other Provisions) Act 1995 NSW.

Recognising the difficulties inherent in directing the jury in respect of some of the counts that it was unsafe to convict on uncorroborated evidence but that it was not unsafe so to convict on the other counts, counsel for the appellant at trial made it clear that he was not seeking such a bifurcated direction. In the words of Gleeson CJ in the Court of Criminal Appeal, "it was accepted on all sides that the trial judge should not get into that area" [14] . Accordingly the appellant makes no complaint of the failure of the trial judge to follow the dictates of s 405C.

The appellant was from 1986 to 1993 a schoolmaster at a boys school in Sydney. In 1986 the complainant, who was then thirteen years old, was a boarder at the school where he remained until 1988. The complainant and the appellant developed a close friendship from 1986 until the complainant's expulsion from the school for fighting in 1988. The complainant alleged that, beginning in 1986 and extending to a time after the complainant had left the school, a sexual relationship developed between him and the appellant. According to the complainant, this relationship involved numerous acts of homosexual intercourse over this period, involving anal penetration and fellatio. The complainant also referred to items that were in the possession of the appellant including a lubricant, called KY jelly and a yellow towel that was kept under the appellant's bed for purposes of masturbation. The complainant said that the appellant used the jelly and the towel in the commission of the offences.

In June 1995 the appellant was tried before a judge and jury, in the District Court of New South Wales, on the six counts. During the course of the trial evidence was given by a schoolfriend of the complainant, W. W gave evidence that supported the existence of a yellow towel and the KY jelly in the appellant's room. W also gave evidence that the appellant had invited him to use his room for the purposes of masturbation and of an occasion in the appellant's bedroom when the appellant had invited him to masturbate, offering him the jelly and the towel for this purpose.

In due course, the jury acquitted the appellant of count 3 and convicted him of the other counts in the indictment.

The grounds of appeal before this Court are as follows:

(i) That the Court of Criminal Appeal erred in holding that in the circumstances of the case, the evidence of the witness W could amount to corroboration of the evidence of the complainant

(ii) That the Court of Criminal Appeal erred in holding that the failure of the learned trial judge to give the jury directions on the question of corroboration did not deprive the Appellant of his right to a fair trial according to law.

(iii) That the Court of Criminal Appeal erred in holding that the failure of the learned trial judge to direct the jury on the use which they could legitimately make of the evidence of the witness W did not deprive the Appellant of his right to a fair trial according to law.

(iv) That the Court of Criminal Appeal erred in not deciding the question of whether the evidence of the witness W should have been admitted in the trial of the Appellant and in failing to hold that the admission of this evidence necessarily rendered the trial proceedings unfair and the convictions unsafe and unsatisfactory.

Although it does not appear with clarity from the manner in which the grounds are formulated, the primary ground advanced by the appellant related to the failure of the trial judge to direct the jury as to the use they might properly make of the evidence of W, with the consequence, it was said, that the appellant did not receive a fair trial according to law. Despite the language of ground (iv), the appellant did not argue that the evidence of W was simply inadmissible. The reason for this possible omission lies in the conduct of the trial.

One of the most dangerous aspects of W's evidence, from the point of view of the defence case, was that it involved the appellant in having to contradict W as well as the complainant. Nevertheless the evidence of W was admitted without objection. Furthermore counsel for the appellant cross-examined W on his evidence.

However the trial judge was clearly concerned about the evidence because, after the cross-examination had proceeded for some time, and in the absence of the jury, the following dialogue ensued:

His Honour: Mr Porter, I realise that this is being led without objection, but I am somewhat perturbed about it.

Porter: Mr [S] has put his character in issue. This is the only matter apparently that can be brought against his character. That is the basis upon which it is led. If it was led as evidence against his bad character, I could not stop it. We took the position that we denied this took place at all.

His Honour: Anyway, it was led without objection and I will not do anything further about it.

Porter: It was a decision we made with considerable thought. It was considered that we should not abandon our good character because of this allegation.

Crown Prosecutor: Your Honour, there is also this; that this evidence goes to another issue and that is corroboration. It is capable of being corroborative of his evidence, insofar as it relates to the towel, the masturbation, the K-Y jelly in [the complainant's] evidence. I do not want to argue it now, but merely flag it as a matter I will argue later.

As is apparent, there was no resolution at that stage of the basis upon which W's evidence was being admitted. Counsel for the appellant spoke of the evidence as going to character. The Crown Prosecutor said that the evidence was also relevant to "corroboration". What is clear however is that the defence took no blanket objection to the evidence.

In the course of his address to the jury, the Crown Prosecutor said that the evidence of W "does tend to support and confirm" and "does tend to confirm and link" the evidence of the complainant and that it "challenges some of the evidence relating to [the appellant's] evidence of good character". Counsel for the appellant told the jury that the evidence of the complainant was "entirely uncorroborated" and made no reference to the evidence of W. The trial judge, in the words of Gleeson CJ, "did not review the evidence in the case, and he directed the jury succinctly upon certain relevant legal principles" [15] . His Honour gave the jury standard directions in relation to the evidence concerning the appellant's good character. He made no mention of the evidence of W as tending to contradict the evidence of the appellant's good character or as possible corroboration of the complainant's evidence or as serving any other purpose. Indeed he did not mention the evidence of W at all. Nor did he say anything about corroboration, even in general terms.

1. Section 405AA of the Act provides that a judge, at the end of a criminal trial before a jury, need not summarise the evidence, if the judge is of the opinion that, in all the circumstances, a summary is not necessary.

Clearly enough, the evidence of W was admissible to rebut evidence of the appellant's "good disposition or reputation". Section 413B of the Act permitted it to be used for that purpose [16] . Furthermore counsel for the appellant accepted at trial its admissibility for that purpose.

1. Section 413B was repealed by Sch 1 of the Evidence (Consequential and Other Provisions) Act 1995 NSW.

Equally clearly there was a risk that the jury might, unless otherwise directed, use the evidence for other purposes. It is at this point that difficulties arise since the trial judge himself said nothing about these matters and he was not asked by either counsel to give further directions to the jury, save as to corroboration as appears later in these reasons. There is thus an air of unreality about the arguments placed before the Court of Criminal Appeal and before this Court. At the same time, the appellant was convicted of serious offences. He was sentenced to concurrent terms of penal servitude, involving a minimum term of four years and an additional term of two years. He was granted bail on the lodging of an appeal to the Court of Criminal Appeal, having then served four weeks of his sentence. He resumed serving his sentence on the day the Court of Criminal Appeal delivered judgment dismissing the appeal, 5 March 1996, and he has been in prison since then. At his trial he gave evidence, strongly denying any acts of intercourse with the complainant.

The grounds of appeal contend that the Court of Criminal Appeal erred in holding that the evidence of W could amount to corroboration of the evidence of the complainant and that it further erred in holding that the failure of the trial judge to direct the jury on corroboration did not deprive the appellant of his right to a fair trial according to law.

The first thing to be said in this regard is that corroboration is not a term of art and that, in the present context, it means no more than confirmation [17] . And so the question is: did the evidence of W confirm that of the complainant in any relevant respect? If it did not, the next question is: should the trial judge have so directed the jury?

1. R v Hester [1973] AC 296 at 325, per Lord Diplock.

At the end of the trial judge's direction to the jury, counsel for the appellant did ask for additional directions. He put it this way

your Honour should give a direction that, two things, (1) that there is no corroboration and there is no corroboration whether by [W] or anyone else and (2) that your Honour should direct the jury that it's a matter that they could take into account in considering their verdict, not that they must but they could. I think your Honour particularly in the light of some of the submissions made by the Crown should direct the jury that in fact there is no corroboration as such of [the complainant's] story.

See, I mean [W's] evidence was purely and simply relevant to refuting good character, nothing else.

As Gleeson CJ noted, the Crown Prosecutor had not submitted to the jury that there was "corroboration as such"; he did not use the word corroboration in his address. He spoke of evidence supporting or confirming or tending to confirm the evidence of the complainant. And, in response to counsel's request for further directions, he took the trial judge to the evidence of W at various stages to show what, in his submission, was capable of being treated by the jury as corroboration.

The view of the Court of Criminal Appeal was that, had the trial judge given the first of the directions sought on behalf of the appellant, his Honour would have been in error. The Court of Criminal Appeal was right in this conclusion. However the issue of corroboration is one of some subtlety. The evidence of W as to the existence and purpose of the yellow towel and KY jelly did relate directly to evidence on those matters given by the complainant. In this respect W's evidence could confirm or tend to confirm the appellant's involvement in the events described by the complainant [18] . Because the appellant, in his evidence, advanced an innocent explanation for the jelly does not prevent the evidence from having the capacity to amount to corroboration [19] .

1. Doney v The Queen (1990) 171 CLR 207 at 211. 2. R v Zorad (1990) 19 NSWLR 91 at 103.

On the other hand W's evidence of the appellant's invitations to him to use his room for masturbation and to masturbate on his bed did not relate directly to any evidence of the complainant and could not constitute corroboration of his evidence in any relevant respect. In this Court counsel for the appellant did not seek to distinguish between elements of W's evidence. He contended that none of the evidence of W was admissible to corroborate the evidence of the complainant. That contention cannot be accepted. Accordingly the focus of attention shifts to the position of the trial judge, in particular whether he was bound to make an explanation to the jury to ensure that they did not use the evidence of W for any impermissible purpose. Of course the matter is further complicated by the failure of counsel for the appellant to object to the evidence of W at trial and by the failure of counsel on both sides to make their positions regarding that evidence clear to the trial judge and to each other. I shall return to this matter.

In this Court, though not in the Court of Criminal Appeal, objection was taken to the evidence of W on the further ground that it might well dispose the jury to a conclusion that the appellant was given to the sort of conduct described by the complainant. As this attack really arose from questioning by the Bench and as we do not have the benefit of the views of the Court of Criminal Appeal, there are real difficulties in the disposition of the attack. Certainly it must be said that counsel for the appellant did not ask the trial judge to direct the jury that they should be careful not to treat the evidence of W as indicating a propensity on the part of his client to the conduct with which he was charged. But, as in the case of corroboration, should the trial judge nevertheless have directed the jury to this effect, particularly as counsel had submitted that W's evidence "was purely and simply relevant to refuting good character, nothing else"?

First of all it is necessary to identify the principles in accordance with which evidence might constitute similar facts. In Hoch v The Queen [20] Mason CJ, Wilson and Gaudron JJ said:

The basis for the admission of similar fact evidence lies in its possessing a particular probative value or cogency by reason that it reveals a pattern of activity such that, if accepted, it bears no reasonable explanation other than the inculpation of the accused person in the offence charged. The evidence must be "relevant to prove the commission of the disputed acts" [21] ; similar fact evidence will meet that requirement if it reveals the pattern of activity referred to in Hoch. 1. (1988) 165 CLR 292 at 294. 2. Hoch (1988) 165 CLR 292 at 295.

The evidence in controversy is that of W regarding the appellant's invitations to him to masturbate. As I said earlier, that evidence did not go to confirm the complainant's evidence. It could only be admissible, if at all, as going to the propensity of the appellant to engage in the sort of conduct which was the subject of the charges. The respondent supported the admissibility of the evidence as falling within this category while urging that the evidence was corroboration of the complainant's evidence.

In my view, the evidence of W as to the invitations was not relevant to prove the commission of the acts with which the appellant was charged, namely, acts of intercourse. It was not admissible as similar fact evidence or indeed on any other ground save as to character.

Because of the view I have taken in regard to similar fact evidence, the possibility of collusion between the complainant and W does not arise in regard to that evidence. It does have some possible application to the extent that W's evidence is corroborative or confirmatory of the complainant's.

In Hoch Mason CJ, Wilson and Gaudron JJ said, in effect, that where there is a possibility of joint concoction, that view

is inconsistent both with the guilt of the accused person and with the improbability of the complainants having concocted similar lies.

the admissibility of similar fact evidence depends on that evidence having the quality that it is not reasonably explicable on the basis of concoction [22] .

1. Hoch (1988) 165 CLR 292 at 296-297.
   1. Hoch (1988) 165 CLR 292 at 296-297.

The appellant faces an initial hurdle. In the Court of Criminal Appeal Gleeson CJ said:

Before examining the nature of W's evidence, it should be mentioned that it was not put to W in cross-examination that he and the complainant had put their heads together to fabricate evidence against the appellant.

In this Court counsel for the appellant argued that the issue had been "obliquely put" to W by reason of questions designed to elicit from him that he had been in the appellant's room when no one else was present and that there were other occasions when he and the complainant were there together, in the absence of the appellant. In my view these questions did not suggest, even obliquely, that there had been collusion between the two boys. The questions could have been asked for a number of reasons, for instance, how W could have known of the presence of the KY jelly. There was also evidence of the close friendship between the two boys at school and of the fact that they continued to see and communicate with each other for some years after the complainant left school. But none of this was made the foundation for any cross-examination suggesting the possibility of collusion. The absence of any reference by the trial judge to collusion did not deprive the appellant of his right to a fair trial.

It is clear that the critical issue in this appeal is whether the trial judge should have directed the jury as to the limited use they might make of W's evidence, notwithstanding the absence of any clear submission on behalf of the appellant that the trial judge should do so and notwithstanding the further absence of any attempt to differentiate between evidence to which other use might be put and evidence to which no other use might be put.

The responsibility of the trial judge in these circumstances was discussed by Barwick CJ in Donnini v The Queen [23] where the issue was the use the jury might make of the appellant's prior convictions where evidence of those convictions had been adduced in cross-examination of the appellant, following questions asked of a witness with a view to establishing his good character. Having spoken of the settled policy of the law that, in general, evidence of a propensity to commit a crime is not admissible, Barwick CJ emphasised that cross-examination of an accused as to previous convictions is not intended to place bad character before the jury as a matter to be considered on the question of his guilt or innocence. He continued [24] :

Therefore, the not unnatural tendency of the juryman and the importance of that policy seem to me to require that the trial judge, when evidence of prior conviction is properly before the jury for the sole purpose of combatting a suggestion of good character or to weaken or destroy an accused's credibility, must assist the jury by expressly and with emphasis telling them that they may not use the fact of prior conviction as tending to the guilt of the accused. In Donnini the appeal was dismissed by majority, for reasons which varied but which did not relate to any absence of directions by the trial judge. Thus the actual decision does not advance the argument in the present appeal. But the judgment of Barwick CJ contains the salutary warning just mentioned. 1. (1972) 128 CLR 114. 2. Donnini (1972) 128 CLR 114 at 123.

In B v The Queen [25] , this Court held that the summing up of the trial judge in relation to charges of sexual offences by the appellant against his daughter was defective. At issue was the use the jury might make of the prior conviction of the appellant of similar offences against his daughter, the appellant volunteering this evidence with a view to showing that he had been made the victim of false allegations by his daughter in the past. In the course of his judgment Brennan J said [26] :

Restrictions on the purposes to which evidence can be put arise when evidence is admissible only for a limited purpose. If evidence, admitted for one purpose, is not admissible for another purpose, it cannot be used for that other purpose. 1. (1992) 175 CLR 599. 2. B v The Queen (1992) 175 CLR 599 at 607-608.

The decision not to object to the admission of W's evidence was deliberately taken, no doubt for good reason as seen by the appellant's experienced counsel. If then this appeal turned only on the reception of that evidence, the conscious decision to allow the evidence in as part of the general running of the defence case would inevitably lead to a dismissal of the appeal. But once the evidence was in, it cannot be said that a failure to clarify the basis on which it went in was part of some strategy on behalf of the appellant. The failure to seek clarification at the time W was called to give evidence or at some stage of his evidence was to a degree exacerbated by the Crown Prosecutor doing little more than "flagging" the basis of admission for later argument.

Certainly the appellant's counsel asked the trial judge, by way of redirection, to direct the jury that "there is no corroboration whether by W or anyone else". His Honour declined to give such a direction and, for the reasons mentioned earlier in this judgment, he was right in refusing to do so. But having refused to do so, it was incumbent on him to tell the jury, not in any technical sense but as a matter of simple language, what evidence might go to confirm the complainant's account and what evidence could not be used for that purpose. The problem was compounded by the failure of the trial judge to warn the jury that they should not treat W's evidence about the invitation to masturbate as in any way confirming the evidence of the complainant as to the acts of sexual intercourse which he described. And while the trial judge was not obliged to summarise the evidence if, in all the circumstances, he was of the opinion that a summary was not necessary, this was a case in which an examination of the evidence would have been helpful to the jury. More than that, it was essential in the areas of corroboration and similar facts or propensity. Absent a clear direction on those aspects, there was a likelihood that the appellant would not receive a fair trial because the jury might regard his actions when with W as showing that he was the sort of person to engage in the conduct described by the complainant. It would have been wrong of them to do so. His Honour did not direct the jury as to the facts at all. In all the circumstances, a failure by the appellant's counsel to ask for directions designed to exclude any reliance on similar facts or propensity is not an answer to the defects in the summing up [27] . It is hard to resist the conclusion that, in all the circumstances, the appellant may not have received a trial according to law.

1. Bahri Kural v The Queen (1987) 162 CLR 502 at 512; Chamberlain v The Queen (1983) 72 FLR 1 at 12; 46 ALR 493 at 501-502.

For these reasons I concurred in the order of the Court made 14 August 1997, allowing the appeal, upholding the appeal to the Court of Criminal Appeal, quashing the appellant's conviction on counts 1, 2, 4, 5 and 6 and ordering a retrial on those counts.

The appellant, BRS, appeals from a decision of the Court of Criminal Appeal of the Supreme Court of New South Wales dismissing his appeal against conviction on five counts of homosexual intercourse with H. H was a boarder at the school at which the appellant was employed as a teacher and in-residence tutor. Three of the offences related to the period in which H was a pupil at the school, and two related to a period after he left.

It is not in dispute that the appellant took a special interest in H who did not adjust well to boarding school. Nor is it in dispute that H was frequently in the appellant's living quarters, sometimes alone and sometimes in company with other students, including a student, W, whose evidence lies at the centre of this appeal.

H gave evidence that on several occasions in 1986 the appellant, while lying on his bed in his living quarters, masturbated in front of him. On those occasions, he said, the appellant cleaned himself with a towel which he kept under the bed. H further deposed that he and the appellant engaged in mutual masturbation later that year. According to his evidence, they engaged in homosexual intercourse the next year, 1987, while he was still a pupil at the school and in 1988 after he left the school. On these occasions, it was said, the appellant used a lubricant, KY jelly.

At his trial, the appellant called evidence as to his good character. His intention to call that evidence was known to prosecuting counsel who led evidence in his case in chief from W. As already indicated, W and H were fellow pupils at the school at which the appellant taught. While at school, W was friendly with H, visiting him on one occasion at his family's country property. There was also some contact between W and H after H left the school but the evidence suggests that there was probably none after 1991 [28] .

1. W's evidence is somewhat confusing on this issue. In his examination-in-chief, W testified that he had not seen H since leaving school in 1991. However, in cross-examination, W agreed that he had seen H several times since leaving school. The context of the cross-examination does not make it clear as to whether W was referring to the period since he left school, or the period after H left the school.

W gave evidence that, in 1988, the appellant offered to let him use his room if he wanted to masturbate, telling W that he kept a yellow towel under his bed which he used to clean himself when he masturbated. Later that year, according to W's evidence, W told the appellant of a sexual liaison between himself and a young lady and, on that occasion, at the invitation of the appellant, W masturbated on the appellant's bed while the appellant masturbated in another section of the room out of W's sight. W said that he was invited on that occasion to use some KY jelly that was beside the bed.

The evidence of W was given without objection, the view being taken that it was admissible to counter the character evidence to be called on behalf of the appellant. However, prosecuting counsel indicated early in the proceedings that he would rely on W's evidence as corroborative evidence "insofar as it relates to the towel, the masturbation, the K-Y jelly". And in his address to the jury, he stated that W's evidence had significance because it "tend[ed] to support and confirm that what [H] said was the initial practices of [the appellant] towards him with relation to masturbation". He also said that it "tend[ed] to confirm and link [the appellant] to the towel, to the "KY" jelly" and that it showed that the appellant "had a predilection and a liking for the company of young boys and for masturbation in their company".

The trial judge summed up, giving no instruction to the jury as to the use that might properly be made of W's evidence. Defence counsel then sought a direction "that there [was] no corroboration and there [was] no corroboration whether by W or anyone else". A little later, the following exchange occurred:

[Defence Counsel:] See, I mean [W's] evidence was purely and simply relevant to refuting good character, nothing else.

[Trial Judge:] When it comes to the question of things like "KY" and the towels, I can't see they are any more corroborative or capable of being corroboration than the fact that it was a [bed room].

[Defence Counsel:] That's precisely my submission your Honour, but I think your Honour should tell the Jury this.

Thereafter, prosecuting counsel argued that W's evidence was capable of corroborating the account given by H and, in the result, the trial judge declined to direct the jury on that issue. No further directions were sought by defence counsel with respect to W's evidence.

It was argued for the appellant in this Court, as it was in the Court of Criminal Appeal, that W's evidence was highly prejudicial and that there was a substantial risk of a miscarriage of justice unless the jury was properly instructed as to its use. In this regard, it was put that the evidence could not corroborate the account given by H and the jury should have been instructed that it was relevant only to the appellant's character. Alternatively, it was put that, if the evidence was capable of constituting corroboration, the jury should, nonetheless, have been directed not to use it in any other way in proof of the appellant's guilt. The arguments give rise to two primary issues, namely, whether W's evidence was admissible as corroborative evidence and, also, whether it was otherwise admissible as evidence of guilt. If either issue is determined in favour of the appellant, two further questions arise, namely, whether a direction should have been given by the trial judge as to the permissible use of that evidence and, if so, whether the failure to give a direction in that regard resulted in a substantial miscarriage of justice.

Corroborative evidence is evidence that " "confirms", "supports" or "strengthens" other evidence in the sense that it "renders [that] other evidence more probable" " [29] . In some cases, notably in the case of accomplice evidence, the question is whether it "tends to show that the accused committed the crime [in question], not merely that the crime has been committed" [30] . In the circumstances of this case, the question is whether W's evidence tended to confirm that the offences were committed, not that they were committed by the appellant.

1. Doney v The Queen (1990) 171 CLR 207 at 211, referring to R v Kilbourne [1973] AC 729 at 758, per Lord Simon of Glaisdale. 2. R v Baskerville [1916] 2 KB 658 at 667. See also Doney v The Queen (1990) 171 CLR 207 at 211.

Evidence is not corroborative if, as is sometimes said, it is "intractably neutral" [31] . In the present case, evidence as to the presence of KY jelly in the appellant's room and of a towel under his bed is, perhaps, of no great corroborative strength: the towel, it seems, was standard school issue; and KY jelly has uses other than those to which W and H deposed. Even so, it is not intractably neutral. Accordingly, in my view, W's evidence as to the towel, the use to which it was put by the appellant and the KY jelly was admissible to corroborate some aspects of the account given by H. That being so, the argument that the jury should have been instructed that it was relevant only to the character of the appellant must be rejected. That does not mean, however, that it was appropriate to leave the jury without guidance as to its use as corroborative evidence. That is a matter to which some additional reference will later be made.

1. See, eg, R v Kerim [1988] 1 Qd R 426 at 447, per Macrossan J; R v Kalajzich & Orrock (1989) 39 A Crim R 415 at 433; R v Zorad (1990) 19 NSWLR 91 at 103; Druett v The Queen (1994) 123 FLR 249 at 284, per Kearney J.

W's evidence that he and the appellant separately engaged in masturbation in the appellant's room stands in a different position from that with respect to the towel and the KY jelly. It is convenient to approach that evidence on the basis that it disclosed the commission of a criminal offence, although that is by no means obvious. If it does not disclose a criminal offence, it nonetheless indicates blameworthy or morally repugnant conduct which, for present purposes, is no different from conduct which offends against the criminal law.

Leaving aside the situation in which character is in issue, evidence that an accused person has on other occasions committed other offences is admissible only in strictly limited circumstances which will be identified shortly. That is because, except in those circumstances, evidence of the commission of other offences is highly prejudicial and of limited probative value [32] . As a general rule, evidence that an accused has on other occasions committed offences of the kind charged proves only that he or she is a person capable of committing offences of that kind, not that he or she committed the offence charged. However, a jury might well reason that, if offences of that kind were committed by the accused on other occasions, he or she is very likely to have committed the offence in question.

1. See, in relation to the prejudicial effect of such evidence, Perry v The Queen (1982) 150 CLR 580 at 593-595, per Murphy J; Sutton v The Queen (1984) 152 CLR 528 at 545, per Brennan J; at 562, per Dawson J; Hoch v The Queen (1988) 165 CLR 292 at 301, per Brennan and Dawson JJ; Harriman v The Queen (1989) 167 CLR 590 at 597, per Dawson J; B v The Queen (1992) 175 CLR 599 at 617-618, per Dawson and Gaudron JJ; Pfennig v The Queen (1995) 182 CLR 461 at 487-488, per Mason CJ, Deane and Dawson JJ; at 512-513, per McHugh J. See also Australian Law Reform Commission, Evidence ( Report No 26, Interim) (1985), vol 1, par 810.

In some circumstances, evidence that an accused has committed other offences on other occasions is admissible because it is of particular probative force or has particular cogency. Its probative force or cogency lies in the fact that it discloses some feature which raises, as a matter of common sense and experience, the objective improbability of its bearing an explanation consistent with the accused's innocence of the offence charged [33] . Its probative value or cogency may derive from its disclosure of strikingly similar facts, some unusual feature common to the events in question or some underlying unity, system or pattern [34] . However, that is not an exhaustive description of evidence that has that special probative value [35] . Nor is it a statement of the underlying rationale for its admissibility.

1. See Pfennig v The Queen (1995) 182 CLR 461 at 481-482, per Mason CJ, Deane and Dawson JJ. See also Hoch v The Queen (1988) 165 CLR 292 at 294-295, per Mason CJ, Wilson and Gaudron JJ; Harriman v The Queen (1989) 167 CLR 590 at 600, per Dawson J. 2. See, generally, Hoch v The Queen (1988) 165 CLR 292 at 294-295, per Mason CJ, Wilson and Gaudron JJ; Pfennig v The Queen (1995) 182 CLR 461 at 482, per Mason CJ, Deane and Dawson JJ. 3. See Pfennig v The Queen (1995) 182 CLR 461 at 482, 484, per Mason CJ, Deane and Dawson JJ.

Evidence of criminal or reprehensible conduct on other occasions is admissible because, when considered in conjunction with other evidence in the case, it supports an inference of guilt, in the sense that that is the only reasonable inference available [36] . Ordinarily, that inference is based on a more immediate inference, namely, that it is objectively improbable "that a person other than the accused committed the act in question, that the act was unintended, or occurred innocently or fortuitously" [37] or that the charge was concocted.

1. See, generally, Hoch v The Queen (1988) 165 CLR 292 at 294, per Mason CJ, Wilson and Gaudron JJ; Pfennig v The Queen (1995) 182 CLR 461 at 482-483, per Mason CJ, Deane and Dawson JJ. See also Martin v Osborne (1936) 55 CLR 367 at 375, per Dixon J; at 385, per Evatt J. 2. Hoch v The Queen (1988) 165 CLR 292 at 295, per Mason CJ, Wilson and Gaudron JJ.

The only issue to which W's evidence as to his and the appellant's masturbating could be directly relevant is the question whether H might have concocted his account of events involving homosexual intercourse with the appellant. Evidence of conduct on other occasions has special probative force or cogency in relation to the possibility of concoction only if it discloses some feature which renders it improbable that two or more persons would give an account of particular conduct if that conduct did not, in fact, occur. Or to put the matter in terms used in Boardman v Director of Public Prosecutions [38] , it has probative force or cogency if there are "features common to the stories which two liars concocting false stories independently of one another would have been unlikely to hit upon".

1. [1975] AC 421 at 461, per Lord Cross of Chelsea.

In Boardman, a school master was charged with homosexual offences involving two students, the offences having been committed on separate occasions. The accounts given by the complainants disclosed similar features in that the boys were woken in their respective dormitories at about midnight or in the early hours of the morning, were admonished not to wake any one else and were thereafter invited to take the active role in homosexual intercourse with the master. The evidence of each was held admissible in proof of the offences charged in relation to the other. However, it was said by Lord Wilberforce and Lord Cross of Chelsea that the case was "border-line" [39] , Lord Wilberforce also expressing the view that, if regarded as an example, the decision "[might] be setting the standard of "striking similarity" too low." [40]

1. Boardman [1975] AC 421 at 445, per Lord Wilberforce; at 461, per Lord Cross of Chelsea. 2. Boardman [1975] AC 421 at 445.

As already indicated, "striking similarity" is not the criterion of admissibility of evidence of criminal or morally reprehensible conduct on the part of an accused person on other occasions. Even so, evidence of conduct on other occasions which is led to establish the objective improbability of concoction must have some feature which gives rise to that inference. Ordinarily, that feature will lay in its disclosing some striking similarity, pattern or system. In the present case, the only similarity, pattern or system to be discerned is that, on several occasions, the appellant masturbated in his room while H was present and on another occasion when W was present, apparently in the sight of H but not in the sight of W.

Subject to one matter to be dealt with later, it may be that, if the appellant had been charged with committing an act of indecency by masturbating in front of H, W's evidence would have had particular probative value or cogency with respect to that charge. But the appellant was not charged with that offence. He was charged with offences involving homosexual intercourse. There is nothing about the incident related by W which makes it inherently improbable that H concocted his account of events involving the offences charged or which, in any other way, makes it inherently improbable that the appellant is innocent of those offences.

One other matter should be noted. It is axiomatic that, in cases such as the present, where evidence of conduct on other occasions is tendered in proof of the objective improbability of concoction, that evidence is not admissible unless the possibility of concoction has been excluded [41] . That possibility must be explored before the evidence is admitted - by means of a voir dire, if necessary. That course was not adopted in this case, doubtless because the view was taken that W's evidence was admissible to counter the character evidence to be called in the defence case.

1. Hoch v The Queen (1988) 165 CLR 292 at 296, per Mason CJ, Wilson and Gaudron JJ; but cf at 300-301, per Brennan and Dawson JJ.

Assuming W's evidence otherwise possessed the requisite probative force or cogency, the possibility of concoction should have been explored by the trial judge in the absence of the jury if it was to be left to the jury as direct evidence of the appellant's guilt. So, too, in the hypothetical situation of W's evidence being proffered as evidence of the appellant's guilt of offences involving masturbation in front of H, it would be necessary for the possibility of concoction to be excluded before it could be admitted on that issue.

The Court of Criminal Appeal dealt with the question of concoction on the basis that the issue to be determined was the possibility of joint concoction and that that issue could be determined by reference to the evidence given at the trial. In some cases, the possibility of joint concoction is the only issue that arises. In other cases, however, there may be a question whether the complainant might have concocted his or her account after becoming aware of some like event or some like allegation involving the accused. And in some cases it may be necessary to consider both questions.

In the circumstances of this case, if W's evidence could otherwise be led as direct proof of the appellant's guilt of the offences charged, it would be admissible only if there was no possibility of joint concoction and no possibility that H became aware of the events to which W deposed before complaining of the events giving rise to the offences charged. And those issues may well depend on evidence other than that in the trial. In this regard, it is sufficient to note that matters may have been elicited on a voir dire examination that defence counsel did not wish to open up before the jury.

It is well settled that where evidence is admissible for one purpose but is inadmissible for another, the trial judge "should direct the jury that they must not use the evidence for the purpose for which it is inadmissible [if] the use of the evidence for that purpose would be adverse to the accused." [42] Certainly, a direction of that kind must be given whenever necessary to avoid a perceptible risk of injustice [43] . And in this case there was a clear risk that, unless instructed otherwise, the jury would make improper use of W's evidence, treating it "as indicative of a disposition to [sexual impropriety with young boys]" [44] and using it as evidence of the appellant's guilt. Indeed, prosecuting counsel virtually invited the jury to do just that, putting that the evidence was indicative of "a predilection and a liking for the company of young boys and for masturbation in their company" and suggesting that it could in some way be used as proof of guilt by cautioning that "[the] events between [W] and the [a]ccused doesn't prove on [their own] that [the accused] did these acts with [H]".

1. B v The Queen (1992) 175 CLR 599 at 619, per Dawson and Gaudron JJ, referring to Donnini v The Queen (1972) 128 CLR 114 at 123, per Barwick CJ. 2. Bromley v The Queen (1986) 161 CLR 315 at 324-325, per Brennan J; Carr v The Queen (1988) 165 CLR 314 at 330, per Brennan J; Longman v The Queen (1989) 168 CLR 79 at 86, per Brennan, Dawson and Toohey JJ; Duke v The Queen (1989) 180 CLR 508 at 515, per Brennan J; McKinney v The Queen (1991) 171 CLR 468 at 480, per Brennan J; Pollitt v The Queen (1992) 174 CLR 558 at 586, per Brennan J; at 605, per Toohey J. 3. Donnini v The Queen (1972) 128 CLR 114 at 123, per Barwick CJ.

Notwithstanding the risk that the jury might have used W's evidence in the manner indicated, the Court of Criminal Appeal took the view that no direction was necessary because "the appellant was better off having the trial judge ignore the evidence of W, than explain to the jury the use they could make of it". Moreover, it held that "[t]he appellant should not have leave to complain about a failure to give directions going beyond those sought by [defence counsel]". The reference to leave is apparently a reference to leave under r 4 of the Criminal Appeal Rules. In general terms, r 4 requires leave if an appellant wishes to raise a matter on appeal that was not raised at tria [45] .

1. Rule 4 of the Criminal Appeal Rules 1952 NSW provides: "No direction, omission to direct, or decision as to the admission or rejection of evidence, given by the Judge presiding at the trial, shall, without the leave of the Court, be allowed as a ground for appeal or an application for leave to appeal unless objection was taken at the trial to the direction, omission, or decision by the party appealing or applying for leave to appeal."

As already indicated, there was a real risk that the jury might use W's evidence as positive proof of the appellant's guilt. In my view, that risk far outweighed any advantage that might have been obtained by the trial judge's failure to give directions as to the use which that evidence could properly be put. Indeed so serious was that risk that, in my view, the trial judge should have instructed the jury with respect to its use notwithstanding the limited directions sought by defence counsel. In this last regard, it is sufficient to note that there is a duty on the part of a trial judge to give proper directions if there is a serious risk of injustice, whether or not those directions are sought [46] . Correspondingly, failure of counsel to seek directions cannot be fatal to the success of an appeal or an application for leave to appeal if it appears there has been a substantial miscarriage of justice [47] .

1. See Bromley v The Queen (1986) 161 CLR 315 at 325, per Brennan J; Carr v The Queen (1988) 165 CLR 314 at 330, per Brennan J; Longman v The Queen (1989) 168 CLR 79 at 86, per Brennan, Dawson and Toohey JJ; Duke v The Queen (1989) 180 CLR 508 at 515, per Brennan J; McKinney v The Queen (1991) 171 CLR 468 at 480, per Brennan J; Pollitt v The Queen (1992) 174 CLR 558 at 586, per Brennan J; at 605, per Toohey J. 2. See Stirland v Director of Public Prosecutions [1944] AC 315 at 327-328, per Viscount Simon LC; Giannarelli v The Queen (1983) 154 CLR 212 at 230, per Brennan J; De Jesus v The Queen (1986) 61 ALJR 1 at 3; 68 ALR 1 at 5, per Gibbs CJ; at 6; 10, per Mason and Deane JJ; Bahri Kural v The Queen (1987) 162 CLR 502 at 512-513, per Toohey and Gaudron JJ.

In the circumstances of this case, the jury should have been instructed that W's evidence could only be used to counter evidence of the appellant's good character [48] , to weaken his credibility [49] and to corroborate the account given by H as to the presence of KY jelly in the appellant's room and a towel under his bed for use by him when he masturbated. In addition, the trial judge should have expressly instructed the jury that they could not otherwise use W's evidence as proof of the appellant's guilt [50] . Had those directions been given, verdicts of guilty may not have been returned. Accordingly, the Court of Appeal should have granted the appellant leave pursuant to r 4 of the Criminal Appeal Rules to argue that the trial miscarried by reason of the want of proper directions with respect to W's evidence. It erred in refusing that leave and in dismissing his appeal.

1. Donnini v The Queen (1972) 128 CLR 114 at 123, per Barwick CJ. 2. Donnini v The Queen (1972) 128 CLR 114 at 123, per Barwick CJ. 3. Donnini v The Queen (1972) 128 CLR 114 at 123. See also R v Stalder [1981] 2 NSWLR 9 at 18, per Street CJ; R v Hamilton (1993) 68 A Crim R 298 at 299, per Hunt CJ at CL.

For the above reasons, I concurred in the orders pronounced by this Court on 14 August 1997 allowing the appeal, setting aside the order of the Court of Criminal Appeal and ordering, in lieu, that the appeal to that Court be allowed, the convictions on counts 1, 2, 4, 5 and 6 be set aside and a new trial ordered on those counts.

The question in the appeal is whether the appellant's convictions constituted a miscarriage of justice by reason of the omission of the trial judge to direct the jury on certain matters even if that omission gave the appellant a better chance of acquittal than if the judge had directed the jury according to law. In my opinion, there was a miscarriage of justice because the jury may have convicted the accused by a reasoning process that a proper direction would have denied to them. It follows that the appeal should be allowed and a new trial ordered.

The appellant, a school teacher at a boarding school, was convicted on five of six counts of homosexual intercourse with the complainant, a boarder at the school. In his evidence the complainant said that on occasions before the offences were committed the appellant had masturbated in his room in the presence of the complainant and had used a lubricant called KY jelly for that purpose. The complainant also said that after the appellant masturbated he had cleaned himself with a yellow towel that he kept under his bed.

At the trial, the Crown called another boarder, W, who gave evidence that the appellant had offered him the use of his room if he wished to masturbate and had told him that he kept a yellow towel under his bed for that purpose. W also swore that on one occasion the appellant invited him to his room and asked him if he would like to masturbate, offering him the use of KY jelly. W then masturbated while he lay on the bed and the appellant masturbated behind an intervening wall in the room.

It was common ground at the trial that W's evidence was admissible because it rebutted evidence that the appellant was a person of good character. In his summing up, the learned trial judge referred to the evidence of good character called on behalf of the appellant. His Honour instructed the jurors that they must take that evidence into account on the question of guilt and in assessing the reliability of the appellant's denials of the offences. However, the learned judge made no reference at any stage of his summing up to the evidence of W.

His Honour erred in not giving the jury directions in relation to the evidence of W. That evidence was admissible not only to rebut the appellant's claim that he was a person of good character but also to corroborate part of the complainant's evidence. The Crown Prosecutor, in his final address, had relied on W's evidence as corroborating part of the complainant's evidence and at the conclusion of the summing up had sought a direction to that effect. However, the learned judge refused to give such a direction. He took the view that, because the appellant admitted that he had KY jelly and a yellow towel in his room, W's evidence was not capable of constituting corroboration of the complainant's evidence. But his Honour's reasoning overlooks the fact that W's evidence, if accepted, showed that the appellant kept KY jelly and the towel under his bed for the purposes of masturbation. It therefore tended to confirm the complainant's account of the sexual activity of the appellant which led up to the oral and anal intercourse which were the subject of the charges against the appellant.

The failure of the learned trial judge to mention the evidence of W in his summing up, therefore, gave the appellant a double advantage. It meant that, although the judge had summed up strongly in favour of the appellant on the question of good character, he had not reminded the jury of W's evidence, evidence, which if accepted, rebutted the appellant's claim of good character and threw doubt on his general credibility. It also meant that the jury considered the case without a direction that W's evidence concerning the towel and the KY jelly corroborated the complainant's evidence that he had had homosexual intercourse with the appellant. On the issues of character and corroboration, therefore, the failure of the learned judge to mention the evidence of W was highly favourable to the appellant. However, the failure of the judge to deal with the evidence of W also had a downside so far as the appellant was concerned, for it left the jury without guidance as to how they could legally use that evidence.

One use that the jury could not legally make of W's evidence was to find that the appellant was "the type of person" who had "a predilection and liking for the company of young boys and for masturbation in their company" and then use that finding to reason that, given this propensity, the appellant had probably done what the complainant alleged. It is a fundamental principle of the criminal law in this country that ordinarily the guilt of an accused person is not to be "inferred from the character and tendencies of the accused" [51] .

It is true, as I pointed out in Pfennig v The Queen [52] , that there are cases where the interests of justice require that evidence should be admitted even though it reveals that the accused has a criminal or reprehensible propensity. It is also true that in some cases evidence may be admitted even though the sole basis for its admission is that it reveals a criminal or reprehensible propensity on the part of the accused which the Crown will rely on to prove the guilt of the accused [53] . Before evidence revealing such a propensity is admitted, however, the judge, according to numerous decisions of this Court [54] , must be satisfied that the probative value of the evidence outweighs its prejudicial value.

1. (1995) 182 CLR 461 at 523. 2. Pfennig (1995) 182 CLR 461 at 526-527. 3. See, eg, Perry v The Queen (1982) 150 CLR 580 at 604, per Wilson J; at 609, per Brennan J; Sutton v The Queen (1984) 152 CLR 528 at 547, per Brennan J; at 560, per Deane J; at 564, per Dawson J; Hoch v The Queen (1988) 165 CLR 292 at 300, per Brennan and Dawson JJ; Harriman v The Queen (1989) 167 CLR 590 at 598, per Dawson J; at 633, per McHugh J; Thompson v The Queen (1989) 169 CLR 1 at 16, per Mason CJ and Dawson J; B v The Queen (1992) 175 CLR 599 at 608, per Brennan J; at 618, per Dawson and Gaudron JJ.

If evidence revealing a criminal or reprehensible propensity is admitted, the trial judge must give the jury careful directions concerning the use which they can make of the evidence. If the evidence is admitted for a reason other than reliance on propensity, the judge must direct the jury that they can use the evidence for the relevant purpose and for no other purpose. In some cases, the judge may need to be more specific. He or she may need to direct the jurors that they cannot use the evidence for an identified purpose. If the evidence is admitted because the Crown wishes to rely on the accused's propensity as an element in the chain of proof, it is especially necessary that the judge give the jurors clear directions as to the manner in which they may use the propensity evidence.

In the present case, the learned trial judge was not asked to rule on the admissibility of W's evidence even though it disclosed criminal or reprehensible conduct on the part of the appellant. This was because it was conceded that it was relevant to rebut evidence that the appellant was a person of good character [55] . Some of it was also admissible, as I have said, because it corroborated the evidence of the complainant. Nevertheless, because the evidence of W did disclose criminal or reprehensible conduct on the part of the appellant, absent some countervailing circumstance of sufficient cogency, the trial judge was bound to direct the jurors as to how they could use the evidence of W.

1. The fact that the trial judge was not asked to rule on it also underlines the point that I made in Pfennig ((1995) 182 CLR 461 at 530) that not all evidence revealing criminal or reprehensible propensity must pass the no rational explanation consistent with innocence test to be admissible.

If the judge had referred to the evidence of W, he would have been bound to tell the jury that they could use the evidence only to rebut the appellant's claim of good character and to corroborate that part of the complainant's evidence concerning the use of the KY jelly and the keeping of a yellow towel under the appellant's bed for the purposes of masturbation. The difficulty of this case lies in the fact that the learned judge did not refer to W's evidence. That being so, a miscarriage of justice has occurred if there is a real chance that the jury may have convicted the accused by a chain of "forbidden reasoning", to use the phrase of Lord Hailsham in Boardman v Director of Public Prosecutions [56] .

Trial judges have no authority to dispense with the directions that the law requires them to give in criminal trials. If the failure to give a direction may have resulted in the conviction of the accused person, the trial has not been conducted according to law and the conviction constitutes a miscarriage of justice. It is no answer that the directions which were given or not given gave the accused a better chance of acquittal than the directions that the law required. A person is entitled to be tried according to law. If the failure to give a required direction may have brought about the accused's conviction, there is a miscarriage of justice.

So a critical question in the present case is whether there is a real chance that the omission to give directions concerning W's evidence may have been the factor that resulted in the appellant's conviction. That question depends on whether the jury might have used W's evidence to conclude that the appellant had homosexual tendencies that made it more probable than not that he had oral and anal intercourse with the complainant.

In my opinion, nothing in the course of the Crown Prosecutor's address invited the jury to convict the accused on propensity. The Crown Prosecutor submitted to the jury:

You've seen [W]. You can make your own assessment, but if you accept his evidence then it has some significance because in my respectful submission it does tend to support and confirm that what [the complainant] said was the initial practices of [the appellant] towards him with relation to masturbation is borne out. Remember that [the complainant] said that he always used "KY" jelly and an old towel under the bed and it seems that that was the practice that [the appellant] or at least that was something that [the appellant] referred to when [W] was with him, there was an old towel, an old yellow towel and a tube of "KY" jelly. And I mean if you accept, let me say this immediately, if you found [W] to be truthful, reliable and accurate and that these events occurred between [W] and [the appellant], that doesn't prove on its own that [the appellant] did these acts with [the complainant]. You don't sort of do it in reverse, you don't say, well I'm satisfied that [W]'s telling the truth therefore he's guilty of the offences with which he's charged concerning [the complainant].

The evidence is of significance for those reasons, I submit. It does tend to confirm and link [the appellant] to the towel, to the "KY" jelly, but perhaps most importantly it meets some argument if you like and provides you with a bit of a picture on the other side of the coin as to the type of person [the appellant] was.

You've heard a lot of evidence from his friends and from teachers at [the school], the only boys that you've heard about that were there at the time with [the complainant], [A] and [W]. And if you accept what [W] says as true then in my respectful submission that to some very large extent challenges some of the evidence relating to [the appellant's] evidence of good character. Because if you accept that then obviously [the appellant] had a predilection and a liking for the company of young boys and for masturbation in their company.

[The appellant] was not a person of good character because it demonstrates that he had this unusual interest and desire in young boys, and he discussed masturbation in their presence, that he masturbated in their presence. These submissions invited the jury to use the evidence of W to rebut the appellant's claim of good character and to corroborate the evidence of the complainant. If in fact the jury used the substance of these submissions to convict the appellant by a forbidden chain of reasoning, they did so without any invitation from the Crown Prosecutor.

Because the Crown Prosecutor did not invite the jury to use the evidence of W in an impermissible way and the learned trial judge did not even mention that evidence, there is nothing in the course of the trial that provides any ground for thinking that the jury may have used a forbidden chain of reasoning to convict the appellant. Moreover, the trial judge has had vast experience in criminal work both at the Bar and on the Bench. It obviously did not occur to him - the man on the spot - that this jury needed a direction not to use the evidence of W as showing that the appellant had a homosexual propensity which made it probable that he committed the offences against the complainant. Furthermore, the appellant was represented by a very able and experienced criminal trial lawyer who did not ask the judge to direct the jury that the evidence of W could not be used as the basis for propensity reasoning. Obviously, it did not occur to the lawyers at the trial, with the advantage of seeing and hearing the witnesses and observing the jury's reaction to the evidence, that the jury might decide the case on a basis other than the issues put to them by the Crown and the appellant in their addresses and by the learned trial judge in his summing up.

The only basis for concluding that the jury may have used a forbidden chain of reasoning in reaching its verdict in this case is that criminal courts take it as axiomatic that, where the evidence reveals the criminal convictions or propensity of the accused, there is a real risk that the jury will reason towards guilt by using the conviction or propensity [57] . Because that is so, the trial judge must give a clear direction to the jury as to the use that they can make of the evidence [58] . However, it does not follow that the failure to give such a direction always constitutes a miscarriage of justice.

1. cf Castro v The Queen (1881) 6 App Cas 229 at 244; Donnini v The Queen (1972) 128 CLR 114 at 123; Sutton (1984) 152 CLR 528 at 541-542; De Jesus v The Queen (1986) 61 ALJR 1 at 3, 7, 8, 9-10; 68 ALR 1 at 4-5, 12, 14, 16. 2. Donnini (1972) 128 CLR 114 at 123.

Furthermore, it is not a proper approach in determining whether there has been a miscarriage of justice in this case simply to ask whether it is possible that the jury used propensity reasoning and whether a direction might have avoided the conviction. The real question is whether, given the trial judge's directions or lack of them, the direction which should have been given would have given the appellant a real chance of acquittal that he did not have because the judge said nothing about W's evidence.

A proper direction would have been on the following lines:

In answer to the evidence of good character called by the accused, the Crown says that W's evidence, if you accept it, shows that the accused is not a person of good character. You will recall that W gave evidence that the accused invited him to use his room if he ever wished to masturbate and that he could use the KY jelly and the yellow towel that he kept under the bed for that purpose. You will also recall that W said that, after he had left a note for the accused, the accused invited him to his room. When W went to the accused's room, the accused suggested that W lie on the bed out of sight and masturbate using KY jelly while the accused also masturbated behind an intervening wall. The Crown says that no schoolmaster of good character would make such suggestions to a pupil. The Crown says that, if you accept the evidence of W, the accused is not a person of good character and, in considering his denials of the complainant's evidence, you will bear in mind that those denials come from a person who is not of good character.

In addition, the Crown says that W's evidence corroborates part of the evidence of the complainant. If you accept W's evidence that the accused told him that he could use the KY jelly and the yellow towel if he wished to masturbate in the accused's room, you are entitled to find that that evidence confirms part of the complainant's evidence. You would be entitled to find that it confirms his evidence that the accused used KY jelly and the yellow towel in the course of the sexual activity which led to the oral and anal intercourse, the subject of the charges.

However, you must use the evidence of W, if you accept it, only in the limited way that I have just described. You can use it to rebut the accused's claim that he is a person of good character and you can use it to confirm the complainant's evidence that the accused used KY jelly for the sexual activities which the complainant described and that he cleaned himself with a yellow towel which he kept under his bed after engaging in those activities. However, except in those two ways, you must not use that evidence in determining whether the accused is guilty of the offences with which he stands charged.

If you accept the evidence of W, you will probably have little difficulty in concluding that the accused is a person of bad character. But you must not use that finding to conclude that he is guilty of these offences because he is a person of bad character. In particular, you must not conclude that W's evidence indicates that the accused has a disposition to engage in improper sexual conduct with young boys and that it is likely therefore that he gave effect to that disposition by having oral and anal intercourse with the complainant. Nor, if you find that the accused made the suggestions and did the things which W said that he did, may you conclude that that makes it likely that he engaged in the sexual misconduct that the complainant described. W's evidence can only be used to rebut the accused's claim that he is a person of good character and to corroborate the complainant's evidence that the accused used KY jelly and a yellow towel that he kept under his bed for the purpose of masturbation. You cannot use W's evidence to conclude that the appellant is a person of bad character who is likely to commit offences of the kind with which he is charged. You can only use W's evidence for the very limited purposes of rebutting evidence of good character and corroborating part of the complainant's evidence.

Directions to this or similar effect would in my opinion have been far more damaging to the appellant than the course which the learned trial judge took. The failure of the judge to mention W's evidence meant that the minds of the jurors were focused on the conflict between the complainant and the appellant in a context where the jury was led to believe that the evidence of the appellant's good character was all one way. Once W's evidence was introduced into the summing up, the focus of the summing up would have shifted from a conflict between the appellant and the complainant to a conflict between the appellant and W on the one hand and a conflict between the appellant and the complainant on the other. Moreover, logically the jury would have had to determine the issue of character before they could resolve the conflict between the appellant and the complainant. Once they accepted the evidence of W, the conviction of the appellant might not have been inevitable but his prospect of acquittal would have been small. If the jury rejected the appellant's denial of W's evidence and his claim of good character, he would have been left with little more than his plea of not guilty. His chance of acquittal would have rested on the jury refusing to act on what would in reality have been the uncontradicted evidence of the complainant corroborated in part by the evidence of W.

The course taken by the trial judge, on the other hand, gave the appellant a much better chance of acquittal. He did not refer to the details of the complainant's evidence. However, he told the jury that "realistically the issue really has come down to a very narrow one". His Honour said there was a:

stark conflict in the evidence that exists here between [the complainant] and the accused The law requires that when you come to confront that you should have regard to the evidence of the prior character of the accused. The judge did not even mention W, whom Gleeson CJ in the Court of Criminal Appeal described as having "no axe to grind", or his very damaging evidence. Moreover, the summing up avoided the need for the jury to determine whether the appellant was a person of good character, the summing up taking it for granted that he was such a person. It also made no mention of the fact that W's evidence corroborated part of the complainant's evidence. All these were formidable advantages for the appellant.

However, despite the summing up, there remains a real chance that the jury not only remembered and accepted the evidence of W but concluded that his evidence combined with that of the complainant made it highly probable that the appellant was a homosexual who preyed on his pupils and had committed the acts with which he was charged. Although I think that the course which the judge took was by far the most advantageous course that could be taken from the appellant's point of view, the fact remains that a proper direction would have avoided the risk of use of the forbidden chain of reasoning. The failure to give the direction means that the appellant lost a real chance of being acquitted. Because there is a real danger that the jury may have used the forbidden chain of reasoning, there has been a miscarriage of justice in the technical sense of that term. Furthermore, although the appellant's chance of acquittal may have been small if a proper direction had been given, it is impossible to conclude that he must inevitably have been convicted.

Since the appellant did not have a trial according to law and may have avoided conviction if he had, I concurred in the order of the Court which allowed this appeal.

In Wigmore on Evidence [59] , commenting on the admissibility of evidence of specific acts of misconduct proffered in a criminal trial to establish bad character, the authors remark:

It may almost be said that it is because of this relevancy of such evidence that it is excluded. It is objectionable, not because it has no appreciable probative value, but because it has too much. Similar remarks have been made from time to time in this Court [60] . 1. 3rd ed (1940), vol 1, par 194, p 646, cited in R v Heydon [1966] 1 NSWR 708 at 738. 2. Harriman v The Queen (1989) 167 CLR 590 at 597, per Dawson J; see B v The Queen (1992) 175 CLR 599 at 608, per Brennan J; Pfennig v The Queen (1995) 182 CLR 461 at 509, per Toohey J; cf Dawson, Recent Common Law Developments in Criminal Law Criminal Law Journal, vol 15 (1991) 5, at p 13 .

This appeal from the Court of Criminal Appeal of New South Wales [61] once again raises the troublesome issue of propensity evidence [62] . It is an area of the law which has been described as "confused and confusing" [63] . Specifically, the appeal concerns the obligation of a trial judge to explain to a jury the way in which such evidence, once admitted, may and may not [4] be used in considering whether the prosecution has proved beyond reasonable doubt the offences alleged in the counts of the indictment. Many other matters were argued in the appeal. But this, in the end, is the consideration which determines its outcome.

1. R v BRS (unreported; Court of Criminal Appeal of NSW; 5 March 1996), per Gleeson CJ delivering the judgment of the Court, Studdert and Dowd JJ concurring. 2. See Pfennig v The Queen (1995) 182 CLR 461. 3. Wright, Case and Comment - Zappala Criminal Law Journal, vol 16 (1992) 190, at p 191 commenting on R v Zappala (unreported; Court of Criminal Appeal of NSW; 5 November 1991). 4. R v Dolan (1992) 58 SASR 501 at 503.

In the late 1980s BRS [5] (the appellant) was employed as a school teacher and resident tutor at a Sydney boarding school for boys. His pupils included the complainant, H, a boarder at the school, and W whose time at the school overlapped with H and who became H's good friend. On one occasion, in January 1987, H and W, together with the appellant, stayed for a week at the rural property of H's mother. H was then thirteen years of age. W was the same age. It was soon after that holiday that, according to H, the first offence by the appellant occurred. This became the subject of the first count of the indictment.

1. With the support of the Crown, the name of the appellant was suppressed pursuant to the Crimes Act 1900 NSW, s 578A in order to protect the identity of the complainant. The names of the complainant and of his witness, W, were suppressed on the same basis.

In January or February 1987, H claimed that, whilst he was in the room in which the appellant resided at the school, the appellant proposed that H should masturbate and he performed oral sex upon him. H stated that the same thing happened in April 1987, also in the appellant's room, although on this occasion he claimed that the appellant procured him to reciprocate the oral sex. In September 1987, during school holidays, when H had just turned fourteen years of age, the appellant again visited the property of H's mother. According to H, on the car journey home, the appellant booked into a motel at Coonabarabran where further sexual acts occurred. When back in Sydney, some time in October 1987 in the appellant's room at night, H claimed that the appellant performed anal sex upon him using a lubricant (KY jelly).

H ceased to be a pupil at the school in February 1988. However, according to his evidence, the association with the appellant continued. In March 1988 the appellant again visited the country property and allegedly engaged in mutual masturbation with H. Shortly afterwards, whilst H was in Sydney, he revisited the school, met the appellant, went to his room and allegedly engaged in mutual oral sex.

It was not until August 1991 that H told his mother of the alleged offences by the appellant. The mother contacted the school chaplain. School officers discussed the complaints with the appellant who denied them. Nothing further was done at that stage. H's mother did not report the matter to H's father, from whom she was separated. However, in January 1993, H told his father. The latter immediately notified the police. They set in train investigations which led, in September 1993, to six charges being brought against the appellant. Four of the charges were brought under the Crimes Act 1900 NSW, s 78N. Two of the charges were brought under s 78K. The former section deals with a circumstance of aggravation, namely that the offender is a teacher who has homosexual intercourse with a pupil. The charges differentiated those incidents which arose whilst H was a pupil of the appellant at the school (counts 1-4) and thereafter (counts 5 and 6).

The appellant has at all times protested his innocence of the allegations. He was tried in the District Court of New South Wales before Judge Court and a jury. The trial lasted six days. The jury retired at 12.55 pm on the fifth day. At 2.35 pm on the sixth day they returned with their verdicts. They found the appellant guilty of five of the six counts, finding him not guilty only of the charge which related to the alleged offence in 1987 in the motel in Coonabarabran. At one stage in the appeal which followed the convictions, a point was argued concerning the suggested inconsistency of the jury verdicts. However, that point was decided against the appellant. It has not been re-agitated.

The appellant had no prior convictions. Evidence was given, including by Crown witnesses, as to his excellence as a teacher and otherwise as to his good character. He gave sworn evidence in which he denied H's complaints. As is usual in offences of this kind, there were no eye witnesses to the alleged events. The appellant's trial counsel (who did not appear in the appeals) is one of the most experienced counsel in criminal trials in New South Wales. His skilful cross-examination appears to have demonstrated that H was a somewhat aggressive and unruly young man, given to fighting: conduct which in 1988 led to his being asked to leave the school.

On the beginning of the fifth day of the trial, the jury sent a note to the judge. The judge's impression at that stage, whilst the prosecutor was cross-examining the appellant, can be gleaned from the fact that he interpreted the note as an inquiry about the jury's right to acquit the appellant after the close of the Crown's case [6] . In the event, the trial was allowed to run its course. The judge gave a very brief summing up, covering little more than eight pages of transcript [7] . There is no complaint about the succinct legal directions which his Honour gave the jury so far as they went. As will appear, the appellant's complaint concerns the failure of the judge to instruct the jury on a number of matters.

1. Transcript (Tpt), 14 June 1995. 2. Crimes Act 1900 NSW, s 405AA provides that the judge need not summarise the evidence at the end of a trial before a jury if the judge is of the opinion that, in the circumstances, a summary is not necessary.

Judge Court told the jury that "realistically the issue really has come down to a very narrow one" [8] . Later he said:

It is probably difficult to imagine a case involving a more stark conflict in the evidence that exists here between [H] and the accused. It is black and white, chalk and cheese. The law requires that when you come to confront that you should have regard to the evidence of the prior character of the accused. [9] 1. Charge to the jury, 15 June 1995 at 3, per Judge Court. 2. Charge to the jury at 8, per Judge Court.

Following the jury's verdicts, Judge Court convicted the appellant. He sentenced him to concurrent terms of penal servitude, comprising a minimum term of four years and an additional term of two years. In the course of his remarks on sentence, he said:

For my part, had I been the tribunal of fact (which I was not), I would have entertained a reasonable doubt as to his guilt.

To understand the jury's verdicts, it is necessary to mention the evidence given by W. At the time of the trial, W was participating in a sporting competition in Europe in which he was representing Australia. He was brought back to Sydney to give evidence. Clearly, his evidence was critical. Of it, the Court of Criminal Appeal said [10] :

We, of course, have never seen either W or [H], and it is difficult to say, merely from a reading of the transcript, what a jury would have made of them. However, the impression created by the transcript is that W would have presented as the more formidable character. He seems to have had no axe to grind. So far as the evidence went, he had never made any allegations to anybody previously about the appellant's behaviour, and he had first mentioned the conduct, of which he gave evidence, when he was approached by a police officer in 1993. 1. R v BRS (unreported; Court of Criminal Appeal of NSW; 5 March 1996) at 8.

According to W, he had been quite a close friend of H when they were at the school together. They had some contact after H left the school and until 1991. However, between that year and the trial in 1995, they had had no contact. It was not suggested that, at the time of the alleged offences, W was aware of any suggested sexual activity between the appellant and H. Nor did W assert that there had ever been any sexual intercourse between himself and the appellant. However, he did give evidence of conduct which was highly inappropriate for a teacher dealing with boys aged thirteen to fourteen years. It is this evidence which raises the issue of propensity and gives rise to the appellant's principal complaint before this Court concerning his trial.

W gave evidence that he had engaged in conversations with the appellant concerning his growing sexual awareness. It was not suggested that this was in any way improper. However, W elaborated two developments which were improper and potentially very damaging to the appellant's defence. The first was a conversation whereby the appellant allegedly invited W to use his room at the school if ever he wished to masturbate. He allegedly told W that he could use KY lubricant which the appellant kept in the room and also a towel which he kept under his bed. This evidence bore a similarity to testimony earlier given by H that, in 1986, before any of the offences charged had occurred, the appellant had also talked to H about masturbation, had masturbated in his presence and had used KY jelly and a towel kept under the bed in the course of such conduct.

Secondly, in 1988, after W left a note for the appellant announcing his first sexual encounter with a young woman, the appellant allegedly invited W to his room. He suggested that W lie on a bed out of sight of the appellant and masturbate whilst the appellant did so behind an intervening wall. W stated that this was the first time he had ever masturbated. He said that the appellant had offered him the use of KY jelly as a lubricant. He stated that, in retrospect, he had come to regard this as inappropriate behaviour on the part of a teacher. But he said nothing about it until 1993 when he was approached by a policewoman investigating the complaints of H. How it became known that W could give evidence of such a kind was not explored.

W was the last witness called in the Crown's case against the appellant. The appellant strenuously denied the truth of W's allegations. W was strongly cross-examined to suggest the falsity of his testimony. However, counsel had little ammunition at his disposal. W agreed that he and H had been together in the appellant's room on as many as 15 occasions, at times when the appellant was absent. Presumably the suggestion was that these opportunities afforded the boys the chance to go through the appellant's drawers, to find the lubricant (which the appellant had admitted was at one time there) and to see a standard issue yellow towel under the appellant's bed. But the basic problem for the appellant lay in the evidence which W gave. To adapt the words of McHugh J in Pfennig v The Queen [11] , it would have required a "superhuman effort by the jury to regard the appellant" in a favourable light, whether or not he was guilty of the offences with H as charged in the indictment, once evidence of such inappropriate conduct was placed before them and they believed it.

Both before the Court of Criminal Appeal and in this Court the principal arguments for the appellant have concerned the admissibility of the evidence of W, the limited purposes for which such evidence could be used, once admitted, and the failure of the trial judge to resolve differences between the prosecutor and the appellant on these questions and to lend the authority of his office to the instruction of the jury about them.

W was also overseas when the committal proceedings took place. Counsel for the appellant gave notice that he required W at the committal for cross-examination. In the event, that was not possible. When, after the trial had commenced, it appeared that W was again overseas, the prosecutor, on the second day, signalled his intention to call him back from Europe. He identified two purposes, namely to provide evidence "that would rebut character evidence" and, as well, evidence which "may afford evidence of corroboration of [H's] evidence" [12] .

1. Statement by the Crown Prosecutor in the absence of the jury on 8 June 1995; see transcript of the trial at 119.

When W gave his evidence, no objection was taken for the appellant as he was led in chief through the testimony just described. Cross-examination of W had begun when the trial judge interrupted. He had obviously become concerned about the potential prejudice to the appellant of the course which the trial was taking. He sent the jury out and raised with counsel the fact that, although the evidence was being led without objection, he was "somewhat perturbed about it". Counsel for the appellant explained that a decision had been taken "with considerable thought" not to abandon the appellant's reliance on his good character because of W's allegations. He asserted that this was "the basis upon which [the evidence] is led", namely to rebut evidence of good character already elicited from Crown witnesses and foreshadowed in the appellant's case which was shortly to begin.

The prosecutor responded to this explanation. He stated that he had a further basis for tendering the evidence which he did "not want to argue now, but merely flag" as a matter to be argued later. This was that the evidence of W was "capable of being corroborative of [H's] evidence, insofar as it relates to the towel, the masturbation the KY jelly". Neither counsel sought a ruling on the admissibility of W's evidence at that stage. The use which might be made of W's evidence once it was admitted, was left unresolved.

Counsel's addresses to the jury were transcribed. Especially because of the brevity of the judge's summing up and the arguments of the parties, it is necessary to note some of what was said.

if you accept his evidence then it has some significance because it does tend to support and confirm that what [H] said was the initial practices of [the appellant] towards him with relation to masturbation is borne out, remember that [H] said that he always used "KY" jelly and an old towel under the bed and it seems at least that was something that [the appellant] referred to when [W] was with him, there was an old towel, an old yellow towel and a tube of "KY" jelly [I]f you found W to be truthful, reliable and accurate and that these events occurred between [W] and the Accused, that doesn't prove on its own that [the appellant] did these acts with [H], you don't sort of do it in reverse, you don't say, [w]ell I'm satisfied that [W]'s telling the truth therefore he's guilty of the offences with which he's charged concerning [H].

The evidence is of significance for those reasons it does tend to confirm and link [the appellant] to the towel, to the "KY" jelly, but perhaps most importantly it meets some argument if you like and provides you with a bit of a picture on the other side of the coin as to the type of person [the appellant] was.

You've heard a lot of evidence from his friends and from teachers And if you accept what he says as true then in my respectful submission that to some very large extent challenges some of the evidence relating to his evidence of good character. Because if you accept that then obviously [the appellant] had a predilection and a liking for the company of young boys and for masturbation in their company.

Counsel for the appellant told the jury that the evidence of H was "entirely uncorroborated" [14] . True to his conception of the limited basis upon which the evidence of W had been admitted, he stated that the accusation of W was evidence before them in refutation of the appellant's claim to be a person of good character [15] :

that's all it does, it doesn't corroborate the charge of course, if anything it shows that [the appellant] was not homosexual because there was nothing homosexual whatever about what [W] alleges. 1. Transcript 15 June 1995 at 3. 2. Transcript 15 June 1995 at 19.

After Judge Court's summing up, counsel for the appellant immediately asked for a redirection on corroboration [16] . His Honour indicated a disinclination to say anything about it. However, at the request of the Crown, the jury were sent out to consider their verdicts whilst debate about further directions took place. Counsel for the appellant acknowledged a complication arising from the statutory abolition, in respect of certain of the counts [17] of the requirement [18] that a warning be given of the danger of convicting of sexual offences on the uncorroborated testimony of the complainant [19] . Counsel acknowledged the complexities. He did not ask the judge to explain "all that" to the jury [20] . However, he did ask for a direction that there was no corroboration, whether by W or anyone else, of H's evidence and that the jury could take that matter into account in considering their verdicts. He said that this was particularly necessary in light of some of the statements made by the prosecutor in his address to the jury.

1. The exact terms are stated as "not transcribable" but the word "corroboration" is used. 2. The first four counts concerned offences which were not "prescribed sexual offences". The statutory provision did apply to counts 5 and 6. See Crimes Act 1900 NSW, s 405B. The provisions are explained in the reasons of Toohey J. 3. See, eg, Kelleher v The Queen (1974) 131 CLR 534 at 553, per Gibbs J. 4. cf Longman v The Queen (1989) 168 CLR 79. The statutory provision does not remove the requirement of the judge to give a warning whenever it is necessary to avoid a perceptible risk of a miscarriage of justice arising from the circumstances of the case. 5. Transcript 15 June 1995 at 10.

Judge Court observed that he could not see anything in W's evidence capable of being corroborative of H's evidence. He said [21] :

When it comes to the question of things like "KY" and the towels, I can't see they are any more corroborative or capable of being corroboration than the fact that it was a [bed in the] room. The appellant's counsel agreed with this comment. He repeated his assertion that the evidence of W was relevant only to refuting good character. He asked the judge to tell the jury so. 1. Transcript 15 June 1995 at 10.

There followed submissions for the prosecutor. He too asked the judge to give the jury a direction on corroboration. However, it was his submission that the evidence of W was admissible as corroboration of the practices of the appellant "in the early stages" both with H and W. He suggested that the evidence tended to link what had occurred to both boys and thereby to confirm the evidence of H. He relied on Doney v The Queen [22] . The prosecutor asked the judge to "explain what corroboration in the eyes of the law is and then simply identify those matters that are capable of being corroborative of [H's] evidence "

At the conclusion of these submissions, Judge Court declined the applications of both counsel. He gave no directions as to corroboration or as to the way in which W's evidence might be used by the jury. At no stage did the appellant's counsel expressly request the judge for a specific direction as to the dangers of reasoning from propensity evidence [23] . Still less did he seek the rejection of W's evidence in the exercise of the judge's discretion on the footing that its prejudicial effect outweighed its probative utility or that it should otherwise be excluded [24] .

1. Hoch v The Queen (1988) 165 CLR 292; Pfennig v The Queen (1995) 182 CLR 461. 2. cf R v Stalder [1981] 2 NSWLR 9.

Following his conviction, the appellant appealed to the Court of Criminal Appeal. Leaving aside the argument about inconsistent verdicts and a suggestion that the convictions were unsafe or unsatisfactory, the appellant's real attack before that Court was that he had not had a trial according to law. This was because the directions given by the trial judge on the question of corroboration and the directions given on the use which the jury might make of the evidence of W were inadequate. The Court of Criminal Appeal rejected these arguments.

As to the suggested inadequacy of the judge's directions on corroboration, Gleeson CJ recounted the submissions of counsel to the jury extracted above and the applications for redirection on corroboration made by both counsel. However, he concluded that, if the trial judge had given the direction sought by the appellant's counsel, he would have been in error. In his opinion, the evidence of W was capable of strengthening the evidence of H by "confirming, or tending to confirm, the appellant's involvement in the events as related by [H]" [25] :

What was significant was that both the complainant and W, apparently independently, said that the appellant told them that he kept [a yellow towel] under his bed, and used it for purposes of masturbation. He also used it, according to the complainant, following their acts of homosexual intercourse The evidence of W was capable of amounting to corroboration of the complainant [26] Gleeson CJ suggested that, by leaving matters as the appellant's counsel had asserted to the jury, viz that there was no corroboration of H's testimony, the trial judge had been unduly favourable to the appellant. Any explanation of the concept of corroboration would not have been to the appellant's advantage. 1. R v BRS (unreported; Court of Criminal Appeal of NSW; 5 March 1996) at 16. 2. R v BRS (unreported; Court of Criminal Appeal of NSW; 5 March 1996) at 17 citing R v Hartley [1941] 1 KB 5 and R v Baskerville [1916] 2 KB 658.

As to the complaint that the trial judge had failed to explain to the jury the use which they could make of W's evidence, Gleeson CJ was of the opinion that leave should be refused to permit such a ground to be raised on appeal as such a direction had not been sought at trial. However, he went on to deal with its merits. The evidence having been admitted without objection, and being concededly admissible at least for the ground acknowledged at the trial, the total omission to mention W's evidence in the judge's summing up was viewed as strongly favourable to the appellant. The failure of trial counsel to seek an explicit direction made it, in the opinion of Gleeson CJ [27] :

impossible to escape the conclusion that the appellant was better off having the trial judge ignore the evidence of W, than explain to the jury the use they could make of it [I]f his Honour had instructed the jury correctly on that subject, he would have told the jury both that the evidence of W, if accepted, contradicted the evidence of good character, and went to nullify, or substantially reduce, the effect of that evidence and, in addition, in certain respects corroborated the evidence of the complainant. 1. R v BRS (unreported; Court of Criminal Appeal of NSW; 5 March 1996) at 20.

It is from the dismissal of the appeal, which followed these conclusions, that the appellant, by special leave, has appealed to this Court.

It is plain from the foregoing description of the trial that the parties were at odds concerning the bases upon which the evidence of W was admitted. At various times, the prosecutor propounded three bases for placing W's evidence before the jury. They were:

(a) To " rebut character evidence ". This was the first purpose, stated on the second day, when the calling of W was foreshadowed. There is a strong inference from the tender of W's statement at the committal and the way the trial unfolded, that the defence had considered this issue and discussed it with the prosecution before the trial began. In his closing address to the jury the prosecutor stated clearly that the evidence challenged "[the appellant's] evidence of good character".

(b) As corroboration of H's evidence. This purpose was also foreshadowed on the second day of the trial when the word "corroboration" was used. Subsequently, before the jury, in his closing address, the prosecutor referred to W's evidence as tending to "support", "confirm" and "link" the appellant to the course of the offences described by H.

(c) As propensity evidence. That is to show, from the similarity of the facts deposed to by H and W that the appellant had a predilection to committing the kind of offences complained of by H and that he was therefore guilty of the offences charged. In his closing address to the jury, the prosecutor suggested that the evidence of W showed "the type of person" the appellant was. He asserted that he had "a predilection and a liking for the company of young boys and for masturbation in their company" [28] . By inference, he was inviting the jury to infer from such " predilection ", if found on the basis of accepting W's evidence, the conclusion that the appellant had committed against H the offences contained in the indictment.

1. Prosecutor's address, 14 June 1995, tpt 20 (emphasis added).

The differences between the parties concerning the evidence of W were plain. The quandary presented to the jury by those differences was apparent, at least as soon as Judge Court took his own initiative to interrupt the trial and to express his acute misgivings about the reception of W's testimony. Obviously, the judge took that course because of his concern about the potential of W's evidence to cause prejudice to the fair trial of the appellant on the matters charged in the indictment.

The appellant's counsel made it plain that, so far as he was concerned, W's evidence was admissible only for purpose (a) above. He never conceded purposes (b) or (c). On the contrary, in his address to the jury, he emphatically denied the existence of corroboration, ie (b). He made no explicit mention of (c). In dialogue with the judge at the end of the summing up, he adhered to his position. He asked for directions concerning the lack of corroboration and the limited use which the jury could make of W's evidence. Despite Judge Court's expressed anxieties about the evidence, and the requests for redirection made by both sides, his Honour neither ruled on the purposes for which the evidence was admissible nor did he instruct the jury upon the use which they might make of it. The ultimate question in this appeal is whether these omissions caused the trial to miscarry in such a way as to occasion the risk of a miscarriage of justice.

The emphasis general rule, to which the common law has long adhered, is that the attention of a criminal trial must be focused throughout on the crimes alleged in the indictment or charge. However relevant to whether the accused committed those crimes might be an examination of his or her general character, past offences and discreditable conduct, evidence of such matters is generally excluded from a trial [29] . Instead, the trial concentrates attention upon the offences alleged against the accused. This approach has consequences for the process by which the offences are specified in the initiating document [30] , the right to particulars [31] , the conduct of the trial (including the admission and rejection of evidence [32] ) and the directions to a jury required of the trial judge [33] .

1. Dawson v The Queen (1961) 106 CLR 1 at 16, per Dixon CJ. 2. Quartermaine v The Queen (1980) 143 CLR 595 at 600-601; King v The Queen (1986) 161 CLR 423; Wilde v The Queen (1988) 164 CLR 365 at 372-373; S v The Queen (1989) 168 CLR 266, noted Criminal Law Journal, vol 14 (1990) 210; R v Jones (1995) 38 NSWLR 652 at 663; Walsh v Tattersall (1996) 188 CLR 77 at 92-93. 3. Johnson v Miller (1937) 59 CLR 467 at 485-486. 4. Johnson v Miller (1937) 59 CLR 467 at 495; B v The Queen (1992) 175 CLR 599 at 607-608. 5. Donnini v The Queen (1972) 128 CLR 114 at 137; B v The Queen (1992) 175 CLR 599 at 619-620.

The reasons for this strict approach have been explained many times. In Pfennig [34] , McHugh J outlined why the law generally excludes evidence of incidents other than those charged, where such evidence would tend to reveal the criminal or discreditable propensities of the accused:

One reason is that it creates undue suspicion against the accused and undermines the presumption of innocence. Another is that tribunals of fact, particularly juries, tend to assume too readily that behavioural patterns are constant and that past behaviour is an accurate guide to contemporary conduct. Similarly, "[c]ommon assumptions about improbability of sequences are often wrong", and when the accused is associated with a sequence of deaths, injuries or losses, a jury may too readily infer that the association "is unlikely to be innocent". Another reason for excluding the evidence is that in many cases the facts of the other misconduct may cause a jury to be biased against the accused Functional reasons also play a part in excluding evidence of bad character. Trials would be lengthened and expense incurred, often disproportionately so, in litigating the acts of other misconduct; law enforcement officers might be tempted to rely on a suspect's antecedents rather than investigating the facts of the matter; rehabilitation schemes might be undermined if the accused's criminal record could be used in evidence against him or her. 1. (1995) 182 CLR 461 at 512-513. Citations omitted.

The exclusion of such evidence, even though relevant, may therefore be seen as a consequence of the high degree of specificity of our criminal procedure. This is itself the product of the endeavour of the courts to avoid the risk of unfairness to an accused person. Such unfairness could arise quite readily if evidence of other conduct, criminal or simply discreditable, were admitted which tended to label the accused with a "bad character" from which a jury might reason to a conviction of guilt of the charges under consideration. Neasey J once observed, extra curially: "even Jeffreys J in Hampden's case in 1684 objected that "to rake into the whole course of a man's life is very hard"." [35]

1. Neasey, Similar Fact Evidence and Propensity Reasoning, Criminal Law Journal, vol 9 (1985) 232, at p 247 . He also referred to Sutton v The Queen (1984) 152 CLR 528 at 564, where Dawson J referred to the danger that a jury might proceed upon suspicion rather than proof.

The foregoing explanations for the strictness of the common law have lately been reinforced by two further developments. The first is the obligation expressed by international human rights law that a person "charged with a criminal offence shall have the right to be presumed innocent until proved guilty according to law" [36] . The application and development of Australian law on this topic should, so far as possible, conform to such principles of international law [37] . Secondly, the need for care in the admission and use of evidence about events prejudicial to an accused has been further reinforced by empirical studies. On the one hand, psychological research has tended to show that assumptions about the stability and "cross-situational consistency" in human behaviour are not validated by objective examination of such behaviour [38] . The research confirms the common tendency to infer from particular conduct character traits which are then used to justify predictions and estimates about other conduct. However, objectively, such predictions are frequently shown to be unwarranted. The need for special care at the point of the admission or rejection of prejudicial evidence is reinforced by other empirical studies. These tend strongly to challenge the presumption that jurors are willing and able to make the limited use of testimony which judges instruct them to do [39] . Once lay decision-makers know facts about the background and character of the accused, the risk is acute that the focus on the particular offences charged will be lost. Instead, the accused may be judged by reasoning that anyone shown to have acted in the criminal or discreditable way proved must be guilty of the offences charged, so long as they bear some similarity to the facts established by the evidence.

1. International Covenant on Civil and Political Rights, Art 14.2. 2. cf Mabo v Queensland [No 2] (1992) 175 CLR 1 at 42; Minister for Immigration and Ethnic Affairs v Teoh (1995) 183 CLR 273 at 291-292. 3. Neasey, Similar Fact Evidence and Propensity Reasoning, Criminal Law Journal, vol 9 (1985) 232, at p 240 ; Piragoff, Similar Fact Evidence (1981), p 292. 4. Schaefer and Hansen, Similar Fact Evidence and Limited Use Instructions: An Empirical Investigation, Criminal Law Journal, vol 14 (1990) 157, at p 179 ; cf Bates, Recent Cases on Corroboration, Criminal Law Journal, vol 11 (1987) 357.

Notwithstanding the foregoing, both by the common law and, now, by statute [40] , evidence of criminal or discreditable conduct on the part of an accused person, other than that precisely relevant to the subject charges, may sometimes be admissible. One such circumstance is to rebut evidence of good character tendered by and for the accused person. In such a case, the prosecution is permitted to call evidence to rebut that testimony in an attempt to give the whole picture. For the prosecution, this right is a shield, not a sword. It may not lead evidence of bad character just to demonstrate that the accused has a propensity to commit the offences charged. But it can bring evidence to rebut an attempt of the accused to place before the jury materials designed to show that there was no propensity (and thus no likelihood) of the accused's committing the offences charged because of proved good character.

1. Crimes Act 1900 NSW, s 413B (now repealed); cf Evidence Act 1995 Cth, ss 97, 104(4), 106(b), (e); Evidence Act 1995 NSW, ss 97, 104(4), 106(b), (e).

For the appellant it was conceded that the evidence of W was admissible against him as evidence for the Crown to rebut the evidence of good character and lack of criminal convictions which he tendered. The evidence of W was therefore placed before the jury. Any suggestion that it ought to have been excluded by the trial judge, acting on his own initiative because of the potential prejudice of such evidence, must be rejected. W's evidence was received without objection. The appellant's counsel estimated the value of adducing evidence of good character. He weighed this against the risks inherent in W's evidence once it got before the jury. His decision binds the appellant. But it was made on an expressed assumption that the use which the law would permit the jury to make of the evidence would be strictly limited and that the judge would so instruct the jury.

Equally clearly the prosecution's submission that the jury might use the evidence of W as corroboration of the evidence of H was stated at the trial. The suggestion that the prosecutor held back from using the word "corroboration" to the jury (and thus that judicial instruction on the meaning and application of corroboration were unnecessary) sits ill not only with the repeated references to "corroboration" in legal submissions to the judge but also with the words used in the prosecutor's address. Those words were "support and confirm" and "confirm and link". Clearly, they are synonyms for corroborate. The prosecutor himself asked the judge for a corroboration direction. Before this Court it was conceded for the Crown that it would have been preferable had a corroboration direction been given.

For the appellant it was objected that nothing in the evidence of W could amount to corroboration. The appellant took up the theme reflected in Judge Court's comments. He submitted that nothing that W had said directly confirmed any of the events recounted by H. W had not witnessed any of the alleged offences. So far as the lubricant and towel were concerned, they were consistent with reasonable hypotheses compatible with the innocence of the appellant. The lubricant is a product commonly available. The boys might have seen a towel, perhaps even under the bed, when they visited the appellant's room.

What is the principle? In Doney [41] , this Court said:

The essence of corroborative evidence is that it "confirms", "supports" or "strengthens" other evidence in the sense that it "renders [that] other evidence more probable" [42] . It must do that by connecting or tending to connect the accused with the crime charged in the sense that it "shows or tends to show that the story that the accused committed the crime is true, not merely that the crime has been committed, but that it was committed by the accused" [43] . 1. (1990) 171 CLR 207 at 211. 2. R v Kilbourne [1973] AC 729 at 758, per Lord Simon of Glaisdale. 3. R v Baskerville [1916] 2 KB 658 at 667.

The mere fact that the evidence of W was not directly related to the offences charged is no answer to the assertion that it was admissible as corroboration. Corroboration may be by circumstantial evidence [44] . The latter is evidence "which proves or tends to prove a fact or set of facts from which the fact to be proved may be inferred" [45] . It is not necessary that corroborative evidence, standing alone, should establish any proposition beyond reasonable doubt. The suggested value of W's evidence was that there were sufficiently common features in the appellant's conduct, as described, in relation to W to confirm, or tend to confirm, the truthfulness of the evidence of H. The commonalities included that both W and H, allegedly: 1. were young male pupils encouraged to come to a teacher's room where talk turned to sexual matters, particularly masturbation; 2. were encouraged to masturbate in the teacher's room; 3. were provided with lubricant for that purpose; 4. were told to use a towel kept under the bed for cleaning themselves, just as the teacher said he did for the same purpose.

1. R v Tripodi [1961] VR 186 at 190-191; Doney v The Queen (1990) 171 CLR 207 at 211 citing R v Baskerville [1916] 2 KB 658 at 667. 2. Doney v The Queen (1990) 171 CLR 207 at 211.

It is true, as Judge Court indicated, that thousands of people would have lubricant and towels in their bedrooms. In this sense, there were reasonable hypotheses for the possession of these objects consistent with the appellant's innocence [46] . However, when the other common features of the testimony of H and W are considered, they combine to afford evidence which a jury could (not must) view as making H's evidence for the prosecution significantly stronger, ie making its truthfulness much more probable.

1. Hodge's Case (1838) 2 Lewin 227 [168 ER 1136]; Plomp v The Queen (1963) 110 CLR 234 at 252; Barca v The Queen (1975) 133 CLR 82 at 104, 109; Knight v The Queen (1992) 175 CLR 495 at 503, 505; Cutter v The Queen (1997) 71 ALJR 638 at 641, 643; 143 ALR 498 at 501-502, 504.

To this extent, I agree with the conclusion of the Court of Criminal Appeal that the submissions of counsel for the appellant, both to the jury and to the trial judge, were wrong. There was a measure of corroboration of H's evidence in the testimony of W. A direction that there was not would therefore have been erroneous. Nonetheless, the corroboration afforded by the testimony exhibited several weaknesses. There were important differences between the conduct described by each of the boys. W recounted conversations with the appellant concerning his heterosexual experience. On W's evidence, the appellant did not attempt to observe him masturbating, still less to have any physical contact with him doing so. The lubricant and towel could be innocently explained as could the boys' knowledge of their presence in the teacher's room. Accordingly, although there was some corroboration, it was circumstantial and not specially strong. Had Judge Court correctly ruled in this way, it might have been expected that submissions would have addressed the proper terms of the instruction to the jury which would follow.

Understandably, counsel for the appellant did not seek any detailed directions on the differentiation between the relevance of corroboration to the several offences of which the appellant was charged. As everyone recognised, that would have been extremely confusing [47] . But both parties did seek a corroboration direction. It ought to have been given. If given properly, it would have upheld the prosecutor's submission that the jury could conclude that corroboration had, in some respects, been shown. But it would also have warned the jury of the limited use which they could make of the evidence of W as corroboration and the important points of distinction between the conduct described by W and the conduct described by H, the latter of which alone had given rise to the charges.

1. At common law the warning about the need for corroboration in cases of sexual offences against children based upon their testimony reached "almost the reverence of a rule of law": Kelleher v The Queen (1974) 131 CLR 534 at 553. For the history on the abolition of the requirement of a corroboration warning in New South Wales in cases of sexual assault see Crimes Act 1900 NSW, s 405C and Warner, Child Witnesses in Sexual Assault Cases, Criminal Law Journal, vol 12 (1988) 286. Notwithstanding the statutory abolition, a judicial warning may be necessary if the particular circumstances of the case so require: Longman v The Queen (1989) 168 CLR 79 at 86-88; cf R v Lubke (1988) 15 NSWLR 318 at 329-330.

So far, I have agreed generally in the approach of the Court below. I now reach the point of divergence. I do not agree that the requirement to direct the jury on the several uses which they might properly have made of the evidence of W could only have harmed the appellant's case so that he did not lose any real chance of acquittal. I reach this conclusion by taking four steps. First, it is not at all uncommon for judges, particularly in criminal trials, to be required to explain to juries the differing uses to which evidence, once admitted, may be put and to which such evidence may not be put. Secondly, this case was complicated by an explicit suggestion for the Crown that the jury could use the evidence of W to establish the "predilection" of the appellant, ie his propensity to commit the offences charged. Thirdly, and in any case, the great risks of the misuse of W's evidence as evidence of propensity required the provision of a firm judicial warning to the jury against the dangers of using the evidence in that way. Fourthly, the judicial obligations to clarify the way in which the evidence might be used was not discharged. This was despite the judge's own expressed misgivings and the submissions for the parties which adequately enlivened the requirement for a redirection about such use. The appellant did not have a trial as the law required. His convictions should therefore be set aside and a retrial ordered.

(a) Differentiated directions on admitted evidence. It is not uncommon for evidence to be admissible for particular purposes in a criminal trial but inadmissible for others. This is a problem "well known to the law" [48] . The problem which is then presented to a judge, conducting a trial with a jury, is not an easy one. Rather than exclude the evidence, the practical resolution of the problem is achieved by requiring the trial judge to instruct the jury concerning the purposes for which the evidence may, and may not, be used. If it were not so, serious risks of injustice could arise. This was recognised by this Court in Donnini v The Queen [49] and in B v The Queen [50] .

1. B v The Queen (1992) 175 CLR 599 at 619, citing Wigmore on Evidence (rev ed, 1983), vol 1, par 13. 2. (1972) 128 CLR 114; see also Attwood v The Queen (1960) 102 CLR 353. 3. (1992) 175 CLR 599 at 607-608, per Brennan J; at 619, per Dawson and Gaudron JJ.

The basis in legal policy for judicial directions to juries on the differential use of evidence admitted in a trial is the judge's obligation to assist the jury in the performance of their task. Without assistance, there could be a risk that a jury will act upon prejudice towards, or revulsion against, the accused [51] . They might fall into the trap of propensity reasoning, ie concluding that because the accused did another act, he or she must be guilty of the acts charged [52] . They might divert their attention from considering whether the prosecution has proved the crimes charged, as distinct from different acts which are not before the jury for trial.

1. R v Schlaefer (1992) 57 SASR 423; cf Wright, Case and Comment - Schlaefer, Criminal Law Journal, vol 16 (1992) 433, at p 435 ; R v Waters (unreported; Court of Criminal Appeal of NSW; 19 October 1990), noted Criminal Law Journal, vol 15 (1991) 222, at p 225; R v Dolan (1992) 58 SASR 501 at 503. 2. R v T (1996) 86 A Crim R 293; R v Johnson (unreported; Court of Appeal of Vict; 27 February 1997) at 8, per Callaway JA.

The judge should not invite the jury to act irrationally for such invitations will be ignored [53] . In a limited number of cases, propensity reasoning will be permitted [54] . But otherwise, the judge must assist the jury in the limited use to which the evidence may be put since the jury, uninstructed, are not likely to be aware of such considerations and of the need for particular care [55] .

1. B v The Queen (1992) 175 CLR 599 at 609, per Brennan J; cf Wright, Case and Comment - Zappala, Criminal Law Journal, vol 16 (1992) 190, at p 191 . 2. Hoch v The Queen (1988) 165 CLR 292 at 297, 301; Pfennig v The Queen (1995) 182 CLR 461 at 483-485, 488. 3. Driscoll v The Queen (1977) 137 CLR 517 at 536; see also Boardman v Director of Public Prosecutions [1975] AC 421 at 451, per Lord Hailsham; cf Bates, Of Beating and Bondage - Sex, Shame, and Similar Facts in Recent Law, Criminal Law Journal, vol 13 (1989) 117, at p 132 .

(b) Proclivity was put to the jury. The Crown denied that it had ever relied upon the evidence of W as evidence of the appellant's proclivity to commit the offences against H. However, such denial is scarcely consistent with the passage in the prosecutor's statement to the jury that W's evidence demonstrated that the appellant "had a predilection and a liking for the company of young boys and for masturbation in their company". This suggestion was reinforced by another, namely that W's evidence showed "the type of person [the appellant] was". The Crown alternatively submitted that no harm was done to the fair trial of the appellant because, in any case, the testimony of W was admissible as propensity evidence in conformity with the decisions of this Court in Hoch [56] and Pfennig [57] . There are a number of difficulties with that submission:

(1) At no stage prior to the suggestion in the closing address did the prosecution submit that the evidence of W was admissible as propensity or similar fact evidence. On the contrary, two different and more limited bases of admissibility were suggested and they were those stated above.

(2) Had it been submitted that the evidence of W was admissible to establish similar facts, from which the jury might infer a propensity or disposition on the part of the appellant to commit the crimes charged, certain procedural protections would have been available to the appellant. He could have sought a voir dire to contest the admission of the evidence on this ground. Notably, he could have sought to explore, as Hoch [58] suggests, whether there was a risk of concoction or collaboration between W and H to explain the striking similarity between their evidence. A voir dire will be appropriate: "to permit an assessment of the probative value of the evidence by reference to the consideration whether, in the light of common sense and experience, it is capable of reasonable explanation on the basis of concoction. It will not be for the purpose of the trial judge making a preliminary finding whether there was or was not concoction." [59] The present case has been approached, doubtless because of the sequence of the evidence and the chronology of the alleged offences, on the footing that the risk of concoction arose from the possibility that H told W about the appellant's sexual conduct. However, another possibility, consistent with the appellant's innocence of the offences, was that W told H about sexual conduct in the appellant's room and that H then adopted and embellished W's assertions to excuse his poor school and other performance. These are matters which might have been examined on a voir dire, had the prosecution clearly indicated its intention to rely upon similar fact evidence to establish a criminal propensity on the part of the appellant.

(3) If it had been suggested at the trial that the evidence of W was admissible for such a purpose, the judge would have been obliged to consider the elements that are prerequisite to the admission of propensity evidence. As the Crown bore the onus of establishing the admissibility of the evidence for the purposes which it propounded, it was the duty of the prosecutor to express those purposes clearly and if necessary to secure a ruling upon them. Had there been an explicit reliance upon similar fact and propensity evidence the appellant would have been afforded the opportunity of challenging the admissibility of the evidence on that basis and of securing a ruling from the judge. Moreover, the submissions to the jury for the appellant would doubtless have taken a different course.

(4) Most importantly, if the evidence of W were admissible to prove similar facts from which it could be inferred that the appellant had a criminal propensity in relation to offences of the kind charged, the trial judge would have been duty-bound, even when admitting such evidence, to instruct the jury on the use that might be made of that evidence and to warn them of the dangers of engaging in a "forbidden chain of reasoning" [60] .

1. (1988) 165 CLR 292. 2. (1995) 182 CLR 461. 3. (1988) 165 CLR 292 at 299-304. 4. Hoch v The Queen (1988) 165 CLR 292 at 297 (emphasis in original). 5. R v Johnson (unreported; Court of Appeal of Vict; 27 February 1997) at 13, per Callaway JA.

In the way the trial was conducted, and particularly because of the suggestions of the prosecutor in his closing address, there was a real risk that the jury were being invited to use W's evidence as demonstrating a criminal propensity on the part of the appellant. This was done without affording the appellant any of the safeguards which have been developed to prevent, or reduce, the risk of miscarriage of justice which propensity evidence can entail.

(c) The risk required a judicial warning. In Pfennig [61] differences appear in the reasoning of the majority of this Court and of McHugh J as to the approach which is to be taken to the admission of propensity evidence. No attempt was made in this case to reargue Pfennig. The facts here would, in any event, not provide a suitable vehicle for doing so, given that the evidence in question, being that of W, was concededly admissible for one other purpose [62] and, as I would hold, for yet another [63] . One common thread can be seen in most, if not all, of the cases where propensity evidence has been admitted. In such cases, the trial judge will ordinarily, even "invariably" [70] require the jury to consider very carefully the use which they make of the evidence of similar facts and to bear in mind the dangers of reasoning from a suggested propensity and their obligation to try the accused upon the counts of the indictment, not for other criminality or moral blemishes which those facts reveal. Thus in Donnini [71] it was the fact that the trial judge had given the jury a strong warning that they would not be entitled to draw any adverse inferences against the accused by reason of the volunteered information about his prior convictions that persuaded the majority of this Court [72] to confirm the conviction. A developing concern in the Court about the dangers of propensity reasoning can, I think, be seen in the course of decisions illustrated by Markby v The Queen [73] , Perry v The Queen [74] and Sutton v The Queen [75] . Perhaps as a reflection of this, a strong warning was given by the trial judge in Hoch [76] . However, it was not enough to save the convictions. In Pfennig strong and clear instructions were given to the jury by the trial judge [77] concerning the use which they might make of propensity evidence. That fact clearly influenced this Court in affirming the conviction in that case. In the present case, there was no such warning. None at all. The jury was left to fend for itself. This course was defended on the footing that counsel for the appellant had not specifically sought a direction, that there had been no undue or improper emphasis given to propensity at the trial [78] and that any elaboration of W's evidence by the judge would have given it an importance which the judge had sought to diminish by ignoring it altogether [79] . I turn finally to these considerations.

1. (1995) 182 CLR 461. 2. That is to rebut evidence of good character given for the accused. 3. That is corroboration of the evidence of the complainant (H). 4. See R v Beserick (1993) 30 NSWLR 510 at 516, per Hunt CJ at CL (Finlay and Levine JJ concurring). 5. (1972) 128 CLR 114. 6. Barwick CJ, McTiernan and Menzies JJ, Walsh and Mason JJ dissenting. 7. (1978) 140 CLR 108 at 117. 8. (1982) 150 CLR 580 at 586. 9. (1984) 152 CLR 528 at 533-534. 10. (1988) 165 CLR 292 at 299. 11. Pfennig (1995) 182 CLR 461 at 474-475, 497-499. 12. cf R v Martin (unreported; Court of Criminal Appeal of NSW; 18 July 1990), noted Criminal Law Journal, vol 15 (1991) 57, at p 59. 13. This was the view of the Court of Criminal Appeal; see R v BRS (unreported; Court of Criminal Appeal of NSW; 5 March 1996) at 20.

(d) The risk of injustice. The Court of Criminal Appeal did not expressly address the risk of injustice arising from the possibility that the jury had reasoned from a propensity on the part of the appellant demonstrated in W's evidence to a conclusion of his guilt upon H's allegations. The Court contented itself by addressing the general ground raised for the appellant that the directions given by the trial judge concerning the use which the jury might make of the evidence of W were inadequate.

It is important to take into account the statutory authorisation for briefer directions to juries of which Judge Court availed himself in this case. It is also necessary to remind oneself that a judge's direction to a jury must be considered in the context of the entire trial and in the light of the issues litigated and the submissions made [80] . This Court should exercise restraint before adding to the catalogue of warnings required of judges conducting criminal trials. If overlooked, or considered unnecessary in a particular trial, they may simply give rise to unmeritorious appeals. But, in certain fundamental matters, the judge carries an unavoidable obligation to provide directions which are necessary to prevent a perceptible risk of a miscarriage of justice [81] The risk must be a real, and not a fanciful one to attract the intervention of an appellate court [82] . It is not enough that a conclusion is reached that a particular warning would have been desirable. It must be shown to have been necessary in order to avoid the risk of a miscarriage of justice. If directions on the jury's legal duties are incorrect or seriously inadequate a new trial will ordinarily be ordered unless the appellate court can affirmatively conclude that the error or omission was in favour of the appellant and that no jury could have failed to convict had they been properly instructed [83] . This was the conclusion which the Court of Criminal Appeal reached. With respect, I disagree.

1. R v Tripodina & Morabito (1988) 35 A Crim R 183 at 191-195; R v Lawrence [1997] 1 VR 459 at 470-472. 2. Pemble v The Queen (1971) 124 CLR 107 at 116-118. 3. R v Johnson (unreported; Court of Appeal of Vict; 27 February 1997) at 17, per Callaway JA. 4. Pfennig v The Queen (1995) 182 CLR 461 at 542; R v Zorad (1990) 19 NSWLR 91 at 108-109; cf Odgers Case and Comment - Martin, Criminal Law Journal, vol 15 (1991) 57, at p 59 .

The risks of propensity reasoning in this case were high. Once W's evidence was placed before the jury, it would have been extremely difficult for them to conclude that the similarities and underlying pattern of conduct recounted both by W and H did not bolster the conclusion which the prosecutor expressly invited them to make. W's evidence, once accepted, tended strongly to show that the appellant was a particular "type of person". It demonstrated that he had the "predilection" described. Without instruction and warning, the risks of prejudice to the fair trial of the appellant upon the charges which the appellant actually faced were overwhelming. The risk was that he would be judged on his suggested character but without any assistance as to the limited way in which such evidence might properly be used. Whilst it would have been preferable had the appellant's trial counsel specifically raised these matters, he did so indirectly by repeatedly expressing his submission concerning the strictly limited basis upon which he suggested that the evidence might be used. It was the duty of the trial judge, then, to make accurate rulings on such use. It was a duty sufficiently enlivened by the conduct and necessities of the trial.

This conclusion is reinforced by several further features of the trial. One was the disquiet which led to the judge's intervention as W's evidence was adduced. Another was presented by the unresolved differences in the submissions of the parties about the use that might be made of the evidence. If that question was unresolved in the submissions of counsel, it was more than possible that it would be unclear in the minds of the jury. Finally, there is the remark on sentencing which indicated Judge Court's concern that a miscarriage might have occurred. Add to these the notorious dangers of propensity reasoning and the need for intervention in this case is plain.

The Crown sought to avoid this conclusion by pointing to the fact that, at the trial, the prosecutor expressly told the jury that they should not "sort of do it in reverse" [84] . However, the voice of the prosecutor in a trial is necessarily partisan. It is the voice of the judge that the jury needed to hear. The dangers of propensity evidence to the fair trial of an accused person are very great. Such evidence commonly introduces a real risk that the jury, uninstructed, will be diverted from attention to the offences charged, which is the mode of trial that, for good reason, our system of criminal justice adheres to. I consider that the warning should be "invariably" given about such dangers and the way to avoid them [85] . This is especially so where, as here, the evidence was admitted for limited reasons and not as similar fact evidence. A warning may not be able to overcome the prejudice of the propensity evidence. But it is the least that our procedures should offer to refocus the attention of the jury on their proper task and to that alone. The remarks of King CJ in R v Dolan [86] should be recalled:

The learned trial judge did not direct the jury as to the use which they could make, and the use which they could not make, of the evidence of the course of conduct. His Honour did not state the basis upon which the evidence of conduct other than that charged was admitted into evidence. I assume that it was admitted as indicating the relationship between the appellant and the alleged victim and as establishing the true context and setting in which the offences were alleged to have occurred. I consider that it was properly admissible on that basis

In all but the most exceptional cases in which evidence of criminal conduct other than that which is the subject of the charges before the court, is allowed to go before a jury, it is encumbent upon the trial judge to direct the jury as to the use which they are entitled to make of such evidence and, even more importantly in most cases, the uses to which such evidence must not be put. In a case such as the present, it is of the utmost importance that the jury should be clearly directed that the accused can be convicted on any count only if the jury is satisfied that the conduct which has been identified as the subject of that count occurred and that it is not permissible to convict the accused on the basis that, although the conduct so identified has not been proved, at least some conduct alleged by the alleged victim has occurred. The reason for allowing evidence of other incidents to go before the jury and the uses to which they can put such evidence should be clearly explained to them.

None of those directions were given in the present case. I think that the result is that the trial was fatally flawed and that the convictions cannot stand.

1. Prosecutor's address, 14 June 1995, tpt 20. 2. R v Beserick (1993) 30 NSWLR 510 at 516, per Hunt CJ at CL; cf R v T (1996) 86 A Crim R 293 at 299-300, per Southwell AJA. 3. (1992) 58 SASR 501 at 503.

Each and every word of King CJ applies to this case. The same result should follow. On 14 August 1997 the Court, having by majority come to this view, made orders allowing the appeal, quashing the convictions on counts 1, 2, 4, 5 and 6 and ordering a retrial. The foregoing are my reasons for joining in these orders.