//
© 2026 Zoe. All rights reserved.
Zoe is a legal information platform. Always consult the Federal Register of Legislation for authoritative text.
[2019] HCA 48
respondent. Appeal dismissed.
No linked legislation citations have been extracted yet.
The High Court decided that a special jury warning known as a Liberato direction is not automatically needed every time an accused person has given a denial to police that is played at trial. The direction is only needed if there is a real chance the jury might wrongly think they have to pick which story they like best or that they can only have a reasonable doubt if they actually believe the accused's version. In this rape case the overall instructions told the jury clearly and repeatedly that they must acquit if not satisfied beyond reasonable doubt, so the lack of the special direction did not create any unfairness and the conviction stands.
AI-generated legal information, not legal advice. Zoe can make mistakes. Check the cited source, and for advice about your situation consult a qualified Australian lawyer.
3,277 words · generated 24/04/2026
Neranjan Agrajith Kalubuth De Silva and the female complainant were acquaintances who independently arranged to stay overnight at the unit of a mutual friend, Neil, after a night spent at bars and nightclubs. Both had consumed alcohol, the complainant to the point of apparent intoxication, while the appellant showed no signs of intoxication. Upon arrival, sleeping arrangements were made: the complainant on the lounge couch, the appellant in the guest room, and Neil and his girlfriend Olivia in the main bedroom.
The complainant retired to the couch wearing a black long-sleeved t-shirt and underwear. The appellant sat with her, gave her a comforting hug as she was upset about a recent breakup, and she then lay down to sleep. The complainant gave evidence that she woke to the sensation of fingers being inserted into her vagina. She described herself as not completely coherent but trying to become fully conscious. The touching stopped, she lay very still, and then it happened a second time. These two sequential incidents formed the basis of two separate counts of rape by digital penetration without consent. After the second incident the complainant jumped up, began yelling, and burst into Neil's bedroom demanding that the appellant be removed. Neil's police statement (he having died before trial) recorded the complainant saying the appellant had "fingered me" while she was asleep. Olivia gave evidence of the complainant yelling at the appellant to get out while wearing only a loose shirt and underwear; the appellant responded by saying he had to go and left the unit. The complainant told Olivia the appellant had "tried to finger her".
In a recorded police interview admitted as part of the prosecution case, the appellant gave an exculpatory account. He said the complainant had spoken to him about her breakup, was naked and crying, and asked if he found her attractive and would date her. He replied that he had a girlfriend, at which point she "sort of freaked out". He denied any act of digital penetration. The appellant did not give or call evidence at his trial in the District Court of Queensland.
The prosecution case on both counts depended entirely on acceptance of the complainant's evidence. The trial judge gave conventional directions on the onus and standard of proof, including that the jury must acquit if left with a reasonable doubt. His Honour directed the jury on the use of the recorded interview, drawing on language from Burns v The Queen, telling them they must first accept that the appellant said the things relied on by the prosecution and that those things were accurate and true, and that it was up to them what weight to give both the incriminating and exculpatory parts. Specifically, the judge said the jury were "entitled to have regard to those answers, if you accept them" that pointed to innocence, while bearing in mind they had not been tested by cross-examination. No Liberato direction was sought or given. The jury returned a verdict of not guilty on the first count and guilty on the second.
The appellant appealed to the Court of Appeal of the Supreme Court of Queensland, arguing that the omission of a Liberato direction and the form of the interview directions caused a miscarriage of justice. Gotterson JA, giving the leading judgment, held that because the appellant had not given sworn evidence there was no conflict of oral testimony and therefore no need for a Liberato direction. The appeal was dismissed. Special leave was granted by the High Court on the question whether a Liberato direction is required if the defendant does not give evidence. The appellant ultimately advanced a second ground that the directions actually given were inadequate. By majority (Kiefel CJ, Bell, Gageler and Gordon JJ, Nettle J dissenting) the High Court held that while such a direction may be appropriate even without sworn evidence, this was not such a case. The summing-up as a whole adequately protected the appellant's right to the benefit of any reasonable doubt, and the appeal was dismissed.
The majority began by revisiting the circumstances that gave rise to the Liberato direction in Liberato v The Queen. There the trial judge had directed the jury in terms that invited them to decide "who do you believe" and had misstated the standard of proof in relation to the defence of honest belief in consent. Brennan J (with whom Deane J relevantly agreed) emphasised that even if the jury preferred the prosecution evidence they must not convict unless satisfied beyond reasonable doubt of its truth, and that even if they did not positively believe the defence evidence they could not reject it if it raised a reasonable doubt.
The majority noted that Murray v The Queen had since made clear that it is never appropriate to present the jury's task as a binary choice between competing accounts; the only question is whether the prosecution has proved the elements beyond reasonable doubt. Against that background, a Liberato direction serves to reinforce the onus and standard of proof where there is a real risk the jury will lose sight of that principle. Critically, the majority held that the direction is not confined to cases of conflicting sworn evidence. Intermediate appellate courts had differed on the point, but the majority stated unequivocally that if the trial judge perceives a real risk that the jury will reason that the accused's out-of-court answers can only create reasonable doubt if believed, or that preferring the complainant is enough to convict, then a Liberato direction should be given regardless of whether the account is on oath.
However, the majority emphasised that whether the direction is required "will depend upon the issues and the conduct of the trial". In the present case the trial was short, completed within two days, and contained repeated, unchallenged directions that the prosecution bore the onus of proving guilt beyond reasonable doubt and that any reasonable doubt required acquittal. The preliminary remarks to the jury after empanelment expressly reminded them of the presumption of innocence. Nothing in the summing-up, the summary of the respective cases, or counsel's addresses invited the jury to decide the matter on the basis of who to believe. Defence counsel's address concentrated on the complainant's intoxicated and emotional state and the consequent unreliability of her evidence, rather than inviting a contest of credibility between sworn witnesses.
The majority acknowledged that the trial judge's directions on the record of interview borrowed language from Burns v The Queen that was strictly applicable only to disputed confessional statements. The instruction to "accept" the exculpatory answers before giving them weight was, on one view, superfluous and potentially confusing. Yet the majority considered it would be reading too much into the passage to conclude that the jury would have understood "accept them" as meaning they must be satisfied of the truth of the answers before any use could be made of them. The immediately following direction—that it was "entirely up to you what use you make of them and what weight you give to them"—cured any ambiguity. Read in the context of the summing-up as a whole, the jury could not have been left with the impression that a preference for the complainant automatically proved guilt or that the appellant's account could only assist him if believed.
The absence of any request for a Liberato direction or redirection was not decisive but tended against a finding that a perceptible risk of miscarriage existed. The majority therefore concluded that the Court of Appeal had not erred; the trial had not miscarried.
Nettle J dissented on the basis that the immediate juxtaposition of the Burns direction concerning the incriminating parts of the interview with the direction on the exculpatory parts created a realistic possibility that the jury would understand the word "accept" in the same strong sense used throughout the summing-up—i.e., substantive acceptance as accurate and true. Because the final instruction left it "entirely up to you what use you make of them", the jury might have thought they could simply disregard the exculpatory account once they did not accept it as true. That would reverse the onus. Nettle J would have allowed the appeal, quashed the conviction and ordered a new trial. The majority, however, regarded that analysis as overly ambitious and not reflective of the overall effect of the charge.
Prior to De Silva the law was unsettled at intermediate appellate level as to whether a Liberato direction was ever appropriate when the accused's version appeared only in a record of interview rather than sworn testimony. Brennan J's original statement in Liberato v The Queen was expressly framed by reference to "a conflict between the evidence of a prosecution witness and the evidence of a defence witness", suggesting sworn evidence on both sides. Some courts had taken the view that the direction was therefore unnecessary where the accused had not become a witness. Others had taken a broader approach, recognising that the underlying risk to the correct application of the criminal standard could arise whenever a jury was presented with two competing versions, one on oath and one not.
The High Court in De Silva resolved that uncertainty. The majority made clear that the direction is not confined to cases of conflicting sworn evidence. The touchstone is the existence of a real risk that the jury will misapply the onus and standard of proof. At the same time, the Court stressed that the direction is not required in every case involving an exculpatory interview. After Murray v The Queen it is rarely appropriate to present the jury's task as a choice between accounts, and the expression "reasonable doubt" itself conveys that uncertainty should produce an acquittal. Therefore, in many cases conventional directions on onus and standard will suffice.
The decision also endorsed, as preferable, the three-step formulation suggested by Wheeler JA in Johnson v Western Australia: (i) if you believe the accused's account you must acquit; (ii) if you do not accept the account but it might be true you must acquit; (iii) if you do not believe the account you put it aside and ask whether the prosecution has proved guilt beyond reasonable doubt on the evidence you do accept. This framing avoids the potential confusion that arises if a jury that has wholly rejected the accused's evidence is nevertheless told it cannot find an issue against the accused if the evidence gives rise to reasonable doubt.
Post-De Silva the law is therefore more nuanced. Trial judges must assess whether, in the particular case, there is a perceptible risk of the prohibited reasoning. If there is, a Liberato (or Johnson-style) direction should be given even where the defence version is unsworn. If there is not—especially where the summing-up is replete with correct onus directions, contains no "who do you believe" language, and counsel does not seek the direction—the omission will not constitute miscarriage.
The foundation of the entire analysis is the passage quoted from Brennan J in Liberato v The Queen:
"When a case turns on a conflict between the evidence of a prosecution witness and the evidence of a defence witness, it is commonplace for a judge to invite a jury to consider the question: who is to be believed? But it is essential to ensure, by suitable direction, that the answer to that question (which the jury would doubtless ask themselves in any event) if adverse to the defence, is not taken as concluding the issue whether the prosecution has proved beyond reasonable doubt the issues which it bears the onus of proving. The jury must be told that, even if they prefer the evidence for the prosecution, they should not convict unless they are satisfied beyond reasonable doubt of the truth of that evidence. The jury must be told that, even if they do not positively believe the evidence for the defence, they cannot find an issue against the accused contrary to that evidence if that evidence gives rise to a reasonable doubt as to that issue."
Plain-English translation: Jurors will naturally ask themselves who they believe. That is fine, but the judge must make crystal clear that believing the complainant more than the accused does not automatically prove the case. The prosecution must still prove every element to the highest legal standard. Even if you do not believe the accused, if his or her version leaves you with a reasonable doubt you must acquit. Belief is not the same thing as proof beyond reasonable doubt.
The majority's key clarification appears in the following passage:
"If the trial judge perceives that there is a real risk that the jury will reason that the accused's answers in his or her record of interview can only give rise to a reasonable doubt if they believe them, or that a preference for the evidence of the complainant over the accused's account in a record of interview suffices to establish guilt, a Liberato direction should be given. Where the risk of reasoning to guilt in either of these ways is present, whether the accused's version is on oath or in the form of answers given in a record of interview, the Liberato direction is necessary to avoid a perceptible risk of miscarriage of justice."
Plain-English translation: The direction is not limited to cases where the accused gives evidence in the witness box. If there is a real danger the jury will only use the police interview to create doubt if they think it is true, or will convict simply because they prefer the complainant's story, the judge must give the direction. The label "on oath" or "not on oath" does not matter; the risk to a fair trial is what counts.
On the facts the majority concluded:
"The summing-up as a whole conveyed that the jury could not convict if the appellant's exculpatory answers in the interview left them with a reasonable doubt as to his guilt."
Plain-English translation: When you look at everything the judge told the jury, it was clear enough that any doubt raised by what the appellant said to the police meant they had to acquit. There was no unfairness.
Nettle J's dissenting concern was that the repeated use of "accept" throughout the summing-up to mean "accept as accurate and true", when applied to the exculpatory answers, risked reversing the onus. That forensic difference of view as to the overall impression left by the charge explains the split outcome.
De Silva will be engaged in any criminal trial, particularly sexual offence prosecutions, where the Crown case depends on the complainant's acceptance and the accused has given an exculpatory version in a recorded police interview that is placed before the jury. The trigger is the existence of a real risk that the jury will subconsciously treat the case as "word-on-word" or will only allow the interview to create reasonable doubt if they positively believe it.
Typical triggers include: (1) the judge or prosecutor uses language that could be understood as inviting a choice between the complainant's evidence and the interview account; (2) the summing-up contains no, or only generic, reinforcement of the criminal standard when dealing with the interview; (3) the exculpatory answers are contrasted with sworn testimony in a way that highlights they have not been cross-examined; or (4) the defence case is left on the basis that the jury must decide whether the complainant is lying or the appellant is lying.
Conversely, the precedent makes clear that the direction is not required where the summing-up is saturated with correct onus directions, the defence address focuses on the inherent weaknesses of the prosecution case rather than a credibility contest, and counsel does not request the direction. The fact that the jury in De Silva returned different verdicts on the two counts was treated as indicating they had carefully applied the beyond-reasonable-doubt standard to discrete factual questions rather than simply choosing one global narrative.
The decision is also relevant wherever an out-of-court exculpatory statement is admitted and the trial judge must decide how to direct on its use. The majority's approval of the Johnson three-step formulation gives trial judges a safe template that can be adapted to interview material.
The source judgment itself records that intermediate appellate courts had previously expressed "differing views" on whether a Liberato direction is appropriate when the defence version is not on oath. De Silva resolves that difference by making the existence of a real risk to the correct application of the standard of proof the decisive criterion, rather than the sworn/unsworn distinction. The majority's careful reading of the summing-up "as a whole" has reinforced the orthodox Australian approach that appellate courts do not parse individual sentences in isolation but assess the cumulative impression on the jury.
The judgment's treatment of Liberato itself is as a source of principle rather than a rigid precedent to be applied mechanically. Brennan J's direction is treated as a starting point that may need adaptation (the Johnson formulation is preferred). The citation of Murray v The Queen is used to reinforce that framing the jury's task as a choice is positively inappropriate, a proposition now cemented.
The distinction drawn between the Burns direction (appropriate for disputed confessions) and directions on exculpatory statements is likely to be influential. The majority's observation that applying the Burns two-step process to exculpatory material is "superfluous and potentially confusing" will caution trial judges against importing confessional-statement language into directions on denials.
Nettle J's dissent, while not carrying the day, highlights a forensic risk that judges must now guard against: the repeated use of words such as "accept" in a summing-up can colour the way a later reference to "accept them" in relation to exculpatory answers is understood. Future courts can be expected to cite both the majority and the dissent when analysing whether subtle linguistic choices created a miscarriage.
The majority expressly left open the precise boundaries of when a trial judge will be taken to have "perceived" a real risk requiring a Liberato direction. Is the test purely objective, or does it turn on the subjective perception of the particular trial judge? The judgment does not resolve that nuance.
A further open question is the significance of the jury's request for a redirection on the meaning of "beyond reasonable doubt". Nettle J thought the failure to give the contrast with the civil standard (as recommended in R v Dookheea) was surprising and compounded the problem. The majority did not address that point. Future cases may test whether a jury's expressed confusion on the standard makes the omission of a Liberato direction more likely to be fatal.
The judgment assumes that defence counsel's failure to seek a direction is relevant but not determinative. Exactly how much weight that failure carries when the risk is said to be "perceptible" remains unsettled. In a case where the risk is borderline, will an appellate court more readily find no miscarriage because counsel, who was present and heard the live atmosphere of the trial, did not ask for the direction?
Finally, the majority's endorsement of the Johnson three-step direction as preferable leaves open how strictly that template must be followed. Can a judge safely depart from it if the particular factual context makes one of the steps unnecessary or confusing? These residual questions mean that while De Silva has clarified the law, trial judges must still engage in a careful, case-specific risk assessment each time an exculpatory police interview is before the jury.
P J Callaghan SC with P Morreau and B P Dighton for the appellant (instructed by Robertson O'Gorman Solicitors)
M R Byrne QC with P J McCarthy for the respondent (instructed by Office of the Director of Public Prosecutions (Qld))
Notice: This copy of the Court's Reasons for Judgment is subject to formal revision prior to publication in the Commonwealth Law Reports.
Criminal practice - Trial - Directions to jury - Liberato v The Queen (1985) 159 CLR 507 ("Liberato") - Where appellant convicted by jury of rape - Where appellant did not give sworn evidence at trial - Where appellant made exculpatory statements in recorded police interview - Where record of interview admitted into evidence - Where appellant did not seek Liberato direction at trial - Where trial judge did not give Liberato direction - Whether Liberato direction required where accused does not give sworn evidence - Whether Liberato direction required where record of interview containing exculpatory statements admitted into evidence.
Words and phrases - "beyond reasonable doubt", "choice between witnesses", "conflicting version of events", "criminal standard", "evidence on oath", "exculpatory answers", "interview with the police", "jury directions", "Liberato direction", "onus and standard of proof", "out-of-court statement", "recorded interview", "summing-up as a whole", "sworn evidence", "who do you believe", "word-on-word".
KIEFEL CJ, BELL, GAGELER AND GORDON JJ. The appellant was arraigned in the District Court of Queensland (Judge Farr SC and a jury) on an indictment that charged him with two counts of rape. Each offence was alleged to have been committed on the same occasion against the same complainant. In each case the allegation was of digital penetration of the complainant's vagina without her consent. The prosecution case on each count was dependent upon acceptance of the complainant's evidence. The appellant did not give, or call, evidence. A recorded interview between the appellant and the police was in evidence in the prosecution case ("the interview"). In the interview, the appellant denied any act of digital penetration. The trial judge gave conventional directions, which are not the subject of complaint, as to the onus and standard of proof. His Honour was not asked to give, and did not give, a direction along the lines of the direction proposed by Brennan J in Liberato v The Queen (a "Liberato direction"). The jury returned verdicts of not guilty on the first count and guilty on the second count. The appellant appealed against his conviction to the Court of Appeal of the Supreme Court of Queensland (Fraser, Gotterson and Morrison JJA), contending that a "miscarriage of justice occurred by reason of the directions given as to how the jury should approach the evidence contained in the appellant's interview with police". The principal deficiency in the directions was said to be the omission of a Liberato direction. Gotterson JA, giving the leading judgment, noted that the jury had not been presented with conflicting oral testimony from the appellant. His Honour said that there had been no need for a Liberato direction. The appeal was dismissed. On 12 April 2019, Bell, Keane and Nettle JJ granted the appellant special leave to appeal on a single ground: "[t]he Court of Appeal erred in finding that a Liberato direction is not required if the defendant does not give evidence". On the hearing, the appellant acknowledged that a Liberato direction is not required as a matter of law. He sought leave to amend his Notice of Appeal to add a second ground: "[t]he Court of Appeal erred in failing to find that the directions given to the jury were inadequate and that as a result there was a miscarriage of justice". The Court heard full argument on the proposed second ground and reserved consideration of whether special leave to appeal would be granted to rely on it. For the reasons to be given, while it may, in some cases, be appropriate to give a Liberato direction notwithstanding that the accused's conflicting version of events is not before the jury on oath, this was not such a case. The Court of Appeal was correct to find that the summing-up as a whole conveyed that the jury could not convict if the appellant's exculpatory answers in the interview left them with a reasonable doubt as to his guilt. In circumstances in which the proposed second ground of appeal is bound up with consideration of the first, it is appropriate to grant special leave to rely on the appellant's Amended Notice of Appeal but the appeal must be dismissed. Liberato Before turning to the terms of the Liberato direction, there should be some reference to the circumstances that gave rise to it. In Liberato, several accused were jointly tried for the rape of the complainant. Each accused admitted to having engaged in the act, or acts, of sexual intercourse with which he was charged, but claimed to have believed that the complainant was consenting. The summing-up contained defects: the jury were directed that an accused was entitled to be acquitted if they were satisfied that he had an exculpatory belief as to consent, and if that "[gave] rise to a doubt" as to the accused's guilt. In addition to this serious misdirection, on three occasions the trial judge identified the issue for the jury's determination as "who do you believe". After full argument, special leave to appeal was refused in Liberato. In the view of the majority, it had been open to the Court of Criminal Appeal to find that, notwithstanding the acknowledged defects in the summing-up, no substantial miscarriage of justice had actually occurred. It is Brennan J's dissenting reasons which are the source of the Liberato direction: "When a case turns on a conflict between the evidence of a prosecution witness and the evidence of a defence witness, it is commonplace for a judge to invite a jury to consider the question: who is to be believed? But it is essential to ensure, by suitable direction, that the answer to that question (which the jury would doubtless ask themselves in any event) if adverse to the defence, is not taken as concluding the issue whether the prosecution has proved beyond reasonable doubt the issues which it bears the onus of proving. The jury must be told that, even if they prefer the evidence for the prosecution, they should not convict unless they are satisfied beyond reasonable doubt of the truth of that evidence. The jury must be told that, even if they do not positively believe the evidence for the defence, they cannot find an issue against the accused contrary to that evidence if that evidence gives rise to a reasonable doubt as to that issue." Deane J agreed with Brennan J that the directions in Liberato were confusing because they left the impression that the jury's task was essentially one of making a choice between the differing prosecution and defence accounts. His Honour also agreed that it was commonplace for judges to invite the jury to consider which of the conflicting accounts they believed. Indeed, his Honour said that express or implied references in a summing-up to a "choice" between witnesses were "sometimes unavoidable and commonly unobjectionable". His Honour did not consider that this posed difficulty, provided that the reference was accompanied by clear and unequivocal directions about the onus and standard of proof. Whatever may have been the practice when Liberato was decided, in Murray v The Queen this Court made clear that it is never appropriate for a trial judge to frame the issue for the jury's determination as involving a choice between conflicting prosecution and defence evidence: in a criminal trial the issue is always whether the prosecution has proved the elements of the offence beyond reasonable doubt. In light of Murray, the occasions on which a jury will be invited to approach their task as involving a choice between prosecution and defence evidence should be few. This is not to say that the occasions calling for a Liberato direction should be few. The Liberato direction serves to clarify and reinforce directions on the onus and standard of proof in a case in which there is a risk that the jury may be left with the impression that the evidence on which the accused relies will only give rise to a reasonable doubt if they believe it to be truthful, or that a preference for the evidence of the complainant suffices to establish guilt. Subject to statute, a Liberato direction should be given in a case in which the trial judge perceives that there is a real risk that the jury might view their role in this way. The Liberato direction is addressed, in terms, to a trial at which there is conflicting sworn evidence. Intermediate appellate courts have expressed differing views as to whether a Liberato direction is appropriate in a case in which the conflicting defence version of events is not given on oath, but is before the jury, typically in the accused's answers in a record of interview. If the trial judge perceives that there is a real risk that the jury will reason that the accused's answers in his or her record of interview can only give rise to a reasonable doubt if they believe them, or that a preference for the evidence of the complainant over the accused's account in a record of interview suffices to establish guilt, a Liberato direction should be given. Where the risk of reasoning to guilt in either of these ways is present, whether the accused's version is on oath or in the form of answers given in a record of interview, the Liberato direction is necessary to avoid a perceptible risk of miscarriage of justice. When an accused gives, or calls, evidence there is a natural tendency for the focus to shift from the assessment of the capacity of the prosecution case to establish guilt to an assessment of the perceived strengths or weaknesses of the defence case. Recognition of this forensic reality suggests that the risk that the jury will reason in either of these ways is more likely to arise in a trial in which the conflicting defence account is on oath. In Johnson v Western Australia, Wheeler JA identified one possible shortcoming in using Brennan J's statement in Liberato as a template for the direction: a jury may completely reject the accused's evidence and thus find it confusing to be told that they cannot find an issue against the accused if his or her evidence gives rise to a "reasonable doubt" on that issue. For that reason, it is preferable that a Liberato direction be framed along the following lines: (i) if you believe the accused's evidence (if you believe the accused's account in his or her interview with the police) you must acquit; (ii) if you do not accept that evidence (account) but you consider that it might be true, you must acquit; and (iii) if you do not believe the accused's evidence (if you do not believe the accused's account in his or her interview with the police) you should put that evidence (account) to one side. The question will remain: has the prosecution, on the basis of evidence that you do accept, proved the guilt of the accused beyond reasonable doubt? Whether a Liberato direction is required will depend upon the issues and the conduct of the trial. At a trial where there has been no suggestion, whether express or implied, that the jury's determination turns on which of conflicting prosecution and defence versions is to be believed, there may be no need to expand on conventional directions as to the onus and standard of proof. As Wheeler JA also observed in Johnson, the expression "reasonable doubt" is apt to convey that a juror who is left in a state of uncertainty as to the evidence should not convict. The facts On the evening of the subject events, the appellant and the complainant, who were acquaintances, made independent arrangements to stay overnight at a mutual friend's unit. Both had attended bars and nightclubs earlier in the evening. By the time they, and their mutual friend, Neil, returned to his unit, the complainant appeared to be intoxicated. The appellant had consumed some alcohol but he did not show signs of intoxication. Neil's girlfriend, Olivia, was present at the unit when the three arrived. It was arranged that the complainant would sleep on the couch in the lounge room and the appellant would sleep in the guest room. Olivia and Neil slept in his bedroom. The complainant got ready for bed, taking off her white t-shirt and putting on a black long-sleeved t-shirt. She removed her skirt but kept on her underwear. She was upset over the breakup of a relationship. The appellant sat on the couch with her and gave her a hug in a comforting manner. She then lay down to go to sleep. The complainant said that she woke to feel fingers being inserted into her vagina. She explained that she "wasn't completely coherent just yet". She had tried to "bring [herself] completely conscious to comprehend what was going on". It stopped and the complainant said she was "very still, trying to comprehend what I thought I had just felt". This was the incident charged in the first count, on which the jury returned a verdict of not guilty. Then the complainant said she felt fingers penetrating her vagina again. This was the incident charged in the second count, on which the jury returned a verdict of guilty. The complainant jumped off the couch and started yelling. Neil had died before the trial. A statement that he gave to the police was in evidence. In this he described the complainant bursting through the door of his bedroom saying "[g]et him out. Get him out". He said that he had walked out of the bedroom and seen the appellant mumbling, "[s]he's crazy. I'm leaving". The complainant was crying and she said "[h]e fingered me. What sort of person does that whilst someone is asleep". Olivia gave evidence of walking into the lounge room, where she saw the complainant yelling and shouting and telling the appellant to "get out". The complainant was wearing a loose shirt and black underwear. The appellant said "I've got to go" and left the unit. Olivia endeavoured to calm the complainant down. The complainant said that the appellant "tried to finger her". In the appellant's interview, he gave an account that the complainant had talked to him about breaking up with her boyfriend. She was naked and crying and complaining that there was no one in her life. He cuddled her to comfort her. She had asked him if he found her attractive and whether he would date her. He replied that he had a girlfriend and at this point, the complainant "sort of freaked out". The summing-up The trial judge gave clear, correct directions on the onus and standard of proof. These included the instruction that: "It is for you to decide whether you are satisfied, beyond reasonable doubt, that the prosecution has proved the elements of the offence. If you are left with a reasonable doubt about guilt in respect of either charge, your duty is to acquit in respect of that charge. That is, to find the defendant not guilty. If you are not left with any such doubt, then your duty is to convict. That is, to find him guilty." His Honour observed that matters which would concern the jury no doubt would include the credibility and reliability of evidence, particularly that of the complainant. In directing the jury with respect to the fact that the appellant had not given or called evidence, his Honour twice reminded them that the onus was on the prosecution to establish guilt to the criminal standard. The instruction respecting the need for separate consideration of each count was accompanied by a direction of the kind discussed in R v Markuleski, namely, that a reasonable doubt as to the truthfulness or reliability of the complainant's evidence in relation to one count may be taken into account in assessing the truthfulness and reliability of the complainant's evidence of another count. The focus of the appeal is on the directions concerning the interview. It is necessary to set them out at some length: "You have also before you the evidence of the defendant's interview with the police officers and the prosecution relies on some of the answers said to have been given by the defendant in that interview as supporting its case against him. … There are two issues, I suppose, that arise from that recording, insofar as those statements that the prosecution rely upon as suggesting it is in some way supportive of its case against him … The first thing, of course, is that you must accept that the defendant said such things. … Well, this is a recording. You can listen to that recording and that is a matter that you can work out for yourselves by simply listening to the tape. The second part of that is that you would have to, of course, conclude that what he said in those statements were accurate and true. So it is up to you to decide whether you are satisfied that those things said by the defendant, which the prosecution says - or submits to you are supportive of its case against him - were said by the defendant, and secondly, whether they were accurate and true. And, of course, it is a matter for you as to whether you - as to what weight you give to them. That is, whether they do support the prosecution case in any way, whether they do not, that is entirely a question of fact for yourselves. ... Now, in that interview he also gave answers which you might view as indicating his innocence. You should know, ladies and gentlemen, that you are entitled to have regard to those answers, if you accept them, and to give them whatever weight you think appropriate. Bearing in mind, of course, that they have not been tested by cross-examination. So in relation to both the answers which the prosecution relies upon as being supportive of its case against him, and those which point to innocence, it is entirely up to you what use you make of them and what weight you give to them." (emphasis added) The trial judge reminded the jury that defence counsel had suggested a possible motive for the complainant to lie in her account of the alleged offences. His Honour directed that if the jury rejected the suggested motive, it did not follow that the complainant was necessarily telling the truth. His Honour again reminded the jury that it was for the prosecution to satisfy them that the complainant was telling the truth because the prosecution bore the burden of establishing guilt beyond reasonable doubt. The Court of Appeal In the Court of Appeal, the appellant complained that the directions concerning the interview were misleading in two respects, each of which had the tendency to undermine the directions on the onus and standard of proof. First, the instruction that the appellant's answers might be viewed as "indicating his innocence" was apt to suggest that the jury's inquiry was directed to innocence as distinct from proof of guilt. Secondly, the instruction respecting the answers in the interview on which the appellant relied that "you are entitled to have regard to those answers, if you accept them, and to give them whatever weight you think appropriate" was apt to reverse the onus. These deficiencies, it was submitted, underscored the necessity that the jury be given a Liberato direction. Gotterson JA noted that Brennan J's statement in Liberato was addressed to circumstances in which the trial judge had summed up the case "as one of choice between prosecution and defence evidence". Such a suggestion had not been made at the appellant's trial. While the appellant's exculpatory answers in the interview were evidence in the case, the appellant had not become a witness and the jury had not been presented with a conflict in oral testimony such as to characterise the case as a "word-on-word" case. In the circumstances, his Honour said that there was no requirement for a Liberato direction. Gotterson JA did not consider that the directions concerning the record of interview were misleading. The reference to answers as "indicating his innocence" was descriptive of the answers as exculpatory, and did not convey that the jury's task was to determine innocence. The instruction "if you accept them" with respect to those exculpatory answers echoed the instruction at the commencement of the directions concerning the record of interview, which, consistently with Burns v The Queen, directed the jury to consider first whether to accept that the answers were given and, secondly, to consider whether they were truthful. Gotterson JA considered that the trial judge's concluding direction concerning the record of interview - that it was entirely up to the jury what use they made of the appellant's answers and what weight they gave to them - aligned with the general directions on the onus and standard of proof. His Honour concluded that the summing‑up as a whole conveyed that the jury could not convict if the appellant's exculpatory answers left them with a reasonable doubt about the appellant's guilt. The submissions The appellant's case in this Court is that it is prudent to give a Liberato direction in most, if not all, cases in which there is evidence of the conflicting defence account of material events. He complains that the instructions given to the jury at his trial were generic and not adapted to the circumstances of the case. Specifically, they did not ensure that the jury understood that a preference for the evidence of the complainant did not preclude a verdict of not guilty. Nor did the directions make clear that disbelieving the appellant's version was no bar to a verdict of not guilty. The appellant maintains that the directions should not have invoked the concept of "innocence", or been expressed in terms that contemplated "acceptance" as a prerequisite to the use of the evidence of his answers given in the interview. The respondent did not take issue below or in this Court with the proposition that a Liberato direction could have been given in a case, such as this, in which the accused did not give evidence. The issue, however, in the respondent's submission, is whether the Court of Appeal erred in concluding that, on the whole of the summing-up, there was no miscarriage of justice in this case. The respondent points to the absence at the appellant's trial of the circumstances that Brennan and Deane JJ suggested called for a specific direction in Liberato: the trial did not involve starkly opposed sworn evidence, nor did the judge invite the jury to consider "who do you believe". Consideration This was a short trial in which the evidence, counsel's addresses and the summing-up were completed by lunchtime on the second day. The trial judge addressed some preliminary remarks to the jury after they were empanelled. These included the instruction that: "A defendant in a criminal trial is presumed to be innocent. So before you may return a verdict of guilty on either charge, the prosecution must satisfy you that the defendant is guilty of the charge in question and must satisfy you of that beyond reasonable doubt." As detailed earlier, the jury were given repeated, correct directions as to the onus and standard of proof in the course of the summing-up. Nothing in the trial judge's summary of the way the respective cases were put, or in the way his Honour summed up, suggests that the jury might have been left with the impression that their verdicts turned on a choice between the complainant's evidence and the appellant's account in the interview. The focus of defence counsel's address was on the suggested incapacity of the prosecution case to support a finding of guilt beyond reasonable doubt in light of the complainant's intoxicated and emotional state. As noted, the directions concerning the use to be made of the appellant's answers in the interview commenced with a direction taken from Burns as to the necessity to be satisfied that the appellant gave the answers and, if so, to be satisfied that those answers were accurate and true. The statements in Burns were made in the context of a dispute as to the making of a confessional statement. The function of the summing-up is to meaningfully assist the jury to decide the real issues in the case. It should not contain formulaic directions unconnected to the determination of those issues. The instruction, that the jury first satisfy themselves that the answers on which the prosecution relied were given, was superfluous and apt only to distract: no party was suggesting that the appellant had not given the answers attributed to him in the electronically recorded interview. Whether, as the Court of Appeal reasoned, the words "if you accept them", when the trial judge came to the answers in the interview on which the appellant relied, would have been understood by the jury as meaning accepting that the answers were given, the direction was superfluous and potentially confusing. The requirement that jurors first satisfy themselves that the accused made an out‑of‑court statement is directed to disputed confessional statements and not exculpatory statements. It is, however, to read too much into the impugned passage to take from it, as the appellant submits, that the jury would have understood the direction as meaning "if you accept the truth of" the answers. The further instruction to give them "whatever weight you think appropriate" would make no sense were the jury to understand that a prerequisite to any use of the answers was satisfaction of their truth. Any risk of confusion was overcome by the further, correct, instruction: "So in relation to both the answers which the prosecution relies upon as being supportive of its case against him, and those which point to innocence, it is entirely up to you what use you make of them and what weight you give to them." In context, the trial judge's reference to answers in the interview "indicating [the appellant's] innocence" was, as the Court of Appeal held, descriptive of the answers as exculpatory. The submission, that the reference to innocence and the inclusion of the words "if you accept them" in these two sentences of the summing-up undermined the clarity of the directions on the onus and standard of proof, is overly ambitious. Defence counsel appears not to have perceived any such risk. There was no request for any redirection on the use the jury might make of the answers in the interview. Nor did defence counsel seek a Liberato direction. The failure of counsel to seek a direction is not determinative against successful challenge in a case in which the direction was required to avoid a perceptible risk of the miscarriage of justice. The absence of an application for a direction may, however, tend against finding that that risk was present. The summing-up made clear the necessity that the jury be satisfied beyond reasonable doubt of the complainant's reliability and credibility. The Court of Appeal did not err in concluding that, when the summing-up is read as a whole, the trial did not miscarry by reason of the omission of a Liberato direction. Orders For these reasons, there should be the following orders:
NETTLE J. As the majority observe, a trial judge's summing up should be tailored to the issues and avoid the recitation of irrelevancies. Presumably in pursuit of that objective, the trial judge gave the jury a Burns direction regarding the use that it was open to them to make of what were said to be incriminatory sections of the appellant's record of interview. That direction was as follows: "You have also before you the evidence of the defendant's interview with the police officers and the prosecution relies on some of the answers said to have been given by the defendant in that interview as supporting its case against him. The prosecution have referred you specifically to the statements that he made that you can hear on the recording, where he speaks about his close proximity to the complainant in the lead‑up to the incident the subject of the charges. As I said, that recording will be with you when you retire to consider your verdicts and you can play it again as often as you wish. There are two issues, I suppose, that arise from that recording, insofar as those statements that the prosecution rely upon as suggesting it is in some way supportive of its case against him on the issue that I have just identified. The first thing, of course, is that you must accept that the defendant said such things. That is, gave answers about his proximity to the complainant in the lead‑up to these two alleged events. Well, this is a recording. You can listen to that recording and that is a matter that you can work out for yourselves by simply listening to the tape. The second part of that is that you would have to, of course, conclude that what he said in those statements [was] accurate and true. So it is up to you to decide whether you are satisfied that those things said by the defendant, which the prosecution says - or submits to you are supportive of its case against him - were said by the defendant, and secondly, whether they were accurate and true. And, of course, it is a matter for you as to whether you - as to what weight you give them. That is, whether they do support the prosecution case in any way, whether they do not, that is entirely a question of fact for yourselves." That direction was unexceptionable in itself. But there are some aspects of it that should be noted for reasons that will later appear. The first is the expression: "you must accept that the defendant said such things". That would have conveyed to the jury that, in order to have regard to the allegedly incriminatory answers, the jury had to be satisfied that the appellant in fact gave the allegedly incriminating answers. The second aspect is the use of the expression: "[t]he second part of that is that you would have to ... conclude that what he said in those statements [was] accurate and true" (emphasis added). That would have conveyed to the jury that the second matter of which they had to be satisfied before they could have regard to the incriminatory answers was that the incriminatory answers were "accurate and true". The third aspect of the direction is the trial judge's explication of the concept of weight, namely: "whether [the allegedly incriminating answers] do support the prosecution case in any way". As so expressed, that would have conveyed to the jury that the concept of weight stands separate and apart from the two conditions earlier explained of accepting that the appellant gave the allegedly inculpatory answers and that those answers were "accurate and true". Hence, as the concept of weight was so explained, the jury would have understood weight in this context to go to probative effect rather than credibility or reliability. Immediately after the Burns direction, the trial judge directed the jury in relation to the exculpatory aspects of the record of interview. That direction was as follows: "Now, in that interview he also gave answers which you might view as indicating his innocence. You should know, ladies and gentlemen, that you are entitled to have regard to those answers, if you accept them, and to give them whatever weight you think appropriate. Bearing in mind, of course, that they have not been tested by cross‑examination. So in relation to both the answers which the prosecution relies upon as being supportive of its case against him, and those which point to innocence, it is entirely up to you what use you make of them and what weight you give to them." There are several aspects of that direction that should also be noted. The first is that it would have conveyed to the jury that they were not entitled to make any use of any exculpatory answers given in the record of interview unless they "accept[ed] them". The second aspect is that, although the judge did not here spell out in terms that, in order to have regard to the exculpatory answers, the jury had to accept that the appellant in fact gave the exculpatory answers and, as a "second part of that", accept that those answers were "accurate and true", it is not unlikely that the jury would have understood the judge's use of the short-hand expression "accept them" as intended to convey exactly the same two conditions as had been outlined in the course of the immediately preceding Burns direction. The third aspect of the direction is that, here, as in the Burns direction, the judge referred to "weight" separately and apart from the notion of "accepting them" and, although his Honour did not here repeat in terms that weight means "whether they [the exculpatory answers] do support the [defence] case in any way", it is not unlikely that the jury would have taken that to be the judge's meaning, just as it was in the immediately preceding Burns direction. Seen, therefore, in the immediate context of the Burns direction, the net result of the direction on the exculpatory aspects of the record of interview may well have been to leave the jury with the impression that they were not permitted to have regard to the exculpatory answers unless they accepted both that those answers were given and that those answers were "accurate and true". The likelihood of that being so is fortified by the fact that, throughout the summing up, the judge repeatedly used the word "accept" to refer to a substantive acceptance, or non‑acceptance, by the jury of particular evidence as being "accurate and true". Thus for example: . "You are to determine the facts of the case based on the evidence that has been placed before you in this courtroom. That involves deciding what evidence you accept." (emphasis added) . "you are not obliged to accept any comment I might make about the evidence". (emphasis added) . "It is for you to decide whether you accept the whole of what a witness says or only part of it or none of it. You may accept or reject such parts of the evidence as you think fit. It is for you to judge whether a witness is telling the truth and correctly recalls the facts about which he or she has testified." (emphasis added) . "Now, many factors may be considered in deciding what evidence you accept ... [Y]ou have seen how the witnesses presented in the witness box when answering questions ... [and] you should consider perhaps the likelihood of the particular witness's account or whether the evidence of a particular witness seemed reliable when compared with other evidence that you accept." (emphasis added) . "the fact that we refer to [some] witnesses as expert does not mean that their evidence has automatically to be accepted, although you might think in this matter there is little reason to not accept their evidence". (emphasis added) . "you are entitled to assess and accept and reject any such opinion evidence as you see fit". (emphasis added) . "It is a matter for you, as the sole judges of the facts, whether you accept the evidence relating to the complainant's distressed condition". (emphasis added) As has been seen, there was very little in or about the direction on the exculpatory aspects of the record of interview that might have conveyed to the jury that the word "accept" was used in relation to the exculpatory aspects of the record of interview in any sense different from the repeated applications of "accept" throughout the summing up as referring to a substantive acceptance, or non-acceptance, of particular evidence as being "accurate and true". Further, as was earlier noticed, at the conclusion of the direction on the exculpatory aspects of the record of interview the judge directed the jury that, if they did "accept" the exculpatory answers, it was entirely up to them what use they made of them and what weight they gave them. In that context, that direction was misleading. The law is that, if the jury believed the appellant's account of what occurred, they were bound to acquit, and, even if they did not accept his account, but considered it was possible that it might be correct, they were bound to acquit. It was only if they rejected his version of what occurred that they were entitled to ignore it. A properly structured Liberato direction delivered in the terms adumbrated in the majority's reasons would have safely conveyed that to the jury and, by doing so, gone far to overcome any misunderstanding, likely to have resulted from the direction on the exculpatory aspects of the record of interview, that the jury were not entitled to have regard to the exculpatory answers unless they accepted them as being "accurate and true". Instead, the admonition that it was entirely up to the jury what use they made of the exculpatory answers, in the context in which it was delivered, very likely compounded the problem. Finally, it might not be without significance that the jury sought a re-direction as to the meaning of beyond reasonable doubt. It is, therefore, a little surprising that, although the judge was aware of this Court's decision in R v Dookheea, his Honour chose not to adopt the response there recommended of contrasting the standard of proof beyond reasonable doubt with the lower civil standard of proof on the balance of probabilities in order to convey to the jury that being satisfied of guilt beyond reasonable doubt does not simply mean concluding that the accused may have committed the offence charged, or even that it is more likely than not that the accused committed the offence charged; that what is required is a much higher standard of satisfaction, the highest known to the law: proof beyond reasonable doubt. By itself, such a direction would not have saved the situation, but it would have assisted. In the result, I consider that there is a realistic possibility that the jury approached their task on the basis that they were to disregard the appellant's exculpatory answers unless they were persuaded that they were "accurate and true". So to approach their task would have been wrong in law and likely to have resulted in a conviction where, if properly directed, the jury may have acquitted. I consider that the appellant was thus deprived of a realistic chance of acquittal to which he was entitled and so subjected to a substantial miscarriage of justice. Conclusion It follows, in my view, that the appeal should be allowed. The order of the Court of Appeal should be set aside and in its place it should be ordered that the appeal to the Court of Appeal be allowed, the conviction be quashed and a new trial be had.
De Silva
The Queen
(1985) 159 CLR 507
Appeal dismissed.