The King v AR

1. [1]

   GAGELER CJ, GLEESON AND BEECH-JONES JJ. The respondent was tried in the District Court of New South Wales on an indictment charging him with committing seven sexual offences against the same complainant ("GC"). GC was 10 years old at the time of the alleged offences. The respondent was convicted of three of the charged offences, found not guilty of two charges, and no verdict was taken on the remainder as they were charged in the alternative.

2. [2]

   At the respondent's trial, the Crown contended that the evidence in support of the charged acts established a tendency on the part of the respondent to have a sexual interest in GC and to act on that sexual interest by committing a particular sexual act against her when she had fallen asleep beside the respondent, while watching a movie. The trial judge directed the jury in terms that identified the evidence said to support the alleged tendency as being the evidence of the charged acts, and told the jury that, if established, the tendency could be used in considering whether it was more likely that the respondent had committed the charged acts ("the tendency direction"). The tendency direction also invited the jury, in determining whether the Crown had established the tendency, to make findings in respect of the charged acts to a lesser standard than the criminal standard. In Director of Public Prosecutions (Vic) v Benjamin Roder (a pseudonym), this Court observed that trial judges "should" avoid directing the jury to that effect to minimise the risk of undermining the jury's understanding of the necessity for the prosecution to prove the elements of the offence beyond reasonable doubt.

3. [3]

   Relevantly, the New South Wales Court of Criminal Appeal granted the respondent leave to appeal against his convictions, allowed the appeal, set the convictions aside and ordered a retrial. That Court concluded that a miscarriage of justice was established because the tendency direction "viewed in the context of the summing up as a whole" was likely to have "deflected [the jury's attention] from applying the required standard of proof in respect of the offences charged". In so concluding, that Court found that the formulation of the tendency in a manner that "replicated the detail of how the offences were allegedly committed" and that relied on the evidence of the charged acts as the basis for the tendency was "inconsistent with the nature of tendency evidence". The Court also found that the tendency direction, reflecting that formulation of the alleged tendency, "encourage[d], if not require[d], the jury to engage in ... impermissible circular reasoning".

4. [4]

   Pursuant to a grant of special leave to appeal to this Court, the Crown challenges the conclusion that there was a miscarriage of justice. For the reasons that follow, the appeal should be allowed and the respondent's convictions restored. The formulation of the alleged tendency and the reliance on the evidence of the charged acts to establish the alleged tendency was not inconsistent with the nature of tendency evidence. Neither the alleged tendency, nor the tendency direction, encouraged, much less required, the jury to engage in "impermissible circular" reasoning. When the tendency direction is considered in the context of the trial judge's summing up as a whole, there was not a real risk that the jury's understanding of the necessity that the elements of the charges on the indictment be proved beyond reasonable doubt was undermined. Standard of proof of tendency in a criminal case

5. [5]

   Section 97(1) of the Evidence Act 1995 (NSW) precludes the admission of evidence of, inter alia, "a tendency that a person has or had ... to act in a particular way, or to have a particular state of mind" to prove that the person has or had that tendency unless "the party seeking to adduce the evidence [has given] reasonable notice in writing" and "the court thinks that the evidence will, either by itself or having regard to other evidence adduced or [sought] to be adduced ... have significant probative value". Section 101(2) precludes the prosecution in a criminal case from using such evidence about an accused "unless its probative value substantially outweighs any prejudicial effect that it may have on the accused". Section 97A(2) of the Evidence Act establishes a presumption in proceedings involving "a child sexual offence" that tendency evidence about the sexual interest a defendant has or had in children and about a defendant acting on that interest has significant probative value for the purposes of ss 97(1)(b) and 101(2).

6. [6]

   The evidence said to demonstrate the relevant tendency on the part of the respondent was also evidence of the acts the subject of the charges on the indictment, and no issue was ultimately pursued as to whether that evidence was cross‑admissible at his trial. Nevertheless, the evidence said to support the existence of the alleged tendency had to satisfy ss 97 and 101(2) of the Evidence Act before the Crown could invite the jury to deploy tendency reasoning or the trial judge could direct the jury that such reasoning could be deployed.

7. [7]

   Tendency evidence is a form of circumstantial evidence. The alleged tendency is an "intermediate fact" to be proved in its own right and, if established, the tendency can then be deployed in aid of proof of the charge(s) to which it relates. Tendency reasoning involves first assessing the strength of the evidence said to establish the tendency and then considering the extent to which the tendency makes it more likely that the elements of the relevant offence or offences charged are established.

8. [8]

   Evidence of either or both charged acts or uncharged acts can be relied on to establish a tendency. Absent legislative intervention and whether the evidence said to demonstrate the tendency is evidence of charged acts or uncharged acts, the standard of proof concerning such evidence is governed by the principles applicable to circumstantial evidence. Unless the tendency is an indispensable intermediate step in reasoning to guilt then, whether the evidence said to establish the tendency is evidence of charged acts or uncharged acts, the jury should not be instructed that before it may act on evidence of those acts to support the finding of an alleged tendency it must be satisfied that those acts have been proved beyond reasonable doubt. In New South Wales, this approach is reflected in s 161A of the Criminal Procedure Act 1986 (NSW).

9. [9]

   Accordingly, in a criminal case, the tribunal of fact, including a jury, may, if it finds that an alleged tendency has been proved to a lesser standard by relying on direct evidence of charged acts, deploy that tendency in determining whether the charged acts have been proved beyond reasonable doubt. Such reasoning is not "circular or incoherent". However, in Roder it was accepted that such reasoning carries a risk of undermining the jury's understanding of the necessity for proof of the elements of the charged offence beyond reasonable doubt. One means of minimising that risk is for the trial judge, when directing the jury in relation to the alleged tendency, to avoid inviting the jury to make findings as to the conduct relied on in proof of the charged offence. Instead, the trial judge should identify the evidence said to support the alleged tendency and invite the jury to consider whether that tendency has been established.

10. [10]

    Ultimately in Roder, this Court observed as follows in relation to a tendency direction where the evidence supporting the alleged tendency includes evidence of charged acts: "Such a direction should not direct or invite the jury to make findings in respect of charged conduct, but instead should indicate the evidence relied on to support the alleged tendency, direct the jury to consider whether they are satisfied of the alleged tendency and then advise the jury that, if they are so satisfied, they can use that tendency in considering whether it is more likely that the accused committed the specific offences with which he or she is charged. Careful directions should be given to the jury as to the requisite onus and standard of proof as well as to the contents of the elements of the offence and the need for separate consideration of each charge." (emphasis added; footnotes omitted)

11. [11]

    It follows from the reference in this passage to "should", as opposed to "must", that the mere fact that a tendency direction invites a jury to make findings in respect of the charged acts does not of itself amount to an error or irregularity, much less a miscarriage of justice. Instead, where a direction to that effect is given, the relevant question on any appeal is whether that direction considered as part of the summing up as a whole gave rise to a "real" or "practical" risk of "undermining the jury's understanding of the necessity for proof of the elements of the [charged offence] beyond reasonable doubt". Only if so, would there be a material misdirection amounting to a miscarriage of justice.

12. [12]

    This is illustrated by JS v The Queen, where, even though the trial judge's tendency direction invited the jury to "decide whether [the accused] did in fact conduct himself in the way the Crown allege[d]" on the occasions the subject of the charge, the directions as a whole were found to have "conformed to the legal requirements". As will be explained, the ultimate issue in this appeal is whether the same can be said of the trial judge's directions in this case. The Crown case

13. [13]

    The seven charges on the indictment arose out of three separate occasions in 2020 when the respondent allegedly committed sexual offences against GC.

14. [14]

    Counts 1 and 2 concerned the first occasion, alleged to have occurred at the respondent's house in West Gosford on 26 September 2020. Count 1 was a charge of having sexual intercourse with GC, being a child between the ages of 10 years and 14 years. Count 2 was in the alternative to count 1 and alleged that the respondent intentionally sexually touched GC, being a child between the ages of 10 years and 16 years. Counts 3 and 4 charged the same offences as counts 1 and 2, respectively, and were also charged in the alternative to each other. They were alleged to have occurred on the second occasion at the house of GC's mother ("JW") in Narara between 24 September 2020 and 8 October 2020. Similarly, counts 5 and 6 were also charges of the same offences as counts 1 and 2, respectively, and were charged in the alternative to each other. They were alleged to have occurred on the third occasion at JW's house between 3 October 2020 and 8 October 2020. Count 7 was a further offence of intentionally sexually touching GC and was alleged to have occurred on that third occasion.

15. [15]

    The respondent pleaded not guilty to counts 1 to 7. He was found not guilty on counts 1 and 2 and guilty on counts 3, 5 and 7. No verdict was taken on counts 4 and 6.

16. [16]

    The respondent was in an occasional relationship with JW from April or May 2019, which ended on 8 October 2020. From time to time, the respondent would spend time at JW's house in Narara, occasionally staying overnight.

17. [17]

    On 20 October 2020, GC participated in a recorded interview with police and described the second and third occasions, on which counts 3 to 4 and counts 5 to 7 on the indictment, respectively, were alleged to have occurred. In describing the acts alleged to have constituted counts 3 and 4, GC said that she was watching a movie with the respondent after JW went to bed and fell asleep with her head in the respondent's lap. She said that when she awoke the respondent's hand was touching both the inside and outside of her vagina (counts 3 and 4). GC said she was awake for "maybe five minutes and then ... realised what was happening and ... sat up". The respondent then removed his hand from GC's pants and placed it on her thigh, and GC finished watching the movie before going to bed. As for the third occasion, GC also described falling asleep while watching a movie and waking up to find the respondent's "hand on [her] vagina" (counts 5 and 6) and her right hand on the respondent's penis, with his hand on top of GC's hand, "squeezing" (count 7).

18. [18]

    In June and July 2022, GC disclosed a further occasion that became the subject of counts 1 and 2; ie, the "first" occasion. GC told the police that she recalled this occasion after a "vivid dream" about being locked in the respondent's apartment at West Gosford. GC said that she attended a sleepover at the apartment without JW. GC said that she fell asleep while watching a movie and awoke to find the respondent's hand inside her vagina.

19. [19]

    There was a difficulty in the evidence as to the timing of this first occasion in that GC said that this was the last occasion when it happened and that she "jumped up" when she awoke "because it [ie, the sexual abuse] ... had happened before". However, by reference to various text messages establishing that the sleepover at the respondent's apartment took place on 26 September 2020, the Crown proceeded on the basis that this occasion was the first in time and thus it formed the basis of counts 1 and 2. The Court of Criminal Appeal referred to the "comparative frailty" of the evidence supporting these counts. The alleged tendency

20. [20]

    Prior to the trial, the Crown served a tendency notice foreshadowing that the Crown would seek to prove and rely on a tendency on the part of the respondent (a) to "have a particular state of mind, namely a sexual interest in [GC]; and (b) to act on that sexual interest by penetrating her vagina with his fingers and/or sexually touching her when she had fallen asleep beside him watching a movie".

21. [21]

    No issue was taken either at trial or in the Court of Criminal Appeal with the use of the phrase "and/or" in the framing of the tendency as revealing an ambiguity in the acts in which the respondent allegedly engaged. Instead, that part of the tendency notice was treated as embracing the (same) sexual contact that GC described as having occurred on the three occasions said to have given rise to counts 1 to 7, each of which having been alleged to have occurred "when [GC] had fallen asleep beside him watching a movie".

22. [22]

    The evidence relied on to establish the alleged tendency was that evidence adduced in support of proof of the counts on the indictment. There was no evidence of uncharged acts and the tendency notice foreshadowed reliance on the alleged tendency in proof of all counts on the indictment. Ultimately, the respondent did not object to the Crown relying on the evidence of the charged acts as tendency evidence and the trial judge ruled that, having regard to the presumption in s 97A(2) of the Evidence Act, the evidence had significant probative value and was admissible as such. The summing up and the tendency direction

23. [23]

    To address the Court of Criminal Appeal's reasoning and the parties' contentions about the tendency direction, it is necessary to describe the summing up in some detail.

24. [24]

    The trial judge commenced the summing up with what her Honour described as a number of "general directions", which included what her Honour described as an "important direction" as to the burden and standard of proof, including that, if the jury was not satisfied that the "Crown has established any one of the essential elements [of] the charges beyond reasonable doubt", then it was the jury's "duty to bring in a verdict of not guilty". Her Honour then explained the elements of each offence on the indictment in the context of the trial. In doing so her Honour stated at least 30 times that the Crown had to prove each element of the charged offences beyond reasonable doubt. Her Honour also gave the jury a direction to consider each count separately, which included a direction that, if the respondent was found not guilty of any count, "particularly if that was because [the jury] had doubts about the reliability of [GC's] evidence, [the jury] would have to consider how that conclusion affected [its] consideration of the remaining counts". The trial judge also informed the jury that the Crown had invited the jury to rely on inferential reasoning, specifically tendency reasoning, reminded the jury (again) of the criminal standard of proof, and directed the jury to be "extremely careful" about drawing inferences and not to "draw an inference from the direct evidence unless it is a rational inference in all the circumstances".

25. [25]

    The trial judge then outlined the Crown case and directed the jury in relation to complaint evidence, the significance of a witness who was not called by the Crown and evidence of good character adduced on behalf of the respondent. In doing so, her Honour referred to the criminal standard of proof eight more times.

26. [26]

    The trial judge then gave the jury the tendency direction, parts of which are numbered for ease of reference: "Part of the Crown case is that the accused had a tendency to have a sexual [interest] in [GC] and to act on that sexual interest by penetrating her vagina with his fingers and/or sexually touching her when she had fallen asleep beside him watching a movie. The Crown says you would be satisfied the accused had this tendency because of his conduct in relation to the three occasions it is alleged that he sexually assaulted [GC]. ... The Crown says that the three occasions when the accused sexually assaulted [GC] reveals the accused had a tendency to have a sexual interest in [GC] and to act on that sexual interest, which makes it more likely that he committed each of the offences charged in the indictment. You will need to consider the evidence relating to the alleged conduct of the accused on the three occasions and decide whether he did in fact conduct himself in the way the Crown alleges. Although you are looking at the three occasions of alleged conduct that are the subject of charges, you are not when considering them as tendency evidence considering whether those episodes of misconduct have been proved beyond reasonable doubt. For the purpose of determining whether the accused has the relevant tendency, you do not consider each of the three occasions in isolation, you look at them collectively to decide what conduct occurred. If you decide that all or at least some of the conduct occurred, you then need to consider whether it enables the inference to be drawn that the accused has the tendency asserted by the Crown. That is, to have a sexual interest in [GC] and to act on that sexual interest in the manner alleged by the Crown. You will recall the direction I gave you earlier about drawing inferences in a criminal trial. You are required to consider whether there are alternative explanations for the evidence and you should not draw an inference from the evidence unless it's a rational inference in all the circumstances. You should bear in mind those directions when you are considering the inferences the Crown asked you to draw in relation to tendency reasoning. If you are not satisfied that any of the conduct the Crown relies upon occurred, then there is no basis upon which the tendency could be inferred. In these circumstances the whole issue is put to one side. If you find the accused did have a tendency to have a sexual interest in [GC], and a tendency to act on that sexual interest in the way alleged by the Crown, then you can use that in considering whether it is more likely he committed the specific offences with which he is charged. However, it is essential you consider in relation to each charge whether the accused acted in that particular way and had that particular state of mind on that specific occasion. Finding the accused did have the tendency the Crown asserts is not enough by itself to prove his guilt beyond reasonable doubt in relation to each charge. It is simply part of the process of reasoning and it may assist the Crown to prove the accused committed each of the offences. The question is whether it makes it more likely the accused conducted himself in the way the Crown alleges on any of the occasions that are subject of the charges. That is the only way the accused['s] tendency to have a sexual interest in [GC] and act on that sexual interest in the manner alleged may be used. Ultimately you must decide whether the specific offences with which the accused has been charged have been proved beyond reasonable doubt. That decision must be based upon the evidence relevant to each charge. This includes the evidence of [GC] about what the accused did, it will include the tendency alleged by the Crown provided you are satisfied it has been established, it will also include all of the other evidence you have heard in both the Crown and defence cases. When considering whether a charge has been proved you will have to decide whether the Crown has proved the essential elements of that charge. I will now summarise the case for the Crown and the case for the accused on this issue of tendency. The Crown argues that the conduct of the accused on each of the three alleged occasions establishes the tendency asserted. The defence argues that you would not be satisfied that any of the conduct alleged on the three occasions has occurred so that the tendency would not be established."

27. [27]

    Six matters should be noted about that direction. First, the form of the alleged tendency identified in the tendency direction reflected the form of the alleged tendency identified in the tendency notice. The tendency direction (correctly) directed the jury that, if established, the tendency (as opposed to the evidence said to support the tendency) could be deployed in aid of proof of the charges on the indictment.

28. [28]

    Second, the tendency direction invited the jury to "decide whether [the respondent] did in fact conduct himself in the way the Crown allege[d]" on the three occasions in determining whether the respondent had the alleged tendency.

29. [29]

    Third, the direction advised the jury that, if it was not satisfied that any of the conduct relied upon to establish the tendency occurred, there would be no basis upon which to infer the existence of the tendency and the "whole issue is put to one side".

30. [30]

    Fourth, the direction told the jury that, to the extent that it was considering the three occasions on which offences were alleged to have occurred for the purpose of determining whether the alleged tendency was established, it was not considering whether the alleged offences had been proved beyond reasonable doubt.

31. [31]

    Fifth, the jury was expressly told that, if it was satisfied that the respondent had the alleged tendency, it could use that tendency in considering whether it was more likely that the respondent committed the offences charged, but the jury nevertheless had to consider whether each offence had been proved beyond reasonable doubt. The jury was also told that proof of the existence of the tendency was not sufficient to prove guilt beyond reasonable doubt in relation to each charge.

32. [32]

    Sixth, the jury was reminded that the Crown had to prove the essential elements of each charge. As noted, the jury had previously been advised many times that each element had to be proved beyond reasonable doubt.

33. [33]

    In the balance of the summing up the trial judge summarised the evidence on each count. The trial judge reminded the jury that, amongst other evidence, the Crown relied on tendency evidence, and her Honour added that the tendency evidence should only be taken into account if the jury had accepted that the tendency had been established by the Crown. Later, in directing the jury as to its consideration of the respondent's evidence, the trial judge referred to the criminal standard of proof a further four times. The Court of Criminal Appeal's reasoning

34. [34]

    The ground of appeal upheld by the Court of Criminal Appeal contended that the trial judge erred by: (i) directing or inviting the jury to make findings in respect of charged conduct in the context of considering whether it was satisfied of the alleged tendency; and (ii) inviting the jury to engage in an impermissible mode of reasoning by directing the jury to consider the conduct "collectively to decide what conduct occurred" for the purpose of determining whether the tendencies were established. The Court of Criminal Appeal addressed both limbs of this ground together.

35. [35]

    In upholding this ground, the Court of Criminal Appeal noted that the tendency direction invited the jury to make findings in relation to the charged conduct in a manner that was "at odds with Roder". However, the Court of Criminal Appeal accepted that this "flaw [was] not necessarily fatal", provided that the summing up as a whole ensured that the jury would not be deflected from the task of determining whether the charges were proved beyond reasonable doubt. Nevertheless, the Court identified what it described as a "fundamental issue" with the framing of the alleged tendency, being that it "replicated" or "coincide[d] precisely with the specific conduct that was alleged against the [respondent] in each of the counts, namely, 'by penetrating her vagina with his fingers and/or sexually touching her when she had fallen asleep beside him watching a movie'". By reference to an earlier decision of a differently constituted New South Wales Court of Criminal Appeal, Kanbut v The King, the Court of Criminal Appeal in this case concluded that "[f]raming the tendency and identifying the evidence to establish it so as to constitute exclusively the offending behaviour itself is inconsistent with the nature of tendency evidence".

36. [36]

    The Court further observed, by reference to Kanbut, that the "intractable problems" caused by the framing of the tendency were illustrated by the direction to the jury that, if the jury was "not satisfied that any of the conduct the Crown relies upon occurred, then there is no basis upon which the tendency could be inferred" and "the whole issue [of the tendency] is put to one side". This direction was found to be "illogical and confusing" because if the jury was not satisfied that the alleged conduct occurred, the jury would have had to acquit the respondent entirely.

37. [37]

    The Court of Criminal Appeal also found that a further difficulty arose from that part of the tendency direction that directed the jury that if it "decide[d] that all or at least some of the conduct occurred, [it] then need[ed] to consider whether it enable[d] the inference to be drawn that the [respondent] ha[d] the tendency asserted by the Crown". The Court of Criminal Appeal held that, in circumstances where the alleged tendency was expressed in the same terms as the conduct alleged, that direction was likely to encourage, if not require, the jury to "engage in the impermissible circular reasoning identified in JS at and Roder at -".

38. [38]

    Ultimately, the Court concluded that, in the circumstances of this case, the direction to the jury to make findings in respect of the charged conduct in order to determine if a tendency expressed in the same terms as the charged conduct was established occasioned a miscarriage of justice. The Court held, applying Kanbut, that such a direction meant that it "was likely that the jury's attention would be deflected from applying the required standard of proof".

39. [39]

    The Crown was granted special leave to appeal from the Court of Criminal Appeal's decision to raise two grounds of appeal: namely, that the Court of Criminal Appeal erred in concluding that a miscarriage of justice was occasioned by a misdirection of law, and that the Court of Criminal Appeal erred in concluding that reliance upon the same evidence to establish the tendency as the evidence relied upon to establish the counts on the indictment was inconsistent with the nature of tendency evidence. Those grounds are best addressed together and by first considering the Court of Criminal Appeal's reliance on Kanbut and that Court's conclusion about the trial judge's tendency direction inviting impermissible circular reasoning, before addressing the ultimate issue of whether the jury was likely to be deflected from applying the correct standard of proof. Framing the tendency: Kanbut

40. [40]

    It follows from the principles stated at the commencement of this judgment that the formulation of the relevant tendency will often be critical to the admissibility and use of the evidence said to support the alleged tendency as well as the reasoning towards guilt that the alleged tendency is said to support. The particularity of the alleged tendency and its capacity to affect the assessment of whether the prosecution has discharged its onus of proof will depend upon the circumstances of the particular case. Other things being equal and leaving aside any consideration of s 97A of the Evidence Act, the greater the specificity of the alleged tendency the more likely it is that the alleged tendency will be significantly probative of the charge(s) to which the alleged tendency relates, but the more difficult it may be to establish the alleged tendency and to relate it to all or most of the charged conduct. Conversely, the more general the alleged tendency, the less likely it is that the alleged tendency will be significantly probative of the charge(s) to which it relates, but the less difficult it may be to establish the alleged tendency and to relate it to all or most of the charged conduct.

41. [41]

    Moreover in some cases the formulation of the alleged tendency may simply fail to engage with the tendency reasoning sought to be deployed. For example, in Kanbut the applicant was (relevantly) convicted of four offences of possessing a slave under s 270.3(1)(a) of the Criminal Code (Cth). Two of the offences concerned one complainant, X, and the other two offences concerned another complainant, Y. "[S]lavery" was and remains defined as the "condition of a person over whom any or all of the powers attaching to the right of ownership are exercised, including where such a condition results from a debt or contract made by the person". The Crown identified eight acts of the applicant in relation to each of X and Y, which comprised "every particular of the conduct of the applicant said to satisfy the definition of slavery and the four offences under s 270.3(1)(a)". Those eight acts included, for example, "to impose on [X] and [Y] a debt of $45,000" and "to take possession of [X's] passport and [Y's] passport until they had paid [off] their debts of $45,000".

42. [42]

    The trial judge in Kanbut gave the jury a tendency direction which identified the alleged tendency of the applicant as being "to act in a particular way, namely" to engage in all of the eight particularised acts against both X and Y. The Court of Criminal Appeal in Kanbut relevantly concluded that this formulation of the alleged tendency, and the tendency direction, gave rise to three related difficulties.

43. [43]

    The first and principal difficulty with the alleged tendency, and the tendency direction, in Kanbut was that framing the alleged tendency by reference to all eight acts and both complainants meant that the direction failed to achieve its purpose in that case of enabling the jury to utilise the evidence of one of the complainants as rendering it more likely to conclude that the conduct directed towards the other complainant occurred.

44. [44]

    This difficulty has no relevance to this case. In Kanbut the alleged tendency, framed as it was by reference to the applicant having a tendency to commit all eight particularised acts against both X and Y, did not have significant probative value in determining whether the applicant in that case committed a particular act against X or committed a particular act against Y. The formulated alleged tendency in Kanbut was not describing features common to the offending against both complainants; it was not a feature or aspect of the alleged offending against X that the applicant in Kanbut committed the same eight acts against Y. The position may have been different had the alleged tendency in Kanbut been formulated in terms of a tendency to commit particular acts amounting to slavery against "vulnerable women generally".

45. [45]

    The tendency notice in this case concerned only one complainant and identified two features of the respondent's conduct in allegedly acting on his sexual interest in GC that were common to all offences: namely, a tendency to act on his sexual interest by committing a particular sexual act or acts against GC, and doing so while GC was asleep beside the respondent watching a movie. Unlike the alleged tendency in Kanbut, the alleged tendency in this case did identify features common to all three sets of alleged offences on the indictment. While it is more common to frame the alleged tendency in a case such as this as simply a tendency to have and to act on a sexual interest in a complainant, a tendency framed in the terms employed in this case, if established, would achieve its purpose of rendering it more likely the respondent committed a particular charged offence against GC.

46. [46]

    The equivalent formulation of the objectionable alleged tendency in Kanbut to circumstances similar to this case would have been an alleged tendency for an accused to commit sexual acts against two victims while both of them were asleep watching a movie together with the accused. Such a tendency might be significantly probative of an allegation that an accused sexually assaulted both victims while both were asleep watching a movie together with the accused. Such a tendency would have little or perhaps no probative value in relation to an allegation that an accused sexually assaulted one of them while that one complainant alone had fallen asleep beside an accused watching a movie.

47. [47]

    Contrary to the reasoning of the Court of Criminal Appeal, nothing in Kanbut supports the conclusion that there was any difficulty with the formulation of a tendency at the level of specificity that was undertaken in this case. In Kanbut the overspecification of the tendency by reference to both complainants and all particularised acts against both complainants robbed the alleged tendency of its probative strength in the assessment of a particular charge against the applicant in respect of a particular complainant. In this case the relative specificity of the alleged tendency was what gave the alleged tendency its significant probative value. While similarity between the conduct said to demonstrate the alleged tendency and the conduct in issue is not a precondition to the admission of tendency evidence, it is a common feature of a tendency which can give that evidence significant probative value. Leaving aside the use of the phrase "and/or" in this tendency notice, to find a flaw in the formulation of the tendency in this case would significantly undermine ss 97 and 101(2) of the Evidence Act.

48. [48]

    The second difficulty with the tendency direction given in Kanbut was its capacity to mislead or confuse the jury. In Kanbut, the jury was directed in effect that "unless [it was] satisfied of each of the precise acts relied on in relation to both complainants, then [it] could not be satisfied that the corresponding tendency [had] been established". The potential confusion arose because, "leaving aside any question of the standard of proof", if the jury was "satisfied of the relevant acts in relation to both victims then the Crown would not need to rely on the existence of any alleged tendency to prove its case as it would already be proved".

49. [49]

    This difficulty also has no relevance to this case. The ultimate error identified by the Court of Criminal Appeal in this case was that the tendency direction was likely to deflect the jury from applying the correct standard of proof. That was not the criticism of the alleged tendency and the tendency direction in Kanbut. In this case and unlike Kanbut it was open to the jury to be satisfied, to a standard lesser than the criminal standard, of only some but not all of the acts said to establish the tendency, and for the jury still to be satisfied that the relevant tendency had been established. The jury could then use that tendency to support a finding to the criminal standard in relation to some of the charges. Thus, unlike Kanbut, a finding in this case that the respondent had the alleged tendency would not necessarily mean that the jury would then have to revisit all of its findings in relation to the charged acts by applying a higher standard, although even if it did, as Roder makes clear, that potential is only problematic to the extent that it may contribute to a conclusion that the criminal standard of proof was undermined.

50. [50]

    The third difficulty with the alleged tendency and tendency direction in Kanbut was that in that case the jury was told that if it could not find that "any of the acts occurred" then the jury should put aside the tendency alleged, whereas if it had not found that any of the acts occurred then the jury would have been obliged to find the applicant not guilty, "as then there would be nothing left of the Crown case". As noted by the Court of Criminal Appeal, the jury was instructed in similar terms in this case. However, the only present significance of that aspect of the direction lies in its capacity to deflect the jury from applying the correct standard of proof, which is addressed below. Allegedly "impermissible circular reasoning"

51. [51]

    As noted, the Court of Criminal Appeal concluded that the framing of the alleged tendency "in the same terms as the conduct alleged" and the tendency direction were likely to "encourage, if not require, the jury to engage in the impermissible circular reasoning identified in JS at and Roder at -".

52. [52]

    The cited passages from JS and Roder did not identify any process of impermissible circular reasoning. Instead, those cases only addressed a complaint that, if the Crown sought to prove the commission of an offence to a lesser standard, then used that "finding [of an offence] to prove the tendency, and then relied on the tendency to prove the offence", such a process of reasoning would be circular. Both Roder and JS did not accept that such a complaint accurately described tendency reasoning. Thus, Roder concluded that "[f]or a tribunal of fact to find that an alleged tendency has been proved to a lesser standard by relying on, inter alia, direct evidence of charged acts, and then deploying that tendency in determining whether the charged acts have been proved beyond reasonable doubt, does not involve circular or incoherent reasoning". The concern identified in Roder with such reasoning is not the risk of circular reasoning but the risk of undermining the jury's understanding of the necessity for proof of the elements of the charge beyond reasonable doubt. Was the standard of proof undermined?

53. [53]

    It follows that the Court of Criminal Appeal erred in concluding that the framing of the alleged tendency, and the reliance on the evidence of the charged acts to establish it, was inconsistent with the nature of tendency evidence, and erred in concluding that the tendency direction encouraged or required the jury to engage in "impermissible circular reasoning". Even so, the ultimate question is whether the Court erred in concluding that the tendency direction, considered in the context of the summing up as a whole, rendered it "likely that the jury's attention would be deflected from applying the required standard of proof".

54. [54]

    The respondent noted that all of the evidence of the charged acts was admissible and available to be used by the jury in determining the respondent's guilt for reasons separate from that evidence's use as tendency evidence, such as to provide context for GC's evidence. The respondent contended that any reliance on the evidence for tendency purposes added little to the Crown case especially as the evidence emanated from the same complainant.

55. [55]

    The alleged lack of strength of the capacity of the alleged tendency to contribute to the jury's assessment of the charges on the indictment does not add to the contention that the direction undermined the standard of proof. There was no dispute that the evidence said to support the alleged tendency met the test of having significant probative value for the purposes of s 97(1)(b) of the Evidence Act (and otherwise satisfied s 101(2) of the Evidence Act). Were it otherwise, the evidence could not have been admitted as tendency evidence and the question that would have arisen was whether a direction precluding its use as tendency evidence (ie, an anti‑tendency direction) was warranted or required.

56. [56]

    The respondent also contended that, by inviting the jury to make findings concerning the charged conduct in order to determine whether the alleged tendency was established, the tendency direction invited the jury to engage in "reasoning" of the kind that was supposedly disapproved of in JS and Roder. The respondent also complains that the tendency direction referred to the alleged tendency "as if it was a discrete body of evidence" when it was based on evidence that emanated solely from GC. These submissions repeat the misconception of the nature of tendency reasoning that was addressed in Roder (and JS). As already noted, Roder did not accept that it was impermissible to find that an alleged tendency has been proved to a lesser standard by relying on direct evidence of charged acts, and then to deploy that tendency in determining whether the charged acts have been proved beyond reasonable doubt. The relevant point made in JS and Roder is that it is the alleged tendency, not the evidence said to demonstrate its existence, that is relied on as circumstantial evidence in proof of the charged conduct. The tendency direction in this case reflected that correct understanding.

57. [57]

    The balance of the respondent's submissions contended that the tendency direction undermined the standard of proof. The respondent's submissions emphasised the second and third of the features of the tendency direction noted above as well as the strong correspondence between the features of the tendency and the charged conduct, the small number of charged incidents on which the alleged tendency was based and the fact that the basis for finding the tendency was only the evidence of one complainant, GC. The respondent contended that, in that context, by inviting the jury to make findings about the incidents the subject of those charges as the basis for finding the existence of the alleged tendency, the tendency direction undermined the standard of proof because the tendency direction made it more difficult for the jury to separate a finding that the alleged conduct occurred for the purpose of finding the tendency from a finding that the offence was proved.

58. [58]

    Much of the force of this submission ebbs away once the misconceived reliance on Kanbut, and the incorrect characterisation of the tendency and the tendency direction as inviting impermissible circular reasoning, are put aside. Even so, it can be accepted that the closer the correspondence between the alleged tendency and the charged conduct and the fewer the instances of charged conduct relied on to establish an alleged tendency, the greater the potential for a direction that invites the jury to make findings in respect of that conduct to a lesser standard in order to determine whether or not the alleged tendency was established to undermine the jury's understanding of the necessity for proof of the elements of the charges beyond reasonable doubt. However, whether there was the relevant risk of that understanding being undermined requires a consideration of the summing up as a whole.

59. [59]

    In this case the tendency direction specifically brought to the jury's attention that, in determining whether the tendency was established based on evidence of the charged conduct, the jury was not considering whether that conduct had been proved beyond reasonable doubt, that to find the respondent guilty of a particular charge the jury had to be satisfied of those elements of that charge beyond reasonable doubt and that proof beyond reasonable doubt could not be established alone by the existence of the tendency. Otherwise, as the above analysis demonstrates, the summing up was saturated with directions to the jury of the necessity to be satisfied of those elements beyond reasonable doubt. The jury received "[c]areful" (and repeated) directions "as to the requisite onus and standard of proof as well as to the contents of the elements of the offence and the need for separate consideration of each charge". A consideration of the entirety of the summing up demonstrates that there was "no [real] risk ... that the onus and standard of proof were not understood and properly applied".

60. [60]

    The Court of Criminal Appeal erred in concluding that the effect of the trial judge's directions was that it was likely that the jury's attention would have been deflected from applying the required standard of proof in respect of the offences charged. The directions did not undermine the jury's understanding of the necessity for proof of each of the elements of the charges beyond reasonable doubt. The grounds of appeal must be upheld. Orders

61. [61]

    The following orders should be made: Appeal allowed. Orders 6, 7, 8 and 9 of the New South Wales Court of Criminal Appeal made on 3 March 2025 be set aside. In place of those orders, the respondent's appeal to the New South Wales Court of Criminal Appeal be dismissed.

62. [62]

    GORDON AND STEWARD JJ. The respondent, AR, was indicted on seven counts, relating to three alleged incidents between late September 2020 and early October 2020 in respect of the same complainant. The Crown case was that the three incidents occurred in the school holidays in September to October 2020 when the complainant was 10 years of age.

63. [63]

    Each incident formed the basis of a count contrary to s 66C(1) of the Crimes Act 1900 (NSW), namely that AR had sexual intercourse with a child between the ages of 10 and 14, and an alternative count contrary to s 66DB(a) of the Crimes Act, namely that he sexually touched a child between the ages of 10 and 16. The last incident was the basis of an additional count, also contrary to s 66DB(a) of the Crimes Act.

64. [64]

    Counts 1 and 2 were in relation to the "West Gosford incident". The Crown alleged that on an evening of the first weekend of the school holidays the complainant had a sleepover at AR's residence at West Gosford. That evening, at a time when AR and the complainant were alone together in the lounge room, AR put on a movie and sat on the lounge with the complainant. The complainant fell asleep. At some stage, she woke up and realised that AR was touching her on the inside and outside of her vagina. The complaint in relation to this incident was delayed. It was recalled in circumstances where the complainant had initially failed to recall the acts themselves, and where the memory of the incident was "triggered" by a dream.

65. [65]

    Counts 3 and 4 were in relation to the "first Narara incident". The Crown alleged that during the second week of the school holidays AR was staying over at the complainant's mother's house in Narara ("the Narara residence"). At one point, AR and the complainant were alone in the lounge room watching a movie. The complainant's mother had gone to bed and her brother was in his room on a computer. The complainant was lying on the lounge with her head on AR's lap. She fell asleep and when she woke up, she realised that AR's hand was on the outside and inside of her vagina. The complainant sat up and AR removed his hand and placed it on her thigh.

66. [66]

    Counts 5, 6 and 7 were in relation to the "second Narara incident". The Crown alleged that during the same week as the first Narara incident AR was again at the Narara residence, watching a movie with the complainant. The complainant's mother went to bed and the complainant fell asleep lying on the lounge with her head on a pillow that was next to AR. When the complainant woke up, she realised that AR had one of his hands inside her pants on her vagina, with his fingers inside her vagina and her right hand was over her head and on the outside of AR's clothes. AR had his hand on the top of her hand and was squeezing it. The complainant could feel AR's penis. She stood up and he withdrew his hand.

67. [67]

    In relation to the two Narara incidents, the Crown relied on the evidence of the complainant in her first record of interview on 20 October 2020, complaint evidence, text messages showing the nature of the relationship between AR and the complainant, and evidence corroborating that the movie nights occurred. In relation to the West Gosford incident, the Crown relied on the evidence of the complainant in her third record of interview on 1 July 2022, as well as complaint evidence and text messages that the Crown said demonstrated the nature of the relationship between AR and the complainant. The most crucial evidence in the Crown case was the complainant's evidence, given that the other evidence was merely corroborative of that evidence.

68. [68]

    The Crown had the benefit of cross-admissibility – that is, the evidence of each of the charges was admissible as circumstantial evidence in support of each of the other charges so that the jury were permitted to use the evidence of each of the incidents in assessing the other incidents. For example, the jury were permitted to reason that, because they were satisfied that AR committed the offences the subject of the first Narara incident, they could, applying ordinary human experience, be more satisfied that the complainant did not simply misinterpret or imagine the second Narara incident.

69. [69]

    Notwithstanding that the evidence of the charged acts was cross-admissible, as part of its case at trial the Crown also alleged that AR had a tendency to have a particular state of mind, namely a sexual interest in the complainant, and "to act on that sexual interest by penetrating her vagina with his fingers and/or sexually touching her when she had fallen asleep beside him watching a movie". The evidence in support of that tendency was the evidence of the charged acts. There were no uncharged acts. Both the tendency and the charged acts depended upon acceptance of the credibility of the complainant. The trial judge ruled that the tendency evidence had significant probative value, having regard to the presumption in s 97A(2) of the Evidence Act 1995 (NSW), and that it was admissible. There was ultimately no dispute at trial (or on appeal) about the admissibility of the tendency evidence.

70. [70]

    After a trial by jury in the District Court of New South Wales, AR was convicted in relation to both Narara incidents but acquitted in relation to the West Gosford incident. On 13 December 2022, AR was sentenced to an aggregate term of imprisonment of 9 years and 6 months with a non-parole period of 4 years and 9 months.

71. [71]

    AR sought leave to appeal against his convictions to the Court of Criminal Appeal of the Supreme Court of New South Wales ("the CCA"). The CCA, relevantly, granted AR leave to appeal and allowed his appeal on the ground that the manner in which the trial judge had directed the jury – to consider whether the charged conduct occurred for the purpose of determining whether the tendency was established – occasioned a miscarriage of justice. In particular, the CCA held that the trial judge was wrong to direct the jury that, in order to be satisfied of the alleged tendency, they needed to consider whether they were satisfied that the charged conduct occurred on the balance of probabilities.

72. [72]

    The CCA concluded that the direction given by the trial judge did not accord with this Court's decision in Director of Public Prosecutions (Vic) v Benjamin Roder (a pseudonym) but that that flaw was not necessarily fatal. The CCA went on to conclude that, in this case, a fundamental issue was that the tendency was particularised as a tendency to act on a sexual interest in the complainant in a way that "coincide[d] precisely" with the conduct that was alleged against AR in each of the counts; and that framing the tendency in this way and "identifying the evidence to establish it so as to constitute exclusively the offending behaviour itself [was] inconsistent with the nature of tendency evidence". The CCA concluded that it was "not possible for the jury to distinguish the tendency from the conduct constituting the alleged offences, which ... invited circular or boot-strap reasoning". The CCA found that there had been a miscarriage of justice.

73. [73]

    The Crown was granted special leave to appeal. The issue in this Court is whether, given how the tendency was framed, and given the summing up as a whole, there was a risk that the standard of proof was undermined. Or, put in different terms, whether the trial judge's directions to the jury, taken as a whole, deflected the jury from their fundamental task of deciding whether the Crown had proved the elements of the charged offences beyond reasonable doubt.

74. [74]

    The Crown accepted that the direction given by the trial judge did not accord with Roder but contended that that deficiency did not necessarily result in a miscarriage of justice by way of misdirection. The Crown contended that the CCA erred in concluding that it was inconsistent with the nature of tendency evidence to rely on a tendency to act in a particular way which is articulated with specificity so that it describes or "replicate[s] the detail" of the type of conduct alleged in the charges on the indictment, while seeking to establish that tendency exclusively by evidence of those charged acts. It is in this context that the Crown contended that the CCA erred in finding that the way in which the tendency was framed coincided precisely with the offending behaviour and erred in concluding that, because proof of the tendency relied exclusively upon evidence of charged acts, no direction could have been given that would not have invited impermissible reasoning.

75. [75]

    For the following reasons, there was a risk that the trial judge's directions to the jury, taken as a whole, deflected the jury from their fundamental task of deciding whether the Crown proved the elements of the charged offences beyond reasonable doubt and that risk amounted to a miscarriage of justice. The appeal should be dismissed. Tendency evidence – principles

76. [76]

    Tendency evidence is a form, and indeed a "special class", of circumstantial evidence. It involves a process of inferential reasoning. The asserted tendency is an intermediate fact that the prosecution seeks to establish in its own right and then relies on as circumstantial proof of each charge. Unless the tendency is an "intermediate fact" which is "indispensable" to proof of guilt, it need not be proved beyond reasonable doubt. As the Court explained in Roder, for a tribunal of fact to find an alleged tendency has been proved to a lesser standard by relying on, among other things, direct evidence of charged acts, and then deploying that tendency in determining whether the charged acts have been proved beyond reasonable doubt, does not involve circular or incoherent reasoning. "[I]t simply means that the jury may consider the same evidence 'at different stages of its deliberations with a different onus of proof and for a different purpose'".

77. [77]

    As was recognised in Roder, a way to minimise the risk of undermining the jury's understanding of the necessity for proof of the elements of the charge beyond reasonable doubt is to avoid giving a tendency direction that invites the jury to make findings as to the conduct relied on in proof of the charge. Instead, the direction should identify the evidence said to support the alleged tendency and invite the jury to consider whether that tendency has been established. Then, if the jury are satisfied of the alleged tendency, they can use that tendency in considering whether it is more likely that the accused committed the specific offences with which he or she has been charged. The observations in Roder remain important and should not be diminished or undermined.

78. [78]

    To be admissible, tendency evidence must meet the threshold of having significant probative value and its probative value must outweigh the danger of unfair prejudice to the accused. A tendency "to act in a particular way" for the purpose of s 97(1) of the Evidence Act must, by its nature, describe how the relevant person acts. The articulated tendency may – at least to some degree – mirror the detail of how the alleged charged acts were committed. That is inherent in the fact that the prosecution is alleging that the accused acted in conformity with the tendency to act in the particular way on those charged occasions.

79. [79]

    The assessment of significant probative value of tendency evidence "involves consideration of two interrelated but separate matters": (1) the extent to which the evidence supports the tendency and (2) the extent to which the tendency makes more likely the facts making up the charged offence or offences. That assessment necessarily involves a comparison between the asserted tendency and the facts in issue at the trial. Significantly, a tendency expressed at a level of particularity or specificity will usually be more likely to provide strong support for the proof of a fact that makes up the offence charged. The specificity of the tendency has a direct impact on the strength of the inferential mode of reasoning. However, a broadly expressed tendency may still have significant probative value, particularly in a single complainant case. Therefore, the prosecution is not compelled to frame the tendency in a particular way. The jury will be directed that it remains a question for the jury to consider whether the asserted tendency is established by the evidence relied upon and, if it is, whether the accused acted in conformity with the tendency on the specific occasions charged. Tendency

80. [80]

    The tendency that the Crown sought to prove was AR's tendency to (a) have a particular state of mind, namely a sexual interest in the complainant; and (b) to act on that sexual interest by penetrating her vagina with his fingers and/or sexually touching her when she had fallen asleep beside him watching a movie.

81. [81]

    A careful reader will notice that the tendency, as framed, replicated the charged conduct. All of the charged conduct involved AR "penetrating [the complainant's] vagina with his fingers and/or sexually touching her when she had fallen asleep beside him watching a movie". It may be accepted that the charged conduct involved more specificity, such as the exact location where the conduct occurred and the specific surrounding circumstances. However, the tendency replicated the essential details of the charged conduct. That is, the tendency replicated all the elements in issue of all the charged acts. It did so with some particularity. For example, the tendency identified the exact nature of the "sexual intercourse" and specified some of the surrounding circumstances, thereby replicating the charges and the alternative charges. The evidence said to support the tendency was entirely co-extensive with the evidence in support of the charged conduct. There was no additional evidence in support of the tendency, separate to the evidence in support of the charged conduct. The sole source of the evidence in support of the tendency and the charged conduct was, in essence, the complainant.

82. [82]

    This gave rise to a risk in this case that the jury would use the evidence of a charged act as proof of a tendency to do that same charged act. That is contrary to this Court's holding in Roder that: "Tendency evidence has been described as evidence that on 'another occasion or occasions' different to the occasion the subject of a charge, the accused 'acted in a particular way'. If so described, the direct evidence of a particular charged act would not be admitted as tendency evidence in relation to that same charged act."

83. [83]

    There was also an associated risk in this case that the standard of proof was undermined. There was one complainant. There were three charged incidents and no uncharged incidents. The only evidence said to support the tendency was the evidence in support of the very same charged conduct. That evidence all came from the one source: the complainant. The tendency was framed with a high degree of specificity. All of the charged conduct involved AR "penetrating [the complainant's] vagina with his fingers and/or sexually touching her when she had fallen asleep beside him watching a movie". The tendency was framed as having a "sexual interest in [the complainant]" and acting "on that sexual interest by penetrating her vagina with his fingers and/or sexually touching her when she had fallen asleep beside him watching a movie". The tendency as framed encompassed all of the elements of all of the charged conduct that was in issue at trial.

84. [84]

    AR's case at trial was that the movie nights occurred largely as described by the complainant, but none of the alleged offending occurred. In other words, what was in issue was whether AR, during those movie nights, while the complainant was asleep beside him, penetrated the complainant's vagina with his fingers and/or sexually touched her. Again, the evidence said to support the tendency encompassed all of the evidence that was in issue at trial – namely, whether the complainant's accounts of the alleged offending were true. In other words, the tendency was to do the (very specific) thing that constituted the charged conduct, in the same circumstances as the charged conduct. Such a tendency was highly probative, if not conclusive.

85. [85]

    In assessing the tendency evidence, the jury were effectively required to consider the evidence of the charged conduct in order to decide whether the tendency to do the charged conduct was established, in order to decide whether the charged conduct was established. The jury were considering the same evidence in order to reach the same conclusion, but at different stages of the analysis and on different standards of proof. This was apt to confuse. The jury were required to make, in effect, a preliminary finding (on an indeterminate standard of proof) as to whether the charged conduct was established, before later considering whether the charged conduct was established beyond reasonable doubt. There was a significant risk that proof of the charged conduct was thereby a foregone conclusion, on a standard of proof below that of beyond reasonable doubt.

86. [86]

    To the extent that the CCA relied on Kanbut v The King to hold that an asserted tendency framed by reference exclusively to the offending behaviour is "inconsistent with the nature of tendency evidence", it was wrong to do so. Kanbut was a decision where the tendency notice was framed by reference to all the charged acts in relation to both of the complainants. As put by Beech-Jones CJ at CL: "There are a number of problems with a tendency direction that identifies the alleged tendency in precisely the same terms as the acts said to give rise to it and which concern both complainants. The utility of a tendency direction in a case such as this is that, if properly formulated, it was a means of enabling the jury to utilise its acceptance of the evidence of one of the complainants to accept the evidence of the other. By grouping the conduct engaged in against X and Y together and then formulating the alleged tendency in a manner specific to both of them, the direction wholly failed to achieve that purpose."

87. [87]

    Kanbut is readily distinguished from the present case because, in Kanbut, the tendency notice was framed by reference to both complainants, and it was for that reason that the tendency direction "wholly failed to achieve [its] purpose". As put by the Crown, were the tendency notice in this case to be analogous to that in Kanbut, the tendency would have been framed as a tendency to commit the entire series of precise acts alleged on all of the three occasions. The tendency in this case was not framed in that way, and so can be distinguished from Kanbut.

88. [88]

    Similarly, it may be accepted that the tendency in this case is not co‑extensive with proof of each and every of the four individual acts charged and that, unlike in Kanbut, it was open to the jury to be satisfied (to an indeterminate standard) of some of the acts, but not others, and still be satisfied that AR had the asserted tendency. However, the point remains that, in the specific circumstances of this case, there was a risk of the kinds of impermissible reasoning described earlier. The reasoning need not be described as "circular". The language of "circular reasoning" may distract from the essential question whether there was a risk of the jury being deflected from their task of finding guilt beyond reasonable doubt. Directions

89. [89]

    It is then necessary to consider whether the trial judge's directions, viewed as a whole, ensured that the jury would not be deflected from their task of determining whether the charges were each proved beyond reasonable doubt.

90. [90]

    It is not disputed that the trial judge's directions erroneously directed the jury to first consider, in assessing the asserted tendency, "the evidence relating to the alleged conduct of [AR] on the three occasions and decide whether he did in fact conduct himself in the way the Crown allege[d]". That direction was contrary to Roder because it directed the jury to decide whether the charged conduct was established, as opposed to whether the tendency was established.

91. [91]

    However, as explained, a direction contrary to Roder does not necessarily result in a miscarriage of justice. The question is therefore, given the summing up as a whole, was there a risk that the standard of proof was undermined in relation to the charged conduct? In the circumstances of this case, there was such a risk, and the trial judge's directions did not ameliorate that risk. If anything, the directions exacerbated it. Directions that were given

92. [92]

    The trial judge gave a number of standard and uncontroversial directions about fact-finding and the standard of proof. First, the trial judge directed that the "onus [of proof] rests upon the Crown in respect of every element of the charges". The trial judge then directed the jury that "[b]efore you can find the accused guilty you must consider all the evidence placed before you and ask yourself whether you are satisfied beyond reasonable doubt that the Crown has made out its case".

93. [93]

    The trial judge followed with a series of directions: "If at the end of your deliberations, having taken into account the evidence both for the Crown and the accused, and after also taking into account the submissions made by each party, you are not satisfied that the Crown has established any one of the essential elements to the charges beyond reasonable doubt, then it is your duty to bring in a verdict of not guilty because the Crown will have failed to do what the law requires the Crown to do in a criminal trial. It is vitally important that you clearly understand that the accused must be found not guilty if his guilt has not been proven to your satisfaction beyond reasonable doubt. It follows from this of course that if you are left unable to decide whether the Crown has proved its case in relation to any essential element, even though you feel that he may be guilty, if you have a reasonable doubt in respect of that matter, the accused is entitled to the benefit of that doubt and you must find him not guilty."

94. [94]

    The trial judge then directed the jury that that standard of proof only applied to the elements of the offences, saying "[t]he Crown does not have to prove, however, every single fact in issue in the trial beyond reasonable doubt. The onus which rests upon the Crown is to prove the elements of each of the seven charges beyond reasonable doubt".

95. [95]

    The trial judge read out the "essential facts" for each offence from an "Elements document". By way of example, in relation to count 1, the trial judge said: "Count 1, sexual intercourse with child between ten and 14 years. The accused is charged with sexual intercourse with the complainant, a child then of the age of ten and under the age of 14 years, namely ten years. The Crown case is that on or about 26 September 2020 the complainant was at the accused's house having a sleepover. It is alleged the complainant fell asleep whilst she was on the couch watching a movie. When she woke up the accused was touching her on the inside of the lips of her vagina. The defence case is that the accused did not touch the complainant as alleged. To prove the accused is guilty, the Crown must prove beyond reasonable doubt each of the two elements which make up the offence. 1. the accused had sexual intercourse with the victim. the victim was then of the age of ten years and under the age of 14 years. ... You can only find the accused guilty if the Crown proves each element beyond reasonable doubt. If the Crown fails to prove any one of these elements you must find the accused not guilty." The "essential facts" for the Narara incidents were expressed in similar terms.

96. [96]

    The trial judge then directed the jury that they needed to consider each charge separately and that there was no legal requirement that the verdicts all be the same. The jury were directed that, if they found AR not guilty of any count, "particularly if that was because you had doubts about the reliability of the complainant's evidence, you would have to consider how that conclusion affected your consideration of the remaining counts".

97. [97]

    Next, the trial judge directed the jury that: "[I]n a criminal trial you must be satisfied of the guilt of the accused beyond reasonable doubt. Amongst other things that means you should be extremely careful about drawing inferences. You should examine any possible inference to ensure that the inference you are drawing is a justifiable inference. In the context of a criminal trial you should not draw an inference from the direct evidence unless it is a rational inference in all the circumstances. And, as I said, in the present case the Crown is going to ask you to draw an inference or conclusion from the facts it seeks to establish that the accused had a tendency to have a sexual interest in the complainant and act on that sexual interest".

98. [98]

    In describing the evidence, the trial judge often reiterated that it did not matter if the Crown could not prove precisely when the offending occurred, "as long as you are satisfied that the events happened and they are established beyond reasonable doubt".

99. [99]

    The trial judge then addressed the tendency evidence. It is in these directions that the risk of confusion about the standard of proof emerges with greatest clarity. Her Honour told the jury that the Crown said the jury would be satisfied that the accused, AR, had the asserted tendency "because of his conduct in relation to the three occasions it is alleged that he sexually assaulted the complainant". The trial judge later directed that "[t]he Crown argues that the conduct of the accused on each of the three alleged occasions establishes the tendency asserted. The defence argues that you would not be satisfied that any of the conduct alleged on the three occasions has occurred so that the tendency would not be established."

100. [100]

     The trial judge warned the jury that "[a]lthough you are looking at the three occasions of alleged conduct that are the subject of charges, you are not when considering them as tendency evidence considering whether those episodes of misconduct have been proved beyond reasonable doubt". The trial judge stated that "[f]or the purpose of determining whether the accused has the relevant tendency, you do not consider each of the three occasions in isolation, you look at them collectively to decide what conduct occurred". The trial judge told the jury to bear in mind the direction her Honour had given earlier about drawing inferences when they were "considering the inferences the Crown asked you to draw in relation to tendency reasoning". The trial judge said, "[i]f you find the accused did have a tendency to have a sexual interest in the complainant, and a tendency to act on that sexual interest in the way alleged by the Crown, then you can use that in considering whether it is more likely he committed the specific offences with which he is charged". The trial judge said, "[i]f you are not satisfied that any of the conduct the Crown relies upon occurred, then there is no basis upon which the tendency could be inferred. In these circumstances the whole issue is put to one side." The trial judge warned that "[f]inding the accused did have the tendency the Crown asserts is not enough by itself to prove his guilt beyond reasonable doubt in relation to each charge. It is simply part of the process of reasoning and it may assist the Crown to prove the accused committed each of the offences."

101. [101]

     The trial judge said: "The question is whether [the tendency] makes it more likely the accused conducted himself in the way the Crown alleges on any of the occasions that are subject of the charges. That is the only way the accused['s] tendency to have a sexual interest in the complainant and act on that sexual interest in the manner alleged may be used. Ultimately you must decide whether the specific offences with which the accused has been charged have been proved beyond reasonable doubt. That decision must be based upon the evidence relevant to each charge. This includes the evidence of the complainant about what the accused did, it will include the tendency alleged by the Crown provided you are satisfied it has been established, it will also include all of the other evidence you have heard in both the Crown and defence cases. When considering whether a charge has been proved you will have to decide whether the Crown has proved the essential elements of that charge."

102. [102]

     Later in the summing up, the trial judge reminded the jury that the Crown relied on tendency evidence and that the tendency evidence should only be taken into account if the jury accepted that the tendency had been established by the Crown. The trial judge reiterated the criminal standard of proof later in the summing up.

103. [103]

     These directions did not ameliorate the risk of flawed or impermissible reasoning. It may be accepted that the directions frequently reiterated to the jury the criminal standard of proof. However, that did not sufficiently ameliorate the risk that the jury would decide AR's guilt on a standard below beyond reasonable doubt. The jury were told that only the elements of the seven charges needed to be proved beyond reasonable doubt, not every single fact in issue. However, that is not necessarily helpful in circumstances where the tendency was framed by reference to the elements of the seven charges.

104. [104]

     The jury were told of the "essential facts", in the "Elements document", in terms broadly replicated in the tendency notice. In other words, the facts that the jury were told were "essential" were the very facts they were also told they needed to be satisfied of in assessing the asserted tendency. The jury were told they should be careful when drawing inferences and then told that the Crown was going to ask them to draw an inference from "the facts it seeks to establish that the accused had a tendency". In other words, the jury were told, in effect, that they would be inferring from the evidence of the charged acts whether the charged acts occurred. This is potentially misleading, given the jury were separately required to do the same exercise on a higher standard of proof, namely beyond reasonable doubt. In describing the evidence, the trial judge often reiterated that, while peripheral details did not need to be proved, the alleged conduct itself needed to be proved beyond reasonable doubt. Again, this is not necessarily helpful in circumstances where the alleged conduct is the very thing said to give rise to the asserted tendency.

105. [105]

     Although the trial judge directed the jury that they were not applying the standard of beyond reasonable doubt to the charged conduct for the purpose of assessing whether the tendency was established but were required to apply that standard to the charged conduct for the purpose of assessing whether AR was guilty, that was insufficient to ameliorate the risk that the standard of proof was undermined. The jury were told, in effect, that they needed to consider whether the charged conduct occurred in order to determine whether the tendency to do the charged conduct was established, in order to determine whether the charged conduct occurred. In other words, the jury needed to consider the same evidence in order to reach the same conclusion, but at different stages of the analysis and on different standards of proof. As has been explained, this was apt to confuse.

106. [106]

     The potential for confusion was highlighted by the trial judge's direction that "[i]f you are not satisfied that any of the conduct the Crown relies upon occurred, then there is no basis upon which the tendency could be inferred. In these circumstances the whole issue is put to one side." In fact, contrary to this direction, if the jury were not satisfied that any of the conduct occurred, they would put not just the issue of tendency aside, but the entire Crown case aside. Directions that should have been given

107. [107]

     It was at least necessary for a Roder direction to be given, namely that: "[the direction] should indicate the evidence relied on to support the alleged tendency, direct the jury to consider whether they are satisfied of the alleged tendency and then advise the jury that, if they are so satisfied, they can use that tendency in considering whether it is more likely that the accused committed the specific offences with which he or she is charged".

108. [108]

     That direction directs the jury to consider whether they are satisfied of the alleged tendency, as opposed to whether they are satisfied of the charged conduct. The effect of a Roder direction is that the jury are not required to make a preliminary finding as to whether the charged conduct occurred, at an indeterminate standard of proof. Rather, the jury are required to make a finding as to whether the accused had the alleged tendency. In other words, a Roder direction effects a clearer delineation between proof of the tendency and proof of the charged conduct.

109. [109]

     However, given the specificity with which the tendency was framed in this case, a Roder direction may not have sufficiently ameliorated the risk of impermissible reasoning. It is not the case that a Roder direction could never be given in a sole complainant case where the tendency is framed by reference to the charged conduct and relies solely on evidence of the charged conduct. To the extent the CCA held otherwise, it was wrong to do so.

110. [110]

     However, had the tendency been framed with less specificity, a Roder direction may have sufficiently ameliorated the risk of impermissible reasoning. As properly accepted by AR's counsel, the tendency could have been framed in the following, less specific, way: "The tendency sought to be proved is [AR's] tendency to: (a) have a particular state of mind, namely a sexual interest in [the complainant]; and (b) act on that sexual interest when she had fallen asleep beside him watching a movie."

111. [111]

     Where the tendency is framed in this way, a Roder direction would require the jury to consider the evidence of the charged conduct to decide whether AR had a tendency to act on his sexual interest, and then to decide whether the charged conduct was established beyond reasonable doubt. This can be distinguished from a chain of reasoning which required the jury to make, in effect, a preliminary finding as to whether the charged conduct itself was established, before later finding whether the charged conduct was established beyond reasonable doubt. The tendency framed in this way would not require the jury to make a preliminary finding as to whether the charged conduct was established on some indeterminate standard of proof. Rather, it simply involves the same evidence being used "at different stages of [the jury's] deliberations with a different onus of proof and for a different purpose", which was considered permissible by this Court in Roder.

112. [112]

     Contrary to the Crown's contention, framing the tendency with less specificity is not inconsistent with this Court's decision in Hughes v The Queen. Hughes was a case involving multiple complainants. This Court in R v Dennis Bauer (a pseudonym) expressly held that, in single complainant cases, where the tendency is framed by reference to the sole complainant, there need be no "particular feature" of the tendency for it to have significant probative value.

113. [113]

     Given that the tendency relied upon by the Crown at trial was framed specifically, as opposed to generally, very clear and careful directions needed to be given to the jury, in addition to a Roder direction, to sufficiently ameliorate the risk of impermissible reasoning. These directions would have needed to explain the risks identified earlier and to direct the jury against being distracted by the tendency from the need to be satisfied of AR's guilt beyond reasonable doubt. Miscarriage of justice

114. [114]

     In this case, there was a miscarriage of justice. The specificity with which the tendency was framed, in light of the circumstances of the case, gave rise to a substantial risk that the jury would not properly apply the correct standard of proof to the charged conduct. The directions given by the trial judge did not ameliorate that risk. Rather, the directions given by the trial judge were apt to exacerbate that risk.

115. [115]

     Given the substantial risk that the jury failed to properly apply the standard of proof (a risk unmitigated by the trial judge's directions), the framing of the tendency and the trial judge's directions could realistically have affected the reasoning of the jury to the verdict of guilt in the trial that was had.

116. [116]

     The Crown submitted that because the jury found AR not guilty of the West Gosford incident, but guilty of the Narara incidents, it follows that the jury must not have been misled from applying the proper standard of proof to each of the charges. That contention cannot be accepted in circumstances where the evidence in relation to the West Gosford incident was not as strong as the evidence in relation to the Narara incidents. The complainant made complaint about the Narara incidents a few weeks after they allegedly occurred. In the course of several complaints and police interviews about the Narara incidents, the complainant did not complain about the West Gosford incident, despite that allegedly occurring around the same time as the Narara incidents and being similar in nature to the Narara incidents.

117. [117]

     The complainant made complaint about the West Gosford incident over a year after complaining about the Narara incidents. The complainant explained her delay in complaining about the West Gosford incident on the basis that she thinks she blocked it out and was prompted to recall it after having a vivid dream. The West Gosford incident was charged as the "first" incident, on the basis of various text messages about when the sleepover took place. However, the complainant's evidence was that the West Gosford incident was the last occasion on which the offending occurred. In these circumstances, the jury's inconsistent verdicts were capable of being explained by the differing evidence in relation to the different incidents. The inconsistent verdicts were not necessarily indicative of the jury properly applying the standard of proof to all charges. Order

118. [118]

     The appeal should be dismissed. Introduction

EDELMAN J.

1. [119]

   Consider a simplified version of the facts of this case. Suppose the only evidence at the trial of two alleged sexual assaults is that of a complainant. The complainant says that two sexual assaults occurred within the space of a few days. In both instances, the sexual assault involved the accused penetrating the complainant's vagina with his fingers when she had fallen asleep beside him watching a movie.

1. [120]

   Can a trial judge direct a jury that they can reason in the following way: (1) consider whether the two alleged assaults (probably) occurred; (2) if the two assaults (probably) occurred, consider whether the accused had a tendency to engage in the offending conduct; and, if so, (3) use that tendency to assist in considering whether to draw an inference beyond reasonable doubt that either of the two alleged assaults was committed?

2. [121]

   As a matter of pure logic, if a juror considers that it is probable that an accused engaged in the same conduct involving the same complainant on, say, 100 occasions then the juror might use the probable occurrence of those 100 offences as a tendency to assist in concluding, beyond reasonable doubt, that the conduct occurred on any of the particular, single occasions. In clear and cogently presented submissions, senior counsel for the Crown properly accepted that "[a]t a level of principle" the effect of the Crown's submission was that this same logic would be equally applicable to the use of tendency reasoning based upon a mere two occasions.

3. [122]

   So it is. In the realm of pure abstract logic, a juror might obtain some support in a reasoning process to a finding of guilt from a finding of a (probable) tendency based on two asserted offences in the same week and involving the same complainant. But that support comes with the grave danger of diluting the burden of proof for each offence. That danger could only be avoided, if at all, by the most careful and powerful directions about how a tendency in such circumstances is to be used.

4. [123]

   The essence of the reasoning of the New South Wales Court of Criminal Appeal in this case was that "[a]pplying the relevant reasoning of Beech-Jones CJ at CL in [Kanbut v The King], the tendency direction in the present case ... did not adequately direct the jury and it was likely that the jury's attention would be deflected from applying the required standard of proof in respect of the offences charged". Similar reasoning was adopted by this Court in Director of Public Prosecutions (Vic) v Benjamin Roder (a pseudonym), when this Court emphasised that in cases which involve a tendency direction "[c]areful directions should be given to the jury as to the requisite onus and standard of proof as well as to the contents of the elements of the offence and the need for separate consideration of each charge".

5. [124]

   In my view, the Court of Criminal Appeal was correct in the application of this aspect of the reasoning of Beech-Jones CJ at CL in Kanbut, which also reflects the reasoning of this Court in Roder. The appeal must be dismissed. Tendency evidence and its two key elements

6. [125]

   The use of tendency evidence involves a particular method of inferential reasoning employed in ordinary, everyday reasoning. In criminal law, as in everyday reasoning, the process involves drawing inferences from an intermediate conclusion that a person has been known to act in a similar manner on other occasions to a final conclusion that the person acted in the similar way on the occasion in question. In criminal law, other than circumstances where the tendency is an indispensable element of proof, or a necessary link in the chain of reasoning to guilt, the existence of a tendency can be established merely as a probability.

7. [126]

   Some relevant aspects of tendency evidence in criminal proceedings under the Evidence Act 1995 (NSW) are not controversial. First, by s 97(1), tendency evidence is not admissible unless it has significant probative value. Secondly, by s 101(2), the tendency evidence cannot be used against a defendant unless the probative value of the evidence outweighs the danger of unfair prejudice to the defendant. Thirdly, for the purposes of ss 97(1)(b) and 101(2), s 97A creates a presumption that there is significant probative value in tendency evidence about "the sexual interest the defendant has or had in children (even if the defendant has not acted on the interest)" and tendency evidence about "the defendant acting on a sexual interest the defendant has or had in children".

8. [127]

   In Hughes v The Queen, four members of this Court said that the assessment of whether evidence has significant probative value within the application of s 97(1) of the Evidence Act, either by itself or having regard to other evidence adduced by the party seeking to adduce the asserted tendency evidence, involves two interrelated but separate matters: (1) "the extent to which the evidence supports the tendency"; and (2) "the extent to which the tendency makes more likely the facts making up the charged offence". Just as Daedalus warned Icarus, the prosecution faces dangers at two extremes when drafting a tendency notice unless careful regard is had to both of these matters.

9. [128]

   The first danger arises if the tendency notice is drafted only to meet (1) in terms that are so general or abstract that the tendency has little value. For instance, if a tendency notice in a sexual assault case involving several counts were drafted merely as a tendency to commit sexual assault, the complainant's evidence might fall precisely within that proposed tendency but unless there are a vast number of complainants the proposed tendency would be very unlikely to have significant probative value in an assessment of whether any of the particular offences had occurred. In other words, a tendency notice that focuses only upon (1) might fail to have significant probative value because of (2).

10. [129]

    On the other hand, the second danger arises if the tendency notice is drafted only to meet (2), in terms that correspond almost precisely with the facts of a charged offence but which do not correspond closely with the evidence of other offences said to support the tendency. For instance, consider a sexual assault case that involved four counts over a period of two years. If a tendency notice were drafted in terms that corresponded with evidence of two of the offences over a two-day period, but not the other two over the remainder of the two-year period, then such a tendency (if it existed) might make those other two offences more likely but the evidence might not provide strong support for the tendency. One reason that the evidence might not provide strong support for the tendency is that the tendency is based only upon two occasions. In other words, a tendency notice that focuses only upon (2) might fail to have significant probative value because of (1).

11. [130]

    The relevance of these two dangers is subject to the provision concerning child sex offence trials in s 97A(5) of the Evidence Act. That sub-section was not in issue in this case. In any event, in all cases where a tendency is relied upon, an example of a Goldilocks tendency notice, which addressed both (1) and (2) to a significant extent, would be a notice that described circumstances that are nearly identical to proposed evidence of many acts by the accused, whether charged or uncharged. The tendency notice would thus express the tendency in terms which are extremely specific and which correspond closely with the evidence of all offences that support the tendency. Far from being a weakness in a tendency notice, a tendency that is expressed in precisely the same terms as the facts making up the charged offence can be highly cogent, with very significant probative value. The matters making up such a tendency might meet the description once used by the common law of showing a "striking similarity". Tendency evidence and a danger of confusing the burden of proof

12. [131]

    Suppose that there is contested evidence from five people that every morning for several months they saw Z perform a highly unusual ritual at the time of waking up each morning. The evidence would support a tendency of Z to perform this ritual. If the five people were absent for one day in that period, the tendency would have significant probative value in supporting other evidence that suggested that, on that day, Z had performed the ritual.

13. [132]

    At the other extreme, suppose that there is contested evidence from only one person that on only one occasion they saw Z perform the unusual ritual. It is a logical and semantic error to use an intermediate conclusion that this contested evidence establishes that Z probably had a tendency to perform the ritual in order to assist in drawing an ultimate conclusion, beyond reasonable doubt, that, on the same occasion, Z performed that ritual. Whether such reasoning is called circular logic or bootstraps reasoning, it involves logical error. The single-act tendency is being used to prove itself.

14. [133]

    Close to this erroneous reasoning, although not a logical or semantic error, consider a third example. Suppose that there is contested evidence from only one person that on two occasions they had seen Z perform this unusual ritual. It is possible, but might be very difficult, to describe the evidence of only two occasions as showing that Z probably had a tendency to perform this ritual. It is also possible, but again might be very difficult, to reason that if Z probably had this tendency, then this tendency provides some support for a conclusion, beyond reasonable doubt, that Z performed this ritual on one of those occasions.

15. [134]

    There are two aspects of the tendency evidence in this third example that might suggest, at first blush, that this reasoning process is cogent. First, the event that is sought to be inferred beyond reasonable doubt is identical to the tendency. Secondly, the tendency involves conduct that is highly unusual. But those matters of potentially increased cogency make the use of such tendency reasoning particularly dangerous without clear emphasis upon two countervailing considerations which will usually make the reasoning process described in the paragraph above very difficult: (1) the evidence is given by only one person and only in relation to two occasions; and (2) the relevant event is one of the very two matters from which the tendency was derived. The point that is fundamental to this appeal is that without a strong emphasis in the trial judge's tendency direction upon these two countervailing considerations there is a grave danger of confusion of the burden of proof. The process of a jury using tendency evidence

16. [135]

    In determining whether evidence supports a tendency, this Court has said that a tendency direction "should not direct or invite the jury to make findings in respect of charged conduct". These remarks were made in the context of a tendency comprised of both charged and uncharged acts, but they are equally applicable to a tendency comprised only of charged acts.

17. [136]

    The statement quoted above, to which I was a party, is, unfortunately, pregnant with ambiguity. Does the "should" mean that the statement is merely a hopeful counsel of perfection (ideally, such a direction should not be given but it might not be an error to give it)? Or is the statement a proposition of law such that "should" means "must"? Is the purpose of such an exhortation or prohibition to prevent such reasoning by a jury? If so, how could a jury ever be expected to reach a conclusion that an accused had a tendency to engage in particular conduct if the jury were precluded from making findings about whether the accused had probably engaged in conduct on some, many, or all of the occasions from which the tendency was said to be drawn?

18. [137]

    My purpose, as party to this passage in Roder, was not to suggest that there was any logical reason for the jury to avoid making findings as to whether the accused had probably engaged in the conduct said to comprise the tendency on any occasion. Indeed, it is impossible for a jury to reach any conclusion about the existence of an alleged tendency without deciding whether: (i) at least some of the conduct said to be the basis for the tendency had occurred on the balance of probabilities; or, (ii) arguably in cases of sufficiently numerous acts, some or all of the conduct had possibly occurred. My purpose for joining in the passage in Roder that emphasised judicial abstention was to avoid "undermining the jury's understanding of the necessity for proof of the elements of the charge beyond reasonable doubt".

19. [138]

    Suppose that a tendency was alleged which relied upon two acts of extremely similar charged conduct. Suppose also that the jury were properly directed, including that they were not given a direction to consider whether those acts had probably occurred when they consider whether the tendency arose. It is extremely difficult to see how the jury could conclude that the accused had a tendency to engage in those acts unless the jury were satisfied that both of the acts had probably occurred. It is extremely likely, therefore, that for the purposes of establishing the tendency the jury will nevertheless consider whether the acts probably occurred. And in that consideration lies danger.

20. [139]

    The danger that lies in that consideration is that the jury will confuse the burden in finding the tendency, as a matter on the balance of probabilities, with the burden of making the ultimate finding beyond reasonable doubt of whether either of the same two acts had occurred. The danger exists even where the judge abstains from directing the jury to make any findings about the conduct for the purposes of finding the tendency. A fortiori, where the judge does direct the jury to consider, for the purposes of establishing the tendency, whether the acts occurred, the danger becomes much more pronounced. This danger can give rise to a miscarriage of justice without powerful, clear, and careful directions on the burden of proof. The decision in Kanbut and the Roder rider

21. [140]

    A heavy focus of AR's submissions in this case was upon the reasoning of Beech-Jones CJ at CL (with whom Adamson and Campbell JJ agreed) in Kanbut. Ms Kanbut was found guilty of offences including four slavery offences against two complainants, X and Y. The Crown had relied upon an alleged tendency to engage in the acts comprising the slavery offences. The tendency in Kanbut, as expressed in the tendency notice served under s 97(1)(a) of the Evidence Act and as directed by the trial judge, was in highly specific terms, relying upon all the particulars of the offending against the two complainants, and nothing more. In other words, the Crown relied upon the evidence of the two complainants constituting the charges against Ms Kanbut as supporting a tendency to engage in that conduct, from which the jury could reason that any of the charges had been proved beyond reasonable doubt.

22. [141]

    The New South Wales Court of Criminal Appeal unanimously allowed the appeal. Beech-Jones CJ at CL identified numerous problems with the tendency direction by the trial judge. Many of those problems are irrelevant to this appeal. But one relevant problem identified by his Honour was that the tendency was expressed in the same terms as the facts which constituted the alleged offence. His Honour set out the two key aspects of tendency evidence by reference to the observation in Hughes (set out above at ) and then said that "[i]mplicit in that observation is that the tendency is not expressed in precisely the same terms as the facts making up the charged offence".

23. [142]

    As explained above, the two key aspects of tendency evidence referred to in Hughes will not always prevent a tendency notice from being expressed in the same terms as the facts making up the charged offence. In some cases, it is entirely appropriate for a tendency notice to do so. If 100 offences were alleged to have been committed in nearly identical, and highly unusual, circumstances then it would be entirely appropriate, and highly probative, for a tendency to be expressed in precisely the same terms as the 100 offences. The relevant problem with the tendency direction in the circumstances in Kanbut, and the real concern of the Court of Criminal Appeal in that case, must have been the possibility for confusion about the burden of proof that a tendency direction creates when the tendency is based on a very limited number of charged acts over a short period of time and involving the same complainant or complainants.

24. [143]

    In Roder, this Court issued the strong cautionary rider for tendency directions: "[I]t follows from the nature of tendency evidence that, in a case where the prosecution relies on both uncharged and charged acts to establish an alleged tendency of the kind under consideration here, a single separate tendency direction should ordinarily be given. Such a direction should not direct or invite the jury to make findings in respect of charged conduct, but instead should indicate the evidence relied on to support the alleged tendency, direct the jury to consider whether they are satisfied of the alleged tendency and then advise the jury that, if they are so satisfied, they can use that tendency in considering whether it is more likely that the accused committed the specific offences with which he or she is charged. Careful directions should be given to the jury as to the requisite onus and standard of proof as well as to the contents of the elements of the offence and the need for separate consideration of each charge."

25. [144]

    The importance of all of the matters in this cautionary rider increases progressively the more that: (i) the tendency is comprised of charged acts rather than uncharged acts; (ii) the number of charged acts reduces; (iii) the number of complainants reduces; (iv) the charged acts are in closely associated circumstances; and (v) the tendency is expressed in particular terms, closely related to the elements of the charged acts. The greatest danger arises where the tendency is comprised only of two charged acts by a single complainant, where the charges involve closely associated circumstances, and with the tendency expressed in terms of the elements of the offence comprised in the charged acts. The circumstances in this case involve very similar conditions. The essential facts and asserted tendency evidence in this case

26. [145]

    The essential facts were set out concisely in the reasons of the Court of Criminal Appeal and can be summarised as follows. From April or May 2019, AR was in an occasional relationship with the complainant's mother. From June 2020, AR and the complainant began texting each other. The complainant was 10 years old at that time. AR was 41 years old. The Crown case against AR was based on three incidents that allegedly occurred during the school holiday period in September/October 2020.

27. [146]

    The "first Narara incident" (count 3 and in the alternative count 4) was alleged to have occurred during the second week of the school holiday period in September/October 2020 when AR was sleeping over at the complainant's mother's house. The complainant's evidence was that her mother had gone to bed and the complainant's brother was in his room on a computer. The complainant was lying on the lounge with her head on AR's lap. She fell asleep and when she woke up, she realised that AR's hand was on the outside and inside of her vagina.

28. [147]

    The "second Narara incident" (count 5 and in the alternative count 6, and count 7) was also alleged to have occurred during the second week of the same school holiday period in September/October 2020, about two or three days after the first Narara incident. The second Narara incident was again alleged to have occurred when AR was present at the complainant's mother's house. AR and the complainant were alone in the lounge room, and were watching a movie on the lounge. The complainant fell asleep on the lounge and woke up to find AR touching her on the inside and outside of her vagina.

29. [148]

    There was another incident (counts 1 and 2) which was alleged to have occurred, which was described by the Court of Criminal Appeal as the "West Gosford incident". The complainant's evidence was that although this complaint was first made in a police interview a year and nine months after the first complaint about the Narara incidents and about three weeks before AR's trial was due to commence, the West Gosford incident was recalled by her after a "vivid dream" about being locked in an apartment which prompted her memory of a time when she slept at AR's home in West Gosford and he had sexually touched her. Although the West Gosford incident was alleged by the Crown to have preceded the other two incidents, the complainant's evidence was that the West Gosford incident had followed the other two incidents. The Court of Criminal Appeal rightly described the counts concerning this incident as having "comparative frailty" when compared with the counts concerning the two Narara incidents.

30. [149]

    As part of the Crown case, the Crown issued a tendency notice in the following terms: "The tendency sought to be proved is [AR's] tendency to: (a) have a particular state of mind, namely a sexual interest in [the complainant]; and (b) to act on that sexual interest by penetrating her vagina with his fingers and/or sexually touching her when she had fallen asleep beside him watching a movie."

31. [150]

    The Crown sought to establish the tendency by "the evidence with respect to the counts for each one of those three occasions (represented by Counts 1-2, Counts 3-4 and Count[s] 5-7) as tendency evidence in respect to Counts representing the other occasions".

32. [151]

    The jury convicted AR of the first Narara incident (count 3, with the alternative count 4 being inapplicable) and the second Narara incident (counts 5 and 7, with the alternative count 6 being inapplicable) but acquitted AR of the West Gosford incident (counts 1 and 2). There is a real possibility that the tendency evidence relied upon by the jury was only the evidence concerned with the two Narara incidents. The trial judge's tendency direction

33. [152]

    The relevant part of the tendency direction given by the trial judge was as follows: "The Crown says that the three occasions when the accused sexually assaulted the complainant reveals the accused had a tendency to have a sexual interest in the complainant and to act on that sexual interest, which makes it more likely that he committed each of the offences charged in the indictment. You will need to consider the evidence relating to the alleged conduct of the accused on the three occasions and decide whether he did in fact conduct himself in the way the Crown alleges. Although you are looking at the three occasions of alleged conduct that are the subject of charges, you are not when considering them as tendency evidence considering whether those episodes of misconduct have been proved beyond reasonable doubt. For the purpose of determining whether the accused has the relevant tendency, you do not consider each of the three occasions in isolation, you look at them collectively to decide what conduct occurred. If you decide that all or at least some of the conduct occurred, you then need to consider whether it enables the inference to be drawn that the accused has the tendency asserted by the Crown. That is, to have a sexual interest in the complainant and to act on that sexual interest in the manner alleged by the Crown. ... If you are not satisfied that any of the conduct the Crown relies upon occurred, then there is no basis upon which the tendency could be inferred. In these circumstances the whole issue is put to one side. If you find the accused did have a tendency to have a sexual interest in the complainant, and a tendency to act on that sexual interest in the way alleged by the Crown, then you can use that in considering whether it is more likely he committed the specific offences with which he is charged. However, it is essential you consider in relation to each charge whether the accused acted in that particular way and had that particular state of mind on that specific occasion. Finding the accused did have the tendency the Crown asserts is not enough by itself to prove his guilt beyond reasonable doubt in relation to each charge. It is simply part of the process of reasoning and it may assist the Crown to prove the accused committed each of the offences. The question is whether it makes it more likely the accused conducted himself in the way the Crown alleges on any of the occasions that are subject of the charges. That is the only way the accused['s] tendency to have a sexual interest in the complainant and act on that sexual interest in the manner alleged may be used. Ultimately you must decide whether the specific offences with which the accused has been charged have been proved beyond reasonable doubt. That decision must be based upon the evidence relevant to each charge. This includes the evidence of the complainant about what the accused did, it will include the tendency alleged by the Crown provided you are satisfied it has been established, it will also include all of the other evidence you have heard in both the Crown and defence cases. When considering whether a charge has been proved you will have to decide whether the Crown has proved the essential elements of that charge. I will now summarise the case for the Crown and the case for the accused on this issue of tendency. The Crown argues that the conduct of the accused on each of the three alleged occasions establishes the tendency asserted. The defence argues that you would not be satisfied that any of the conduct alleged on the three occasions has occurred so that the tendency would not be established." Four errors or irregularities in the tendency direction

34. [153]

    There were four errors or irregularities in the tendency direction given by the trial judge.

35. [154]

    First, the reference in the direction to the possibility of a tendency being established if "at least some of the conduct occurred" and to there being no basis to infer a tendency only if the jury were not satisfied that "any of the conduct the Crown relies upon occurred" permitted: (i) the semantic distension of a tendency being established on the basis of finding that a single charged act probably occurred; and (ii) the logical impossibility of using that probable single-act tendency to prove that the same single act occurred beyond reasonable doubt. In other words, by referring to "some of the conduct" and "any of the conduct", the direction permitted the logically impossible approach by which the jury could reason that: (i) it was probable that one of the charged acts had occurred; (ii) the probable occurrence of that act meant that AR had a tendency to engage in conduct of that nature; and (iii) the probable tendency of AR to engage in conduct of that nature assisted in proving beyond reasonable doubt that AR engaged in conduct of that nature on that same occasion. This error can, however, be put to one side because, in light of the convictions on counts 3, 5 and 7, the jury must have been satisfied that the conduct occurred on at least two of the occasions.

36. [155]

    Secondly, the direction does exactly what this Court in Roder said should not be done: "a direction should not direct or invite the jury to make findings in respect of charged conduct". Although that statement was made in the context of a tendency comprised of a combination of charged and uncharged acts, it applies equally to a tendency based only upon charged acts. Nevertheless, as the Court of Criminal Appeal rightly observed in this case, consistently with the reasoning of Basten A-JA in JS v The Queen, a direction to this effect will not necessarily amount to a miscarriage of justice. Whether such a direction amounts to a miscarriage of justice needs to be considered in light of the summing up as a whole. In particular, the whole of the summing up needs to be scrutinised in light of the potential for confusion of the burden of proof.

37. [156]

    Thirdly, in addition to the failure of the direction to conform with the approach suggested in the remarks made (after the trial in this case) by this Court in Roder, the trial judge did not direct the jury about two important points that might have ameliorated the dangers of confusing the burden of proof: (1) that the jury should bear in mind that the tendency evidence is given by only one person and only in relation to three alleged occasions; and (2) that establishing a tendency by reliance on a limited number of alleged incidents might not, by itself, be sufficient to find beyond reasonable doubt that any of those incidents occurred.

38. [157]

    Fourthly, although the trial judge did tell the jury elsewhere in the directions that they should consider each count separately, the trial judge did not tell the jury to bear in mind, before using any conclusion about tendency in the separate consideration of each count, that the tendency conclusion had not been reached beyond reasonable doubt. The confusion for the jury and the miscarriage of justice

39. [158]

    Although the tendency upon which the trial judge directed was comprised of three events, the comparative frailty of the evidence of the West Gosford incident meant that AR was correct to submit on this appeal that there is a very real likelihood in this case that any tendency reasoning of the jury was based only upon two events. The conviction of AR upon each of counts 3, 5 and 7 makes it a real possibility that the jury, following the directions of the trial judge, found the tendency to be established upon being satisfied that the two incidents probably occurred.

40. [159]

    The two events upon which the jury's tendency reasoning may have been based were closely related in time and comprised of substantially identical allegations for the purposes of the tendency. The jury were confronted with the cognitive struggle of using a conclusion that the two incidents had probably occurred as part of an inference that the two incidents had occurred beyond reasonable doubt. They were given little assistance to guide them through the mental confusion that might ensue and the consequent danger of diluting the burden of proof.

41. [160]

    The jury were not told of the limitations of the tendency evidence, namely that it was given by only one person and only in relation to three alleged occasions. The jury were not reminded that although the alleged facts comprising each of the elements of the offence were closely related to the tendency, those elements still needed to be proved beyond reasonable doubt. In other words, the jury were not reminded that any tendency found might not, by itself, be sufficient to establish that any charge had been proved beyond reasonable doubt.

42. [161]

    The directions did not sufficiently remove this potential for confusion and miscarriage of justice. Although it is unnecessary to decide, it may be that this cognitive struggle could have been overcome by a direction to the following effect: "In a case of this nature where the tendency evidence is based upon the evidence of a single complainant and involves the facts of a limited number of closely related allegations, you must be very careful always to bear in mind when using the tendency evidence that the evidence does not, by itself, establish that each of the counts is proved beyond reasonable doubt. You should not allow your findings as to tendency to distract you from your ultimate duty before convicting the accused of any charge to find that particular charge proved beyond reasonable doubt." Conclusion

43. [162]

    Although in dissent, I have considered it necessary in these reasons to set out a large part of the trial judge's directions in relation to the tendency evidence. I have not done so to suggest that the direction is a model direction. On any view, there were serious errors or irregularities in the direction. On any view, although the jury did not need to follow one aspect of the direction (finding a tendency based on a single act and using that single-act tendency to prove itself beyond reasonable doubt), that aspect permitted a logically impossible form of reasoning. On any view, the direction also encouraged reasoning that amounts to what this Court subsequently said in Roder should not be done. On any view, the directions given to the jury should have been more cautious and conscious of the dangers that can arise by reliance upon a tendency in such circumstances.

44. [163]

    The appeal should be dismissed. Overview of the appeal

JAGOT J.

1. [164]

   The Crown appeals on two grounds, that the Court of Criminal Appeal of the Supreme Court of New South Wales (Wright, Ierace and Rigg JJ) erred in concluding, first, that a miscarriage of justice was occasioned by the trial judge (Judge Bright in the District Court of New South Wales) misdirecting the jury in law and, second, that the Crown's reliance upon the same evidence to establish the tendency of the respondent as the evidence relied upon to establish the counts on the indictment against the respondent was "inconsistent with the nature of tendency evidence". These grounds of appeal must be allowed and orders must otherwise be made as proposed by Gageler CJ, Gleeson and Beech-Jones JJ.

1. [165]

   The respondent was charged on indictment of sexually assaulting a child, GC, who at the time was 10 years old, on three separate occasions: 26 September 2020 at West Gosford (counts 1 and 2), between 24 September 2020 and 8 October 2020 at Narara (counts 3 and 4), and between 3 and 8 October 2020 at Narara (counts 5, 6 and 7). The jury convicted the respondent of counts 3, 5 and 7 at Narara and acquitted the respondent of counts 1 and 2 at West Gosford. GC was the child of the respondent's former partner, JW.

2. [166]

   The Crown gave notice of its intention to adduce evidence to prove that the respondent had a tendency as required by s 97(1)(a) of the Evidence Act 1995 (NSW). The notice asserted that the respondent had a tendency to: "(a) have a particular state of mind, namely a sexual interest in [GC]; and (b) ... act on that sexual interest by penetrating [GC's] vagina with his fingers and/or sexually touching her when she had fallen asleep beside him watching a movie". The notice identified the evidence relied upon to prove the asserted tendency as "the evidence with respect to the counts for each one of [the] three occasions (represented by Counts 1-2, Counts 3-4 and Count[s] 5-7) as tendency evidence in respect to Counts representing the other occasions".

3. [167]

   The trial judge admitted the tendency evidence, about which there is no complaint in this appeal. In summing up the trial judge gave a direction to the jury about the tendency evidence which, in part, said that the jury "need[ed] to consider the evidence relating to the alleged conduct of the [respondent] on the three occasions and decide whether he did in fact conduct himself in the way the Crown allege[d]". The trial judge stated that the jury, when considering the tendency evidence, was not required to consider "whether those episodes of misconduct ha[d] been proved beyond reasonable doubt", and that if the jury decided that "all or at least some of the conduct occurred, [the jury] then need[ed] to consider whether it enable[d] the inference to be drawn that the [respondent] ha[d] the tendency asserted by the Crown" and, if so, the jury "[could] use that in considering whether it [was] more likely [the respondent] committed the specific offences with which he [was] charged".

4. [168]

   The Court of Criminal Appeal recognised that the direction, in referring to the jury deciding that the conduct in support of the alleged tendency had occurred rather than deciding if the alleged tendency had been established, was contrary to the recommendation in Director of Public Prosecutions (Vic) v Benjamin Roder (a pseudonym) ("Roder"). The Court of Criminal Appeal correctly acknowledged that such contrariety did not mean that a miscarriage of justice had occurred, because the question of miscarriage has to be decided in the context of a trial judge's summing up to the jury as a whole. The Court of Criminal Appeal concluded that the trial judge's tendency direction had occasioned a miscarriage of justice as it considered that: (a) "[f]raming the tendency and identifying the evidence to establish it so as to constitute exclusively the offending behaviour itself is inconsistent with the nature of tendency evidence", consistently with the reasoning in Kanbut v The King ("Kanbut"); (b) the trial judge's tendency direction was "illogical and confusing"; (c) where, "as in this case, the tendency is expressed in the same terms as the conduct alleged, such a direction is likely to encourage, if not require, the jury to engage in the impermissible circular reasoning identified in JS [v The Queen ('JS')] at and Roder[] at -"; and (d) the trial judge's summing up as a whole "was not designed to, and could not adequately ... address the difficulties that arose as a result of the tendency being expressed in precisely the same terms as the acts relied upon to establish the tendency and relied upon to establish the guilt of the [respondent] in respect of the relevant counts".

5. [169]

   For the reasons given below, the Court of Criminal Appeal's conclusion that the trial judge's tendency direction occasioned a miscarriage of justice is unsustainable. Accordingly, the Crown's appeal must be allowed and the respondent's convictions on counts 3, 5 and 7 must be reinstated. Inconsistency with the nature of tendency evidence?

6. [170]

   The reasons of Gageler CJ, Gleeson and Beech-Jones JJ explain why Kanbut does not establish any principle capable of general or specific application to a case in which the evidence said to establish the alleged tendency is the same evidence as that said to prove beyond reasonable doubt the charged acts constituting the offences. Leaving Kanbut aside, the first problem the Court of Criminal Appeal perceived with the trial judge's tendency direction, that identifying the alleged tendency and identifying the evidence said to establish the alleged tendency as the same evidence relied upon to prove the offences charged is "inconsistent with the nature of tendency evidence", does not exist.

7. [171]

   An alleged tendency can be proved by evidence of either or both uncharged and charged acts. In New South Wales, s 161A of the Criminal Procedure Act 1986 (NSW) regulates tendency and coincidence directions. Section 161A provides that: "(1) A jury must not be directed that evidence needs to be proved beyond reasonable doubt to the extent that it is adduced as tendency evidence or coincidence evidence. (2) If evidence is adduced as both tendency evidence or coincidence evidence and as proof of an element or essential fact of a charge before the jury, the jury may be directed that the evidence needs to be proved beyond reasonable doubt, but only to the extent that it is adduced as proof of the element or essential fact. (3) Subsection (1) does not apply if a court is satisfied – (a) there is a significant possibility that a jury will rely on an act or omission as being essential to its reasoning in reaching a finding of guilt, and (b) evidence of the act or omission has been adduced as tendency evidence or coincidence evidence."

8. [172]

   Section 161A ensures that, in proving an offence (and subject to statutory modification), it is only each element of the offence and any fact essential or indispensable to proof of an element of the offence which must be proved beyond reasonable doubt. The trier of fact in a criminal trial, jury or judge, need not satisfy itself to that standard of proof in respect of the existence of other facts and circumstances relevant to the ultimate finding of guilt.

9. [173]

   Where an accused's alleged tendency is formulated as a tendency to act as identified in charged acts and that alleged tendency is sought to be proved by the evidence said to establish the charged acts, the potential problem is not that the alleged tendency or the evidence said to prove it is or may be "inconsistent with the nature of tendency evidence". Evidence that an accused tended to act in one way on one occasion (whether that occasion involves a charged or uncharged act) is likely to have significant probative value that the accused tended to act in the same way on another, similar occasion. Such evidence is quintessential tendency evidence, as the reasoning in R v Dennis Bauer (a pseudonym) makes clear. Tendency direction "illogical and confusing" and encouraged or required "circular reasoning"?

10. [174]

    The second problem that the Court of Criminal Appeal perceived with the trial judge's tendency direction, that the direction was "illogical and confusing" and encouraged, if not required, the jury to engage in "circular reasoning", is also not sustainable. This is so even if the purportedly offending part of the tendency direction is considered in isolation from the balance of the trial judge's summing up to the jury.

11. [175]

    The relevant authorities propose that it is better for a trial judge to direct a jury to decide only if the evidence said to support an alleged tendency rationally proves that the accused has the tendency (the accused having the tendency not needing to be proved beyond reasonable doubt) rather than if the evidence rationally proves that the conduct said to disclose that the accused has the tendency occurred (the conduct not needing to be proved beyond reasonable doubt). The potential problem identified is that, if the evidence said to prove the alleged tendency includes charged acts, the latter form of direction, focusing on the occurrence of the conduct, might cause a jury to misunderstand the standard of proof ultimately required to prove criminal guilt of those charged acts. The potential misunderstanding might arise from the jury being directed to consider the evidence of conduct twice – once for the purpose of considering if the accused has the tendency and then for the purpose of considering if the charged acts have been proved – applying two different standards of proof (a rational inference standard in respect of proof of the accused having the alleged tendency and beyond reasonable doubt in respect of proof of the charged acts). Whether this risk can be said to be a real and tangible risk in any case will always depend on the trial judge's summing up as a whole in the circumstances of the case.

12. [176]

    In the present case there was nothing illogical, confusing or that encouraged, let alone required, circular reasoning in the purportedly offending part of the trial judge's tendency direction, even if that part is considered in isolation from the balance of the summing up. The trial judge directed the jury to consider the three occasions said to found a rational inference that the respondent had the alleged tendency collectively, rather than in isolation, to decide "what conduct occurred" on the basis that the standard of proof of beyond reasonable doubt did not apply to this decision. The trial judge further directed the jury that if it was satisfied that "all or at least some" of the conduct occurred, it had to consider if this enabled a rational inference to be drawn that the respondent had the tendency the Crown alleged and, in so doing, had to consider whether there were alternative explanations (that is, alternative to the existence of the alleged tendency) for such conduct. The reference to "all or at least some" of the conduct is innocuous because it was said in the context of the jury being directed to decide if it could draw a rational inference as to the respondent having the alleged tendency. Such an inference would not be rational if, for example, the jury was satisfied that the relevant conduct occurred on one occasion only or on more than one occasion but not because of the alleged tendency. The trial judge then directed the jury that if it was not satisfied that any of the conduct occurred, it had to put the issue of the respondent's alleged tendency to one side; it was only if the jury found the respondent did have the alleged sexual interest in GC and the tendency to act on that interest in the way the Crown alleged that it could use the tendency in considering if it was more likely that the respondent had committed the offences with which he was charged.

13. [177]

    It may be accepted that if, for the purpose of deciding whether the respondent had the alleged tendency, the jury considered that the evidence did not give rise to a rational inference that any of the conduct said to disclose the tendency occurred, then it would have followed that the jury had to acquit the respondent of the charged offences. That is because it would then be impossible for the jury to be satisfied beyond reasonable doubt that the charged acts constituted by that conduct had been committed. The real point the trial judge was making, however, is clear – that the jury could use the tendency as evidence to support it being satisfied beyond reasonable doubt that the respondent had committed the offences only if it had first found, on a rational basis, that the respondent in fact had the tendency in question.

14. [178]

    Contrary to the reasoning of the Court of Criminal Appeal, nothing in this part of the trial judge's tendency direction encouraged or required the jury to engage in illogical, confused or circular reasoning. JS at and Roder at - do not support such a conclusion. The point made in JS at was merely a rebuttal of an accused's submission that the Crown had "sought to prove the commission of an offence (at a standard of proof less than beyond reasonable doubt), and then relied on that finding to prove the tendency, and then relied on the tendency to prove the offence". As Basten A-JA explained in JS at , that is not the process of reasoning involved in respect of an alleged tendency in which a jury has regard to all the evidence relied on in proof of the tendency as evidence of the tendency alleged and, if satisfied of the existence of the tendency, thereafter relies on the tendency as circumstantial evidence aiding in proof of the charge beyond reasonable doubt. Roder at - did no more than explain and then reinforce that the actual process of reasoning in respect of tendency evidence involves nothing more than a jury being directed to consider the same evidence "at different stages of its deliberations with a different onus of proof and for a different purpose". It is not to be merely assumed in any case that a jury is incapable of faithfully applying the substance of such a direction, whatever the precise terms in which the direction is given, without any real risk arising of the jury misunderstanding the ultimate onus of proof of beyond reasonable doubt which the jury must apply to each element of an offence as charged.

15. [179]

    The submissions for the respondent effectively accepted that, if the trial judge had referred to the existence of the alleged tendency rather than the occurrence of the alleged conduct, the respondent could not or at least was unlikely to be able to establish a miscarriage of justice. This exposes the error involved in assuming that there exists some strict logical divide between a jury's rational satisfaction of the existence of the alleged tendency (said to be proved by the alleged occurrence of certain conduct on more than one occasion) and a jury's rational satisfaction of the alleged occurrence of certain conduct on more than one occasion (said to prove the existence of the alleged tendency). There is no such logical divide. There is simply a recognition in the authorities that instructing a jury in terms of the jury's satisfaction of the existence of the alleged tendency rather than in terms of the jury's satisfaction of the alleged occurrence of certain conduct reduces a risk, which might or might not arise depending on the facts of the case and the trial judge's summing up as a whole, of the jury misunderstanding the required reasoning process. Summing up as a whole did not rectify problem?

16. [180]

    Even if that part of the trial judge's tendency direction that the Court of Criminal Appeal considered offended against principle and logic had done so, the balance of the trial judge's tendency direction, as well as the balance of the summing up, would have ensured that no miscarriage of justice occurred in the present case.

17. [181]

    The existence of some correlation between the number of occasions said to demonstrate the alleged tendency and the likelihood of evidence of those occasions having probative value outweighing any unfairly prejudicial effect that evidence may have on an accused (as required by s 101(2) of the Evidence Act for that tendency evidence to be admissible if adduced by the prosecution) may be accepted. But everything depends on the evidence in the particular case, which the jury is required to consider by applying "its collective experience of ordinary affairs" and "common sense". Even if the evidence is confined to only two occasions said to support the alleged tendency and the tendency is expressed in terms that reflect the elements of the offences charged, that does not mean that the evidence is incapable of satisfying the test in s 101(2). Nor does it mean that the jury is necessarily more likely to be confused as to the standard of proof it must apply to each element of the offences as charged of beyond reasonable doubt. It would be wrong to assume that a jury reasons by or must reason by reference to a logic that depends on the number of occasions the alleged tendency is said to have been demonstrated rather than the nature and quality of the evidence said to demonstrate the alleged tendency.

18. [182]

    This is why in the present case the respondent's submissions to the following effect must be rejected: (a) the alleged tendency was of minimal if any logical relevance; (b) because the evidence relating to the first occasion of the alleged conduct said to demonstrate the alleged tendency was (as may be accepted to have been the case) weaker than the other two subsequent occasions, there were only two occasions of the alleged conduct said to demonstrate the alleged tendency (in itself an illogical proposition as the jury might have been satisfied that this occasion demonstrated the existence of the alleged tendency despite not being satisfied beyond reasonable doubt that the alleged conduct occurred on this occasion); and (c) because the evidence of each occasion was that of the complainant, GC, alone, rather than being from a source external to GC, it followed that the evidence could be relevant only to the credit of the complainant and not to the alleged tendency, and the problem identified by the Court of Criminal Appeal both existed and was "acute" in the present case. The evidence of a complainant alone of alleged conduct of an accused on more than one occasion can be tendency evidence, the probative value of which outweighs any unfairly prejudicial effect such evidence might have on the accused, being circumstantial evidence relevant to the proof that the accused committed the offence and not merely to the credit of the complainant. Such evidence does not give rise to the problem of illogical, confused or circular reasoning as the Court of Criminal Appeal proposed (or, as submitted for the respondent, "bootstrapping" reasoning). The most that could be said is that, depending on a trial judge's tendency direction in the circumstances of the case and in the context of the trial judge's entire summing up, a tendency direction focusing on the occurrence of the conduct said to demonstrate the alleged tendency rather than on the existence of the alleged tendency might (not will) give rise to a miscarriage of justice by reason of a real risk of the jury being distracted from the requirement that it be satisfied as to each element of each charged offence beyond reasonable doubt.

19. [183]

    In the present case, as discussed, no such real risk arose from the supposedly problematic part of the trial judge's tendency direction. Moreover, the balance of the trial judge's tendency direction comprehensively eliminated any such putative risk. The trial judge, in completing the tendency direction, continued by saying that: "Finding the [respondent] did have the tendency the Crown asserts is not enough by itself to prove his guilt beyond reasonable doubt in relation to each charge. It is simply part of the process of reasoning and it may assist the Crown to prove the [respondent] committed each of the offences. The question is whether it makes it more likely the [respondent] conducted himself in the way the Crown alleges on any of the occasions that are subject of the charges. That is the only way the [respondent's] tendency to have a sexual interest in the complainant and act on that sexual interest in the manner alleged may be used. Ultimately you must decide whether the specific offences with which the [respondent] has been charged have been proved beyond reasonable doubt. That decision must be based upon the evidence relevant to each charge. This includes the evidence of the complainant about what the [respondent] did, it will include the tendency alleged by the Crown provided you are satisfied it has been established, it will also include all of the other evidence you have heard in both the Crown and defence cases."

20. [184]

    In circumstances where the ultimate question is whether the trial judge's tendency direction involved a real risk of undermining the jury's understanding that it needed to be satisfied of each element of each offence beyond reasonable doubt, it is impossible to give that question an affirmative answer in the face of these instructions to the jury. To this may be added the numerous occasions throughout the summing up, before and after (as well as during) the tendency direction, in which the trial judge directed the jury that it could find the respondent guilty of an offence only if the Crown proved beyond reasonable doubt each element of the offence and that the jury must find the respondent not guilty if the Crown failed to prove any such element beyond reasonable doubt. It may also be added that the fact that the jury acquitted the respondent of counts 1 and 2 (the relatively weaker counts) is itself inconsistent with the existence of any real risk that the jury misunderstood the requirement to apply the standard of proof of beyond reasonable doubt to the counts relating to the other two occasions of offending. Conclusion

21. [185]

    For these reasons the appeal must be allowed and the respondent's convictions reinstated.